1 15 10 https://www.nal.usda.gov/exhibits/speccoll/files/original/18e316bef07da71f9df02c5519c300e1.pdf 9cbfa08f7cf70f8ad0a01bd3e24864b4 PDF Text Text Item D Number °5598 D Not Scanned Lathrop, George D. Epidemiology Division, Data Sciences, Division, USAF S Report/ArtldO TitlB Pr °iect Ranch Hand II: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Baseline Mortality Study Results Journal/Book Title Year 1983 Month/Day June3 Color ° D Number of Images ° DOSCrlptOD NOtQS Also included are summaries of the baseline mortality study results (June 30,1983), the baseline morbidity study results (February 24, 1984), and the mortality update (December 10, 1984. Alvin L. Young filed these documents together with others in a folder labeled, "Agent Orange Working Group Science Panel, Current Folder." Tuesday, March 19, 2002 Page 5598 of 5611 �I PROJECT RANCH HAND II AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES BASELINE MORTALITY STUDY RESULTS 30 JUNE 1983 Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 Approved for public release; distribution unlimited EPIDEMIOLOGY DIVISION DATA SCIENCES DIVISION USAF SCHOOL OF AEROSPACE MEDICINE (AFSC) BROOKS AIR FORCE BASE, TEXAS 78235 �Unclassified SECURITY CLASSIFICATION OF THIS PAGE (When DatafEntered)i REPORT DOCUMENTATION PAGE 1. REPORT NUMBER «. Tl T L t 2. GOVT ACCESSION NO READ INSTRUCTIONS BEFORE COMPLETING FORM 3. RECIPIENT'S CATALOG NUMBER 5. TYPK OF REPORT A PKRIon C O V K R I n («"</ S AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES: BASELINE MORTALITY STUDY RESULTS Annual Report, Initial Report 6. PERFORMING OHG, REPORT NUMBER 7. AUTHORfs) 6. CONTRACT OR GRANT NUMBERS; 9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT, PROJECT, TASK AREA ft WORK UNIT NUMBERS George D. Lathrop, Colonel, USAF, MC; Patricia M, Moynahan; Colonel, USAF, NC; Richard A. Albanesei M.D.; William H. Wolfe, Lt Colonel, USAF, MC USAF School of Aerospace Medicine (EK) Aerospace Medical Division (AFSC) Brooks Air Force Base, Texas 78235 11. CONTROLLING OFFICE NAME AND ADDRESS The Surgeon General United States Air Force Washington, D.C. 20311* 1Z. REPORT DATE U. MONITORING AGENCY N'AME ft ADDRESSfff dltterent Irom Controlling Ottice) 15. SECURITY CLASS, (ol thti report) 30 June 1983 13. NUMBER OF PAGES 61 Unclassified I5«. DECLASSIFI CATION/DOWN GRADING SCHEDULE 16. DISTRIBUTION STATEMENT (oi this Report) Approved for public release; distribution unlimited. 17. DISTRIBUTION STATEMENT (of the abstract entered In Block 20. It different from Report) IB. SUPPLEMENTARY NOTES 19. KEY WORDS (Continue on reverse side H necessary and Identity by block number) Epidemiologic investigation Air Force Health Study Mortality study RANCH HAND Matched cohort design Nonconcurrent prospective design '0. ABSTRACT (Continue on reverse side if necessary and identify by block number) In 1979 the United States Air Force (USAF) made the commitment to Congress and to the White House to conduct an epidemiologic study of the possible health effects from chemical exposure in Air Force personnel who conducted aerial herbicide dissemination missions in Vietnam (Operation RANCH HAND). The purpose of this epidemiologic investigation is to determine whether long-term health effects exist and can be attributed to occupational exposure to herbicides. This study uses a matched cohort design in a nonconcurrent prospective setting, incorporating mortality, morbidity, and follow-up studle; DD , F O R M73 <|.|« EDITION OF 1 NOV 65 IS OBSOLETE JAN WO Unclassified 7 SECURITY CLASSIFICATION OF THIS PAGE (When Data Entered) �unclassified SECURITY CLASSIFICATION OF THIS PAGE(HTi«n Data Entertd) The purpose of this report is to present the baseline mortality study results. As of December 31» 1982, 50 Ranch Hand and 250 comparison subjects had died (certified on/before April 27, 1983). Data analysis showed that the mortality experience of the Ranch Hand group is nearly identical to that of this comparison group. However, this mortality report can in no way be regarded as conclusively negative since the study population may not yet have reached the latency period. Subsequent mortality reports will include additional anaylses and will be updated annually for the next 20 years. Unclassified S E C U R I T Y C L A S S I F I C A T I O N OF T U I ' PAGEfWien Dalf Entered) �EXECUTIVE SUMMARY Baseline Mortality Study The Ranch Hand II epidemiclogic study uses a matched cohort design in a nonconcurrent prospective setting, incorporating mortality, morbidity, and follow-up studies. The purpose of this report is to present the baseline mortality study results. Since 1979, a detailed population ascertainment process has enumerated a total of 1269 Ranch Hand personnel who served in Vietnam during the period of 1962-1971. As described in the protocol, this total is believed to comprise, the entire exposed study population. The eligibility of each Ranch Hander was verified by a hand review of his personnel record, A comparison group was formed by identifying all individuals assigned to selected Air Force organizational units with a mission of flying cargo to, from, and in Vietnam during the same period. All Ranch Hand and comparison subjects designated as killed in action were removed from the study population. By a computerized nearest neighbor selection process, up to 10 comparison individuals were matched to each Ranch Hander by job category, race, and age to the closest month of birth. A hand record review of the matched comparison sets revealed that on the average, 8.2 comparison individuals were fully suitable for study. From each matched comparison set, five individuals were randomly selected for the mortality study, yielding a 1:5 design. Every Ranch Hander and his set of comparisons will be the subjects of annual mortality updates throughout the entire 20 years of the follow-up study so that emerging mortality patterns or disease clusters may be detected with maximal sensitivity. Each living Ranch Hander and his first and willing comparison match were selected to participate in a comprehensive physical examination and an in-home interview; the results of this study will be presented in a subsequent report in late 1983. A mortality determination on 1,247 Ranch Handers and 6,171 comparison subjects was made, sequentially using the data sources of the Air Force, Veterans Administration, Social Security Administration, Internal Revenue Service, and personal contact efforts. As of December 31, 1982, 50 Ranch Hand and 250 comparison subjects had died (certified on/before April 27, 1983). Death certificates were obtained on all 300 deceased subjects and were coded by an Air Force nosologist (ICD, 9th ED). All codings were verified by the National Center for Health Statistics. Autopsy results are currently being sought for future analyses. Statistical analyses of noncause specific death emphasized survival curve estimates, linear rank procedures, relative risk estimates, and standardized mortality ratios (SMRs). Cause specific analyses were limited to relative risk estimates because of small cell sizes. In addition to these approaches, three other data bases were contrasted to the Ranch Hand population, where possible; the 1978 US White Male Mortality experience, the 1978 Department of Defense (DoD) Nondisability Retired Life Table, and the mortality experience of the West Point Class of 1956. These additional comparison groups have substantial comparability or sample size limitations, rendering conclusions to the weakest order. Analyses with these "external" comparison groups were accomplished to crudely define the healthy worker effect and to determine if the Ranch Hand group mortality was drastically out of line with that of other military populations. �Data analysis showed that the mortality experience of the Parch Hand group is nearly identical to that of the comparison group. Analyses showed that, officers are living longer than enlisted personnel in both Ranch Hand and comparison groups. This difference between officers and enlisted personnel was statistically significant in the comparison group whereas it was not in the Ranch Hand cohort. A contrast of the Ranch Hand and comparison group to the 1978 DoD Life Table showed significantly less mortality for Ranch Hand officers, comparison officers and comparison enlisted men, however, there was not a statistically significant favorable mortality rate for Ranch Hand enlisted personnel. This pattern of mortality was also seen in a contrast of the Ranch Hand and comparison groups to the 1978 U.B. white male mortality experience. That is, highly favorable mortality differentials for Ranch Hand officers, comparison officers and comparison enlisted personnel were observed, but not for Ranch Hand enlisted. This trend is consistent with the self perceptions of differential herbicide exposures reported by many of the Ranch Hand subjects. The reason(s) for these observations are speculative at present, but may include the related items of sample size, socioeconomic differences, access to medical care, and health education and possible herbicide effects. Cause specific analyses were statistically nonsignificant. The Ranch Handers showed a relative paucity of overall cancer but an excess of digestive disorder deaths, both statistically nonsignificant. No soft tissue sarcoma deaths were detected in either group. Analyses of both the Ranch Hand and the comparison groups to the 1978 US White male mortality experience showed highly significant favorable findings. Most of these differences are speculatively attributed to the healthy worker effect. A contrast of the Ranch Hand and comparison groups to the 1978 DoD Life Table showed significantly less mortality for Ranch Hand officers and comparison officers and enlisted men. The West Point comparison showed nonsignificant SMRs of 0.530 and 0.778 for the Ranch Hand officers and the comparison group officers, respectively. Overall, the limitations of the statistical power calculations in most of these analyses were substantial in most analyses due to 1) the low mortality rate (4%) in the Ranch Hand and comparison groups to date, 2) the inherently small group of Ranch Handers (as described in the study protocol), and 3) the observed relative risks which approached unity in most categories. This baseline mortality report can in no way be regarded as conclusively negative because this small, young, and relatively healthy cohort may not have yet reached the latency period wherein attributable fatal disease might be expected and detected within limited power boundaries of this study. Future commitments for the annual mortality updates include detailed covariate analyses for disease risk factors, herbicide exposure, and confounding industrial chemical exposures. Further, subsequent morbidity reports will include full spectrum, disease specific analyses, e.g., cancer (fatal, ongoing, cured) in an effort to enhance study sensitivity to emerging herbicide effects, if they occur. ii �PREFACE In October 1978, the United States Air Force (USAF) Surgeon General made the commitment to the Congress and to the White House to conduct an epidemiologic study of the possible adverse health effects arising from the herbicide exposure of Air Force personnel who conducted aerial dissemination missions in Vietnam (Operation Ranch Hand). The purpose of this epidemiologic investigation is to determine whether long-term adverse health effects exist and whether they can be attributed to occupational exposure to herbicides and their contaminants. The study protocol (1) for this effort incorporates a matched cohort design placed in a nonconcurrent prospective setting. The study approach includes mortality, morbidity, and follow-up elements linked tightly in time, in order to produce the most data in the shortest period of time. The study addresses the question: Has there been, or are there currently, or will there be any adverse health effects among former Ranch Hand personnel caused by repeated occupational exposure to 2,4,5-T containing herbicides and the contaminant, TCDD? At the request of the Principal Investigators (see Appendix I) the study protocol was extensively and independently peer reviewed. The review agencies included: The University of Texas School of Public Health, Houston Texas; the USAF Scientific Advisory Board; the Armed Forces Epidetniological Board; and the National Research Council of the National Academy of Sciences. In 1980, the Science Panel of the Agent Orange Working Group was created as an additional peer review agency. This group, redesignated as the Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants, has consented to the oversight responsibility of the Ranch Hand study and continues to monitor the conduct of this epidemiologic investigation (see Appendix II). The approved and official protocol for this effort is available to the public through the National Technical Information Services, 5285 Port Royal Road, Springfield, Virginia 22161. The Ranch Hand II Study protocol heralds the suboptimal statistical power of the mortality study. The mortality study was motivated by the desire to use a full spectrum epidemiologic approach to the herbicide question. Additionally, the investigators were scientifically obliged to pursue the mortality study because of previous and emerging studies (some with small sample sizes) which suggested the possibility of a soft tissue sarcoma end point (2,3,4). Within the inherent sample size limitation of the Ranch Hand population, detection of such a rare condition will be missed unless there is marked case clustering and correspondingly high relative risks. Also, because of sample size limitations as well as the myriad of proposed clinical end points, a case-control design was not entertained. The investigators have attempted to enhance statistical power and analytic sensitivity where possible by using ] a large comparison group, 2) precise ) matching procedures, 3) annual mortality updates, A) mortality-morbidity linkages, 5) a lengthy follow-up study, 6) external comparison groups, and 7) state-of-the-art statistical methodology. A final assessment of overall mortality must necessarily await substantially more data and covariate approaches to identify and isolate unusual emerging mortality patterns, if they occur. iii �This report is primarily directed to individuals with statistical and epidemiologic backgrounds. It also assumes that the reader has a familiarity with the herblcide/dioxin issue and a detailed 'knowledge of the protocol of the Air Force study. In the interest of brevity, the reader is referred to the protocol published as US Air Force School of Aerospace Medicine Technical Report 82-44. iv �Page Executive Summary Preface ............................. ............................... Table of Contents Chapter I. . . .iii ............................. THE MORTALITY STUDY DESIGN 1 . The Study Population 2. The Mortality Population Chapter II. i .......................... ........................ THE MORTALITY DETERMINATION PROCESS 1. Introduction ....................... . ...... 2. United States Air Force Data Sources .................. 3. Veterans Administration Death Beneficiary Identification and Record Location Subsystem ........................... 1. Other Governmental Data Sources .................... 5. Morbidity Population Tracking ..................... 6. Receipt and Coding of Death Certificates ................ 7. Results ................... ............. Chapter III. RANCH HAND VERSUS COMPARISON GROUP ANALYSES 1. Introduction v 1 1 2 4 4 4 6 6 6 7 8 9 9 2. Overall Comparisons 9 3. Noncause Specific Occupational Comparisons 15 4. 17 Cause Specific Ranch Hand Versus Comparison Mortality Chapter IV. NONCAUSE SPECIFIC COMPARISONS WITH EXTERNAL POPULATIONS 1. Background and Motivation 2. Adjustment Difficulties A. Adjustment for Calendar Year of Death 20 20 20 20 B. Adjustment for Military Status (Active Duty, Separated, Retired). . 20 C. Adjustment for Selection D. Adjustment for Branch of Service 21 21 �3. Comparisons with 1978 DoD Life Tables ................. 4. Comparisons with U.S. 1978 White Male Life Table ......... 21 . . . 24 Chapter V. COMPARISONS WITH THE WEST POINT STUDY GROUP 1 . Background and Motivation ..... . ................. 2. Noncause Specific Comparisons of Ranch Hand and Comparison Subgroups with the West Point Study Group .................... 3. Cause Specific Comparisons ....................... Chapter VI. STATISTICAL ASPECTS 1. 2. 3. 4. 5. 6. 28 28 31 34 Purpose ................................ 3^ Survival Curve Estimates and Confidence Bands ............. 34 Linear Rank Procedures ......................... 35 Relative Risk Estimation ....... . ............ .... 36 Indirect Standardization ........... . ............ 37 Comparing Observed Life Table Data with a Known Survival Curve. .... 37 Chapter VII. CONCLUSION 1. 2. 3. 4. 5. 28 39 Introduction .................. . ...... . . . . . 39 Internal Comparison Group . . . . . . . ..... . .......... 42 External Comparisons. . . . . ......... . . . . . . . . . . . . 42 Power Considerations. . . . ....... . . . . . . . . ....... 42 Consistency Patterns ...... . ................... 43 APPENDICES I. Ranch Hand II Principal Investigators, Coinvestigators, Contributors, and Management Personnel 45 II. Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants . . 47 III. Matching Results in the Mortality Population 48 IV. Year of Birth, Occupational and Race Specific Mortality 51 V. The Extended E j i gou-McHugh Relative Risk Estimator ..53 VI. Survival Curves and Confidence Bands for Ranch Hand and Comparison Subgroups 54 REFERENCES 59 VI �Chapter I THE MORTALITY STUDY DESIGN 1. The Study Population The exposed study population, termed "Ranch Hand", was defined as those individuals who were formally assigned to the USAF organizations responsible for the aerial dissemination of herbicides and insecticides in the Republic of Vietnam from 1962 to 1971. These individuals were identified from historical data sources at the National Personnel Records Center (NPRC), St. Louis, Missouri and the USAF Human Resources Laboratory, Brooks Air Force Base, Texas. A total of 1,269 Ranch Hand personnel were eventually identified through this process. The comparison population was defined as those individuals who were assigned to a variety of cargo mission organizations throughout Southeast Asia during the same time period. Cargo mission aircrew members and support personnel were selected because of sufficient population size, similar training and military background experiences, and psychologic similarities to the Ranch Hand group. The comparison population was not occupationally exposed to herbicides or insecticides in the Republic of Vietnam. Identification of this population was completed using the same historical data sources as were used with the Ranch Hand population; 2*1,971 individuals were so identified. In preparation for matching the study and comparison populations, all subjects killed in action (KIA) were removed from the data base. The rationale for this action is the assumption that combat death in the Ranch Hand group was not caused by the immediate effects of herbicide exposure; KIA's were removed from the comparison group for comparability purposes. A KIA analysis will be performed in a subsequent report. The Ranch Hand KIA subgroup, numbering 22 individuals, although not matched, was maintained in the data base but was deleted from the mortality analysis, leaving 12M7 Ranch Hand subjects. The Ranch Hand population was matched to the comparison population with an iterative nearest-neighbor computer program (1). Up to 10 comparison subjects were matched to each Ranch Hander by year of birth, race (Black versus non-Black), and occupational category (officer pilot, navigator and other; enlisted flight, engineer and other), thus creating matched sets of one study subject and up to 10 comparison subjects. All subjects are males. The mean age of the study subjects is *J5 years. Following the original match, the majority of Ranch Handers had 10 comparisons. The exceptions were the group of non-Black pilots who had a mean of only 9.5 comparisons per exposed subject due to the extreme ages of several individuals, and the strata of Black pilots and other Black officers who only had means of 2.7 and 5.0, respectively. In December 1981, the USAF Principal Investigators learned that several morbidity study comparison subjects had reported no experience in Southeast Asia, suggesting that overselection of the comparison population had occurred (1). Manual review of the comparison subjects' military personnel records revealed that 18 percent of the 12,193 comparison individuals in the original match were ineligible for study. The inadvertent inclusion of several non-Southeast Asia organizations resulted in the selection of these inappropriate individuals. These ineligible subjects were found to be randomly distributed throughout the matched sets and were removed from the study. Following the removal of the �ineligible subjects, the study was reduced to a 1:8 design. Also during this period, five Ranch Hand subjects were identified through personnel record sources and Veterans Administration Education Benefits and Financial Records. These five individuals had not been identified earlier because the majority of their military personnel records had been destroyed in a fire at the NPRC in St. Louis. Three of these five were newly found Ranch Handers and two were comparisons subsequently identified as Ranch Handers. No attempt was made to match comparisons to these five new Ranch Handers. During the removal of ineligible subjects, one Ranch Hander, a Black officer pilot, lost his only comparison and remains unmatched, giving a total of six unmatched Ranch Handers. All six of these unmatched Ranch Handers are included in the morbidity and mortality studies. They were used in the analyses where appropriate, in order to improve statistical power. 2. The Mortality Population Five comparisons per exposed subject were considered more than adequate for mortality analyses; this estimate has recently been verified under a multiplicative model by Breslow, et al. (5). Up to five comparisons in each matched set, were identified from the 1:8 cohort as the mortality comparisons. Since the positions of the individuals in the matched sets had already been randomized in the data file, the selection of the first five positions in each matched set array for membership in the mortality comparison resulted in a random selection of the mortality comparison cohort. If a Ranch Hander had at least one but no more than five comparisons after removal of the ineligibles, then all of his matched set were used in the mortality component of this study. The mortality population is, therefore, defined as the 1241 matched Ranch Handers and their randomly chosen mortality comparisons (6171 individuals) and the six unmatched Ranch Handers. Table 1 summarizes the mortality population by occupational category and race. Here, and elsewhere in this report, non-Black is defined as Caucasian, Mexican or Oriental. Table 1 MORTALITY POPULATION SUMMARY BY OCCUPATION AND RACE Occupation, Race Officer-Pilot, Non-Black Officer-Pilot, Black Ranch Hand Counts Comparison 349 6 1740 13 Officer-Navigator, Non-Black Officer-Navigator, Black 80 2 390 10 Officer-Other, Non-Black Officer-Other, Black 25 1 123 2 189 15 935 75 528 52_ 1247 2628 255 6171 Enlisted-Fit Eng, Non-Black Enlisted-Fit Eng, Black Enlisted-Other, Non-Black Enlisted-Other, Black �The overall match ratio, 6171/1247=4.95, reflects the lack of suitable controls in some strata, the subsequent removal of ineligible comparisons and the addition of five unmatched Ranch Handers. A detailed description of the matching results is given in Appendix III. Those Ranch Handers having fewer than five matched mortality controls are summarized in Table 2. Table 2 RANCH HAND SUBJECTS WITH LESS THAN FIVE COMPARISON SUBJECTS Counts Occupation, Race Officer-Pilot, Non-Black Ranch Hand Comparisons-^ Notes Officer-Navigator, Non-Black Officer-Other, Non-Black Officer-Other, Black Enlisted-Fit Eng, Non-Black Enlisted-Other, Non-Black Enlisted-Other, Black 2 3 1 1 0 1 1 2 1 Officer-Pilot, Black 1 1 1 2 3 1 1 1 2 2 1 2 12 1 0 4 2 0 J» 0 2 1 1 2 1 2 27 Note 1. Lack of suitable comparison subject or loss due to ineligibility. Note 2. New Ranch Hander, no attempt to match. Note 3. Comparisons per Ranch Hander 1 �Chapter II THE MORTALITY DETERMINATION PROCESS 1. Introduction The mortality status of the Ranch Hand group and their mortality comparisons are, and will continue to be, ascertained using four major data sources: USAF, Veterans Administration (VA), other Governmental and morbidity population tracking. The mortality determination process using these data sources is presented in Figure 1. Figure 1. RANCH HAND II MORTALITY DETERMINATION ALGORITHM RANCH HAND II MORTALITY STUDY ENTIRE STUDY POPULATION 1 U.S. AIR FORCE ACTIVE DUTY ^ RETIRED PERSONNEL CENTER ACCOUNTING t FIKAHM CASUALTY BRANCH WSC FACILITY IHFUR. BENEFICIARY^ DIED ACTIVE DIED AF HQSP^ ALL I i DEATH BFJiEFIT ' SOCIAI SECURITY AOVIS. CLOSED BY DEATH INTEMIAL RLTOUE ACTIVE ACCT. CLOSED BY DEATH 3E 5 JE 3 X Ut ALIVE f ,__ TRACKING CONTACT UTTER LOWS HARRIS DEAD EAO - REOtiESf DEATH CERTIFICATE VETERml'S ADMIN. (GIRLS, SGI.I) The entire study population was matched or checked against the first three sections of this algorithm while only the morbidity population was contacted and tracked. A description of the data sources within the algorithm follows. 2. United States Air Force Data Sources The USAF data sources include the USAF Military Personnel Center (MFC) records, the USAF Accounting and Finance Center records, and the USAF Medical �Service Center Facility Use Data. The USAF MFC records include the individual's military personnel record and the data accumulated by the Casualty Branch of the MFC. Individual military personnel records are created at the time of induction into the USAF, and reflect a chronological history of the individual's military career. Epidemiologically, these records are an invaluable data source as they can be used for the development of occupational histories, identification of race, sex, and date of birth as well as for location of personnel, and for determining vital status. Hard copy records of these data are maintained at the individual's base of assignment while on active duty; a computer copy of these records is maintained at the USAF military personnel center, Randolph AFB, Texas. Following retirement and/or separation from the USAF, these records are forwarded to the National Personnel Records Center (NPRC), St Louis, Missouri, the record repository for all military personnel records. They are indexed by Social Security Account Number or Air Force Serial Number at the NPRC. If an individual should die while on active duty, after retirement, or within 120 days of separation from active duty, it is the responsibility of the Casualty Branch of the USAF Military Personnel Center to update the hard copy military personnel record and the MFC computer data base and to inform the USAF Accounting and Finance Center of this fact. At the same time, USAF MFC personnel initiate a copy of the USAF Form 1312, Report of Retired Casualty, or Department of Defense DD Form 1300, Report of Casualty. The selection of the appropriate form is based on the current status of the individual concerned. The DD Form 1300 also clarifies an individual's casualty status which can be either battle or nonbattle. Copies of the appropriate death form are sent to appropriate agencies while the original is placed in the individual's military personnel record. Since the initial review of military personnel records, a system has been established with the Casualty Branch of the Military Personnel Center wherein all active duty and retired death forms are forwarded monthly to the Occupational Epidemiology Section of the USAF School of Aerospace Medicine Epidemiology Division. In this way, the mortality status of all active duty and retired study subjects is systematically determined on a continuing basis. The USAF Accounting and Finance Center data base was used as a resource to update individual Air Force serial numbers to Social Security numbers. The Social Security number is required for all other aspects of the mortality algorithm. The Air Force Medical Service Center (AFMSC) Facility Use Data is a computer data base containing information regarding all active duty and retired deaths that .occur in Department of Defense (DOD) Medical Facilities. This data base identified no additional deaths in the mortality population, but did verify the deaths known to have occurred in DOD hospitals. In addition to the USAF data bases, the Ranch Hand Association, a reunion association of approximately 850 Ranch Handers, has contributed to the success of this study. This group has assisted the Principal Investigators in the ascertainment of the exposed population, and in the determination of the current location and the mortality status of the group. The association contacts all of its members yearly through newsletters and provides updated information to the Air Force investigators. �3. Veterans Administration Death Beneficiary Identification and Record Location Subsystem The Beneficiary Identification and Record Locator Subsystem (BIRLS) is a Veterans Administration data base generated by the Veterans Administration for determination of funeral allowance. If the family of the deceased informs the funeral director that the deceased served in the US military, the funeral director submits the required data to the Veterans Administration. In January 1981, August 1982,and January 1983,the BIRLS data base was searched for Ranch Hand and comparison deaths. In addition to these searches, the Department for Veterans Benefits, Veterans Administration, coordinated the gathering of death certificates from VA regional offices. 4. Other Governmental Data Sources A. Internal Revenue Service Public Law 96-126, Section 502, 28 November 1979, authorized the use of Internal Revenue Service (IRS) addresses for individuals who had been exposed to occupational hazards in order to determine the status of their health. The National Institute for Occupational Safety and Health (NIOSH) coordinated the USAF requests for these IRS addresses. This system is based on the address shown on individual tax returns and is corrected once a year. The addresses are verified by NIOSH through use of a post card mailed to the post office responsible for the individual's mail delivery. NIOSH assumes that the individual is alive if he files a tax return and if the verification scheme confirms his address for mail delivery. The IRS assumes an individual ia dead if the individual is so reported on a joint tax return. The IRS data base search provides an incomplete mortality determination, however, since absence of an individual tax return does not necessarily imply death of that individual. B. Social Security Administration The Social Security Administration (SSA) is a source of mortality information based on data maintained by the Office of Remuneration and Earnings. The basis for this data is employer-reported earnings. The SSA assumes that an individual is living if there is no indication of death on the individuals record and earnings are recorded for the last calendar year or retirement, disability, black lung or supplemental security income payments are being made. The SSA did inform us that they do not conduct an exhaustive search, and all deaths are not necessarily reported to SSA. Therefore, this mortality information may not be complete. 5. Morbidity Population Tracking Individual tracking techniques apply only to the morbidity population, defined as those selected and compliant to questionnaire. The morbidity population for this effort is defined as all Ranch Handers and their morbidity comparisons. The morbidity comparisons are, in general, also mortality comparisons. The selection procedure for the morbidity study is presented in Figure 2. �Figure 2. SELECTION PROCEDURE FOR THE QUESTIONNAIRE, PHYSICAL EXAMINATION, AND FOLLOW-UP STUDY Living Ranch Hand Comparison Individuals Randomly Ordered Mortality Comparisons r -A—"^ r~ 1 t i 1 11 +-* #* 1:1 Morbidity Match + Dead - Unwilling * Volunteer ** Replacement candidates In this figure, the first randomly ordered comparison was found to be dead. The second was contacted but was unwilling to participate and the third volunteered to participate in the questionnaire component of the morbidity effort. This contacting process for the morbidity effort was the final step in the baseline mortality determination. The original contact was made by certified mail. Each Ranch Hander and a random living comparison were sent an introductory letter and fact sheet signed by the USAF Surgeon General. A Louis Harris and Associates (LHA) interviewer then accomplished an in-home interview. LHA identified two Ranch Handers and nine comparisons who could not be located. All eleven unlocatable subjects were assumed living and remain included in the mortality study. 6. Receipt and Coding of Death Certificates Death certificates were ordered from the vital statistics department of the appropriate state, trust territory, or foreign country. Death certificates or their equivalent were obtained on all appropriate subjects. All death certificates were coded by two individuals, trained by the National Center for Health Statistics (NCHS) in underlying and multiple cause of death coding procedures, using the International Classification of Diseases, Ninth Edition (1977) coding system. Classification of the underlying cause of death was in accordance with NCHS decision tables. Each coder independently classified the underlying and multiple, causes of death and gave the coding worksheet, with each corresponding death certificate, to the coding supervisor, a trained nosologist, for reconciliation. Following reconciliation, one of the coders placed the death code information, by computer terminal, in the death certificate mortality file via a blind verification program designed to mimic the NCHS underlying multiple cause of death coding sheet. At the �conclusion of this initial input of the death codes, a copy of the death certificate was forwarded to NCHS for further validation. The NCHS returned coded death certificates, which were then compared with the Air Force classification. Discordances were resolved in cooperation with NCHS and entered into the data base. 7. Results Chapter II has reviewed the comprehensive, cohesive, sequential ascertainment process of death in the study populations. This process has resulted in the identification of 50 dead Ranch Hand subjects and 250 dead comparison subjects. Although it is understood that early differential ascertainment occurred in the Ranch Hand members (because of detailed knowledge of the study group), it is judged that the overall comprehensive ascertainment process is currently balanced with respect to the two groups. Table 3 and Appendix IV contain summary counts by age, job, and race category for all Ranch Handers and their mortality comparisons; these counts reflect mortality as of 31 December 1982, as known on 27 April 1983. In the stratified analyses, the term "at risk" is defined as simply the number of subjects within a specific stratum, and in life table analyses, as the number of subjects entering a specific age bracket. The term "rate" is the proportion of those individuals "at risk" who are dead. Table 3 OCCUPATIONAL AND RACE SPECIFIC MORTALITY Race Occupation Non-Black Officer-pilot Officer-navigator Officer-other Enlisted-fit eng Enlisted-other Black Officer-pilot Officer-navigator Officer-other Enlisted-fit eng Enlisted-other TOTAL Ranch Hand At Risk Dead Rate Comparisons At Risk Dead Rate 3*»9 80 25 189 528 12 2 1 6 25 .03* .025 .040 .032 .047 1740 390 123 935 2628 72 13 3 46 97 .041 .033 .024 .049 .037 6 0 0 0 2 2 50 .000 .000 .000 .133 .038 .040 13 10 2 75 255 6171 0 0 0 9 10 250 .000 .000 .000 .120 .039 .041 2 1 15 52 1247 �Chapter III RANCH HAND VERSUS COMPARISON GROUP ANALYSES 1. Introduction Overall survival comparisons, without regard to cause of death, were made via survival curve estimation, linear rank procedures, relative risk estimation and standardized mortality ratios. Survival curves were estimated and plotted using the method of Kaplan and Meier (6); 95% confidence bands (7) for each survival curve estimate were also plotted on each graph. Linear rank testing was carried out using the logrank test and Prentice's censored data extension of the Wilcoxon test (8). All linear rank tests were carried out with matched sets merged when Ranch Hands differed by less than one year relative to date of birth, within each stratum of job and race (9). These merged matched sets were regarded as separate strata for testing purposes (9, 10, 11). Relative risk estimates and confidence intervals were computed using an extension of the method of Ejigou and McHugh (12) to variable length, one-to-many matched sets (see Appendix V). Here, due to the one-to-many limitation of the algorithm, matched sets were not merged as when testing procedures were performed. Standardized mortality ratios and associated tests and plots were carried out as in Gail (13). These analyses are fully adjusted for the matching variables, age, race and occupation, but are unadjusted for other variables of interest, such as length of time in Vietnam or Southeast Asia, herbicide dose, time since exposure, time in active duty military, and other medical or occupational risk factors. Some of these variables, such as herbicide dose and time since exposure will be adjusted for in the next analyses, after such data become available. In particular, latency analyses cannot be undertaken at this time but will be included in the next mortality report. In these analyses, we have used summary statistics for which underlying modeling assumptions can be tested. For this reason, we have used the Breslow-Day (13) approach to SMR calculation, rather than the more traditional person-years method. A detailed explanation of this choice is given in Chapter VI. 2. Overall Comparisons Survival time in these analyses was regarded as independent of censorship, if any, and was taken to be age at death. All subjects not certifiably dead, as of 31 December 1982, at the time of analysis, were considered censored at their age on that date. Contact has been lost with two Ranch Handers and nine comparisons as described in Chapter II, but these are not assumed lost to follow-up for the purpose of mortality determination. They are assumed to have been alive on 31 December 1982. With this assumption, no subjects were lost to mortality follow-up before 31 December 1982 in this study. Ranch Hand and comparison group survival curve estimates and their associated 95% confidence bands are shown in Figure 3 and Appendix VI for the five groups: pooled, officers, enlisted, flying and ground personnel, as defined in Table 4. The curves for the pooled groups are shown in Figure 3 with the 95% confidence interval bands deleted in the interest of legibility, but they are included in the group1 specific curves in Appendix VI. Review of �Ranch Hand operations has strongly suggested that Ranch Hand enlisted personnel were more heavily exposed to herbicide than Ranch Hand officers. Further, there is a perception of possible exposure differential between flying and ground Ranch Hand personnel. These notions prompted the above groupings and analyses seen in this and subsequent chapters. Analyses of latency are not possible at this time due to the as yet incomplete nature of the military service data base. These analyses will be performed after the hand review of military tour records has been completed. Figure 3 SURVIVAL CURVE ESTIMATES FOR POOLED RANCH HANDERS AND COMPARISON SUBJECTS 1 0.9 0.8 Z 0.7 t—i > > 0,6 a: co 0.5 0 « * 0.4 •-• 0.3 0.2 0-1 0 R - RANCH HAND C - COMPARISON 10 20 30 YEflRS 70 �Table 4 GROUP DEFINITIONS Group Definition Officer Enlisted Officer-pilot, navigator, other Enlisted-flight engineer, other Flying Officer-pilot, navigator Enlisted-flight engineer Ground Officer-other Enlisted-other Pooled All occupational categories Summary counts by group are shown in Table 5. Ignoring the matching, interaction between officer-enlisted categories and Ranch Hand membership, and interaction between flying-ground categories and Ranch Hand membership was evaluated using log-linear models. No statistically significant interactions were detected. Table 5 SUMMARY COUNTS BY GROUP Group Ranch Hand At Risk Dead Rate Comparisons At Risk Dead Rate Officer Enlisted 463 784 15 35 .032 .045 2278 3893 162 .039 .042 Flying Ground 641 606 22 28 .03** .046 3163 3008 140 110 .044 .037 Pooled 1247 50 .040 6171 250 .041 88 Linear rank procedures were carried out on the same five groups. The results, summarized by test statistics and two-sided P-values, are shown in Table 6. Small P-values, less than .05, indicate significant differences, at the 5% level, between the two groups. These procedures are designed so that the statistic will be positive when the Ranch Handers are dying before the comparison subjects and negative when the comparisons are dying prior to the Ranch Handers. The null hypothesis is that the actual survival distributions of Ranch Handers and their matched comparisons are identical. Each statistic is approximately null distributed as a standard normal random deviate. 11 �Table 6 TEST RESULTS AND P-VALUES FOR OVERALL COMPARISONS Group Logrank (Value) P-Value Wilcoxon (Value) P-Value Officer Enlisted (-0.631) (0.383)' .526 .702 (-0.722) (0.33D .470 .741 Flying Ground (-1.021) (1.023) .307 .306 (-1.116) (0.950) .264 .342 Pooled (-0.047) .962 (-0.123) .902 There is no significant difference, based on these data, between the Ranch Handers and their mortality comparison group. This means that, in particular, the mean ages-at-death of the Ranch Handers and their matched comparisons are not significantly different. In some groups, pooled, officer and flying, the statistics are negative, indicating that the Ranch Handers are living longer than the comparisons, but the differences are, again, insignificant, as evidenced by the large P-values. The situation is reversed for enlisted and ground personnel. These findings are consistent with the observation that, within each group, the comparison confidence bands are contained within the Ranch Hand confidence bands. When matched sets are stratified by five year intervals on year of birth, the same procedures give larger P-values than those in Table 6. Relative risk estimates, the associated 95$ confidence intervals, twosided P-values for testing the null hypothesis of relative risk equal to unity and the associated power are given in Table 7. Here, the power of the test is defined as the conditional probability of rejecting the null hypothesis at the 5% level of significance, given that the relative risk is equal to its estimated value. Table 7 RELATIVE RISKS AND 95$ CONFIDENCE INTERVALS, P-VALUES AND POWER Group Relative Risk 95% Confidence Interval P-Value Power Officer Enlisted 0.763 1.065 (.320- 1.207) (.660- 1.471) .373 .742 .105 .072 Flying Ground 0.734 1.232 (.387 - 1.081) (.694- 1.769) .211 .337 .197 .195 Pooled 0.964 (.658- .819 .051 12 1.269) �The confidence intervals and P-values in Table 7 indicate no significant difference, at the 5% level, between the mortality of the Ranch Handers and comparisons in each of the five groups. Year-of-birth specific mortality rates for each of the five groups are given in Tables 8 through 12, with the corresponding standardized mortality ratios (SMR). In each group, the comparisons are the internal standard. The SMR estimates relative risk in these comparisons if the year-of-birth specific relative risks are all equal (13). A likelihood ratio test for the hypothesis of equal year-of-birth specific relative risks was carried out for each comparison; its P-value is denoted by P1 . In addition, the hypothesis that relative risk is unity, given that relative risk is constant across strata, was tested via a likelihood ratio procedure (13); its P-value is denoted by P2. The SMR and both P-values are given with each comparison. Here, and elsewhere in this report, the denominator of the SMR where n^j is the number of individuals for the ith stratum population and r^ is the death rate, per person, in the standard for the ith stratum. In these calculations the data is stratified birth. Table 8 POOLED SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR - .996; P1-.389, P2-.955) Birth Year Ranch Handers At Risk Dead Rate Comparison At Risk Dead Rate 1905-14 5 2 .1100 1*1 2 .143 1915-19 17 4 .235 96 11 .115 1920-24 1925-29 48 84 3 2 .063 .024 241 501 24 40 .100 .080 1930-31 1935-39 304 207 15 7 .049 .034 1389 1020 67 33 .048 .032 1940-44 1945-54 208 374 5 j_2 50 .024 .032 1096 1814 23 50_ 250 .021 .028 13 is of the jth population on year of �Table 9 OFFICER SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR - .827; P1-.233, P2-.490) Ranch Hand Officers Rate At Risk Dead Birth Year 1910-24 1925-34 1935-39 1940-44 191)5-49 1*1 194 93 90 45 3 4 4 2 2 15 Comparison Officers At Risk Dead Rate 205 930 458 495 190 .073 .021 .043 .022 .044 17 49 11 6 5 88 .083 .053 .024 .012 .026 Table 10 ENLISTED SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR = 1.074; P1=.733, P2=.722) Birth Year Enlisted Ranch Handers At Risk Dead Rate 1905-14 1915-19 1920-24 1925-29 1930-34 1935-39 1940-44 1945-54 4 9 16 41 153 114 118 329 2 1 3 2 11 3 3 10 35 Enlisted Comparisons At Risk Dead Rate .500 .111 .188 .049 .072 .026 .025 .030 12 54 80 211 749 562 601 1624 2 7 11 22 36 22 17 45 .167 .130 .138 .104 .048 .039 .028 .028 162 Table 11 FLYING SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR = .769; P1-.678, P2-.238) Birth Year 1915-24 1925-34 1935-39 1940-44 1945-49 Flying Ranch Handers At Risk Dead Rate 44 272 142 120 63 4 8 6 2 2 22 .091 .029 .042 .017 .032 14 Flying Comparisons At Risk Dead Rate 220 1316 698 653 276 23 71 22 14 10 140 .105 .054 .032 .021 .036 �Table 12 GROUND SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR = 1.257; P1-.535, P2=.302) Birth Year Ground Ranch Handera At Risk Dead Rate Ground Comparisons At Risk Dead Rate 1905-14 1915-24 1925-29 1930-34 5 21 31 85 2 3 2 7 .1400 .143 .065 .082 14 117 151 423 2 12 19 17 .143 .103 .126 .040 1935-39 1940-44 1945-54 65 88 311 1 3 1£ 28 .015 .034 .032 322 443 1538 11 9 40 110 .034 .020 .026 These SMR comparisons are in agreement with the preceding relative risk and linear rank analyses; there is no significant difference in mortality, based on these data, between the Ranch Hand group and the comparison group. 3. Noncause Specific Occupational Comparisons Within-group comparisons by occupation via SMR's, with P-values for testing constant relative risk across year of birth strata (P1) and for testing relative risk equal to unity (P2) are given in Tables 13 through 16. The enlisted and ground personnel are the internal standards in these comparisons. Comparisons via the logrank procedure are given in Table 17. Table 13 RANCH HAND OFFICERS VERSUS RANCH HAND ENLISTED MORTALITY BY YEAR OF BIRTH (SMR «= .544; P1-.280, P2= .087) Birth Year 1905-24 1925-34 1935-39 1940-44 1945-54 Ranch Hand Officers At risk Dead Rate 41 194 93 90 45 3 4 4 2 _2 15 .073 .021 .043 .022 .044 15 Ranch Hand Enlisted At Risk Dead Rate 29 194 114 118 329 6 13 3 3 lp_ 35 .207 .067 .026 .025 .030 �Table 14 RANCH HAND FLYING PERSONNEL VERSUS RANCH HAND GROUND PERSONNEL MORTALITY BY YEAR OF BIRTH (SMR - .581; P1-.382, P2-.100) Birth Year Ranch Hand Fliers At Risk Dead Rate 1905-24 1925-34 1935-39 1940-44 1945-54 44 272 142 120 63 4 8 6 2 _2 22 Ranch Hand Ground At Risk Dead Rate .091 .029 .042 .017 .032 26 116 65 88 311 5 9 1 3 1P_ 28 .192 .078 .015 .034 .032 Table 15 COMPARISON GROUP OFFICERS VERSUS COMPARISON GROUP ENLISTED MORTALITY BY YEAR OF BIRTH (SMR - .697; P1-.640, P2-.015) Birth Year 1905-19 1920-24 1925-29 1930-34 1935-39 1940-44 1945-54 Comparison Officers At Risk Dead Rate 44 161 290 640 458 495 190 4 13 18 31 11 .091 .081 .062 .048 .024 .012 .026 6 5 Comparison Enlisted At Risk Dead Rate 66 80 211 749 562 601 9 11 22 36 22 17 1624 45 T52 M .136 .138 .104 .048 .039 .028 .028 Table 16 COMPARISON FLYING PERSONNEL VERSUS COMPARISON GROUND MORTALITY BY YEAR OF BIRTH (SMR - .930; P1-.305, P2-.867) Birth Year 1905-19 1920-24 1925-29 1930-34 1935-39 1940-44 1945-54 Comparison Fliers At Risk Dead Rate 45 175 350 966 698 653 276 6 17 21 50 22 14 10 140 .133 .097 .060 .052 .032 .021 .036 16 Comparison Ground At Risk Dead Rate 65 66 151 423 322 443 1538 7 7 19 17 11 9 40 110 .108 .106 .126 .040 .034 .020 .026 �Table 17 LOGRANK WITHIN GROUP COMPARISONS Comparison Logrank P-Value RH Officer vs RH Enlisted RH Flyers vs RH Ground Comparison Officer vs Comp Enlisted Comparison Flyers vs Comp Ground -1.468 -1.455 -2.597 -0.363 0.1*12 0.146 0.009 0.717 The SMR and logrank analyses are somewhat in agreement, with both procedures finding significant differences between comparison officers and comparison enlisted, with the officers living longer. The two methods approximately agree on the Ranch Hand fliers versus ground personnel and on Ranch Hand officer versus enlisted personnel with the logrank result near significance at the .10 level; the fliers appear to be living longer than the ground personnel within the Ranch Hand group. 4. Cause Specific Ranch Hand Versus Comparison Mortality Cause specific mortality, relative risks, two-sided P-values for testing relative risk equal to unity, power and 95? confidence intervals for relative risks are summarized in Table 18 for the 1241 matched Ranch Handers and their mortality comparisons. Mortality data for the six unmatched Ranch Handers were not used in this analysis. Of the six, one has died of an accident and the rest are still alive. In some categories, the data were too sparse for relative risk estimation. Table 18 CAUSE SPECIFIC MORTALITY AND RELATIVE RISKS Cause Dead Relative RJH Comparison Risk 95% Conf Int. Accidental 18 Suicide 3 Homicide 2 Infectious, Parasitic 0 Malignant Neoplasm 4 Uncertain Neoplasm 0 Endocrine 1 Mental Disorder 0 Nervous System 0 Circulatory 16 Respiratory 0 Digestive 5 Genitourinary 0 111 Defined 0 Unknown _0_ 92 14 3 49 250 P-value Power .959 1.071 3.333 (.466 - 1.453) (0 - 2.407) (0 - 9.297) .875 .913 .099 .047 .061 .489 .503 (0 - 1.024) .205 .153 5.000 (0 -18.859) .102 .562 1.002 (.411 - 1.594) .994 .050 2.273 (0 - 4.675) .085 .457 3 39 2 1 1 2 70 4 11 3 2 3_< 17 �The low powers in Table 18 reflect the sparseness of data or the fact that some of the observed relative risks approach unity. However, two categories do stand out as deserving further attention: malignant neoplasms and digestive system deaths. It should be noted that if matched sets are ignored and relative risk is estimated using the method of Mantel and Haenszel (14), these results remain essentially unchanged; using this'approach, the relative risk for malignant neoplasms, for example, is .506 with a P-value of .195 and power equal to .254. The 95/& confidence interval for relative risk using this approach is .180 to 1.419. The Mantel-Haenszel relative risk for the digestive system comparison is 2.254, with a P-value of .132 and a power equal to .325; the 95$ confidence interval for relative risk is .782 to 6.501. The digestive system deaths are further defined in Table 19. There has been an increase in deaths due to liver disease among the Ranch Handers; however, this observed difference is not statistically significant. These data are also based on death certificate diagnoses and will be subjected to verification and validation from medical record and autopsy reports. When all deaths from liver disease are considered as a whole, a relative risk of 2.50 is found, with a 95% confidence interval of 0 to 5.501. The P value is 0.083. Similarly, the relative risk for pancreatitis is 2.50 with a 95% confidence interval of 0 to 8.501.; the P value is 0.386. These observations are of interest and will be pursued in depth in subsequent reports. Table 19 DIGESTIVE SYSTEM MORTALITY Deaths Ranch Hand Comparison ICD Code (9th Ed) Pancreatitis (5770) Alcoholic cirrhosis (5712) Nonalcoholic cirrhosis (5715) Nonalcoholic fatty liver (5718) Chronic liver disease (5728) Alcoholic liver disease (5711) Duodenal ulcer (5325) 1 0 3 0 0 1 0_ 5 2 3 3 1 1 0 _J_ 11 Table 20 SITE SPECIFIC MALIGNANT NEOPLASM MORTALITY Deaths Ranch Hand Comparison Site ICD Code (9th Ed) Lip, oral cavity, Pharynx (140-149) Digestive organs, peritoneum (150-159) Respiratory, intrathoracic (160-165) Bone, connective tissue, skin, breast (170-175) Genitourinary organs (179-189) Brain (191-192) Lymphatic and hematopoietic tissue (200-208) No site specification (199) 18 0 0 2 4 8 15 0 1 0 0 J_ 4 1 3 3 4 J_ 39 �The malignant neoplasms are detailed in Table 20, the cell types or the neoplasms, as recorded on the death certificates, are summarized in Table 21. Table 21 MORPHOLOGY OF NEOPLASMS ICD Code 9th Ed. M800 M801-80*» M805-808 M8HJ-838 M872-879 M905 M938-948 M959-963 M965-966 M986 Deaths Ranch Hand Comparison Nomenclature Neoplasms not otherwise specified (NOS) Brain Bronchus and Lung Colon Intestinal Tract Epithelial neoplasms (NOS) Bronchus and Lung Esophagus Kidney Nasopharynx Pancreas Unspecified site Papillary and Squamous Cell Nasal Sinus Lip Tongue Tonsil Andenomas and Adenocarcinomas Appendix Bronchus and Lung Colon Kidney Stomach Nevi and Melanomas Skin (NOS) Mediastinal Mesothelioma Bronchus and Lung Gliomas Frontal Lobe Brain (NOS) Lymphomas NOS and Diffuse Lymphomas (NOS) Hodgkins disease Hodgkin's (NOS) Myeloid Leukemias Acute Myelocytic Leukemia 0 0 0 0 1 3 1 1 1 0 1 0 0 1 8 1 1 1 2 1 0 0 0 0 1 1 1 1 0 0 0 0 0 1 2 1 2 1 0 1 1 0 0 1 0 0 1 1 0 1 0 2 0 4 Epithelial, papillary and adenomas account for 6H% of the comparison neoplasms. Three Ranch Hand neoplasms arose from epithelial cells. There were no tumors in either group which were classified as soft tissue sarcoma. 19 �Chapter IV NONCAUSE SPECIFIC COMPARISONS WITH EXTERNAL POPULATIONS 1. Background and Motivation It is important to know, not only how the Ranch Handers and their matched comparisons relate to each other, but also how they compare with general military and male United States populations. Pitfalls inherent in these comparisons are well known and are briefly reviewed below for specific comparisons with 1978 DoD period life tables for nondisability retired military officer and enlisted personnel (15) and the 1978 U.S. White Male Life Table (16). Although there are difficulties in the use of these comparisons, their use does provide an additional indicator of trends in mortality when viewed in the context of the total analytic process. 2. Adjustment Difficulties Mortality rates in any military population are strongly dependent upon 1) calendar year of death, 2) military status (active duty, separated, retired), 3) selection and retention, and 4) branch of service. Adjustment for these effects was not made in these comparisons because published select Air Force life tables, by calendar year and by status, are not available. In addition, there is also a problem with the statistical method used, since the Gail and Ware (17) procedure assumes constant relative risk with respect to age; the selection effect has been shown to diminish sharply with time making this assumption untenable in these comparisons. The adjustment difficulties (1-4), and their likely consequences, are detailed below. These difficulties apply to all of the comparison groups, but these concerns have less effect on the comparisons of the Ranch Hand group to their matched cohort since these two groups are generally equivalent, relative to these key factors. A. Adjustment for Calendar Year of Death Due to the continuing decrease in overall mortality in the military (18) and in the United States (19), the referenced external age-specific rates are appropriate only for the calendar period of the referenced external life table, that is, 1977-79 for the 1978 period military table used in this analysis. The 1977-79 period rates would, for example, be too low for comparison with subjects dying in 1970 at the age of 40. These subjects would more properly be compared with the death rate for 40 year olds in a 1970 period life table or with a death rate for 40 year olds in a cohort military life table for subjects born in 1930. Calendar time is not taken into account in this analysis because period life tables covering the three decades from 1950 to 1980, for the the active duty, separated and retired Air Force subpopulations, are not currently available. This discrepancy is serious because the decline in death rates in the active duty Air Force during the period 1966 to 1980 has been very substantial (18). B. Adjustment for Military Status (Active Duty, Separated, Retired). The only published military life tables available at this writing are 1978 period tables for DoD nondisability retired officer and enlisted personnel (15) and a series of yearly abridged tables for the active duty Air Force, the first covering the period 1966-1968 and the last, 1978-1980 (18). With these data limitations, adjustment for military status is not possible. It is 20 �olear, however, that there are substantial differences between active duty and retired death rates with the active duty rates being lower than retired rates (15). C. Adjustment for Selection Entry into the military carries with it an effect known as selection, a lengthening of life expectancy due to health prerequisites upon entry into select status and periodic health checks thereafter. This effect is well known to insurance actuaries who have observed that, in insured populations, the effect diminishes as time passes unless there are continued checks on the state of health of the insured persons (20). If selection is to be adjusted for in this analysis, it would be necessary to know Air Force death rates as a function of both age and of time elapsed since entry into the Air Force. It would also be necessary, therefore, to know enlistment and discharge or retirement dates for all study subjects. It is the lack of these data that makes this adjustment impossible at this time. The consequences of this lack of adjustment are not known at this writing. D. Adjustment for Branch of Service Age specific active duty Air Force death rates are substantially lower than the corresponding rates for other services (18). Nonservice specific death rates are therefore too high for appropriate comparison with these two study groups. 3- Comparisons with 1978 DoD Life Tables In Tables 22 and 23, Ranch Hand officers and comparison group officers are contrasted to a 1978 DoD nondisability retired officer life table (15) and in Tables 24 and 25, Ranch Hand and comparison group enlisted personnel are compared with a 1978 DoD nondisability retired enlisted life table (15). In each table, the column labeled "At Risk" lists the number of subjects entering each five year age interval, the column labeled "Deaths" tabulates the numbers of deaths in the age intervals and the column labeled "Expected Deaths" gives the expected numbers of deaths in the age intervals if the study subjects had experienced the same death rates as those specified by the DoD table. The value of the test statistic (17) for testing the null hypothesis of equality is denoted by T; its two-sided P-value is denoted by P. While each table summarizes the findings with five year age intervals for ease of presentation, one year age intervals were used for the computation of the statistic T. All comparisons are conditioned on survival to age 35, since the DoD tables begin at that age. All comparisons are unadjusted for race since the DoD tables are not race specific. 21 �Table 22 RANCH HAND OFFICER VERSUS DOD NONDISABILITY RETIRED OFFICER LIFE TABLE (T - -3.962, P < .001) Age At Risk Deaths Expected Deaths 35-39 40-44 45-49 50-54 456 366 288 173 2 1 1 1 4.183 4.244 4.578 3.099 55-59 57 1 2.043 60-64 30 2 .823 65-68 1 £ 8 .076 19.046 Table 23 COMPARISON OFFICERS VERSUS DOD NONDISABILITY RETIRED OFFICER LIFE TABLE (T - -2.402, P = .016) Age At Risk Deaths Expected Deaths 35-39 40-44 45-49 50-54 2264 1822 1365 842 12 13 24 12 20.837 20.703 21.920 15.901 55-59 60-64 308 145 9 4 10.265 4.377 65-68 .19 JD .601 74 94.604 Table 24 ENLISTED RANCH HANDERS VERSUS DOD NONDISABILITY RETIRED ENLISTED LIFE TABLE (T = -.239, P = .811) Age At Risk Deaths 35-39 668 6 40-44 45-49 50-54 55-59 60-64 392 287 140 41 20 5 5 5 2 2 65-69 6 0 70-71 1 J_ 26 22 Expected Deaths �10UJ.C C-J ENLISTED COMPARISON SUBJECTS VERSUS DOD NONDISABILITY RETIRED ENLISTED LIFE TABLE (T = -3.214, P = .001) Age At Risk Deaths Expected Deaths 35-39 40-44 45-49 50-54 55-59 60-64 65-59 70-74 3299 1945 1437 695 203 103 35 5 21 20 31 14 12 3 1 0 102 33.370 27.681 31.450 20.076 10.980 7.515 2.593 .646 134.311 These findings suggest that, if the effects discussed in section 2 are assumed to be negligible, Ranch Hand officers and comparison officers and comparison enlisted personnel are living longer than expected relative to their respective external populations. Enlisted Ranch Hand personnel are not different from DoD enlisted personnel. In the above DoD comparison there is a suggestion of interaction between officer-enlisted categories and Ranch Hand versus comparison group membership. If matching and time of death are ignored, the following table can be constructed. The term "rate" is as defined on page 8 of this report. Table 26 DEATH AFTER 35 YEARS Ranch Hand Alive Dead Rate Officer Enlisted 448 642 8 26 Alive .018 .039 Comparison Dead Rate 2190 3197 74 102 .033 .031 Analysis using log-linear models shows a statistically significant interaction with pS 0.05. It appears that Ranch Hand officers have a lower mortality after age 35 than Ranch Hand enlisted or comparison officers or enlisted. However, the converse situation is noted considering mortality prior to age 35 and is significant with pSO.05. The data for this analysis of mortality prior to age 35 is set out below. Table 27 DEATHS BEFORE AGE 35 YEAR Alive Officer Enlisted 456 775 Ranch Hand Dead Rate 7 9 > Alive .015 .011 2264 3833 23 Comparison Dead Rate 14 60 .006 .015 �These interactions will require further detailed analysis and evaluation, with specific consideration of medical covariables including risk taking, other life patterns and herbicide. 4. Comparisons with U.S. 1978 White Male Life Table Non-Black Ranch Handera and non-Black comparisons are compared in this section with the population of White males, as represented by the 1978 U.S. White Male Life Table (16). Two serious and well known problems with the use of this table are the lack' of adjustments for the calendar year and selection effects just described; when comparing occupational cohorts with national populations, the selection effect is known as the "healthy worker" effect. The pitfalls of these kinds of comparisons are well documented (21, 22, 23). In Tables 28 and 29, non-Black Ranch Handers and non-Black comparisons are compared, via the method of Gail and Ware (17), with the 1978 U.S. White Male Life Table (16). In Tables 30 through 33, non-Black officers and enlisted personnel in both study groups are compared with the same 1978 U.S. White Male Table. Table 28 NON-BLACK RANCH HANDERS VERSUS 1978 U.S. WHITE MALE LIFE TABLE (T--4.588, P <.001) Age At Risk Deaths Expected Deaths 21-24 25-29 1171 1169 2 6 9.003 9;783 30-34 35-39 1163 1054 7 7 9.396 9.256 40-44 722 5 10.381 45-49 50-54 549 304 6 5 12.085 8.114 55-59 60-64 65-69 70-71 98 50 7 1 3 4 0 1 W 5.039 2.790 0.669 0.089 76.605 24 �Table 29 NON-BLACK COMPARISONS VERSUS THE 1978 U.S. WHITE MALE LIFE TABLE (T = -11.230, P <.001) Age At Risk Deaths Expected Death 19-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 5816 5815 5799 5772 5245 3593 2675 1487 509 248 54 5 1 16 27 23 31 29 50 26 20 7 1 0 231 10.325 55.444 48.592 46.719 46.124 51.041 58.810 40.529 25.210 11:161 3.403 0.601 354.540 Table 30 NON-BLACK RANCH HAND OFFICERS VERSUS 1978 U.S. WHITE MALE LIFE TABLE (T «= -4.575, P < .001 ) Age 25-29 30-31 35-39 40-44 45-49 50-54 55-59 60-64 65-68 At Risk Deaths 151 151 147 362 285 172 57 30 1 3 1 2 1 1 1 1 2 _0 15 25 Expected Deaths 3.794 3.710 4.42Q 5.304 6.370 4.541 3.019 1.302 0.110 32.570 �Table 31 NON-BLACK COMPARISON OFFICERS VERSUS 197$ U.S. WHITE MALE LIFE TABLE (T - -7.923, P < .001) Age At Risk Deaths Expected Deaths 25-29 2253 9 18.880 30-34 2244 5 18;530 35-39 2239 12 22.137 40-44 45-49 50-54 55-59 60-64 65-68 1801 1352 834 308 145 19 13 24 12 9 4 _0 25.841 30.468 23.328 15.157 6.923 0.887 88 162.151 Table 32 NON-BLACK RANCH HAND ENLISTED PERSONNEL VERSUS 1978 U.S. WHITE MALE LIFE TABLE (T - -1.753, P - .080) Age At Risk Deaths 21-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-71 717 715 712 607 360 264 132 41 20 6 1 2 3 3 5 4 5 4 2 2 0 J_ 5.510 5.988 5.686 4.836 5.077 5.716 3.573 2.020 1.488 0.588 0.089 33 40.571 26 Expected Deaths �Table 33 NON-BLACK COMPARISON ENLISTED PERSONNEL VERSUS THE 1978 U.S. WHITE MALE LIFE TABLE (T =-5.923, ? < .001) Age At Risk Deaths Expected Death 19-19 20-21 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 3563 3562 3546 3528 3006 1792 1323 653 201 103 35 5 1 16 18 18 19 16 26 14 11 3 1 0 6.325 33-938 29.713 28.189 23.987 25.200 28.341 17.201 10.053 7.538 2.515 0.601 T4~3 213.601 Given the cautions just described, these findings suggest that the non-Black Ranch Handers and comparisons are living much longer than expected relative to the 1978 U.S. White Male Life Table. The ratios of the observed to the expected deaths described in Tables 28 and 29 reveal that the Ranch Hand and comparison subjects are experiencing death at only 60 to 6556 of the rate of the U.S. White male population. The ratio is 0.461 for the subset of Ranch Hand officers, 0.543 for comparison officers, 0.813 for enlisted Ranch Handers, and 0.669 for enlisted comparison subjects. The healthy worker effect is very likely a major contributor to the undoubtedly real differences between these study groups and the general population. 27 �Chapter V '. COMPARISONS WITH THE WEST POINT STUDY GROUP 1. Background and Motivation The statistical and epidemiological literature is replete with warnings against the uncritical use of the SMR and related summary measures for comparing study groups with published vital statistics for national populations or subpopulations (5), (24), (25). Those cautions are based on the adjustment difficulties described in Chapter 4, Section 2, and departures from the assumption of constant relative risk across age intervals between the study group and the external population. These drawbacks can be avoided by not referencing an external standard at all, by using one of the study groups as the standard (13). or by using as an external standard a group of military personnel, born during approximately the same years, with the same mortality follow-up, as the Ranch Hand and comparison groups. An external group of sufficient size for meaningful statistical comparisons is not available at this time. Mortality and year of birth data are available, however, on a small group of West Point graduates, the subjects of the West Point Follow-up. Study. Although this group is too small for all but very crude statistical comparisons (1), it is the only known external data available at this time. The following comparisons are, therefore, primarily descriptive. The West Point Study Group consists of W members of the West Point graduation class of 1956. These men have been followed up since then for morbidity and mortality. All members of that class were, or still are, officers in the U.S. Armed Forces. The purpose of the West Point study is to investigate the relationship between blood lipid levels and cardiovascular disease. Each study subject is physically examined biennially and blood samples are obtained for lipid and lipoprotein analyses at the USAF School of Aerospace Medicine (26). 2. Noncause Specific Comparisons of Ranch Hand and Comparison with the West Point Study Group Subgroups For the purpose of these mortality comparisons, 15 of the 36 known West Point deaths occurring on or before 31 December 1982 were deleted, 9 of the 15 were killed in action, one was killed in 1959 in the line of duty and 5 were killed in automobile crashes prior to 1962. These deletions imitate the deletion of personnel killed in action from the Ranch Hand and comparison groups. Noncombat or accidental deaths prior to 1962 were deleted because death prior to 1962 would have precluded membership in the Ranch Hand or comparison groups. In addition, one West Pointer who is also a Ranch Hander, was deleted; that individual was alive on 31 December 1982. 28 �A summary of the remaining 21 deaths among the 458 West Point subjects used in these analyses is given by year of birth in Table 34 and by age in Table 35. In Table 35 the column headed "censored" lists by age, the number of West Pointers alive on 31 December 1982. Table 34 WEST POINT DEATHS BY YEAR OF BIRTH Year of Birth At Risk Dead 1930 1931 1932 1933 1934 1935 20 59 90 136 141 12 458 0 2 6 8 4 J_ 21 Table 35 WEST POINT DEATHS BY AGE Age 25-29 30-34 35-39 HO-44 H5-H9 50-52 At Risk Censored 458 456 451 4 4 8 lji)6 162 Dead 0 0 0 0 276 161 437 2 5 3 2 8 J_ 21 In this analysis, non-Black Ranch Hand and comparison officers are compared, without regard to cause of death, with the West Point study group; all of the West Point subjects are non-Black. Non-Black Ranch Hand Officers were matched, one-to-one, by year of birth, to West Point subjects. Due to the relatively small number of Ranch Hand officers and the limited year of birth range imposed by the age of the Class of 1956, only 283 of the 458 West Point subjects received a matched Ranch Hander. Matched sets with West Pointers having the same year of birth were then merged to create six matched sets, corresponding to the six years of birth, 1930 through 35, of the West Pointers. To compare West Pointers with comparison officers, two non-Black comparison officers were matched to each West Pointer by year of birth. All West Pointers received two matched comparison individuals. Matched sets with West Pointers having the same year of birth were merged, giving six. matched sets containing a total of 916 comparisons. 29 �Logrank teats were carried out on these two matched data sets, and the results are summarized in Table 36. In these analyses, survival time is age at death. Censorship is due to survival to 31 December 1982. For those still alive on 31 December 1982, censoring time is age on that day. Table 36 STUDY GROUP VERSUS WEST POINT GROUP LOGRANK COMPARISONS WITH TWO-SIDED P-VALUES Comparison P-Value Ranch Hand officer versus West Point Comparison officer versus West Point .218 .528 An SMR analysis, with the West Pointers being the standard, is summarized in Table 37. Table 37 SMR COMPARISON OF NON-BLACK RANCH HAND AND COMPARISON OFFICERS WITH THE WEST POINT STUDY GROUP (SMR = .530) Birth Year 25-31 32 33-34 35-40 (SMR = .778) Ranch Hand (SMR = 1) Comparison West Point At Risk Dead Rate At Risk Dead Rate At Risk Dead Rate 95 35 60 93 2 1 1 1 8 .021 .029 .017 .043 272 164 257 223 19 7 6 _5 37 .070 .043 .023 .022 79 90 277 12 2 6 12 J_ 21 .025 .067 .043 .083 The test for constant relative risk across year of birth strata gives a P-value of .229. Further, a likelihood ratio test suggests that these SMR's are not different (P = .392). �j. Cause Specific Comparisons The <VHi.se .sponi f \.r. death count,0, for the West Point Study Group are givon in Table Table 38 WEST POINT MORTALITY BY CAUSE Cause Count Accidents Infectious disease Malignant neoplasms Circulatory Digestive Genitourinary 111 defined 6 1 6 5 1 1 1 21 Cause specific comparisons are carried out with three causes, cancer (malignant neoplasms), other diseases, and nondisease (accidents, suicides, homicides and ill-defined), with an adjustment for year of birth by stratification on year of birth. Relative risks are calculated using the method of Mantel and Haenszel (14). These results, based on the counts in Tables 39 and 40, are shown in Table 41. Table 39 CAUSE SPECIFIC COMPARISONS RANCH HAND OFFICERS VERSUS WEST POINT Ranch Hand At Risk Dead West Point At Risk Dead Cause Birth Year Nondisease 1925-1933 1934-1940 166 117 1 M 305 153 5 1 Cancer 1925-1930 1931 1932 1933 1934 1935-1940 72 23 35 36 24 93 0 0 0 0 0 0 20 59 90 136 141 12 0 1 3 1 1 0 Other diseases 1925-1934 1935-1940 190 93 2 1 446 12 8 1 31 �Table 40 CAUSE SPECIFIC COMPARISON COMPARISON OFFICERS VERSUS WEST POINT Birth Year Cause Nondisease Comparisons Dead Number Other diseases 1 11 2 1 2 79 90 136 153 2 2 1 272 164 148 332 2 2 1 79 90 136 153 3 1 1 436 148 109 223 9 1 1 169 136 141 12 1 5 2 1 1929-1931 1932 1933 193^-1937 272 164 148 1929-1931 1932 1933 1934-1937 Cancer West Point Dead Number 1929-1932 332 1933 1934 1935-1937 2 3 1 Table 41 CAUSE SPECIFIC RELATIVE RISKS, P-VALUES 95$ CONFIDENCE INTERVALS FOR LOG RELATIVE RISK 95% Conf Interval Cause Comparison Nondisease RH vs WP Comp vs WP 1.072 0.841 (-1.504 - 1.643) (-1.354 - 1.009) .931 .775 .051 .059 Cancer RH vs WP Comp vs WP 0.690 (-1.634 - .891) .564 .089 Comp vs WP 0.474 0.779 (-3.540 - 2.047) (-2.367 - 1.867) .600 .817 .082 .056 RH vs WP Comp vs WP 0.539 0.728 (-2.191 - .954) (-1.940 - .306) .441 .702 .120 .067 Other diseases RH vs WP All causes RR for Log Rel Risk 32 P-value Power �While the Ranch Hand versus West Point cancer comparison cannot be assessed using the Mantel-Haenszel procedure, the absence of Ranch Hand cancer deaths in this analysis is of interest. This finding is consistent with the apparent but nonsignificantly decreased Ranch Hand cancer mortality noted in the Ranch Hand versus matched comparison group analysis (Chapter III). 33 �Chapter VI STATISTICAL ASPECTS 1. Purpose The purposes of this chapter are 1) to briefly describe each statistical procedure used in the preceding chapters 2) to state the underlying assumptions of each procedure and 3) discuss the validity of those assumptions in this study. The procedures used in this analysis were survival curve estimates and confidence bands, linear rank tests, relative risk estimation and standardized mortality ratios. Points 1-3 are addressed for each procedure in Sections 2 through 5. 2. Survival Curve Estimation and Confidence Bands The survival function of a homogeneous population, S(t), is defined as the probability of surviving t years. The problem is to estimate S(t) and make a confidence statement about that estimate based on randomly censored data. Randomly censored data occur in survival studies since analyses are usually carried out before all subjects have failed. In the present application, failure is defined as death and censorship occurs because most subjects are still living at the time of analysis. Other causes for censorship in this kind of epidemiological study are loss to follow-up or death from causes other than those of interest. Thus far in this study, there have been no subjects lost to follow-up, and all causes of death are of interest. The survival function is estimated here by the product limit estimate K(t), also called the Kaplan-Meier estimate (6). This estimate is derived under the assumption that, in a life testing experiment with n subjects on test, exactly k subjects, with k less than n, are observed to fail; the other n-k remaining are observed only until they are censored. The subjects are assumed drawn randomly from a homogeneous population. Censorship is assumed to be independent of failure. The Kaplan-Meier estimator is asymptotically unbiased and reduces to one minus the empirical distribution function in the absence of censoring. In the present application, the homogeneous populations are the Ranch Handers, the comparisons and various subgroups of these two groups. Death time is taken as age at death measured to the nearest month; censoring time is age" on 31 December 1982, measured to the nearest month. Survival time is age at death or age on 31 December 1982 for those subjects still living. The process n[K(t)-S(t)] converges weakly to a zero mean Gaussian process, as n tends to infinity, under random censorship when the underlying survival function S(t) and the censoring distribution are continuous on a bounded interval (27). This convergence is the theoretical basis for the confidence band algorithm (7) used in Figures 2 and 3, Chapter III and Appendices VI. �The independence of death and censorship can be assumed to hold horc since censorship (survival to December 31, 1982) is not being invoked on individuals because they appear to be at unusually high, or low, risk of death (28). Direct contact has been lost with two Ranch Handers and nine comparisons as described in Chapter II, but these are assumed to be alive, and hence censored at their age on 31 December 1982. The reason for this assumption is that the extensive death ascertainment system is believed to be thorough enough so that, had any of these subjects died, the death would have been detected. Hence, while contact has been lost, loss to follow-up for the purpose of mortality determination has not occurred (29). All other subjects still alive on 31 December 1982 are censored at their age on that date. The validity of inferences based on the estimate K(t) and its associated confidence band depends on the sample size and the observed number of deaths. The sample sizes and numbers of deaths in every stratum used in these analyses exceed the minimum requirements for these procedures (7). The survival curve estimates and confidence bands displayed in Figures 2 and 3 and Appendix VI are not adjusted for year of birth. To do so would have required stratification on year of birth, creating many small strata with associated sample size difficulties. Some year-of-birth adjusted plots in the larger occupational strata will be presented in the next report. 3. Linear Rank Procedures The hypothesis of interest in this analysis is that the actual survival distributions of the Ranch Handers and their matched comparisons are identical. The procedures of choice for testing equality of the two unknown survival distributions based on the matched and censored data in this study are the censored data extensions of the exponential scores and Wilcoxon tests, due to Prentice (8). The first of these is widely known as the logrank test. The test statistics, T, are of the form given by equation 6-23 of (28), where the summands are calculated on matched sets consisting of survival information on one Ranch Bander and his matched mortality comparisons. The statistic T, for either logrank or generalized Wilcoxon summands, is approximately standard normal under the null hypothesis (9). tions other study favor The large sample normal approximation for T will hold when all distribuare continuous and all censoring times are mutually independent of each and independent of death. These assumptions are well satisfied in this since the censorship mechanism, survival to time of analysis, does not one group over the other. In these procedures, the sampling unit is a matched set, so that these tests are adjusted for all matching variables. Prior to calculation, matched sets with Ranch Handers in the same race and job classification having the same year of birth are merged. 35 �The logrank and extended Wilcoxon tests are locally most powerful when the logarithm of the survival times are distributed as extreme value or logistic random variates, respectively. While the efficiency of these procedures peaks at these two underlying distributionsf they have been shown to be robust against departures ( ) These distributional assumptions, however, are 8. not viewed as strictly valid in this study since there is good evidence in the literature that survival time due to certain cancers and other diseases is log normally distributed (30, 31, 32, 33). A linear rank procedure of the Prentice form, whose efficiency peaks under the lognormal distributional assumption, can be constructed" (3*0, but this algorithm is not available at the present time; it will be included in the next analysis. The effect of this departure from the assumptions is considered mild. It should also be noted that these distributional assumptions cannot be checked since these match sets are small and the observations in the combined samples of all matched sets cannot be assumed to have a common distribution. Therefore, reliance must be placed on historical data to determine which linear rank procedure to use. The logrank and Wilcoxon procedures are used here because they are powerful and widely accepted in epidemiology and statistics. 4. Relative Risk Estimation Two relative risk estimators are used in this analysis, a generalization of the Ejigou-McHugh estimator for one to many matched data (12) and the Mantel-Haenszel estimator for stratified data (14). The Ejigou-McHugh estimate was chosen because it allows full adjustment for the one-to-many year-of-birth matching in this study, it is asymptotically as efficient as the maximum likelihood estimator and it is noniterative. The Mantel-Haenszel estimate was chosen because of its ease of calculation, efficiency (35), and general acceptance. It's variance is estimated according to the advice of Anderson et al. (36). Recent work suggests that the variance of the Mantel-Haenszel statistic might be better estimated by a jack-knife procedure (37); this newer method will be carried out in the next mortality report. The Ejigou-McHugh estimator in its published form is suitable only for 1 to R matched designs in which the number, R, of controls matched to each case is the same for all cases. Since the number of controls matched to each Ranch Hander is not the same for all Ranch Handers, the Ejigou-McHugh estimate and its variance was extended to a one-to-many matched design in which the number of comparisons is allowed to vary from case to case. Since this extension is unpublished it is stated in Appendix V for reference. The extended estimate and its variance reduces to the Ejigou-McHugh estimate and variance when all matched sets contain an equal number of comparisons. It is asymptotically efficient and consistent and is noniterative. The Ejigou-McHugh estimate and the Mantel-Haenszel estimate are based on the assumption that relative risk is constant across levels of the matching variable. Some indication that this assumption holds in this study when the data is grouped, by stratifying on year of birth, is furnished by likelihood ratio testing; there is no evidence in this study to suggest that relative 36 �risk is not constant across levels of the matching variables when the event of interest is death from any cause. Therefore, the Ejigou-McHugh and Mantel-Haenszel estimates are appropriate for these data. 5. Indirect Standardization With either an external or internal standard, the SMR is a good summary mortality index for comparing two or more populations, provided the product model, P.jj=ripj, holds, where PJJ is the probability of death in stratum i of population j, r^ is a set of standard stratum specific rates and PJ characterizes the mortality of population j, 1=1,2, ..., I, j=1,2, ..., J, (38, 13). If standard rates are known from some external source and if the product model holds, the best estimate of pj is proportional to the SMR. If J=2, the product model holds, and if one of the two groups is used as the standard, the SMR estimates relative risk. In any case, any SMR summary of mortality data should be preceded by analytical and graphical tests of fit of the product model. Because one of the study groups was always used as the standard in these analyses, the test of fit of the product model was, equivalently, a test of constancy of relative risk across year of birth strata. The fit of the model was verified in each analysis. Further, a likelihood ratio test for equality of population was carried out as described by Gail (13). The results of both tests are summarized by their P-values in each application. The sample sizes in every application are large enough so that chi-square approximations hold; these analyses are, therefore, valid and appropriate. The expected number of deaths in the SMR used in these analyses was calculated as In^rj, where n^j is the number of subjects in the ith stratum of the jth population. The person-years SMR was not used here for two reasons. First, its validity as an estimator of relative risk is dependent upon the fit of the proportional hazards model for which an omnibus test is not currently available. Secondly, the person-years calculation is typically carried out from entry into follow-up (5); in this study, follow-up begins at first entry to Vietnam or Southeast Asia and these entry dates are being verified at this writing. 6. Comparing Observed Life Table Data with a Known Survival Curve The procedure of Gail and Ware (17) is used in these analyses to compare Ranch Hand and comparison group survival data with published period life tables. The basic assumptions of this procedure are that death and censorship are independent competing risks and that the reference curve is a survival distribution for some external population. The test is of the form e v wnere an e are ^°j~ j }/(I j ^ ^» °j $ j observed and expected numbers of deaths in age interval j, and Vj is the variance of OJ-BJ. The statistic is not an omnibus goodness-of-fit test consistent against all alternatives to the null hypothesis that the observed sample comes from a known survival distribution. Rather, it has good power against proportional hazards alternatives or, more loosely, against alternatives for which the observed survival is better (or worse) in every interval than predicted by the known survival curve. 37 �The independence of death and censorship assumption is well satisfied in these data, as discussed in Section 2 of this chapter. The life tables used in these analyses do not, however, represent the survival distribution of any population since they are period, not cohort, life tables. The appropriateness of this procedure is, therefore, dependent upon the extent to which these period life tables approximate the survival distribution of some relevant reference population. These period tables were used because the more appropriate cohort life tables were not available at the time of analysis. �Chapter VII CONCLUSION 1. Introduction The mortality analyses described in this report have not revealed any adverse death experience in the herbicide/dioxin exposed cohort. The results of the analyses, regardless of the source of the comparison data, were consistent: at this time, there is no indication that operation Ranch Hand personnel have experienced any increased mortality or any unusual patterns of death in time or by cause. They are not dying in increased numbers, at earlier ages, or by unexpected causes. The fact that only a relatively small number of Ranch Hand deaths were available for analysis is reassuring in itself. However, the fact that adverse effects have not yet been detected does not imply that an effect will not become manifest at a future time or after covariate adjusted analyses. For this reason, further analyses are intended and mortality in the study population will be ascertained annually for the next 20 years. A summary of the statistical techniques applied to each source of comparison data is presented in Table 42. It should be noted here that these analyses have been carried out without knowledge of covariate information, such as herbicide exposure, industrial chemical exposure, or other risk factors and that these analyses were carried out at a time when approximately 96? of Ranch Handers and their matched comparison subjects were still living. The data, therefore, must be viewed as preliminary to more definitive analyses, which will be performed over the next 20 years. Table i»3 summarizes the results of the noncause specific analyses by source of the comparison data, and Table M presents the results of the cause specific analyses. 39 �Table 42 SUMMARY OF STATISTICAL PROCEDURES USED IN ANALYSIS Comparison Database Internal Comparison Group 1978 1978 DoD U.S. Life West Point White Males Tables Class of 1956 Noncause Speoifio Analyses Logrank & Wilcoxon Procedures Ejigou-McHugh Relative Risk Mantel-Haenszel Relative Risk SMR/Breslow-Day Product Model Gail-Ware Procedure + + •«• + + + + + + Cause Specific Analyses Ejigou-McHugh Relative Risk Mantel-Haenszel Relative Risk + + + Procedure usage is indicated by a "+" symbol. 40 �SUMMARY OF NONCAUSE SPECIFIC MORTALITY ANALYSES BY SOURCE OF COMPARISON DATA Internal Comparison 1978 US Group White Males^ Ranch Hand Group RH = C RH0 = C0 RHE = CE RHp = Cp RHG = CG Comparison Group RH <«US RH0 «<US RHE SUS C <«US C0 <«US CE <«us Internal Occupational Group Specific « £ < « <« P P P P P 1978 DoD West Point Life Tables1 Class of 19562 RH0 «<DoD0 RHE = DoDE CQ < RH0 - WP0 C0 = WP0 RH0 $ RHE RHp $ RHG C0 < CE valued greater value equal to value equal to value equal to value equal to than .10 or less than .10 or less than .05 or less than .01 or less than .001 RH Ranch Hand Group C Comparison Group 0 Officers E Enlisted F Flying G Ground 1 2 Validity of these comparisons is questionable (see Chapter H) Statistical inference is limited by small sample size 3 All P value symbols are based upon SMR and Gail-Ware analysis Table W SUMMARY OF CAUSE SPECIFIC ANALYSES BY SOURCE OF COMPARISON DATA RH Versus Internal Comparison No significant difference in cause specific relative risks RH Versus West Point* No significant difference in cause specific relative risks * Statistical inference is limited by small sample size �2. Internal Comparison Group Based on these early results, there appears to be no significant difference between Ranch Handers and comparisons as regards mortality. This null finding holds for both cause specific and noncause specific comparisons. One within group comparison did yield a significant difference, however. The non-Black comparison officers are living significantly longer than the non-Black comparison enlisted personnel. This may reflect the underlying health care and socioeconomic differences between these two groups. Non-Black Ranch Hand officers also appear to be living longer than non-Black Ranch Hand enlisted personnel, but this finding cannot be viewed as significant, with a P-value of .142 (Table 17). This lack of significance in the Ranch Hand analysis might be attributed to the smaller group sizes within the Ranch Hand cohort in contrast to the comparison cohort. 3. External Comparisons As outlined in the study protocol, considerable effort was expended in the selection of the study comparison group. While the chosen comparison group appeared closest to the Ranch Hand cohort except for herbicide exposure, it seemed appropriate to also contrast the Ranch Hand mortality experience to that of additional comparison groups. Three additional comparison data sets were then selected: mortality data from the West Point Class of 1956, the DoD Nondisability Retired Officer and Enlisted Life Tables for 1978, and the U.S. White Male Life Table, also for 1978. These data sets were chosen in a hierarchical fashion with the expectation that, in the absence of a herbicide effect, the Ranch Handers would have: 1) a mortality pattern comparable to the West Pointers, 2) a lower mortality than the DoD group due to the healthy worker effect, and 3) a still lower mortality than the U.S. male cohort due to healthy worker and military selection effects. These expectations were reassuringly fully realized with respect to overall mortality. Additionally, interesting officer-enlisted differentials emerged. As discussed below, these officer-enlisted differentials may have resulted from sample size effects or from covariable effects, potentially including herbicide exposure. 4. Power Considerations The power limitations of this study, specifically regarding mortality from rare conditions, such as soft tissue sarcoma, were fully acknowledged and described in the protocol (Ref 1, page 67). For example, a fatal disease with an incidence of .001 would require an approximate risk of 4 for a power of 0.8. Power calculations, while desirable for planning and study design, are also revealing at analysis. They are, however, sometimes difficult to carry out without further assumptions. The powers of the logrank and Wilcoxon tests �and the likelihood ratio tests in the SMR analyses are not calculable at this time due to the lack of appropriate methodology. The powers of the tests for cause specific mortality were calculated at the estimated relative risk. The values were low because the estimates of relative risk were close to unity and/or the data were sparse. The null findings in this report are unlikely to have been observed by chance had the true group differences been substantial. For example, if the true overall relative risk were in fact equal to 2, a crude calculation gives a probability of .0007 of observing a relative risk smaller than the observed .96*1 (Table 7). This probability is less than .001 if the true relative risk is 1.5. These findings are, therefore, very likely reflective of a near overall equivalence between Ranch Handers and their matched comparisons. Finally, these unadjusted findings do not preclude the possibility of the emergence of significant differences after adjustment for risk factors. 5. Consistency Patterns When the analysis of each external comparison data base is considered separately, the restrictions inherent in each source limit the strength of the inferences which can be made. However, when the results of all internal and external comparison data bases are considered in context, some patterns of consistency emerge. While some of these patterns may not have firm statistical underpinnings, they still may provide epidemiologic clues to the dynamics of the mortality process. The Ranch Hand officers exhibit a very consistent and predictable pattern across all analyses. As shown in Table ^43, their mortality is nearly the same as that of their most equivalent comparison groups (the matched comparison group officers and the West Point group). As the comparison groups become progressively less equivalent to the Ranch Hand group, the relative mortality of the Ranch Hand officers improves, presumably due to selection comparability (healthy worker effect, etc.). Their mortality is lower than that of their enlisted counterparts; however, this difference is not as striking as is the statistically significant comparable analysis between the matched comparison officers and the matched enlisted personnel. Unfortunately, the cross-comparison trends for the enlisted Ranch Handers are not as clearcut. Their mortality is greater,though not significantly different from their matched comparisons. The enlisted comparison group had a highly significant underrepresentation of mortality against both the DoD and US life tables, whereas the Ranch Handers are equivalent to the DoD group and only marginally better than the 1978 US White males. The consistent observation that the enlisted Ranch Handers appear to demonstrate less of a difference in relative mortality than do their matched comparisons is intriguing. This may reflect an actual increase in mortality due to herbicide exposure or some other factor, or it could be an artifact of small sample size created by the 1:5 matching or basic comparability problems as previously described. The inclusion of substantially more subjects in one group than another can have a profound effect on the significance level of a �statistical technique. Nevertheless, these observations are of interest, and will continue to be subjected to detailed analysis throughout the course of the follow-up study. This trend is consistent with self-perception of herbicide exposure held by many of the Ranch Hand group. Covariate analyses will be conducted, the herbicide exposure index will be applied to these data, and the effects of interaction will be assessed to determine whether the Ranch Hand enlisted findings are real or artifactual. The next mortality assessment will include analyses by person-year of follow-up, adjusted for age in an effort to better address the issue of latency. As the number of deaths in the study population increases with the passage of time, all of the statistical approaches outlined in the protocol (1) will be applied to the data. �Appendix I RANCH HAND II PRINCIPAL INVESTIGATORS COINVESTIGATORS, CONTRIBUTORS, AND MANAGEMENT PERSONNEL A. Principal Investigators George D. Lathrop, MD, MPH, PhD, FACPM Colonel, USAF.MC Chief, Epidemiology Division Patricia M. Moynahan, RN, MS Colonel, USAF, NC Chief, Occupational Epidemiology Section Richard A. Albanese, MD, GM-15 Chief, Biomathematical Modeling Branch Data Sciences Division William H. Wolfe, MD, MPH, FACPM Lt Colonel, USAF,MC Chief, Epidemiology Services Branch B. Coinvestigators Joel E. Michalek, PhD, GS-13 Mathematical Statistician Data Sciences Division Richard C. McNee, MS, GM-13 Chief, Advanced Analysis Branch Data Sciences Division Alton J. Rahe, MS, GS-13 Mathematical Statistician Data Sciences Division �Appendix I (Continued) C. Contributors William J. Besich, BS, GS-12 Computer Systems Analyst Data Sciences Division Vincent V. Elequin, BS, RRA, GS-11 Medical Record Librarian Occupational Epidemiology Section William E. Nixon, BS, GM-13 Computer Systems Analyst Data Sciences Division Thomas J. White, MA Senior Subject Matter Specialist Data Sciences Division i Alvin L. Young, BS, MS, PhD Major, USAF Special Assistant in Military Herbicides D. Management Personnel Project Director: Roy L. DeHart-, MD, MPH, MS, FACPM Colonel, USAF, MC Commander, USAF School of Aerospace Medicine Project Resource Manager: Melvin B. Dobbs Colonel, USAF, BSC Director, Systems Acquisition Research, Development and Acquisition Aerospace Medical Division Air Force HQ Systems Command Coordinator: Ronald D. Burnett Colonel, USAF, BSC Command Bioenvironmental Engineer Office of the Command Surgeon Air Force Surgeon General Coordinator: Robert A. Capell Major, USAF, BSC Assistant for Bioenvironmental Engineering Office of the Surgeon General 46 �Appendix II SCIENCE PANEL Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants. Dr. G. W. Comstock Johns Hopkins Research Center Box 2067 Hagerstown MD 217^0 Dr. John Doull Professor Department of Pharmacology and Toxicology University of Kansas Medical Center Kansas City KA 66103 Dr. John A. Moore (Chair) Deputy Director National Toxicology Program P.O. Box 12233 Research Triangle Park NC 27709 Dr. Richard Monson Professor of Epidemiology Harvard School of Public Health 677 Huntington Avenue Boston MA 02115 Dr. Norton Nelson Professor and Chairman Department of Environmental Medicine New York University School of Medicine New York NY 10016 Dr. Alan Poland Associate Professor of Oncology McCardle Laboratory University of Wisconsin Madison WI 53706 Dr. Irving Selikoff Director, Environmental Sciences Laboratory Mt Sinai School of Medicine 5th Avenue and 100th Street New York NY 10029 47 �Appendix III MATCHING RESULTS IN THE MORTALITY POPULATION The matching results are described here for the mortality population consisting of 1241 Ranch Hands, their 6171 matched mortality comparisons, and the six unmatched Ranch Hands. The matching procedure is described in the Protocol (Ref. 1, pages 23~26). All study subjects were matched perfectly on job category. Three mismatches occurred on race because the recorded race designations for three study subjects were found to be incorrect at the LHA interview. These three subjects were comparisons, two were in the enlisted-other stratum (one was originally recorded as Black and was discovered to be non-Black, the other was originally recorded as non-Black and was discovered to be Black), and one was in the enlisted-flight engineer stratum (he was originally recorded as Black and was discovered to be non-Black). Matching on date of birth was carried out by first expressing date of birth in months from 1 January 1900, to the nearest month; the result is termed month-of-birth. Six discrepancies occurred in matching on month-of-birth due to erroneous months-of-birth for one Ranch Hand and one comparison. These were discovered at the LHA interview. The Ranch Hand, in the non-Black enlisted-other stratum, was discovered to be 72 months older than was recorded prior to the matching. The comparison, in the non-Black officer-pilot stratum, was found to be 15 years younger than was originally recorded. The erroneous Ranch Hand month-of-birth put all five of his matched comparisons 12 months out of range since he was originally perfectly matched to all five mortality comparisons. The erroneous comparison month-of-birth put that comparison 119 months out of range. Given the very small number of mismatches on age and race relative to the number of subjects, their effect was assumed negligible. The matching by month-of-birth, overall, and within each of the ten job and race categories within the mortality population is summarized in this Appendix. The column headed "Age Difference" lists absolute differences of months-of-birth of Ranch Hands and comparisons. The column headed "Number of Comparisons with RH younger (older)" gives, at each level of age difference, the number of comparisons within the level of age difference and older (younger) than the Ranch Hand to whom they are matched. The column headed "Total Count" gives the total numbers of comparisons having the absolute age differences with their matched Ranch Hand given in the first column; in "Total Percent", these counts are expressed as percentages of 6171. These are cumulated in the last two columns. �Appendix III MATCH SUMMARY FOR THE MORTALITY POPULATION Strata Age Difference Comparisons with RH Total Younger Older Count Cumulative Total % * Overall Officer-pilot Non-Black Officer-Pilot Black OfficerNavigator Non-Black 0 1-6 7-12 13-18 19-21 25-30 31-36 37-42 13-18 19-51 55-60 72 179 0 1-6 7-12 13-18 19-21 25-30 31-36 37-12 13-18 19-51 55-60 179 0 1-6 7-12 31-36 37-12 13-18 19-51 55 0 1-6 7-12 713 77 10 22 12 16 10 9 13 17 0 0 706 102 36 22 19 11 19 13 7 7 5 1 272 33 20 8 9 13 7 7 11 11 0 259 32 17 12 11 10 18 11 7 7 1 3 3 0 2 1 0 0 1 0 0 0 0 2 1 70 6 71 0 19 1261 1119 179 76 1 1 31 30 29 22 20 21 5 1 1261 5710 69.0 23.5 2.9 1.2 0.7 0.5 0.5 0.5 0.1 0.3 0,1 0.1 0.0 5889 5965 6009 6010 6070 6099 6121 6111 6165 6170 6171 961 531 65 37 20 20 23 25 18 18 21 1 55.2 30.5 3.7 2.1 1.1 1.1 1.3 1.1 1.0 1.0 1.2 0.1 961 1192 1557 1591 1611 1631 1657 1682 1700 1718 1739 1710 89.5 91.6 92.8 93.9 95.2 96.7 97.7 98.7 99.9 100.0 0 3 3 1 0.0 23.1 0.0 23.1 16.2 53.9 69.2 76.9 92.3 100.0 61.5 98.5 100.0 2 1 2 1 7.7 15.1 7.7 15.1 7.7 0 3 6 7 9 10 12 13 210 111 6 61.5 36.9 1.5 210 381 390 23.1 69.0 92.5 95.1 96.7 97.1 97.9 98.1 98.8 99.2 99.5 99.9 100.0 100.0 55.2 85.8 �Appendix III (Continued) MATCH SUMMARY FOR THE MORTALITY POPULATION Strata OfficerNavigator Black Number of Total Comparisons with RH Age Younger Older Count % Difference 0 1-6 7-12 13-18 19-21 25-30 31-36 Officer-Other Non-Black 0 1-6 7-12 13-18 19-21* 25 Officer-Other 13-18 1 0 2 0 0 0 0 2 1 1 1 1 38 2 1 1 0 57 8 1 0 1 2 0 Cumulative Total % 1 1 1 1 1 1 1 10.0 10.0 10.0 10.0 10.0 10.0 10.0 1 2 6 7 8 9 10 10.0 20.0 60.0 70.0 80.0 90.0 100.0 11 95 11.1 77.2 10 2 1 1 8.1 1.6 0.8 0.8 11 109 119 121 122 11.1 88.6 96.8 2 98.1 123 99.2 100.0 100.0 2 100.0 516 306 55.2 32.7 55.2 87.9 6.7 516 822 885 Black EnlistedFlight 0 1-6 Engineer Non-Black 7-12 13-18 19-21 25-30 31-36 37-12 46 EnlistedFlight Engineer Black 111 34 11 0 2 2 1 1 7 6 1 0 0 63 30 7 8 3 1 1 26 7 22 5 0 0 10 18 12 3 2 13.3 61.0 16.0 16 0 1-6 Non-Black 7-12 13-18 19-21 25-30 31-36 37-12 13-18 72 0 1-6 7-12 3.2 91,7 97.9 98.6 99.5 99.8 99.9 3 2 116 1 1 91 11 0.7 915 922 0.9 930 0.3 933 931 935 100.0 1.0 2.7 10 58 70 73 75 13.3 77.3 93.3 97.3 100.0 2382 0 1-6 7-12 19-21 55-58 EnlistedOther EnlistedOther Black 165 29 90.6 2382 7.9 0.5 0.2 0.5 0.0 0.1 0.0 0.1 0.2 2589 90.6 98.5 99.0 99.1 99.6 99.6 53.7 11.7 10 1 1 0 0 0 2 0 1 1 2 5 207 12 1 12 1 2 1 2 5 19 0 65 1 137 111 1 3 50 0.1 0.1 1.6 2601 2605 2617 2618 2620 2621 2623 2628 99.7 99.7 99.8 100.0 137 251 255 53.7 98.1 100.0 �Appendix IV YEAR OF BIRTH, OCCUPATIONAL AND RACE SPECIFIC MORTALITY Birth Ranch Hand Job Category, Race Year At Risk Dead Death Rate Officer-Pilot, Non-Black .375 1 915-19 1 920-21 1 925-29 1 930-31 1 935-39 1 910-11 1 915-19 8 31 31 113 66 60 10 3 0 0 3 3 1 Comparison At Risk Dead 1 13 11 Death Rate 23 8 5 5 .103 .081 .060 .050 .025 .011 .028 TOTAL 319 Officer-Pilot, Black 2 .027 .015 .017 .050 39 155 232 156 326 351 178 12 .031 1710 72 .011 0 1 3 2 0 0 0 0 3 1 6 0 0 0 0 0 6 0 13 0 9 35 21 13 17 163 105 67 8 3 7 3 0 0 .061 .013 .029 2 0 1 1 0 0 80 2 390 13 .033 1 1 0 0 6 1 0 0 2 0 10 0 1 0 1 3 2 1 0 0 0 0 0 0 2 3 6 11 0 0 0 1 1 0 1 0 1 930-31 1 935-39 1 910-11 1 915-19 TOTAL Officer-Navigator 1 925-29 Non-Black 1 930-31 1 935-39 1 910-11 1 915-19 TOTAL Officer-Navigator 1 930-31 Black 1 935-39 TOTAL Officer-Other, Non-Black 1 910-11 1 915-19 1 920-21 1 925-29 1 930-31 1 935-39 1 910-11 1 915-19 TOTAL 13 1 .029 .018 .025 12 .077 19 66 1 .010 1 123 0 25 51 .091 .083 .015 .021 �Appendix IV (Continued) Job Category, Race Officer-Other, Black Birth Ranch Hand Year At Risk Dead Death Rate Comparison At Risk Dead Death Rate 1940-44 TOTAL Enlisted-Fit Eng Non-Black 1 0 0 3 2 0 0 1925-29 1930-34 1935-39 1940-44 TOTAL Enlisted-Other Non-Black 1905- 9 1910-14 1915-19 1920-24 1925-29 1930-34 1935-39 1940-44 1945-49 1950-54 1 6 5 3 0 1 0 1 15 2 0 4 8 12 28 76 52 67 270 11 0 2 0 3 2 6 1 2 9 0 25 Enlisted-Other Black .333 .200 .049 .049 .041 .033 .056 935 46 .049 1 5 1 .333 10 34 16 15 2 .100 .150 .063 .133 .133 75 9 .120 2 10 48 60 140 0 .167 .500 .250 .071 .079 .019 .030 .033 376 263 340 1333 56 2 5 7 18 14 8 7 36 .200 .104 .117 .129 .037 .030 .021 .027 0 52 52 2628 97 .037 .143 35 40 35 .034 145 2 3 1 4 .057 .075 .029 .028 .038 1 0 0 1 TOTAL 2 4 3 15 10 7 5 .032 Enlisted-Fit Eng Black .0*17 .042 6 20 61 304 243 211 90 1.000 255 10 .039 .047 �Appendix V THE EXTENDED EJIGOU-McHUGH RELATIVE RISK ESTIMATOR Let R k , k=1,2, ..., K, denote the distinct numbers of comparisons matched to the cases and let nk denote the number of matched sets with exactly Rk comparisons. A matched set is defined as the case and his matched comparisons. Let n=ni+n2+ ... +nj( denote the total number of matched sets. Define Zk«i,T» k-1 .2 ..... K, i-0,1, by Zk 0 T * tne number of matched sets, among those having exactly Rk comparisons, in which the case is alive and exactly T of the R k comparisons have died, T=1,2, ..., Rk Zk,1 T * the number of matched sets, among those having exactly Rk comparisons in which the case has died and exactly T of the Rk comparisons have died, T=0,1,2, ..., Rk-1 . The extended estimate, \p, is given by K 1 Rk I zk,0,Tzk,1 ,T-1/(zk,0,T+zk,1 ,T-l) K Rk Z * TZ2k,0,T/(Rk-T+1)(Zk,0,T+Zk,1,T-l) k-1 T-1 and its variance is estimated by K Rk Z Z Zkf0>T/[i|;+(Rk-T+1)/T] k-1 T-1 Since K is finite, this estimate has the same distributional properties as the Ejigou-MoHugh estimator; it is asymptotically efficient and unbiased. The underlying assumptions used in its derivation are that the disease under study is of low incidence and that relative risk is constant over the levels of the matching variables. The Ejigou-McHugh estimate and the above extension are equivalent in asymptotic efficiency to maximum likelihood estimation (12). 53 �Appendix VI Figure 4 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR POOLED RANCH HANDERS C.9 0.3 Ci) z 0.7 - 14 - O.G - CO C'.V> - 2: C3 C.4 CT ce: 0 . 3 C.I . 0I 0 <Q 10 bO 70 RGF. Figure 5 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR POOLED COMPARISONS I 0.9 0.0 § 0«7 »—i > £ 0.6 W Q.& 2 0.4 I— o CE or. ° •3 u. 0.2 0.1 0 0 J_ _. 10 :ji"" fiC-f 54 I '"0 70 �Figure 6 SURVIVAL CURVE ESTIMATE AND'95% CONFIDENCE BANDS FOR RANCH HAND OFFICERS 0.9 O.Q 0.7 0.6 9s 0.5 •z. CJ 0 0.3 0.2 0.1 0 0 10 20 30 60 70 ROE Figure 7 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR COMPARISON OFFICERS I 0-9 o.e 0.7 0.6 z H- 0.4 0.3 0.1 0 . IK 30 60 ROE 55 70 �Figure 8 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR RANCH HAND ENLISTED PERSONNEL 1 0.9 0.8 g 0.7 oc to 0.5, z 2 Q.4 Iu S 0.3 u. 0.2 O.I 0 10 30 60 70 RGE Figure 9 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR COMPARISON ENLISTED PERSONNEL I 0.9 O.Q g 0.7 o.e cn z S 0.4 0.3 0.2 0.1 0 1— 10 no f. 56 60 70 �Figure 10 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR RANCH HAND FLYING PERSONNEL 1 0.9 0.0 0.7 > 0.6 or O.b z o £ 0.3 U_ 0.2 0.1 0 10 20 30 E>0 60 70 flGF. Figure 11 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR COMPARISON FLYING PERSONNEL 1 0.9 0.0 Z 0.7 U 0.6 W 0-b CJ 0.3 0.2 0.1 0 10 20 30 bO 57 60 70 �Figure 12 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR RANCH HAND GROUND PERSONNEL 0.9 0.9 g 0.7 £ 0.5 to 0.5 z •-• 0.4 ho £ 0.3 u_ 0.2 0.1 10 20 30 60 70 ROE Figure 13 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR COMPARISON GROUND PERSONNEL 1 0.9 0.8 Si 0.7 > £ 0.6 W 0.5 Z •-< 0.4 o £ 0.3 u_ 0.2 0.1), 0 0 10 30 40 flOE 58 £.0 70 �REFERENCES 1. Lathrop, G. D., Wolfe, W. H., Albanese, R.A. and Moynahan, P. M. Epidemiclogical investigation of health effects in Air Force personnel following exposure to herbicides: Study Protocol. SAM-TR- 82-4*1, Dec 1982. 2. Honchar, P. A., and Halperin, W. E. 2,4,5-T, trichlorophenol, and soft tissue sarcoma. Lancet Jan 31:268-269 (1981). 3. Hardell, L., and Sandstrom, A. Case-control study: soft tissue sarcomas and exposure to phenoxyacetic acids or chlorophenols . Br. J. Cancer 39:711-717 (1979). 4. Eriksson, M. , Hardell, L., Berg, N. 0., Moller, T., and Axelson, 0. Case-control study on malignant mesenchymol tumors of the soft tissue and exposure to chemical substances. Lakartidningen 76:3872-75 (1979). 5. Breslow, N. E., Lubin, J. H., Marek, P. and Langholz, B. Multiplicative models for cohort analysis. Journal of the American Statistical Association 78:1-12 (1983). 6. Kaplan, E. L. and Meier, P. Nonparametric estimation from incomplete observation. J. Am. Stat. Assoc. 53:1»57-II81 (1958). 7. Hall, W. J, and Wellner, J. A. Confidence bands for a survival curve from censored data. Biometrika 67:133-1^3 (1980). 8.. Prentice, R. L. Linear rank tests with right censored data. Biometrika 65:167-179, (1978). 9. Michalek, J. E. and Mihalko, D. Linear rank procedures for matched observations. SAM-TR-83-16 (In Press), May 1983. 10. Michalek, J. E. and Mihalko, D. Matched survival analysis (MSURV). SAM-TR-83-15 (In Press), May 1983. 11. Michalek, J. E. and Mihalko, D. Linear rank procedures for litter matched data. Biometrics (to appear). 12. Ejigou, A. and McHugh, R. Relative risk estimation under multiple matching. Biometrika 68:85-91 (1981). 13. Gail, M. The analysis of heterogeneity for indirect?- standardized mortality ratios. Journal of the Royal Statistical Society, Series A (1978). . Mantel, N. and Haenszel, W. Statistical aspects of the analysis of data from retrospective studies of disease. Journal of the National Cancer Institute 22:719-748 (1959). 59 �15. Valuation of the Military Retirement System FY 1980. Office of the Actuary, Defense Manpower Data Center, 300 North Washington Street, Alexandria, Virginia 22314. 16. Vital Statistics of the United States, 1978 Vol II - Section 5, Life Tables; U.S. Dept of Health and Human Services, DHHS Publication No (PHS) 81-1104; Hyattsville, Maryland:1980. 17. Gail, M. H. and Ware, J. H. Comparing observed life table data with a known survival curve in the presence of random censorship. Biometrics 35:285-391 (1979). 18. Annual Report, 1966-1981, Servicemen's Group Life Insurance Program, Insurance Service, Department of Veterans Benefits, Veterans Administration. 19. Cohort Mortality and Survivorship: United States Death-Registration States, 1900-1968. Public Health Service. DHEW Publication No (HSM) 73-1400. 20. Elandt-Johnson, R.C. and Johnson, M.L. New York: John Wiley, 1980. Survival Models and Analysis 21. McMichael, A. J., Standardized Mortality Ratios and the "Healthy Worker Effect": Scratching Beneath the Surface. J. of Occupational Medicine, Vol 18, No. 3, March 1976. 22. Tabershaw, I. R., Special Communications, "What Do We Expect from an Occupational Cohort?", J. of Occupational Medicine, Vol 17, No. 2, February 1975. 23. Enterline, P. E. ; Pitfalls in Epidemiological Research; J. of Occupational Medicine, Vol 18, No. 3, March 1976 24. Gaffey, W. R. A critique of the standardized mortality ratio. Journal of Occupational Medicine 18:157-160 (1976). 25. Wong, 0. Further criticisms of epidemiological methods in occupations studies. Journal of Occupational Medicine 19:220-222. 26. Clark, D., Allen, M. and Wilson, F. Longitudinal study of serum lipidstwelve year report. American Journal of Nutrition 20:743-752 (1967). 27. Breslow, N. and Crowley, J. A large sample study of the life table and product limit estimates under random censorship. The Annals of Statistics 2:437-453 (1974). 28. Kalbfleisch, J. D. and Prentice, R. L. The Statistical Analysis of Failure Time Data. New York: John Wiley, 1980. 29. Monson, R. R. Occupational Epidemiology, CRC Press, Inc., Boca Raton FL (1980). 60 �30. Feinleib, M. A method of analyzing log-normally distributed survival data with incomplete follow up. Journal of the American Statistical cal Association 55:53*1-535 (1960). 31. Feinleib, M. and McMahon, B. Variation in the duration of survival of patients with chronic leukemias. Blood 15:332-3*19 (1960). 32. Nice, C. M. and Stenstrom, K. W. Radiology 62:6*11-652 (195*1). Irradiation therapy in Hodgkins disease. 33- Osgood, E. E. Methods for analyzing survival data, illustrated by Hodgkin's disease. American Journal of Medicine 2*4:*IO-*I7 (1958). 3*1. Mehrotra, K. G., Michalek, J. E., Mihalko, D. and White, T. Score computation for linear rank procedures. Journal of Statistical Computation and Simulation 16:201-211 (1982). 35. Breslow, N. Odds ratio estimators when the data are sparse. Biometrika 68:73-84 (1981). 36. Anderson, S., Auquier, A., Hauck, W., Oakes, D., Vandaele, W., and Weisberg, H. Statistical methods for comparative studies, techniques for bias reduction. New York: John Wiley, 1980. 37. Breslow, N. E. and Liang, K. Y. The variance of the Mantel-Haenszel estimator. Biometrics 38:9-43-952 (1982). 38. Breslow, N. E. and Day, N. W. Indirect standardization and multiplicative models for rates, with reference to the age adjustment of cancer incidence and relative frequency data. Journal of Chronic Diseases 28: 289-303 (1975). 61 �PROJECT RANCH HAND II AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES BASELINE MORTALITY STUDY RESULTS 30 JUNE 1983 Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 Approved for public release; distribution unlimited EPIDEMIOLOGY DIVISION DATA SCIENCES DIVISION USAF SCHOOL OF AEROSPACE MEDICINE (AFSC) BROOKS AIR FORCE BASE, TEXAS 78235 �EXECUTIVE SUMMARY Baseline Mortality Study The Ranch Hand II epidemiologic study'uses a matched cohort design in a nonconcurrent prospective setting, incorporating mortality, morbidity, and follow-up studies. The purpose of this report is to present the baseline mortality study results. , ,>••• Since 1979, a detailed population ascertainment process has enumerated a total of 1269 Ranch Hand personnel who served in Vietnam during the period of 1962-1971. As described in the protocol, this total is believed to comprise the entire exposed study population. The eligibility of each Ranch Bander was verified by a hand review of his personnel record. A comparison group was formed by identifying all individuals assigned to selected Air Force organizational units with a mission of flying cargo to, from, and In Vietnam during the same period. All Ranch Hand and comparison subjects designated as killed in action were removed from the study population. By a computerized nearest neighbor selection process, up to 10 comparison individuals were matched to each Ranch Hander by job category, race, and age to the closest month of birth. A hand record review of the matched comparison sets revealed that on the average, 8.2 comparison individuals were fully suitable for study. From each matched comparison set, five individuals were randomly selected for the mortality study, yielding a 1:5 design. Every Ranch Hander and his set of comparisons will be the subjects of annual mortality updates throughout the entire 20 years of the follow-up study so that emerging mortality patterns or disease clusters may be detected with maximal sensitivity. Each living Ranch Hander and his first and willing comparison match were selected to participate in a comprehensive physical examination and an in-home interview; the results of this study will be presented in a subsequent report in late 1983. A mortality determination on 1,247 Ranch Banders and 6,171 comparison subjects was made, sequentially using the data sources of the Air Force, Veterans Administration, Social Security Administration, Internal Revenue Service, and personal contact efforts. As of December 31, 1982, 50 Ranch Hand and 250 comparison subjects had died (certified on/before April 27, 1983). Death certificates were obtained on all 300 deceased subjects and were coded by an Air Force nosologist (ICD, 9th ED). All codings were verified by the National Center for Health Statistics. Autopsy results are currently being sought for future analyses. Statistical analyses of noncause specific death emphasized survival curve estimates, linear rank procedures, relative risk estimates, and standardized mortality ratios (SMRs). Cause specific analyses were limited to relative risk estimates because of small cell sizes. In addition to these approaches, three other data bases were contrasted to the Ranch Hand population, where possible; the 1978 US White Male Mortality experience, the 1978 Department of Defense (DoD) Nondisability Retired Life Table, and the mortality experience of the West Point Class of 1956. These additional comparison groups have substantial comparability or sample size limitations, rendering conclusions to the weakest order. Analyses with these "external" comparison groups were accomplished to crudely define the healthy worker effect and to determine if the Ranch Hand group mortality was drastically out of line with that of other military populations. �Data analysis showed that the mortality experience of the Ranch Hand group is nearly Identical to that of the comparison group. Analyses showed that officers are living" longer than enlisted personnel in both Ranch Hand and comparison groups. This difference between officers and enlisted personnel was statistically significant in the comparison group whereas it was not in the Ranch Hand cohort. A contrast of the Ranch Hand and comparison group to the 1978 DoD Life Table showed significantly less mortality for Ranch Hand officers, comparison officers and comparison enlisted men, however, there was not a statistically significant favorable mortality rate for Ranch Hand enlisted personnel. This pattern of mortality was also seen in a contrast of the Ranch Hand and comparison groups to the 1978 U.S. white male mortality experience. That is, highly favorable mortality differentials for Ranch Hand officers, comparison officers and comparison enlisted personnel were observed, but not for Ranch Hand enlisted. This trend is consistent with the self perceptions of differential herbicide exposures reported by many of the Ranch Hand subjects. The reason(s) for these observations are speculative at present, but may include the related items of sample size, socioeconomic differences, access to medical care, and health education and possible herbicide effects. Cause specific analyses were statistically nonsignificant. The Ranch Handers showed a relative paucity of overall cancer but an excess of digestive disorder deaths, both statistically nonsignificant. No soft tissue sarcoma deaths were detected in either group. Analyses of both the Ranch Hand and the comparison groups to the 1978 US White male mortality experience showed highly significant favorable findings. Most of these differences are speculatively attributed to the healthy worker effect. A contrast of the Ranch Hand and comparison groups to the 1978 DoD Life Table showed significantly less mortality for Ranch Hand officers and comparison officers and enlisted men. The West Point comparison showed nonsignificant SMRs of 0.530 and 0.778 for the Ranch Hand officers and the comparison group officers, respectively. Overall, the limitations of the statistical power calculations in most of these analyses were substantial in most analyses due to 1) the low mortality rate (4%) in the Ranch Hand and comparison groups to date, 2) the inherently small group of Ranch Handers (as described in the study protocol), and 3) the observed relative risks which approached unity in most categories. This baseline mortality report can in no way be regarded as conclusively negative because this small, young, and relatively healthy cohort may not have yet reached the latency period wherein attributable fatal disease might be expected and detected within limited power boundaries of this study. Future commitments for the annual mortality updates include detailed covariate analyses for disease risk factors, herbicide exposure, and confounding industrial chemical exposures. Further, subsequent morbidity reports will include full spectrum, disease specific analyses, e.g., cancer (fatal, ongoing, cured) in an effort to enhance study sensitivity to emerging herbicide effects, if they occur. ii �AIR FORCEHEALTH STUDY (PROJECT RANCH HAND II) AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES BASELINE MORBIDITY STUDY RESULTS 24 FEBRUARY 1984 Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 George D. Lathrop, Colonel, USAF, MC William H. Wolfe, Lieutenant Colonel, USAF, MC Richard A. Albanese, M.D., GM-15 Patricia M. Moynahan, Colonel, USAF, NC Approved for public release; distribution unlimited. .£>. USAF SCHOOL OF A E R O S P A C E MEDICINE Aerospace Medical Division (AFSC) Brooks Air Force Base, T e x a s 7 8 2 3 5 �.* EXECUTIVE SUMMARY BASELINE MORBIDITY STUDY The Ranch Hand II epidemiologic study uses a matched cohort design in a nonconcurrent prospective setting, and .incorporates mortality, morbidity, and follow-up studies. -The purpose of this report is to present the baseline morbidity study. The morbidity study design matched each living Ranch Hander (by age, Job, and race) to the first living and compliant member of a randomly selected comparison mortality set of 5 individuals, producing a 1:1 contrast. The comparison group was formed from numerous flying organizations which transported cargo to, from, and within Vietnam, but were not involved in aerial spray operations of Herbicide Orange. Of the potential study participants, 99.5$ were located. Early in the physical examination phase of the study, it was discovered that 18$ of the entire comparison group was ineligible to participate because of inappropriate selection. Thereafter, study eligibility was certified only after a hand-review of personnel records. Next-in-line compliant comparisons entered the study as replacements after fully completing the questionnaire and physical examination. Statistical analyses of these replacement individuals later showed that they differed from the original comparisons in a variety of subtle and often opposite ways. As a conservative measure to avoid possible bias by the inclusion of the replacements in the analyses, a management decision was made to base the statistical tests in this report primarily upon contrasts of the Ranch Hand group to the original comparison group. The preponderance of data was obtained from the in-home interviews and the physical examination, each conducted under contract to the Air Force by Louis Harris and Associates, Inc., New York NY, and the Kelsey-Seybold Clinic, P.A., Houston TX, respectively. All contacts with the participants were carried out with utmost professionalism and sensitivity. Other morbidity data sources included reviews of medical records, military personnel documents, and birth certificates; in-home questionnaires and telephone questionnaires of the study participant's wives, former wives and, occasionally, their next-of-kin. All aspects of the study were voluntary. As a contract requirement, data collection personnel were blind as to the exposure status of the participants. Ninety-seven percent of the Ranch Handers and 93$ of the comparisons participated in the in-home interview. For the physical examination, 87$ of the Ranch Handers and 76$ of the comparison group participated, a total of 2,272 individuals. This differential attendance at the examination may have introduced a potential participation bias that, in a military population predominantly engaged in flying duties, is multifactorial and complex. All study phases were monitored by stringent quality control standards. Statistical analyses of the data consisted primarily of log-linear models, logistic regression techniques, generalized linear models, matched covariate analyses, and Kolmogorov-Smirnov, chi-square, and t tests. The physical examination and the in-home questionnaire data were analyzed by major organ system. In terms of general health, more Ranch Handers perceived themselves to be in fair or poor health than did their comparisons. No �group differences were detected for hematocrit or percent body fat determinations. Unadjusted group differences in sedimentation rate were not observed; however, significantly more young comparisons had abnormalities in sedimentation rate than did their Ranch Hand counterparts. There were no statistically significant differences in the occurrence of malignant or benign systemic tumors between the groups. One case of soft tissue sarcoma was found in a comparison member. Significantly more nonmelanotic skin cancer was noted in the Ranch Hand group, but these - analyses have not yet considered (adjusted for) sunlight exposure, the prime etiology of these cancers. Such nonmelanotic skin cancer (predominantly basal cell carcinoma) is the most common neoplasm in the White population of the United States. Up to the statistical limits of the study there were no consistent data that showed that the Ranch Handers were developing uncommon cancers, or cancer in unusual sites, or at an unusual age. Measures of fertility and reproductive outcome showed mixed results. It is emphasized that the fertility and reproductive results are preliminary at this time as they are based largely upon subjective self reports that await full medical record and birth certificate verification. Four measures of fertility: number of childless marriages, couples with the desired number of children, the infertility index and the fertility index, showed no difference between the Ranch Hand and comparison groups. A semen specimen obtained from those willing and able to provide one showed no group differences with respect to total sperm count or percent abnormal sperm. There were no significant findings in conception outcomes for miscarriages, stillbirths, induced abortions, or live births. For live birth outcomes no differences were observed for prematurity, learning disability, or infant deaths. There was no significant disparity between groups for the classifications of severe or moderate birth defects. By parental history, however, Ranch Hand offspring showed significantly more minor birth defects (birth marks, etc). Reporte'd neonatal deaths and physical handicaps were also significantly excessive in the Ranch Hand group when contrasted to the total comparison group. All fertility and reproductive findings in the Ranch Hand group showed inconsistent relationships to the herbicide exposure index. Medical records and birth certificates are currently being chronicled for complete verification of all historical findings. A comprehensive neurological examination showed no consistent abnormalities in the cranial nerves, peripheral nerves or central nervous system function of the Ranch Handers. As expected, there was a profound influence of diabetes and alcohol in both groups upon numerous neurological tests. Detailed psychologic data were obtained on all participants at both the in-home interview and the physical examination. It is emphasized that the majority of psychological data was derived from self reported responses during interview and has not been fully assessed for the effect of differential reporting. A variety of subjective deficits (fatigue, anger, fear, anxiety, etc) were significantly more common in the high school educated Ranch Handers. Educational level significantly and consistently influenced most subjective test results. In sharp contrast, more objective performance testing by the Halstead-Reitan battery and IQ testing did not reveal any significant intergroup differences. The roles of overreporting and the Post Vietnam Stress Syndrome In these analyses have not as yet been assessed. Liver function tests and clinical history data showed mixed results. Ranch Handers had some elevated liver enzyme tests and lower cholesterol levels. More Ranch Handers were found to have hepatomegaly and verified histories of prior hepatic disease than their counterpart comparisons. Exposure to alcohol, degreasing chemicals, and industrial chemicals in general, influenced ii �the liver test results. Ranch Handers reported significantly more symptoms resembling porphyria cutanea tarda than the comparisons-, but these data have not been verified by medical record reviews nor were they substantiated by laboratory testing or by physical examination. Exposure index analyses were essentially negative. In the dermatologic evaluation, no cases of c'hloracne were diagnosed clinically or by biopsy. A thorough questionnaire analysis of acne showed that the incidence, severity, duration, and anatomic location did not differ between groups, and suggested that the historical occurrence of chloracne was highly unlikely in the Ranch Handers. Evaluation of the cardiovascular system showed equal proportions of abnormalities in blood pressures, electrocardiograms, past electrocardiograms, and heart sounds in both groups. Ranch Handers are not having premature heart attacks or generalized heart disease. However, the Ranch Handers showed significant deficits in 2 specific peripheral leg pulses and all leg pulses as a group. These puzzling findings were highly correlated with age and smoking patterns, and verified past heart disease. The assessment of the immune system by laboratory testing was compromised by excessive test variability. An independent review committee determined which test data were suitable for statistical analysis. As an unexpected finding, the test data were significantly influenced by the age and smoking history of the participant; no group differences were detected after adjustment for these factors. A hematologic test battery revealed three red cell abnormalities in the Ranch Hand group, but these were difficult to place into a clinical or epidemiologic context. Evaluation of renal, pulmonary, and endocrine functions generally disclosed small and inconsistent proportions of abnormalities between groups, and were deemed clinically unimportant. An unrefined assessment of all summed and weighted organ system abnormalities by group did not show an aggregation of multisystem disease or malfunction. Any interpretation of these study data, in whole or in part, must carefully consider the methodical steps required for a proper inference of causality. It is specifically pointed out that many group differences were largely based upon subjective data, and that a subtle effect of differential reporting is suggested but has not been fully evaluated. For objective data, group differences were generally within normal ranges and were not correlated to the herbicide exposure index, nor fell within the expected latency periods following Vietnam service. The proposed clinical end points of dioxin exposure, chloracne, soft tissue sarcoma, and porphyria cutanea tarda, were not found in the Ranch Hand group (study power limitations recognized). Overall, substantial credence is given to the objective study findings, particularly after observing the consistent duplication of the classical effects of risk factors such as age, smoking, alcohol, etc., in almost all clinical areas. Additional work with these baseline data is still required in the areas of data base refinement, statistical testing and bias analysis, exposure index refinement, establishment of the follow-up examination requirements, and collaboration with other dioxin research studies. This baseline report concludes that there is insufficient evidence to support a cause and effect relationship between herbicide exposure and adverse health in the Ranch Hand group at this time. The study has disclosed numerous medical findings, mostly of a minor or undetermined nature, that require detailed follow-up. In full context, the baseline study results should be viewed as reassuring to the Ranch Handers and their families at this time. iii �PROJECT RANCH HAND II AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES MORTALITY UPDATE -1984 10 DECEMBER 84 William H. Wolfe, Colonel, USAF, MC Joel E. Michalek, PhD Richard A. Albanese, M.D. George D. Lathrop, Colonel, USAF, MC Patricia M. Moynahan, Colonel, USAF, NC Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 EPIDEMIOLOGY DIVISION DATA SCIENCES DIVISION USAF SCHOOL OF AEROSPACE MEDICINE (AFSC) BROOKS AIR FORCE BASE, TEXAS 78235 �Project Ranch Hand IT Mortality Update - 1984 EXECUTIVE SUMMARY HAOKCR<WNI> The purpose of the Ranch Hand II Study is to determine whether those Individuals involved in the aerial spraying of herbicides in Vietnam during the'Ranch Hand operation have experienced any adverse health effects as a result of their participation in that program. The study evaluates both mortality (death) and morbidity (disease) In these Individuals 6ver a 20-year period of time after the studies were initiated. The baseline mortality study was released in June 1983 and the baseline morbidity study in February 1984. Neither study demonstrated health effects which could be conclusively attributed to herbicide or dioxln exposure. The reader is referred to reports of the studies for further details (1, 12). METHOD The present study report describes the second mortality analyses. Deaths in the 1256 Ranch Hand and 6171 comparison subjects were determined, using the data sources of the Air Force, Veterans Administration, Social Security Administration, Internal Revenue Service, and personal contacts. As of 31 December 1983, 54 Ranch Handers and 265 comparison subjects had died. Death certificates were obtained on all subjects. Autopsies were conducted on 157 of the individuals who had died. Results have been obtained for 104 of these autopsies to confirm the death certificate findings. Autopsy reports for the 53 others have been requested, but have not yet been obtained. Extensive statistical analyses were accomplished, as detailed in the report, to compare the death experience in the Ranch Hand population with the comparison group. In addition, death experience in these groups was compared to the 1978 U.S. White Male Mortality experience, the 1978 Department of Defense Nondisability Retired Life Table, the mortality experience of the West Point Class of 1956, the USAF active duty personnel, and the active U.S. Civil Service population. RESULTS As was the case in the first mortality report, the current mortality analyses did not reveal any statistically significant differences in mortality between the exposed and comparison groups. The percentages dead in each major category are summarized below. Percent Deaths Ranch Hand Comparison Rank Officers Enlisted 3.2 4.9 4.0 4.5 Occupation Flying Ground 3.6 5.1 4.7 3.9 Note: None of the above differences between the Ranch Hand and Comparison groups are statistically significant. �Ranch Hand Total Overall 4.3 Comparison 4.3 As was reported In the baseline mortality study, the Ranch Hand officers had a nonstatlstically significant though slightly lover death rate than their comparisons, Ranch Hand flyers had a nonstatlstically significant though slightly lower death rate than comparisons, and Ranch Hand ground personnel had a slightly higher but nonstatlstlcally significant death rate than the comparisons. The herbicide/dioxln exposure index described In the morbidity report was applied to the data, and no relationship between exposure and mortality experience was identified. As was also noted in the baseline mortality study, analyaes consistently demonstrated significantly better survival in the Ranch Hand officers than Ranch Hand enlisted members, as was the case with comparison officers and comparison enlisted personnel. Cause-specific analyses did not demonstrate any Increased Ranch Hand mortality for accidents, suicide, homicide, malignancy or circulatory system disease. No unusual patterns of malignancy were observed in either the Ranch Hand or comparison groups, a finding which would be expected from the small number of deaths to date. When compared to the 1978 U.S. White male population, the Ranch Rand officers, comparison officers, and comparison enlisted are living significantly longer than expected. Although Ranch Hand enlisted are also living longer, the difference is not Significant. A similar pattern was seen in analyses using the DOD retired population. All groups bad a mortality experience similar to the civil service population. As would be expected from the fact that individuals In the active duty population who develop severe chronic disease are medically retired, all groups in this study had an increased mortality when compared to the Air Fore* population currently on active duty. Both Ranch Hand and comparison officers had mortality similar to the West Point group. I i CONCLUSION AND RECOMMENDATION Continued mortality surveillance is recommended, since the study groups are still relatively young and healthy. While sufficient time may have elapsed for some clinically significant conditions to occur, additional time Is necessary for other conditions, which; nay possibly be attributable to herbicide exposure, to develop. At this time, however* there Is no evidence of increased mortality as a result of herbicide exposure in those individuals who accomplished the Ranch Hand spray operations in Vietnam. �ABSTRACT OF REPRODUCTIVE OUTCOME DRAFT REPORT, 17 December 1984 Abstract: This report was written in response to requests from the Office of Science and Technology Policy, the Chair - Agent Orange Working Group and by direction of the Air Force Deputy Surgeon General. It was intended as an "interim" analysis of the data and not as a definitive answer to the birth defects issue. Since the release of the baseline morbidity report in February 1984, birth defects and neonatal deaths reported by study participants during the baseline questionnaire have been verified by record review. This verification was accomplished by the review of birth and other medical records, birth certificates and death certificates. Since not all data can be verified for at least an additional 24 months, these interim analyses were based on the v e r i f i c a t i o n of the p o s i t i v e reports only. Verification of negative responses to the birth defect and neonatal death questions have not as yet been completed. The draft report was submitted to the Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants. Their comments have been received and will be discussed with them in a forthcoming visit to the Ranch Hand Office. This interim assessment will then be forwarded to the OSTP through the Chair, AOWG. b-s � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 199 Folder The folder containing the original item. 5598 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. Patricia M. Moynahan Richard A. Albanese William H. Wolfe Description An account of the resource <strong>Corporate Author: </strong>Epidemiology Division, Data Sciences, Division, USAF School of Aerospace Medicine (AFSC), Brooks Air Force Base, Texas Date A point or period of time associated with an event in the lifecycle of the resource June 30 1983 Title A name given to the resource Project Ranch Hand II: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Baseline Mortality Study Results ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/fe93542fe9f1ceee027678eef4d5f7bb.pdf fba26b8edca7a530691dc9d16c535ee4 PDF Text Text Item ID Number 01586 Author Lathrop, George D. Corporate Author RODOrt/ArtlClO TitlO Air Force Health Study: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Volume II, First Followup Examination Results, January 1985-September 1987 Journal/Book Title Yoar Month/Day Color 1987 October n Number of limps 496 DOSOrlDton NOtOS Contract no. F41689-85-D-0010 and SAIC Project no. 2816-XX-195/254-XX. Wednesday, May 23, 2001 Page 1587 of 1608 �Air Force Health Study An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides SAIC Team Air Force Team George D. Lathrop, M.D., M.P.H., Ph.D. Stella G. Machado, Ph.D. Theodore G. Karrison, Ph.D. William D. Grubbs, Ph.D. Wanda F. Thomas, M.S. COL William H. Wolfe, M.D., M.P.H. Joel E. Michalek, Ph.D. LTC Judson C. Miner, D.V.M., M.P.H. LTC Michael R. Peterson, D.V.M., M.P.H., Dr.P.H. Project Manager: Program Manager: W.F. Thomas SCIENCE APPLICATIONS INTERNATIONAL CORPORATION 8400 Westpatk Drive McLean, Virginia 22102 EPIDEMIOLOGY DIVISION US AF School of Aerospace Medicine Human Systems Division (AFSC) Brooks Air Force Base, Texas 78235 October 1987 VOLUME II First Followup Examination Results January 1985 to September 1987 Contract Number F41689-85-D-0010 SAIC Project Number: 2-816-XX-195/254-XX (Distribution Unlimited) R.W. Ogershok �APPENDIX A Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants �Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants Committee Members Robert W. Miller, M.D., M.P.H., Dr. P.H. Chairman Chief, Clinical Epidemiology Branch 8C41 Landow Building National Cancer Institute National Institutes of Health Bethesda, Maryland 20892 Tel: (301) 496-5785 Charles C. Brown, Ph.D. Division of Cancer Prevention and Control Blair Building 3A07A National Cancer Institute National Institutes of Health Bethesda, Maryland 20892 Tel: (301) 427-8720 Jan M. Friedman, M.D., Ph.D. Director, Division of Clinical Genetics University of Texas Health Science Center 5323 Harry Hines Blvd. Dallas, Texas 75235 Tel: (214) 688-2143 Julianne Byrne, Ph.D. Executive Secretary Clinical Epidemiology Branch 8C41 Landow Building National Cancer Institute National Institutes of Health Bethesda, Maryland 20892 Tel: (301) 496-4947 Kathleen Kreiss, M.D. National Jewish Hospital and Research Center 3800 E. Coifax Avenue Denver, Colorado 80206 Tel: (303) 398-1525 Leonard T. Kurland, M.D., Dr. P.H. Professor and Chairman Dept. of Medical Statistics and Epidemiology Mayo Clinic 200 First Street, S.W. Rochester, Minnesota 55905 Tel: (507) 284-3216 George W. Comstock, M.D., M.P.H., Dr. P.H. Professor of Epidemiology Johns Hopkins University Johns Hopkins Research Center Box 2067 Hagerstown, Maryland 21740 Tel: (301) 791-3230 A-l �Richard R. Monson, M.D., Sc.D. Department of Epidemiology Harvard School of Public Health 677 Huntington Avenue Boston, Massachusetts 02115 Tel: (617) 732-1050 Herbert Pardes, M.D. Professor and Chairman Department of Psychiatry N.Y. State Psychiatric Institute 722 West 168th Street, Room 1411 New York, New York 10032 Tel: (212) 960-2500 Norton Nelson, Ph.D. Department of Environmental Medicine New York University School of Medicine New York, New York 10016 Tel: (914) 351-2566 Craig T. Ramey, Ph.D. Director of Research Frank Porter Graham Child Development Center University of North Carolina 54 Bypass Chapel Hill, North Carolina 27514 Tel: (919) 966-4121 A-2 �APPENDIX B Questionnaire Methodology �TABLE B-l. Elements of the Interval Questionnaires Type Method Elements Participant Interval Questionnaire In Person Demographic, educational, occupational, toxic exposures, reproductive experience, medical Spouse* Interval Questionnaire By Mail Comprehensive reproductive history Baseline Participant Questionnaire In Person Demographic, educational, occupational, medical, compliance, toxic exposures, and reproductive experience Baseline Spouse* Questionnaire In Person Comprehensive reproductive history Telephone Survey By Telephone Self-perception of health: current medications, severity of recent illness, absenteeism, income level *Present, former, or partner. B-l �B-2 �Participant Interval Questionnaire BKUIH UEC:K 01 DECKS OI-O2 442BSec-1 TIME BEGAN SECTION 1i AM | 1O-13X PM 3.INTERVIEHERI INTRODUCTION This part of the study asks about the health of current and foraer Air Force MeBbera and their faBillea. At various points in the questionnaire, we w i l l be using the terB "biological" to describe faally relationships. For exaaple, we Bight ask about your 'biological* children. When we uae this terB, we do not Bean your step-children or step-parent* or people related to you through adoption. He Bean people related to you by blood. You Bay refuae to answer any question you chooee. However, we and the Air Force aak that you anawer as Bany of the queatlona aa you can, so the results will accurately and fully tell your atory. I'd alao like to eBphaeise that we need aa accurate a picture aa you can reBBBber. So when we ask you about the date* of event* In your life, please think carefully. Hy records Indicate that your data of birth la (DATE or BIRTH. FROM mroRMATION t, ITEM I). la that correct? Ho (CORRECT 1NIURIIATIOR 2. B. C. Was It In 1981 or 1982? (RECORD ON INFORMATION SHEET. Naa It in the Spring, Busier, rail or Winter? (INTERVIEW! IF SPRING, RECORD ON INFO SHEET AS MARCH IF SUMMER, RECORD ON INFO SHEET AS JUNE IF FALL, RECORD 'ON INFO SHEET AS SEPTEMBER IF WINTER, RECORD ON INFO SHEET AS DECEMBER COHTIMUB TO FIND DOT FROM TBB MISSING nrroRiiAnoH FROM TMB INFORMATION INtUDMnlKJBj SU&I* I5-I6/ O2 17-18/ .OS 21-22/ 1b which of the following racial or ethnic groups does your biological father belong? (CODE ALL THAT APPLY) (PROBEi what others?) English/Heleh....01 SI-S2/ Engliah/Welsh....o1 Scottish 02 S3-54/ Scottish .03 5S-S6/ 19-2O/ .04 Irish 16-17/ 02 18-19/ .03 20-21/ Irish O4 S7-58/ Irish O4 22-23/ OS 23- 24/ Scandinavian OS S9-6O/ Scandinavian.....OS 24-2S/ 06 25-26/ Polish O6 61-62/ Polish O6 2S-27/ Russian O7 27-2B/ Ruaaian O7 Ruasian 07 28-J9/ Other Slavic OS 29-3O/ Other Slavic O8 6S-66/ Other S l a v i c . . . . . O H 30-31/ .....09 3I-32/ Jewish 09 67-68/ Jewish O9 32-33/ French. .10 33-34/ French 10 69-7O/ French 10 34-3S/ Italian it 3S-36/ Italian it 71-72/ Italian II J6-17/ Spanish 1} 37-3B/ Spanish 12 73-74X Spanish II 38-39/ 13 7S-76/ Mexican 13 40-41/ 14 77-7B/ Greek 14 42-43/ Jewlah (RECORD ON INFORMATION SHEET) What Booth did the Interview take place? CAN'T REMEMBER EXACT MONTH ASK C) INTERVIENERi 1b which of the following racial or ethnic groups doea your biological Bother belong? (CODE ALL THAT APPLY) (PROBE• What others?) Polish ADD GO ID Q.2)...2 Do you reateBber the Bonth and year you were interviewed the last tiae for thia study? We are talking about the interview in your hoaw before you went for your physical exaB. A. Englleh/Helah....O1 C. FATHER Scandinavian Yea w B. MOTHER To which of the following racial or ethnic qroupa do you belong? (CODE ALL THAT APPLY) (PROBEi What others?) Scottish First 1 have a few background queationa to aak you* 1. ASK A POR RESPONDENT, THEN ASK B AND C FOR PARENTS A. RESPONDENT IF R Mexican 13 39-40/ Mexican Greek 14 41-42/ Greek American Indian..is 43-44X taerlcan Asian 16 4S-46/ «»lan .17 47-4S/ African African Other (SPECIFY) 63-64/ Indian..15 79-8OX BEGIN DECK O2 16 10-11/ 17 12-1 3/ Other (SPECIFY) .18 49-SO/ American Indian..IS 44-4S/ Asian 16 46-47/ African 17 4B-49/ ..IB SO-S1/ Other (SPECIFY) ..18 14--15/ �DECK 03 URCKS 02-03 SECTION 2l 1. 2. EDUCATION My records show that when you were lent interviewed you had received a (DEGREE LAST OBTAINED FROM ITEM 2. IMFOMUTIOD SHBKT). Have you received any additional regular school certificates, diplomas or degrees since that Lime, that is, since (DATE OF LAST INTERVIEW)? Since (DATE OF LAST INTERVIEW) have you participated in any civilian job training program (other than the for ml schooling that we discussed), that prepared you for a Major change in your occupation? Tes (ASK No Yes (ASK A AND B) A) 1 (SKIP TO Q.I) 2 I 52/ Ho (SKIP TO Q.2) 2 1ST PROGRAM INTERVIEHERt 2ND PROGRAM 3RD PROGRAM FOR EACH DEGREE CODED IN A, ASK B. B. A. What certificates, diplomas, and/or degrees did you qet? (CODE ALL THAT APPLT) High school diploMa For what kind of work was your first civilian training program preparing you? In what year did you receive (DEGREE IN A. RECORD TEAR O1 SJ-S4/ "LJLJ E. for what kind of work was your second civilian training prograsi preparing you? For what kind of work waa your third civilian training prograst preparing you? 55-5C/ Tear High school equivalency diploam........02 57-SS/ Associate of Arts (A.A.I. 61-6J/ .01 "LJLJ Tear "LJLJ 59.-60/ 63-64/ IJ-IS/ 03 *- B. Bachelor of Arts (B.A.) or Bachelor of Science (B.S.I O4 6S-M/ "LJLJ Tear 67-C8/ Masters (H.A. or M.S.) OS 69-70/ «L_LJ Tear 7S-76/ In what Month and year did you start this training? 71-72/ "LJLJ Tear 25-I7/ Doctorate (Ph.D., M.D., M.O., Sc.D.)..O6 T3-74/ Month «»L_LJ Month K. L_L_L_I_J Month Tear Month 20- 2V In what Month and year did you start this training? I_J__I_J I Month Tear 40-4V Tear 28-31/ In what Month and year did you complete this training? 79-BO/ Tear 77-7S/ BEGIN DECK OJ 08 IO-I1/ J. l_L_J_LJ Tear I8-I9/ year did you complete this training? . .07 No certificate, dlploM, or degree (volunteered) In what month and year did you a tart this training? C. In what Month and Others (SPECIFY) 37-19/ In what Month and year did you cosplete this training? I__I_J_J Tear J2-JV Month I Tear 44-47/ CURRENTLT IN TRAINING...1 CURRENTLT IN TRAINING...1 CURRENTLT IN TRAINING...1 O. H. L. Have you participated In any other civilian job training prograM that prepared you for a Major change In your occupation? Tes..(ASK r) I No.(SKIP TO 0.1)..2 24/| Have you participated In any other civilian job training prograM that prepared you for a Major change in your occupation? Yes..(ASK I) 1 Ho.(SKIP TO p. 31..2 36/ Have you participated in any other civilian job training prograM that prepared you for a Major change In your occupation? Tes(GO TO HEH OUEX).t Mo..(GO TO p.J) 2 48, �5. 3. Hava you served in the military (at a l l ) since (DATE OF LAST INTER V I E W ) Tea No 4. I (SKIP TO SECTION 3> 49/ BHCIN OFXK O4 (Continual! A. 2 3RD PROGRAM 2NU PROGRAM 1ST PROGRAM For what kind of wi»rk was your third M i l i t a r y training prograM preparing you? For what kind of work w«« your second M i l i t a r y training prograM preparing you? For what kind of work was vmir f i r s t M i l i t a r y training prograM preparing you? Are you currently serving in the Military? so/ If.*.., No*••• 5. Now, let'a talk about any military and specialised training program* that prepared you for a Major chanqe in your occupation. Since (DATE OF LAST INTERVIEW), (and besides the forMl schooling, and job training program you've told M shoot), have you participated in any Military technical or specialized training prograa* ttiat prepared you for a Major chanqe in your carssr? Tea....(ASK A-EI 1 B. Hhat is the AFSC for that Job? G. Hhat Is the AFSC for that Job? L. SI/ C. CO In what Month and year did you atart this training? I I J I Hontfi Year H. In what Month and year did you start this training? I I 57-60/ l>. I. In what Month and year did you coMptrt** this train!nq? L.LJ _J _I_J I 4/ 10- LJL_UL_LJ 66-70Y _ _ _ _ 52-S6/ Ho.(SKIP TO SUCTION 31..2 H* In what, month and y«?ar. did you start thin L r a t n i n i ? I__I__J_J Month Yenr 7I-74/ iS-in/ N. In what Month und year did you conplPte thin training? In whal -aonlh yf*<ir did you cowpl^l" this LJ_I__I__I Honlh Y*-ar UL_LJ_I Month Year l_LJLJLJ Month Year What is the AFSC for that Jol>? SI-64/ ?S-7e/ CURRENTLY IN TRAINING..I CURRENTLY IN TRAINING..I E. Hive you •participated In any other M i l i t a r y Job training prograM that prepared you for a Major change in your occupation? Y e a . . I ASK F ) . . I 6S/ Have you participated in any other M i l i t a r y Jon training prograM that prepared you for a Major change in your occupation? Yes..(ASK K ) . . 1 ?9/ CUItRraiTLY IN TRAIHIII'i. .» O. Hav-s you participated in any oti'*r M t U l . i r y Job tmlninq prtxirAM that prepared you for a major chan<i* In your occupation? Yes<NEH 24-27/B �DECKS 04-05 RUCTION 3: KHPI.OYHFJIT 1. (Continued) E. How I have now* questions about working. Please tell me ahoiiL any jobs you've had that lasted for 3 Months or longer sine*. (DATE OF LAST INTERVIEW). If you had nore than one Job at the SAM« tine* please tell Me about each joh separately. Count chanqes of jobs for the saMe employer as separate jobs. Do not include jobs in the Military. Let's start with the s*>st recent regular job you've had and work back in tiMe to (DATE OF LAST INTERVIEW), 1. Please look at this card and tell Me which niuber best describes the kind of industry you (work/worked) in? ( W R I T E IN NUMBER) ENTER NUMBER, In what Month and year did you start your Most recent job that lasted 3 Months or longer? P. I Year A. I I Month HO CIVILIAN JOBS (SKIP TO SECTION 4) I J2/ CURRENT JOB What (is/was) the n«Me of your employer? G. | 65-66/ | In what Month and year did this job end or is Job* 2B-3I/ Month | I this your current I 67-70/ Tear (SKIP TO O.2> » 1 1V What was the Main reason you stooped working on your job? 33-S7/ T B. 7Z-73/ (Is/Was) this a full-ti»e or part-tisn job? i. Ful 1-tiMe Pa r t- ttMe C. I 2 S8/ What kind of business (is/was) that—what (do/did) they Make or do there? , ^ FOR EACH SUBSTANCE CODED IN p. 2, ASK A. Mhils working at (EMPLOYER) (do/did) you COMB in contact with any of the substances on this card? By contact I Mean that you inhaled, tasted, had skin contact with these fibers snd chemicals or were exposed to ionising or nuclear radiation. CODE ALL THAT APPLY A. In general, how nany days a Month did you co«e in contact with (SUBSTANCE)? Less than once a Month 59-61/ Asbestos 01 I 74-7S/ | BEGIN DECK OS Ionizing or nuclear radiation O2 10-11/ I D. What (do/did) you actually do on the job—what (are/were) sane of your Main duties? 62-6V 1 Industrial cheoicals | Insecticides or pesticides Degreaslng che*lcel Defoliants or herbicides O3 14-1S/ I O4 OS O« None of the above (SKIP TO O.S)...07 95 7S-77/ I days 95 12-13/ | days 95 16-I7/ »8-1»/ | 22-21/ | 26-J7/ | 30-3I/ day | | | days | days | | days 95 2O-2V 95 J4-25/ 95 I8-29/ �III'.Ill HI.' I 3. All of the tUe C. 1 Hhat kind of business was that—what did they «ake or do there? What did you actually do on the Job—what were some of your Bain duties? 2 Never (Continued) D. t Same of the time 4. 6. Hlille you "ere on that Job, how often (do/did) you wash to renove the (SUBSTANCES) or use protective gear — would you say all of the tlxe, ao«ie of the tine, or never? (SKIP TO p.5) 12/ Which of the following (do/did) you use on that job? (CODE ALL THAT APPLI) 13-IS/ » Air filter Ol 3J-34/ Goggles 02 3S-36/ Face .hleld 0) 37-J8/ Special clothing 04 39-4O/ Waahlng facilities...... OS 4I-42/ Self-contained or supplied air breathing apparatus 0« 4J-44/ Moo. 07 4S-4«/ E. HAND CARD E HAND ENTER NUMBER I | F. 1 I. (SKIP TO Q.21) L_J Hontli 16-1 47/ L_J I8-21/ Year CURRENT JOBi....(SUP TO 0.7) OOO1 2 G* 6. | In what Bonth and year did this Job end? I Did you have another job before the Job with CHAME IN O. IA) but. since (DATE or LAST INTERVIEW) that lasted ) mntha or longer? No | CARD C 00 T, Please look at this card and tell He which number beat describes the kind of Industry you worked In? (WRITE IN NUMBER) Hhat was the nain reaaon you stopped working on your Job? In what snath and year did you start that Job? 22-2V I A. I I I Honth Tear I 4B-S./ Hhat was the naM of your employer? _______ B. Has this a full-tin* or part-tin* Job? rull-tlM (Continued) I Part-tlM 6. S2-76/ .2 77/ III, �I>H( K (H, UECKS 06-07 7. FOR EACH SUBSTANCE While workinq al (EMPLOYER) did you come in contact with any of th<* substances on this card? By contact I mean that you inhaled! tasted* had skin contact with these fibers and chemcials, or were exposed to ionlzlnq or nuclear radiation. (CODE ALL THAT APPLY) Asbestos Ol .03 Industrial chemicals A. In general how many days a month did you come in contact with (SUBSTANCE)? Less than once a month 24-2V | Ionizing or nuclear radiation....O2 IO. conni IN p. 7. ARK A. | 12-11/ | | 6S/ No II. I | day* 95 O4 36-37/ ( oegreaslng chemicals OS Defoliants or herbicide* O6 44-4S/ | J None of the above (SKIP TO p.tO).O7 8. 4O-4I/ | | | | days Some of the time. 95 46-47/ Ha* this a full-time or part-time job? rul 1- time 1 Part-time 2 Never BEGIN DECK O7 «O-34/ B. 1 66-69/ 42-41/ 95 4B-49/ All of the time L_l Year What was the name of your employer? While you were on that job, how often did you wash to remove the (SUBSTANCES) or use protective gear — would you say all of the time, some of the time, or never? CO I 39-19/ | | days 2 14-35/ A. Insecticides or pesticides I Month 30-3I/ | (SKIP TO 0.21) In what month and year did you start that job? 26-27/ |__| 2B-29/ | Did you have another job before the job with (NAME IN Q.6A) but, since (DATE OF LAST INTF.RVIEH)? SO/ 3 C* Miat kind of business was that what did they make or do there? J6-38/ 9. (SKIP TO O..1O) Which of the following did you use on that job? " ottle." ^ 'CtU*Ulr *» - «• Job-what -ere some of your main ' (CODE ALL THAT APPLY) Air filter 01 S1-S2/ Goggles O2 53-S4/ Face shield Ol S5-S6/ Special clothing O4 S7-58/ Hashing facilities OS 59-60/ Self-contained or supplied air breathing apparatus O6 6I-62/ None 07 6J-64/ 39-41/ HAND I CARD E I Please look at thia card and tell me which number best describes tn* kind of Industry you worked In? ENTER NUMBED, | II. (Continued) | | 42-43/ �II. 14. (Continued) F. Mhlch of the following did you use on that Job? In what Month and year did thla Job end? (CODE ALL THAT APPLY) I I I I O1 ID-It/ Goggles I Air f i l t e r O2 12-1 3/ 44-47/ CURRENT JOB G. HAND CARD E what was the naln reason you stopped working on your Job? 4B-49/ 12. While Hording at (EMPLOYER) did you cone In contact >.wltli any of the •ubatancea on thla card? By contact I >ean that you Inhaled, tasted, had skin contact with these fibers and cbeMlcals, or were exposed to Ionizing or nuclear radiation. CODE ALL THAT APPLY A. Asbestos 01 SO-51/ | Ionizing or nuclear radiation....O2 HAND CARD O Industrial chemicals | 54-5S/ | O3 | days | 5S-S9/ | | days | 16-I7/ O5 18-19/ O6 2O-21/ O7 22-2J/ (DATE Or LAST INTERVIEH 17 1 Yes Ho 95 52-SV 16. | days I 04, 62-63/ I I I days 95 «4-«S/ OS «6-«7/ I I I days ......95 OS I I days 95 2 Year I ZS-28/ «•-«»/ Defoliants or herbicides I__LJ Honth 95 «O-««/ Insecticide* or pesticides (SKIP TO 0.21) 24/ In what Month and year did you start that Job? 95 56-S7/ Degreastng cheMlcals 72-7J/ None of the above (SKIP TO Q.1S).O7 13. 14-IV O4 Old you have another Job before Job with (NAME IN g.HA), but since In general how Many days a Month did you COMS tn contact with (SUBSTANCE)? Leas than once a Month W vO 15. rOft EACH SUBSTANCE COOED IN O.I2, ASK A. O3 Self-contained or supplied air breathing apparatus tOf.lt TO O.12)....OOOI Face shield Hashing f a c i l i t i e s Year Special clothing None Month 7O-7I/ I A. what was the nans of your employer? 29-S3/ 74-7S/ B. Mas this a futl-tlMe or part-tlM Job? While you were on that Job, how often did you wash to remove the (SUBSTANCES) or uee protective gear — would you say all of the Han, mourn of the tls*, or never? All of the tlM 1 SOM of the tlMe..... 1 1 Part-tlMe 2 54/ 2 Never Ful 1- tiMe (SKIP TO Q.1S) 76/ C. what kind of business was that—what did they Make or do there? S5-57/ 16. (Continued) �16. (Continued) D. IB. what did you actually do on the Job—what were duties? of your Ha In While you were that Job, how often did you wash lo remove the (SUBSTANCES) or use protective qear — would you say All of the tine, some of the time, or never? All of the time I Some of the til.-..; 2 Never 24/ 3 58-60/ E. Please look at this card and toll me which number best describes the kind of Industry you worked In? (WRITE IN NUMBER) 19. ( S K I P TO Q.2O) Which of the following did you use on that Job? CODE ALL THAT APPLY Air f i l t e r ENTER NUMBER I | F. | | I I I Month I I 20. 67-6B/ i—* O 29-JO/ O4 J1-J2/ O5 JJ-34/ Of. JS-36/ None G. , What was the Mln reason you stopped working on your Job? Cd O3 Self-contained or supplied air breathing apparatus CURRENT JOBI....ISKIF TO Q.1D....OOOI 27- 28/ Mashing facilities 63-6S/ O2 Special clothing Tear 25-2V Face shield HAND I CARD E \ In uhat month and year did this Job and? 01 Goggles 61-62/ 07 37- 38/ Did you have another job before the Job (NAME IN O. ISA), but since (DATE OF LAST INTERVIEW)? . Yes 17. Hhlle working at (EMPLOYER) did you cone In contact with any of the substances on this card? By contact I aiean that you Inhaled, tasted, had skin contact with these fiber* and chemicals, or were exposed to tootling or nuclear radiation. (CODE ALL THAT APPLY) Asbestos tOK EACH SUBSTANCE CODED IN 0.17, ASK A. No 21. In general, how Many days a Month did you COM In contact with (SUBSTANCE)? Less than onth 1 During the past six Months, did Illness or injury keep you from, work, not counting work around the house? 1 No 69-7O/ days .95 71-72/ 7J-74/ days .95 79-60/ days .95 (SKIP TO SECTION 4 . . ).) 7S-76/ .95 40/ (SKIP TO SECTION 4>...2 Retired Ol )9/ 2 Yes Ionizing or nuclear radiation....Ol 12-I3/ Unemployed..(SKIP TO SECTION 4 . . ).4 22. HAND CARD D (USE HEM QUEX) Industrial chemicals Insecticide's or pesticides OJ 77-7B/ BEGIN DECK O9 .04 to-tt/ Altogether, how tuny days did Illness or Injury keep you from work during the past six months? (REFERS TO -WORKING DAYS* OHLY) 41-43/ Degreaslng chemicals.... ...OS 14-1S/ days .95 16-I7/ Defoliants or herbicides OC 18-19/ days .95 2O-21/ None of the above (SUIT TO Q.2OI.O7 22-23/ ENTER NUMBER OF DAYSi 21. What illnesses or Injuries caused you to alas work? 44/ �DECK O9 SECTION 4i 1. 3. HII.ITMtY Now I urn qoinq to ask you about aone of your experiences in the M i l i t a r y . First, which of the followinq statements best describes your asaiqnoenl durinq the vietaan Mar? (Continued) C. Followinq your (retireMent/separatlon/dlacharqe) in (DATE IN B.), did you reenter the at Bed forces? Yes. A crew Member in Vietnam who was on • flying status 1 45/ No.. Not a crew Member, hut flew one or Bore Missions in Vietnam. .2 A crew Member, but did not log flyinq tim» in VietnaM 3 4. Not a crew nearer. . . . . . . . . • • • • • • . . • • • . • * • . . . . . . . . . . • • . . . . • • • .4 2. INTERVIEWER! HAS R SERVED IN MILITARY SINCE LAST IHTERV1CM7 IN SECTION 2, COOED "»ES"?) YES (GO TO p.3) (IS O.3 I would like to ask you the na>ea of all the countries. Including the United States, you have been stationed in since (DATE OF LAST INTERVIEW) When last Interviewed you were stationed In (COUNTRY FROH mrORMATIOH ITEH 3), and your assignment began In (DATE OF ASSIGNMENT FROM ITEH 3). Is that correct? 46/ 1 Yes....(ASK • THROUGH K> 3. (SKIP TO SECTION *) No A. 47/ (ASK A THROUGH C). (SKIP TO Q 4 . . .).. Were you retired, discharged or separated? Retired t Discharged/Separated B. 48/ 2 In what aonth and year were you (retlrad/discharged/separated) fro the (BRANCH OF SERVICE FROM IHKNMKnoil SMCT ITEM 3)? 4»-52/ Month Year 2 2 Since (OATE OF LAST INTERVIEW) have you retired, been discharged or separated fro. the (BRANCH OF SERVICE FROM INFORMATION 8HEKT ITEM 3)7 Yes 1 No...(CORRECT nrOMMTIOM SHUT, THEN ASK B THROUGH K) NO s«/ �IIIOCKS 09-11) 4. HECKS 10-11 BEGIN DECK IO (Continued) 4. Since (DATE OF LAST INTERVIEW), In what country were you next ntatloned while on active duty? Include temporary duties of greater than 9O days. Since (DATE OF LAST INTERVIEW), tn what country were you next stationed while on active duty? Include temporary duties of greater than 9O days. (Continued) O. (Do/Old) your duties In (COUNTRY) Include flying* 60/ BEGIX DECK I I O. 1 Yes COUNTRY B. H. In what stonth and year did you begin and end active duty In (COUNTRY)? E. 3S-36/ IO-II/ In what month and year did you begin and end active duty in (COUNTRY)? X. How Many flight hours did you log while In {COUNTRY)? In what Month and year did you begin and end active duty In (COUNTRY)? I Yes No...(SKIP TO GI....2 55-S6/R (Do/Did) your duties In (COUNTRY) Include flying? IO/ P. No..(SKIP TO CO....2 How aiany flight hours did you log while in AA. How Many flight hours did you log while In (COUNTRY)? 11-M/ l__l__l I Honth Year I I I_JL_J Honth Year I r. 37-40/ 57-60/ END EMO L_I_J_J I L_I_J_J_J Honth Year Year 4I-44/ 6I-64/ Current Cur r n . . . . . . . et......I I C. What specific Job assignments (do/did) you have in (COUNTRY)? Can you give me the Air Force Speciality Code? (PROBEt Nhat others?) N. What specific job assignments (do/did) you have in (COUNTRY)? Can you give a* the Mr Force Speciality Code? (PROBEi What others?) V. What specific job assignments (do/did) you have in (COUNTRY)? Can you give SM the Air Force Specialty Code? (PROBEs Nhat others?) 1. I '• 1. I I I_J 6S-69/ I_I_J_J_J_J 20-I4/ I I I I I | 7S-79/ 3 . !_J_J_J_J_J 30-34/ I I I I I 4S-49/ 2. 3. 3I-34/ DDDD Hour* l_l I Yes BEGIN BEGIN Honth (Do/Did) your duties In (COUNTRY) Include flying? 30/ No...(SKIP TO RI....2 (COUNTRY)? 6I-64/ Z. I_J_J I_JL_| 50-S4/ what specific letter and numerical designation^) did each aircraft have? Nhat specific letter and nunerical designation^a) did each aircraft have? BB. What specific letter and numerical designation^) did each aircraft have? �DFCKS 11-12 4. H. (Continued) R. in your joH assignin (COUNTRY), (do/did) you come in contact with any of the substances on thin card? By contact 1 mean that you inhaled. tasted, had skin contact with these fiber a and chemicals. or were exposed to ionizing or nuclear radiation. CODE ALL THAT APPLY Ionising or nuclear radiation.... 02 66-«7/ degreasing chesiicaia.OS 58- 59/ degreaaing chesdcala.OS 72-T3/ defoliant* or herbicides 74-7S/ defoliants or O6 60-6 1/ none of the above 62-6 3/ none of the above 7«-77/ O« 20-21 / none of the above 22-2J/ asbestos | | | 50- b . less than less than ionizing or nuclear radiation ionictnq or nuclear radiation | | | 26- 27/ | I | 39- 40/ less than less than industrial chemicals | || 28- 29/ | | | S2-53/ | | | 41-42/ industrial chemicals f leas than inaectlclde* or pe*Ucide* | | I 4 3- 4V lea* than insecticides or pesticide* | degreaaing cheMlcal* decreasing chemicals | I 54-5S/ less than degress! nq che»ical* | I| 32-S3/ | | | 4S-46/ | | 56-57/ less than | I | 58- S9/ less than I. leas than less than defoliant* or herbicide* | || 34-3S/ lea* than degreasing chemical*. OS !•-!»/ defoliants or herbicides In general, how nany days a iftonth (do/did) you come in contact w i t h (SUBSTANCE)? insecticides or pesticides | | | 30-H/ less than inaectlclde* or pe*ticide*. . . ......O4 IS-17/ 70-7 I/ | | | J7-18/ ID. lea* than induatrlal cheMical*.O3 14-IS/ insecticide* or asbestos iesa than Ionizing or nuclear radiation. .. .02 12-1 I/ Insecticides or In general, how Many days a Month (do/did) you COMC in contact with (SUBSTANCE)? ionizing or nuclear radiation BEGIN DECK 12 industrial chenlcal*.O3 6B-69/ 56- 5 7/ | || 24- 2V industrial chemical* industrial chemicals. O3 54-S5/ S. less titan 10-1 1/ 64-6V ionizing or nuclear radiation. ..»O2 52-5V asbestos contact with these fibers and cheMicala, or were exposed to ionising or nuclear radiation. CODE ALL HAND CARD 0 D 50-5I/ CC. In your job aaaiqnMents while stationed in (COUNTRY), (do/did) you COMA in contact with any of the aubntancea on thla card? By contact I Mean that you inhaled. tasted, had skin THAT APPLY HAND CARD D I you coMe in contact with any of the substances on this card? By contact I Mean that you inhaled. tasted, had akin contact with theae fibera and cheMicala, or were exposed to ionicing or nuclear radiation. CODE ALL THAT APPLY HAND CARD I—' OJ In your job assignaients whi le stationed in (COUNTRY), (do/did) In general, how Many days * Month (do/did) you cove in contact with (SUBSTANCE)? defoliants or herblcidea | || 47-4B/ le*a than defoliants or herbicides | | | Did you waah to re*ove the (SUBSTANCES) or did you u*e protective clothing or gear when stationed in (COUNTRY) all the tine, BOMB of the tine, or never? T. Did you wash to reMove the (SUBSTANCES) or did you use protective clothing or gear when stationed in (COUNTRY) all the tiae. BOMS of the tine, or never? SOMC of the ti«...l SOMB of Never. (SKIP TO K ) . . 3 Never. (SKIP TO V K . J * the U.MC...2 * 60-6 l/ less than EE. Did you wash to remove the (SUBSTANCES) or did you use protective clothing or gear when stationed in (COUNTRY) All the time, some of the tine, or never? Some of the time ... 2 Never. ( S K I P TO G G > . 3 * �DECKS 12-1) OFCi442asec-5 -24- DBCK I BRGIN DECK 1 ] J. Which of the following did you use on that Job? CODE ALL THAT APPLY Air f i l l e r CARD 1 E | II. Which of the following did you use on that job? CODE ALL THAT APPLY 01 63-A4/ Air filter Ol 67-6D/ Face Shield. ...O3 14-IS/ SECTION 5: 1. Yen Face Shield. ...0) Special Special Washing facilities Hashing facilities 29- JO/ No A. 7»-72/ Self contained or supplied air breathing 05 18-19/ Self contained or supplied air breathing INTERVIEWER I A) 1 (CORRECT IHrO MEET, THEN ASK A) 40/ 2 HAS STATUS 'MARRIED" AT LAST INTERVIEW? YES Self contained or supplied air breathing (SKIP TO C) t NO OS 33-34/ B. (ASK B) 2 INTERVIEWER) SEE INFORMATION SBEBT ITEM S. WITH PARTNER' AT LAST INTERVIEW? YES W (ASK Special Hashing facilities. ....05 K. MARITAL AND FERTILITY HISTORY Now 1 would like to ask you about your personal relationships. when we lalk»rt w i t h you last, you said you were (READ MARITAL STATUS FROM INFORMATION SHEET ITEM 4). is that correct? 1O-1I/ Face Shield.... O3 FF. Which of the following did you use on that Job? CODE A1.L THAT APPLY Are there any other countries that you have been stationed In since (DATE OF LAST INTERVIEW)? Yea. (GO TO L)...» NO. (SKIP TO SECTION 5J..2 V. 77/ Are there any other countries that you have been stationed In since (DATE OP LAST INTERVIEW)? Yes. (GO TO NK..I NO. (SKIP TO SECTION S)..J HO CG. Are there any other countries that you have been stationed In since (DATE OF LAST INTERVIEW)? M/ Tea. ..(USE HEW OUEX)....l C. 4I/ HAS RESPONDENT "LIVING (ASK C> 42/ ....(SKIP TO Q What Is the current f u l l nane of the person you were living with in (DATE OF LAST INTERVIEW)? J9/ HO. (GO TO MCTION 5)., 2 FIRST NAME INTERVIEWER I D. MIDDLE NAME RECORD NAME ON INFORMATION SHEET, ITEM OS In what Month and year did you start living with (NAME)? ENTER MONTH AND YEAR |_ | MO | |_ | _ | IK 4J-46/ (GO TO 0.2) E° According to our records, you were married to (NAME AT INFORMATION SHEET ITEM 6). Is that correct? Yes .................................. . ...... 1 No. .(CORRECT INFO SHEET, THEN GO TO V - 2 K . . . 2 47/ �OFC:442BSec-5 1. DECKS 11-16 (Continued) F. 2. In what month and year did you get married to (NAHEl? ENTER MONTH MO YEAR | | | | NO | (Continued) E. | 4B-SI/ Durinq this (Barriaqe/relalionahip), (aincr the (DATK OF LAST INTERVIEW)|, 4id you ever have a problem conc*lvinq a child because of prolonged Reparation? YR Yea. >2. Have you stopped living Kith (NAME)? Tea (ASK No.. At 1 F. Ho A. (SKIP TO C) How did this (marriage/relationship) and? • HAND CARD And what is (NAME OF SPOUSE/PARTNER)'a present street address? 1 " STREET ADOS ESS Separation*••••*«••••••••••••••••••«•••••••.1 BEGIN DECK 14 10- 34/ S3/ G. And what city, state and sip coda does (SPOUSE/PARTNER) live in? Dlvorce....•••••••••••••••••••••••••••••••••2 G Death of •poiiM or partner.. (ASK f-mt 0.3) I CITT i 3S-54/ H. I I I I I STATE S5-S6/ I ZIP I And what is (NAME of SPOUSE/PARTNER) 's present telephone number? I B. I CODE S7-61/ I In what month and year did till* occur? ENTER MONTH AMD YEA*! | | | | | HO I S4-S7/ I I AREA CODE I I I I - I NUMBER I I I I 62-7I/ TK H-» Ul C. During thia (marriage/relationship), how many times war* you living apart tram (NAME) (or 3 months or mra alnca (DATE OF LAST INTERVIEW)? ENTER MUHBE* OP TIMES I Nona D. ............ |_ !• S8-S9/ (_ | Thinking of all the people you know, who would be the one person who would be most likely to know where (NAME OF SPOUSE/PARTNER) is? ENTER FULL NAME OF PERSON BELOW AND ASK J-H. BEGIN DECK IS 1O-29/ (LAST), 3O-44/ (SKIP TO F) For how Mny montha did yon »!»• »f*rt the C first/next) (MIDDLE) time? J. Flr.t tlM, | | I 60-61/ Second time. | | | 62-63X Third UMI | | | 64-6S/ Fourth time I | | I 66-6T/ What Is (PERSON'S) relationship to (NAME OF SPOUSE/PARTNER)? L_L_L_LJ_L_LJ_J_LJ_J_LJ_J K. Where does (PERSON) live? 47-71/ APT! STREET ADDRESS BEGIN DECK 16 L.LJ STATE CITT 10-29/ 30-3I/ I I IZIP I CODEI I 32-16/ �DECKS 2. (Continued) L. 4. (Continued) Hhal ia (PERSON'S) telephone nunber? A. AREA CODE What la the nil name of (thia person/your w i f e ) ? LJ_I PHONE NUMBER NO PHONE H. 16-17 ID I ..< IF (PERSON! HAS PHONE. In whoae naan la the phone Hated? (PERSON'S) name (SKIP TO H) OTHER 77-7B/ 1 (SPECIFY BELOM) (LAST) 47/ 2 DON'T KNOH.. (FIRST) B INTERVIEWER! 48-7 2/ _^__^ (LAST), (NIDDLE) RECORD FULL NAME OF (SPOUSE/PARTNER) ON INFORMATION SHEET, ITEM O7 AND RECORD IDI ABOVE. BEGIN DECK 17 What waa her full Balden nane? (FIRST! (MIDDLE) J N. INTERVIEHER> HAS F STOPPED LIVING WITH SPOUSE OR PARTNER? IIS "TES* CODED AT Q.2T) what waa her blrthdate? .,-t Ye* RECORD DATEi | I | I I | | ( HO 73/ DA I IO-29/ | | JO-35/ YR W No (SKIP TO p.lO) 2 B. 3. In what aontn and year did you (reconcile/get aarrled to/start living with) (NAHE)? Since (DATE OF LAST INTERVIEW), have you reconciled, lurried (again) or have you lived together •• • partner (or 3 Bonthe or aiore with aomone to whoa you Keren't Mrrled? Tea No (ASK A) 1 (SKIP TO Q.IO) ENTER HONTH AND YCARi HO Co t Have you •topped living with (NAHE)? How Bany tlewa did you get Harried or live aa a partner wlUi aowone (or 3 auntha or Bore el nee (DATE OF LAST INTERVIEW)? RECORD NUMBER OF TIHESi | | No 75/ D. (ASK D-P)... 40/ (SKIP TO 0.4F). Haw did Uils (Harrlage/relatlonehlp) end? Separation 4. HAND CARD Thjnklnq of (that/the tlrat) relatlonablp alnce (DATE OF LAST INTERVIEW), did you a*rry thla peraoB? Tea. Ho RECONCILED. 36-19/ YR 74/ Yea A. | | ( | | i Dt vorca 2 6 76/ Death of apouae or partner. . I ASK «-B, THEN •KIP TO O.5I...3 417 �DECKS 18-JO 4. (Continued) E. 4* In what month and year did this occur? ENTER MONTH MID YEAR. (Continued) L. | | | _| HO | 42-4S/ YR Thinking of all the people you know, either around here or elsewher*. who would be the one person who would be moat l i k e l y to know where (NAME OF SPOUSE/PARTNER) is? ENTER FULL NAME OF PERSON BELCH AND ASK H-P. 47-6V F. (LAST). During this (marriage/relationship), how riany timea were you living apart from (NAME) for 3 months or more since (DATE OF LAST INTERVIEW)? 66-8O/ (FIRST) (MIDDLE) BEGIN DECK ENTER NUMBER OF TIHESl | | | 19 46-47/ M. What Is (PERSON'S) relationship to (NAME OF SPOUSE/PARTNER)? OR 10-II/ Hone (GO TO I) OO N. G. "here does (PERSON) live? For how many months did you live apart the (first/next) time? I2-36/ First time, | | | 48-49/ Second time, | | | SO-SI/ Third time, l__l__l S2-S3/ Fourth time, I_J__J 54-SS/ STREET ADDRESS APTI CITY H. 17-56/ O. I i I I ZIP CODE S9-63/ - I. I I I I 1 In whose name la the phone listed? (SKIP TO Q.S) 1 OTHER (SPECIFY BELOW) 2 DON'T KHOM And what is IF (PERSON) HAS PHONE. (PERSON'S) name I. P. 64-73/ I PHONE NUMBER NO PHONE S6/ I (PERSON'S) telephone number? I AREA CODE L-JLJ During this (marriage/relationship), (since the (DATE OF LAST INTERVIEW)!, did you aver have a problem conceiving a child because of prolonged separation? Yes Hhat is L_LJ I STATE S7-58/ (SKIP TO O.5) 8 7V (NAME or SPOUSE/PARTNER)'* present street address? S7-80/ BEGIN DECK 2O STREET ADDRESS BEGIN DECK 18 J. L_L_I STATE CITY 10-29/ K. L_L_L_L_J_I XIP CODE 30-1I/ 32-36/ And what is (NAME at SPOUSE/FARTHER I • s present telephone number? LJLJLJ - I_J__LJ_J "-«•/ IO-34/ (LASTI, And what city, state and sip coda does (SPOUSE/FARTHER) live In? (FIRST) S. INTERVIEWER, (MIDDLE) IS THERE A SECOND RELATIONSHIP SINCE THE DATE LAST INTERVIEW? (1C NUMBER OP TIMES RECORDED Hf O.3A EODAI. TO 2 OR MORE?) YES NO (GO TO O.6) (SKIP TO Q.1O) I 2 3V �DEI'K 6. Thinking of the next relationship since (DATE or LAST INTERVIEW), did you »arry this person? Yes. 6. (Continued) E. In what month and year did this occur? 36/ ENTER MONTH AND YEAR: | No.. A. Mhat is the current full MUM | | | HO of this person? 2D-2I F. 7I-74/ | YR During this relationship, how many times were you l i v i n g apart from (NAME) for 3 months or more since (DATE or LAST INTERVIEW)? ENTER NUHRER OF TIHESt |_ 7S-76/ |_ | (LAST) 37-38/ (FIRST) INTERVIEWED (HIDOLE) None ...... (GO TO I) RECORD FULL NAME OF (SPOUSE/PARTNER) ON INFORMATION SHEET, ITEM O7 AND IDf ABOVE. G. First tlmei.........|_ | _ | I I I I I I I LJLJLJ Hhat was her blrthdate? OO For how many months did you live apart the (first/next) time? Hhat was her full Maiden nans? I .................. RECORD DATEt I I I | ) [ | HO DA | | »-»/ | 77-7B/ Second time i........|_ |_ | 79-BO/ BEGIN DECK 21 Third timei.........|_ |_ | IO-II/ S9-64/ Fourth tlmei........|_ |_ | YH I2-13/ W *—• B. oo In what month and year did you (get married to/start living with) (NAME)? EHTER HOHTH AND YEAR I | | | NO | | N. During this (marriage/relationship), (since the (DATE OF LAST INTERVIEW)), did you ever have • problem conceiving a child because of prolonged separation? 6S-68/ Yes No ..... . C. ......................... I M/ YH .................... 2 Have you stopped living with (HAME)? I. Yes , (ASK And what is (NAME OF SPOUSE/PARTNER) "a present street address? I O-P) 15-39/ No D. (SKIP TO r) 2 STREET ADDRESS ' ' ~ How did thia (marriage/relationship) end? Separation..................I HAND CARD G to/ 3, And what city, state and zip code does (SPOUSE/PARTNER) live in? |_ I_ I STATE Divorce.....................2 40-59/ Death of spouse or partner (ASK «-o, TB» SKIP TO B.71 3 1_ I_ I_ I_ I _ ZIP CODE 62-66/ �OWK 21-iJ -34- 6. DECKS 21-24 (Continued) 7. K. And what l« (NAME Or SPOUSE/PARTNER) *s present telephone IMTEKVlEWEKi r? IS THERE A THIRD RELATIONSHIP SINCE THE DATE LAST INTERVIEW? lie Hunan or TIMES UCOHDCD n O.JA BOOAI. TO I OH HORE7I LI—I—I AREA COOS LJ 67-7*/ I NUMBER IES .....(CO TO g.8» HO PHONE. L. .1 I n / HO.....(SKIP TO 0.10) Thinking of all the people you know, ulther around here or elsewhere, who would be th« on* person who would be mom*, llfcsly to know where (NAME Or SPOUSE/PARTNER) Is? ENTER fUU. NAME OT PERSON BELOW AND ASK H-P. BEGIN DECK 22 *. Thinking of the n«xt relationship since (DATE Or LAST 1MTERV1EM), did you starry this person? Yes. (LAST). Ho.. 30-44/ (MIDDLE) M. *. Hhat Is the current full oaM of this per person} What Is (PERSON'S) relationship to (HAHC OT SPOUSE/PARTNER)} 7S-76/ ..a .rn H. Cd I Mhere does (PERSON) Ural INTERVIEHERi (MIDDLE) RECX)RD rOLL HAHE Of (SPOUSE/FARTHER) ON IHFORMATIOH SHEET. ITEM O7 AMD ID I ABOVE. 47-7»/ M»t IMS her full svldsn naa*7 KPTl STREET ADDRESS BEGIN DECK 24 •COIN DECK 33 CITY Hhat was her blrthdate} IO-2»/ RECORD DATEl | | |__l_| HO O. DA I 3O-3S/ IR What Is (PERSON'S) telephone nwber? B. IAREAICODE LJ l__l_l__l - I L_L_LJ PHONE HUHBER HO PHONE P. I 32-3C/ 30-3I/ IF (PERSON) HAS PtlOHEi ENTER MONTH AND »EA»« I 1 OTHER (SPECirt MELON) I DON'T KHOf (SKIP TO Q.7) | | MO In whoee nan* Is the phone listed} (PERSON'S) MM....(SKIP TO P-7) • 47/ C. | | | J6-3S/ n Have you stopped living with (NAME)} Yes Ho 4B-72/ (LAST). In what aonths and year did you (get Married to/»tart living with). <MHE)} (ASK D-K....I (SKIP TO D . . . 2 40/ �DKCKS 24-2S -15- < Coun 11 ntied) O. B. How did this (.tarrlaae/relatloMhlp) end? I 4I/ I I I I AREA O»E I I_J I - I NUMBER I I I_J «-4«/ HO PHONE Death of spouse or partner •AM m-m. no L. •Kir to g-») 1 In what «ionth and year did title occur? ENTER HOHTH MO TEAR I LJ_LJ__I HO T. And xhal is (HAHR OF SPOIIsr./PARTNeR) *a present tvlnphone nuohnr? Divorce**......*...*.******.*2 G f. (Oantlniind) r. Separation HAND CARD DECKS «-27 Thinking of all the people you know, either around here or elnewher<s, who would be the one peraon wno would be neat likely to know where (HAHC OF SPOUSE/PARTNERS la? ENTER FULL MAKE OF PERSON BEUW AND ASK H-P. «*-«*/ 47-7I/ (LAST*, m During thle (»arrlage/relattonet>tp). how siany tinea were you living apart f t am (NAME) (or 1 Matha or «wre a luce (DATE or LAST (MIDDLE) (FIRST) INTERV1EM)? ENTER NUMBER OP TIMES. | | | 4S-47/ H. I OH Hone (GO TO Nhat t> (PERSON'S) relationship to (NAHE Or SPOUSE/PARTNER)? ft.9) I I I ! I I I I I I I I I I BEGIN DECK 26 OO H. Mhere does (PERSON) live? 00 ts} O G. IO-J4/ For hov nany souths did you live apart the ttlrat/nemtl tine? STREET ADDRESS First tlM | | | 4B-49/ Second tlMi | | | 50-5I/ I I S2-SV L_l_J *«-**/ third tie* fourth tine H. ....| During thle (•arrteae/relatlof.ehlp). etnce the (DATE or LAST INTERVIEW}, did you ever have a problen conceiving a child becauee of prolonged eeparationT Tea Ho I. I L_LJ STATE 3S-44/ O. L_L_LJLJLJ ZIP CODE SS-56/ Mhat Is (PERSON'S) telephone number? LJ_LJ AREA CODE LJL-LJ -NUMBER L_LJLJ_J PHONE HO PHONE P. 5?-tt/ IP (PERSON) HAS PHOHEi 2 S6/ In whoss naiM ia 62-7(/ I the phone lilted? (PERSON'S) »«•«....(SKIP TO 0.9) OTHER And what la (HAKE OP SPOUSE/FARTHER)*e pceaent street eddreea7 I (SPEC I FT BELOM) 2 (SKIP TO 0.9) H DON'T KNOW 72/ S7-8O/ BEGIN DECK 27 STREET ADDRESS •EGIH DECK 25 J. (LAST). 10-J4/ And vhat city, state end sip code does (SrOUSE/PARTHER) live In? L_LJ STATE C1TT IO-29/ 3O-1I/ (FIRST) SIP CODE 12-I6/ (MIDDLE) �-J7- IUX'ICS 2 J - 2 K 9. INTERVIEWER I ~ IS THERE A FOURTH RF.I.ATIONSHIP SINCE TIIE DATE LAST INTERVIEW? (IS MMBBIt Of TINES M9CORDED IH Q. 3A EQUAL TO 4 Oil HDRB?) Y E S . . . . ( G O TO HEM QUESTIONNAIRE) 1 NO II. INTERVIEWS; ASK THIS p"F.STION FOR EACH CHILD LISTED OM CRILmUM'S RECORD FORM FOR WHOM THERE IS NO DEATH DATE. What is (NAME OF 1ST CHILD/NAME OF 2ND CHILD. ET<:.»'a current »qe? RECORD OH CHILDREN'S RECORD FORM. 35/ 2 IF DECEASED SINCE LAST INTERVIEW. ASK A-C. OTHERS CO TO O. 12. A. When did (CHILD) die? RECORD FORM. RECORD DAT, MONTH. AND YEAR ON CHILORJ9TS B. What was the cauae of death? C. Where la (CHILD)'a death registered? VERIFICATION OP BIOLOGICAL CHILDREN USING caiLMors RECORD row 10. INTERVlEWERi ARE CHILDREN LISTED OH CBILDMM'S RECORD FORM? CHILD ID ' I YES (ASK A) 1 NO (ASK Bt 2 36/ I I 39-457 RECORD BELOW. CHILD IDI I BEGIN DECK 28 In what city and atate? | | 60^-617 CHILD ID, I I I io~-TT7 CAUSEi RBGISTRATIOHt A. W I'd like to read information about your (child/children) froa our last interview lo check our records. Aa of {DATE or LAST INTERVIEW), our record* show that you have had (NUMBER OP CHILDREN). . .(READ EACH CHILD'S PULL NAMJE, SEX. AND BIRTHDATE AND MOTHER'S NAME). la that correct} IF INFORMATION IS CORRECT (GOTOQ.11> 1 IF INFORMATION It INCORRECT, HAKE CORRECTIONS ON CHILDREN'S KBCORD FORM (THEN GO TO Q.I1) (CITT) IF INFORMATION IS INCORRECT. ASK FOR (CHILD/CHILDREN) *S FULL NAME, SEX, BIRTHDATE AND MOTHER'S HAHP.. RECORD BEGINNING AT LINE 01 OM CBIUHIBt'8 ....(THEN GO TO 0.11) 63-7B/ (CITI) IJ-28/ SB-59/ (STATE) 79-8O/ (STATE) 29-JO/ INTEBVlrWEHi HAS R BEEN HARRIED OR HAD A PARTNER FOR J MONTHS OR MORE SINCE (DATE OF LAST INTERVIEW)? 1CS HO Our recorda show that you had not had any children of your of (DATE OP LAST INTERview). la that correct? IF INFORMATION IS CORRECT (GO TO 0.12) 12. (CITt) (STATE) 2 3V 4Z-S7/ 1 A. Ho B. (SKIP TO SECTION 6) SI/ 2 Haa/Have (your wife/any of your partners) bee since (DATE OF LAST INTERVIEW)? lea 3B/ , (ASK A) pregnant by you (ASK B . . ).. 32/ (SKIP TO p.25). How K«ny preqnancies (has your wife/have your partners) had with you since (DATE OF LAST INTERVIEW)? ENTER NUMBER OF PREGNANCIES! (CO TO 0.11) I I I 33-J4/ �DECKS 2B-29 13. Mien did ( Ihat/lhr f i r s t , etc.) pregnancy begin? RHTER MONTH AND YEAR. A. IHTKRVIEMKR; What Month and year? |_ |_ |_ |_ | HO >H II. F.. 3S-38/ HAS R HAD MORE THAN ONE RELATIONSHIP SINCE DATE OF LAST INTER VI EX? (SEE INFORMATION SHEET. ITEMS O5, O6 AND O7) (Continued) Please look at this card and tell *e all the lumbers of the ty|>es of birth control you or (NAME) were using when she hecane pregnant, cone ALL THAT APPLY. I (GO TO C) ......... 2 Jelly, cream, suppository. 39/ Hhich (spouse/partner) had thi» pregnancy? RECORD 40-64/ (LAST) 74- 75/ 76- 7 7/ 78- 79/ O4 BBC IN DECK 29 IO-I1/ S. 6. 7. 8. 9. 10. II. Condom Diaphr Diaphr RhythM Rhyth* Hi thdr 12. HO 8. .......... Douche 4. H (ASK B) 7J-73/ 1. 2. HAND CARD Other (SPECIFY) 1 2- 1 3/ I4-I5/ 16-I7/ lfl-19/ 20-21/ 22-23/ oa 24- 25/ 12 TFIRST) (MIDDLE) 26- 27/ DON'T INTERVIEWER. RECORD ID I FROM INFORMATION SHEET, ITEM OS, O6, OR 07 6S-66/ L_J_I 00 C. N) N) How many Months did it pregnant (this tine)? RECORD HONTHS AND/OR YEARS | | HOS | AND/OH | Hain't trying | YRS | Still pregnant ..... (SKIP TO 0.25) OO 69-70/ A. «a Hera either you or (NAME or SPOUSE/PARTNER) oalnq birth control at the tlM ahe becaa* pregnant? tea...........(ASK E> COO TO Q. 14) 1 28/ live birth... ......... (ASK A-J) ............. I Miscarriage ........ (SKIP TO p. 16) ........... 2 Stillbirth ....... ..(SKIP TO 0.16)...........3 Abortion...........(SKIP TO Q.I6)........ . 4 .. 67-ea/ OH Ho Did that pregnancy result in a live birth, or in a Miscarriage, stillbirth, or abortion, lor is (NAME) still pregnant)? take (NAME OF SPOUSE/PARTNER) to becoM DON'T KNOW. D. 14. ........... S What is the f i r s t and last naM of the child as it appears on the birth certificate? RECORD OH CBIU>KE*)'S RBCOW) PORH OH MirrUMEMTART CHILDREN'S RECORD POWf. IRTERVIEHERl RECORD ID FROM OHLDRBI'SftBOOHDPORH |_ |_ | 29- JO/ When Has (CHILD) born? ENTER BIRTHDATE OH CHIUHIEH'S RECORD PORN Oft MPPLEHKNTART CHILDRBI'S MKCOPD FORM. 71/ 2 C. Was (CHILD) nale or fenale? RECORD OH CnUDRKH'S HECORD PORH OR SOPPLEMBTTMT CHILDREN'S RECORD POXH. How Much did (CHILD) weigh at birth? ENTER POUHDSt AND OUNCES! OR Don't know | | j 3I-32/ | | | 33-34/ 98 �14. 14. (Conlinued) E. (Conlinup.1l K. When did (CHILD) die? RECORD DAY. HOMTH. AM) YEAR OH CHILDREN'S RI9CORD POKH OR SUPPLEMENTARY CHILDREN'S RECORD PORH. L. Has (CHILD) a twin? What was the cause of death? Ye*. Mo.. RECORD BELOW. 41/ F. Has (CHILD) preMture, full te», or overdue? PreMture Pull tera Overdue Don't know t 2 1 ..8 J6/ N. Where is (CHILD'S) death registered? In what city and state? (CITY) 42-S7/ G. How old (STATE) SB-59/ was (MANE OF MOTHER) when (CHILD) was born? 15. RECORD AGEi |___L_J Don't know IHTERVlrWERi 37-J8/ IS THERE A SECOND PREGNANCY SINCE THE DATE OF LAST INTERVIEW (IS NUHBER OF PREGNANCIES IN 0.1 2B EQUAL TO 2 OR HORE? I 98 YES H. 16. DOCTOR'S HUME -BTT I NO What is the MM and address of the doctor or Medical facility who has (CHILD) •• birth registration records? RECORD BELOM (SKIP TO 0.17) (SKIP TO 0.25) 60/ 2 When did that pregnancy end? FACILITY HAKE RECORD DATEl to | | | | HO | | DA | 6I-66/ YR STREET ADDRESS A. J_L_I IMTERVIEMERi I. ENTER NUMBER OF WEEKS I (STATE) (CITY) Don't know RECORD NAHE MD ADDRESS OH MEDICAL CONSEHT rORH | | | 67-68/ 9g IF MISCARRIAGE OR STILLBIRTH, ASK B-C| OTHERS GO TO D. what is the nave and address of the doctor or.Mdlcal facility who has (CHILD) •• current awdlcal records? RECORD BELOW. B. DOCTOR'S HAHE How Mny weeks had (NAHE) been pregnant when that happened? FACILITY NAME Did a doctor tell you why this (olscarriaqe/stillhirth) .iqht have occurred? Yes (ASK C) t No STREET ADDRESS (GO TO D) J 69/ JLJLJI (CITY) IHTERVIEHERi (STATE) C. What did the doctor say caused the (Miscarriage/stillbirth)? RECORD VERBATIM. RECORD HAHB AMD ADDRESS OH MEDICAL CONSENT FORM 70/ J. Hhat is (CHILD)'a current aoe? RECORD IH CHILDREN'S RECORD FORM OR SOPrLEHENTARY CHIUHUOI'I RBOOfD tORN. IP DECEASED. ASK K-H. OTHERS GO TO Q. 15. �DECKS 24-3(1 16. (Continued) D. 17. INTERVIEWERl (Continued) IS THERE A SECOND PREGNANCY SINCE DATE OF LAST INTERVIEW} (IS NUMBER OT PREGNANCIES IN Q. »2B EQUAL TO 3 OK MORE?) YES (GO TO 0.17) 1 NO 17. OF.CK ] 0 (SKIP TO Q.2M 2 When did the next pregnancy begin? types of birth control you or (NAME) were using when she became pregnant. CODE AU. THAT APPLY. 7I/ 1. 2. 5. 4. S. 6. 7. HAND CARD H What Month «nd year? 42-437 44-457 46-47/ Douche Jelly, 72-7S/ (_JH_I I HO YR e. 9. A. INTERVIEWERl HAS K HAD MORE THAN ONE RELATIONSHIP SINCE DATE OF LAST INTERVIEW? (SEE INTO SHEET, ITEMS OT. O* AND Oil YES (ASK B) I (GO TO C l . . . NO.. Which (apouse/partner) had this pregnancy? 18« Did that pregnancy result In • live birth) or In a piscarriage, stillbirth, or abortion, (or is (NAME) still pregnant)} Live birth Hiscarrlage Stillbirth Abortion... Still pregnant rvJ (MIDDLE) INTERVIEWER) RECORD ID I FROM INFORMATION SHEET. ITEM. OS. O6, OX O7 A. L_L_I RECORD HOHTHS AND/OR YEARSt 35-36/ take (NAME OT SPOUSE/PARTNER) to bscom | | MOS | AND/OR I Hasn't trying OR DON'T NKM D. I I YRS I 2 3 4 5 687 What Is the first and last naa* of the child as It appears on the birth certificate? RECORD ON CHILDREN'S RECORD FORM OR SUPPLEMENTARY CHILDREN'S RBCORD PORN. INTERVIEWERi RECORD ID FROM CHILDREN'S RECORD FORM OR SOTPLEMIMTARY CHILDREN'S RBCORD FORM. | | | 69-70/ B. Mien vas (CHILD) born? ENTER BIRTHOATE ON CHILDREN'S RBCORD PORN OH SUPPLEMENTARY CHILDREN'S RBCORD PORN. C. Has (CHILD) Bale or fenale? RECORD ON CULDRnTS RECORD PORN OR BOrPLEMBITARY CHILDREN'S RBCORD FORM. D. How mich did (CHILD) weigh at Mrth? 39-40/ 98 were either you or (NAME OP SPOUSE/PARTNER) using birth control at the tlaie ah* became pregnant? No (ASK A - J ) . . . (SKIP TO Q.2O) (SKIP TO p.20) (SKIP TO O.2O) (SKIP TO O. 25) 37-3S/ OO Yee 64-6S/ 66-67/ 12 IO-34/. How Many snaths did it pregnant (thla tljw>? <tg 58- 59/ 60-61/ 62-63/ 2 RECORD NAHEt C. O9 S2-53/ 54-557 S6-57/ 07 It/ BEGIN DECK 3O B. 10. tl. 11. Condos) Diaphr Diaphr Rhythm Rhythai Nlthdr awal Other (SPECIFY) DON'T ENTER MONTH AND YEARi 48- 49/ 50-5 1/ OS (ASK E) (SKIP TO p. 18) 1 2 41/ ENTER POUNDS I AND OUNCES I OR Don't know | _ |_ | 71-727 73-747 98 �BEGIN DECK IB. (Continued) E. 18. Mas (CHILD) a twin? 31 (Continued) K. When did (CHILD) (lie? RECORD DAY, MONTH. AND YEAR ON CHILDREN'S RECORD FORM OR SUPPLEMENTARY CHILDREN'S RECORD FORM. L. Hhal was the cause of death? 7S/ RECORD BELOW. 10/ F. Has (CHILD) preMature, full tern, or overdue? PreMature. Full terM Overdue Don't know G. How old I 2 3 8 76/ | Don't knov | 19. INTERVIEHERl | I1-26/ I I (STATE) 27-2B/ IS THERE A THIRD PRPGNANCY SINCE THE DATE OF LAST INTERVIEW? (IS NUMBER OF PREGNANCIES IN O. 128 EQUAL TO 3 OR MORE?) YES 20. (SKIP TO p. 21) I NO What t» the tame and address of the doctor or Medical facility oho ha» (CHILD)*• birth registration records? RECORD MOON NJ (Jl In what city and state? I 98 DOCTOR'S MAHE Where is (CHILD'S) death registered? (CITY) was (NAME OF (WINER) when (CHILD) wa» born? RECORD AGE I H. (SKIP TO O.25) 29/ 2 Hhen did that pregnancy end? FACILITY NAME RECORD DATEi ' * ' MO ' • DA ' I JO-JS/ YR STREET ADDRESS J (CITT) INTERVIEHERl I. I I A. (STATE) F.NTER NUMBER OF WEEKS: RECORD MMUt AMD ADDRESS OH HEDICAL CONSENT FORH Don't know Hhat is the msM and address of the doctor or Medical facility who haa (CHILD) 'a currant Medical records? RECORD BELCH OR | | | 16-J7/ 98 IF MISCARRIAGE OR STILLBIRTH, ASK B-C| OTHERS GO TO D. B. DOCTOR'S NAME How many weeks had (NAME) been pregnant when that happened? FACILITY NAME Did a doctor tell why this (Miscarriage/stillbirth) Might have occurred? BO/ Yes (ASK C) 1 No (GO TO D) 2 STREET ADORES* (CITT) INTERVIEHERl J. L_I_J (STATE) RECORD NAME AND ADDRESS ON HEDICAL CONSENT FORH Hhat is (CHILD) "8 current age? RECORD IH CnUMBN'B RECORD FORM OR SUFFLCMEHTARf CBILDRBT* RBOORD FORH. IF DECEASED. ARK K-H. OTHERS GO TO 0.19. C. Hhat did the doctor say caused the (Miscarriage/stillbirth)? RECORD VERBATIM. IS/ �HECIN UEX'K 32 2O. 21. (Continued) INTERVIEWER: YES 21. IGO TO 0.21) I (SKIP TO 0.2S) 2. 3 4 5. 6 . 7 8 9 HAND Mhat Month and year? ENTER MONTH AND YEARI INTERVIEWER; 4O/ Please look at thla card and tell M all the numbers of the types of birth control you or (NAME) were tialng when she becaae pregnant. CODE ALL THAT APPLY. 2 When diet the next pregnancy begin? A. E. IS THERE A THIRD PREGNANCY SI NOB DATE OF LAST INTERVIEW? (IS HUHBKR OF PREGNANCIES IN p. 12B EQUAL TO 3?) NO (Continued) | | || | HO VR 4J-44/ to. I NO 12. 12 22. RECORD NAHEt Live birth 46-7O/ A. INTERVIEMERt RECORD ID « FROM INFORMATION SHEET, ITEM OS, O6, OR O7 L_L_J 71-72/ How nany Months did It tak* (NAME OF SPOUSE/PARTNER) to becoM pregnant (this tlrM)7 AND.OR 7S-74/ .00 OR 75-76/ Tea (ASK E) No...(SKIP TO p.22) I 2 77/ 4 5 (SKIP TO Q.24) (SKIP TO O.2S) RECORD ID FROH CUJUMIEH'S RJX3ORD FORM OR SUPPLBHENTUII OHLDRJM'l |UKXNU> FORM. |_ | 37-3B/ B. When %>aa (CHILD) born? C. Nat (CHILD) Bale or fe»ale? RECORD ON CHILDREN'S RBCORD FORM OR SOmaMENTARI CHILDREN'S MCORD FORM. D. flow such did (CHILD) weigh at birth? .8 9 Were either you or (NAME OF SPOUSE/PARTNEK) using birth control at the tine >he becaiw pregnant? 36/ 2 3 What is the f i r s t and last naa* of the child as it appears on the birth certificate? RECORD ON CHILDREN'S RBCORD FORM OR SOPFLKMEHTAMY CHILDREN'S RBCORD FORM. INTER VIEWER i D. t Abortion Still pregnant (MIDDLE) DON'T KNOW.... (ASK A-J) Miscarriage (SKIP TO O..24) Stillbirth.........(SKIP TO O.24) rv> wasn't trying. 32-3J/ 34-3S/ Did that pregnancy result in a live birth) or in a Biscarrlage, stillbirth, or abortion, lor ia (NAME) still pregnant!? iLAST) RECORD MONTHS AND/OR TEARSt Other (SPECIFY) 45/ Which (spouse/partner) had this pregnancy? C. 12-S3/ 14- IS/ 16-17/ 18-19/ 20-21/ 22-Z3/ 24-2S/ 26-27 / 20-29 / 3O- 3 1 / 2 (ASK Cl (FIRST) 1O 11, HAS R HAD MORE THAN ONE RELATIONSHIP SINCE DATE OF LAST INTERVIEW? (SEB IHPO SHEET, ITEMS OS, O6 AND O7) IBS ,O4 OS ENTER 8IRTHDATE ON CHILDREN'S RECORD FORM OR (OPPLaMBirTART CBUDRKN'8 RECORD FORM. ENTER POUNDS I | | | 39-40/ | | | 41-42/ AND OUNCES, OR Don't k n o w . . . . . . . . . . . . 9 8 �22. (Continued) 22. E. Was (CHILD) a twin? L. Where is (CHILD'S) death registered? How old INTERVIEUERt " Don't know | | J 90 (CO TO HEH QUESTIONNAIRE) NO (SKIP TO 0.2i) 1 68/ 2 Mhen did that preqnancy end? RECORD DATE, DOCTOR'S NAME 66-67X IS THERE A FOURTH PREGHAHCY SINCE THE DATE Of LAST INTERVIEW? (IS BUMPER OF PREGNANCIES IN O.I 26 EpUAL TO 4 Oil MORE?) »ES 24. I 5O-65/ 45-46/ Hhat is the name and address of the doctor or medical facility who has (CHILD!*s birth registration records? RECORD MELON I (STATE) (CITY) was (NAME OF MOTHER) When (CHILD) was born? RECORD AGEi T In what c i t y and slate? I I 2 3 8 23 H. RECORD BELCH. Has (CHILD) premature, full term, or overdue? Premature roll term Overdue Don't kno« G. What was Ihe cause of dealh? M. Yes. No.. F. (Continued) I L_LJ_LJ ' HO DA YR 69-74/ FACILITY HAME 47Y H. How many weeks had (HAMFJ been pregnant when that happened? STREET ADDRESS ENTER NUMBER OF WEEKS. JLJLJ (STATE) (CITY) INTERVHMERI I. Don't know | 7b-76/ 98 B. Did • doctor tell why this miscarriage mitlht have occurred? ' DOCTOR'S NAME 1 IF MISCARRIAGE. ASK B-C| OTHERS GO TO D. RECORD HAKE AMD ADDRESS OH HEDICAL COHSEHT FDRH Mhat is the name end address of the doctor or medical facility who has (CHILD) <s current medical records? RECORD BELCH | Yes (ASK C> » (GO TO D) 2 "/ FACILITY MAKE Ho STREET ADDRESS C. (CITY I INTERVIEHERi What did the doctor say caused the miscarriage? RECORD VERBATIM. to/ (STATE) RECORD NAME MO ADDRESS OH MEDICAL CON8EMT FORM J. Hhat Is (CHILD)'s current aq«7 RECORD IM CBOOKEM'* RBCORD FORM OH SUPPLEMENTARY CTILORJ9VB RBCCMD nMM. IF DECEASED, ASK r-H. OTHERS 00 TO p. 23. *.. When did (CHILD) die? RECORD DAY. MONTH, ANT) YEAR OH CHILDREN'S D. INTERVIEWER: ' IS THERE A FOURTH PREGNANCY SINCE DATE OF LAST INTERVIEW? (IS NUMBER OF PREGNANCIES IN 0.12B FOUAI. TO 4 OR MORE?) jES NO (GO TO NEW QUESTIONNAIRE) .. .I 2 79/ �BEGIN DECK 2S. )} since (DATE OF LAST INTERVIEW) have you ever tried for a period of one year or More, to conceive m child and were not able to do eo? Yea 32. INTERVIEWER- IS THERE A SECOND PERIOD OF INFERTILITY SINCE DATE OF LAST INTERVIEW? (IS O.Z6 CODED "THO" OR HORE7) IO/ No YES (SKIP TO SECTION 6>. BO 26. (SHIP TO SECTION 6). For how many periods of one year or Bore did this happen? 33. One... Two... Three. • I/ Since (DATE OF LAST INTERVIEW), in what aonth and year did the aecond period begin? And in what annth and year did It end? Begin Since (DATE OF LAST INTERVIEW), in what exmth and year did the flrat period begin? And in what Month and year did it and? Begin 12-IS/ LJL_L_LJ MO 28. CO End L_LJ_LJ YR MO 34. I6-19/ ««•«»> U_I_U MO YR During thla aecond period what waa your wife or partner'a flrat naa«? RECORD «LON. <a 6</ a*"0 '-WT-' YR 3S. During (PERIOD) what waa your wife or partner*a Mrat naa«7 RECORD 8ELCM. Mow old waa (MANE) in (BEGINNING DATE OF PERIOD)? JO-33/ 34-3S/ M CO RECORD AGEi M. 29. How old waa (NAME) In (BEGIMHING DATE OF PKRIODI7 RECORD KG* 30. ( | | 37. 64-6S/ 66/ CODE 'PERIOD 2" AND HAND SELF-ADMINISTERED FORM 1. There are many reaaona that eone couple* find it difficult or l«r>oa*lble to conceive a child. Pleaae read thia card and circle the nunber on Side A for each reaaon which applied to you for thla period. Side B provide* reaaona appropriate for your epouee. Circle aa Mny reaponaea aa appropriate. CODE •PERIOD I" AND HAND SBLr-AOHIMISTERED FOMi 1. Nov pleaae f i l l out thla card and place it in Ute envelope when you are finlahed. | Ho.. During (PERIOD), did either of you aea a doctor to dlacuaa any difficultly in conceiving children? There are awny reaaona that eo«e couplea find it difficult or inpoaalble to conceive a child, pleaae read thla card and circle the nuMber on side A tor each reaaon which applied to you for thla period, side B provldea reaaona appropriate for your apouae. circle aa eumy reaponaea aa appropriate. | Yea. No.. 31. | During tbta aecond period dl'd either of you aee a doctor to dlacuaa any difficulties In conceiving children? 36-J7/ Yea. 44-47X End 40-41/ L-LJU-J HO YR. 27. 19 / (GO 10 0 . 3 ) ) . . . . Now pleaae fill out thla card and place it in the envelope when you are finlahed. 38. INTERVIEW HI. IS THERE A THIRD PERIOD OF INFERTILITY SINCE DATE OF LAST INTERVIEW? (IS Q.26 CODED "THREE"?) YES NO (<X> TO Q.19) (SHIP TO SECTION 6) 1 2 67/ �DECKS 33-34 39. Begin 68-71/ End OR HAS NOT L_I_LJ_J HO YR U_l_l_l MO YR 4O. 0000 During (PERIOD) what waa your wife or partner 'a flrat MM? '-A- I L-LJLJLJLJL | | | INTERVIEWER; 1O-23/ 24-2S/ HAS RESPONDENT HAD ANY BIOLOGICAL CHILDREN? YES NO (ASK A) 1 (SKIP TO 0.22) 2 A. 26-I1/ ME CHILDREN RECORDED OH CHILDREN'S RECORD FORM? • YES 42. (ASK B) I NO (SKIP TO 0.2) 2 30/ During (PERIOD), did either of you see a doctor to dlacuea any dlfflcultlea In conceiving children? B. Tea. FOR EACH CHILD LISTED ON CHILDREN'S RECORD FORM ASK: Our records Indicate that (CHILD)(had/did not have) a birth defect in 1982, when you were laat interviewed. Is this information correct? No.. IF INFORMATION IS CORRECT..(GO TO O.2). ro vo 43. 2t First, I would like to verify whether (any of) your biological child(ren) have |had| a birth defect. HOH old waa (NAME) In (BEGINNING DAT* OF PERIOD)? RECORD AGBl Child and Family Health How I would like to ask you some questions about birth defects in your family. By birth defects I mean a physical abnormality present (though not necessarily noticed) at the time of birth. Birth defects ranqe in severity from unusual birthmarks to a missing or mishappen limb. Birth defects can affect any part of the body, including bones, body organs such as kidneys or the heart, reproductive and respiratory systems, blood, and the skin. I. RECORD BELOH. 41. Section 6: Since (DATE OF LAST INTERVIEW), In what Month and year did the next) period begin? And In what Booth and year did it end? .1 IF INFORMATION IS INCORRECT, MAKE CORRECTIONS OH CHILDREN'S RECORD FORM (Then go to O.2) There are Mny reaaona that some conplca find It difficult or impossible to conceive • child. Pleaae read thla card and circle the number on Side A for each reaeon which applied to you for thla period. Side B provide* reaeone appropriate foe your apouae. Circle a* Many reeponaea aa appropriate. 2 Don* t know. CODE 'PERIOD 3* AND HAND SELF-ADMINISTERED FORM 1. B .(GO TO Q.2) nut EACH CHILD (EXCEPT CHILDREN WHO DIED BEFORE 1982) ASKi Has a(n additional) defect been identified in (CHILD)(since 1982)? RBCORD OH CHILDREN'S RECORD FORM OK SUPPLEMENTARY CHILDREN'S RECORD FORM. Now pleaaa fill out this card and place It In the envelope when you are finished. 3. INTERVIEWERi HAVE DIED. ASK QUESTIONS 4-2O FOR EACH CHILD, INCLUDING ALL WHO MAY (Now I would like to ask about (NEXT CHILD)). �-56- orCi442BSec-6 2«> CHIIO 1ST CHI 10 CHIIO'S WNfi CHIIO'S lot HOIHCR'S 10* 4. | | 1 32- 5 V J4-3V OM (CHIIOI «v«r k*v« • r»> 49- 50/ SI-52/ No »•» No ..i .. 2 m ..1 .. 2 M/ «...! 2 5V CATftOMtl... pky»lc*l or Kitor k»p«tr»«iit.l 3. 2 3>/ MM (CHIlOt »»«r <l»gno<«d ..(A PLEASE GO OH TO NEXT PAGE &B . I I. M/ ...(0.6I...2 40-4V •. «tat Ho< of Don't kno. DtFCCI m WPAIIHEHT KIN iKNTirifo m iCHiioit CHtCK CHI LOOT'S MOOR) ram «MD QS. t,i, wo 4. ». IMIHWIiKilli IS THfRE WIOfHdl CHIIOI ns. .(».?) NO.. TIS ....... ASK 0.4 MO » I 2 ....... (SKIT ID 9.221 46/ i ..1*1 imTOo.j». NO ?. R «;/ 2 .« AM> 3)..I 10.22 2 i «/ («.»» I 2 <0.ll 2 o« tblrth 4*l«ct<l>/_ 4I/ ICHIIOI 014 you (or MMVOA* •!»•> *UcM» ICHItD'I) Cklrth «*t*ct<»l/l»*«lriw«tl «ltk • doctor* r.» No (so ID «.«> <SXM> TOQ.IIt 6V �-57- DFX-KS 34-35 CHHD'S WW(: 3RD CHltD 4TH CHILD J 66-6T/ ' 6B-69/ No 2 70/ »«S I 1 2 2 I4-I3/ 16- 1?/ t.» I Ho 2 IB/ I 72/ I 1 6TH CHILD 1 3I-32/ 1 33-34, 1 | 1 | CHUO'S ID « MOTHER'S Ip I 4B-4./ 30-3I/ ». 2 20/ N> ».« Y« No 7I/ 1 . 3TH CHILD ..LI J_L I L t JL.L IMTER»IE.ntR! IF DCriCIISI. DISABILITY. OR HPAIFMEN1 FOUND SINCE IW2 ASK 0.»; OTHERS 00 TO 9.10. Klxt It tko ••«• mkd nddrot* of tk* voctor «fco nlM|iK»«d CHILD •• ko»l»n * tlrtk «of oet/lBnalrmMit t DOCTOR'S HME DOCTOR'S NME CD OR FACILITY MM! FACIIITT NME STREET ADDRESS ....(A I B)....l 7V ....(A I I . . B..I 2I/ ....(A t B)....l 3B/ STDfET AOORESS ....(A 1 B)....l 3V .....<9.A>.....2 10. 74- 7 7/ 22- 2V L_l__l_l_l HO m HD TR MO MD TR DM tko doctor >*y th.l (CHILD) nootf(l/o«) My toitlnn, •odlentloo, trootnonf. t«ro*ryv or 3*-3*/ 3V-42/ Yd koenno of • Iblrtli ••loct/leoolnmtlt (By 2*/ Don't mo-.....» Don't kM».....B 60/ 4V • wlwolctiolr, ••Ikor. •rtltlcUl Itak, body lt or crwtckos). Don't kno.....A Don't kno».....B II. BESIM DEOK 33 (9.7) 1 IQ/ ....(ASK A). ...2 (9.> * »»/ (9.7 1 44/ ....(ASK AI....2 ....(ASK AI....2 9.7 1 «!/ ....(ASK AJ....2 ..19.4 AND 3 . I I. 4V ..19.4 AND 31. . 1 DU (CHILD) *vnr r*e«l«o •»y t*ttln||, M4lcotlon. tr**l*on*B ••rnory or •ooelnl onvlniw of • (klrlk d«f< Don't lwo> 12. ..(9.4 AND 3>..t ll/ ..(9.4 AMD 31.. 1 2B/ Don't Knov.................B *2/ • •••(9.221.... .2 At »y tl<M. did (CHILD'S) (klrtk «ofoet(i>/loo«lrMiit> Intorloro In pay «ny «ltk ICMILO'*) ckyilol or KKUI >o»*lo*M«tf For OKMinIo, I Yot (GO TO 9.13) No I ASK A) I 6»/ 2 .(O.III .(ASK A) , Jo. or -HI*, tv I2/ 46/ Don't kno> •V A. IMTEI«lt»€Rl DM TMtM * TtS COO60 AT g.10 or t (00 TO 0.1» ...... I »•/ .(0.13) O.I It (9.*l (O.I It 1 2 IV (».«) 1 (9.IH...J 30/ I9.*> 1 47/ 1 . 1 ... 01 1 . 2 I9.*> 19.11 2 «....( ASK Bl.. ...... .2 1 ««/ •. l»I[Xlt«t»l IS THEHC ANOTHEH CHILD1 TES....IOO BACK TO O.4 FOR NEXT CHILD! . . I .. T00.« 1 .IASK «> .(0.4) .(0.221 »!/ �lit* t :. JHD CHILD 4IH CHILD » • • - lit 61 H CHILD MH CHILD 2ND CHILD ISf CHIID H H I 1 DOCTOR'S NAME OR DOCTOR'S NAME ?)/ 1 1 1 IJ/ CHIID'S NMC: CHILD'S ID 1 NDIICR'S ID 1 OR FACIlIt NAHE 1 1 DOCTOR'S NAHE OR DOCTOR'S NAHE OR FACILITY HAMC I It/ FACILITY NAME 25/ 1. 1 FACILITY NAME OH (CHIID'SI doctor lay ttot U«r o(l (CHIID'S) (klrtk «*r«ct(«)/liv«lr««iiti ««t/«w* llf»-tlir««t«nlfig If Utt ••trfltftt m, It/ )J/ SIREET M»RESS 1 1 STATE CITY STREET ADDRESS 1 1 CU< 1 1 STATE SIREET ADDRESS 1 CUT 1 1 STATE 1 CITY 1 MtrMl<4 1 mtrnn II (CHILD) iH Ml r.c.lv. xiroM-r. mftlettlan, » >p«cl«l «l.t. or >o»« o«hw Mdlcil Inter rwvtloo.) 1 StATl A. INTEMIEICRi (ASK A) 1 Ho (ASK A) J (ASK M.....2 Don't kno»...IA9( Al * (ASK A>.....« !»/ (ASK A) IS CHILD WOER DIE MFORC HE/SHE MS I HO fEAKS OlOt 14. t.t JZ/ STREET ADDRESS OK (CHIIO) •«•!• >••< k»lp • Itk Mtlng. «r»»lng. k»tkl«t> or Mlxg tk« tolut HO J 3V )9/ )4/ <0/ JV «!/ J6/ 42/ l*V*lr«*attMH»l» locltido w tsJ rtlkM- tkM >i«t itxxtl.g kr t» «»llt l« KM««d.l .2 ... ..J IS, ••CWM ol • (klrtk tf«Uct/ l«V*lr«mtlc 4H (CHILOI «»«r K6IH DECK it ««• or iio*4 *ny ••chMtlctl •hMlcMlr. nllwr. bo«y (ASK A) 2 (ASK Al J cr«tck»* to corry out •<twf <«r ecti'iti.tt 1. * (9.l» 1 M/ (9.111 * 1 " (9.111 1 21/ <9.l)> M» (CHIlOt ovwr un.bl. <o tcto »«rt «t >ll In 1 2»/ cklloron k*c«i*>« of • (ASK » J (ASK •) J CASK 01 J» A. (9.4) ..1 I2/ (9.41 1 I*/ (9.4 1 24/ *<M 1 IkM (CHUO) ovw ll»lt«d SO/ •toy k«/»k* co«l4 do k»c»wt* fiiwlraw>t>l (9.22) ..2 (9.22 a <Q.22> a �2HO CHIIO CHIIO'S MAKE: 3RD CHIIO 4TH CHIIO SIH CHIIO •IH CHILD CHIIO'S ID I itoiMtH't ibT" H 1 11 11i III 1 1 1 U. I 1 014 (CHILD'S) (birth 4.toctl»l/ l*o«lr».nt> ovor k..p (*!•/ <»/ (ASK A) (ASK A) (ASK Al 1 I Jt (ASK Al SV 1 • •••••(A9C Ala,***,*! • •••••(ASK Ai v > C ( 4 B 3 (ASK Al (ASK Al J J» s*/ ...l»* A) 1 6V (0.111 ..... 1 'V 1 014 (CHILD) ov.r KKV.J to 90 to o cwtaln typo ol school, or t>« In o Yo*. (GO TO 0.1*1 ...... 1 M/ (ASK •) of (kll/kwl(blrtk ••••••CASK Al* •••**2 IASK AI....2 .2 ........ ...<o.ie> i '«/ ...(ASK 81...J J ••••••(AW Al •••*••• vu A. • I/ ASK A .(GO TO 0.191 No.. 43/ t.». korl Iron going to school! Ml (CHIIO) .v.r ll.lt.4 •7/ .I . »/ .J III . . . (Q.20I.......I ... ••....•..•.•.. SCkOOl •ttOAdOIICO OT !• tol«9 >kl* to loom kocow* ol Ikli/horKklrtk 4«/ Jt I . Bocovw ol IkU/horl (klrtk I 4.f»ctUI/l«*«lrB.iit> 414 SI/ 43/ SI/ (CHItO) ov»r HM« • lot inro hot* tkM othor Ckll4ro« (CHILD'S) ••> In oolag oiittloo. •V 7 fO*tl«f «o (ckool. 901119 to «ko »toro. Mi4 ottMr ovoryvoy oclKltl.l Ilko tK.ll It. >•/ SI/ 4«/ •«/ of • (klrtk tfofoct/ I, 414 ICHIIOI •vor (woo1 tko kolo of •wlkor pwioo for ovory4«y 4ctl«l*lol »ck «« tckliit cwo ol tho kooto T.I. .I . HO.. .J 4 1/ ..••••(Q.I7>.......t (ASK Al J <«/ SV IO.ITI (ASK Al S9/ 1 J 9V I *v *.....<Q.I'I.*** . 1 . IASK AI a BEGIN DECK 31 Tot (80 TO 0.2 II Ill (klB/kor fro> -jrkl«9 on • Jok for »oyt A. IASK Al*..,..2 tot MIII/Ho.14 ICHIIO'S) klrtk «ofoct(tl (koop/b*.* lt.pt) 6V B. «IM/»>«I4 (CHItOI (b»/h«y. kOMl Ibltod In IM »l"4 ol «ock (h./th.) co.14 Uo/h»>o «o»o) fcocouM ol (hl«/korl klrtk 4«f«ct(ill Mll/lto«l4 (CHILD) (b>/kovo k*M> Ih(to4 In IM Moo*t of >ork lko/lh*l ctx!4 (4o/M«o 4oo.) k*c«u>. ol (kli/korl klrtk 4*loctltil 7»/ J or »r*Mrliif iwoltl 1 . 0 I »!/ .(O.III ID/ .(ASK AI....J It, (COTOQ.2I) I IV HD........IASK Bi...........J V.«. ..I No., .J .(g.2i) i .(ASK BI....J >V ii/ �OETK 17 WO CHILD 4IM CMItO MM CHIID «H CHIIO 2ND CHILD 1ST CHILD 1 1 I I ..I .B ..9I I 1 I)/ ....(ASK A>. ...2 9.iB> i 2i/ IASK AI....2 ..1 . * .. 91 1 ( CHILD'S MM4EI 1 1 2V IO.IBI ....IASK AI...J IASK Al . . I .* ..9l > 1 (O.I* I/ I CHILD'S ID* NOIHtlt'S 10 * 21. J INItmiiWH: ODES RtSPOHDCNI MA»t 'ANOtHtR CHILOI .I.. «S....(6O BMX IO 9 I . . I 4V a ....iQ.iat i M/ ....(ASK 6>...J . . I9.l*> .. 1 JJ/ ....IASK Bl... J 1 JO/ ....(ASK Bl... J O 9J21 2 19.22 22. e«****|A5K BI»« ( ***I Did you ever have • birth detect? tern (ASK It) Mo ft. What kind of birth defect « • it? • T 21. US t 2 Any otheri? Do yon have any biological brother* or eictere? Include any brothers or BlBtern who Bay have died before the age of 1. Tee ........... (ASK 0.24) .......... 1 No .......... (SKIP TO P. 25)........2 Don't Know.. (SHIP TO O.25)........8 »*•* •*» •• •• • •! 4I/ . 1 4Z/ ... ..J ....49.211 1 IV ....(9.211 1 MS ....IASK AI....2 ....<ASK AI....2 ..1 . 1 .. 92 ) ..I . I ..»2I 1 I*/ IASK BI....2 1 2J/ ....IASK Bl... J ....19.2*1 1 W *SK A)...J ..1 . 1 .. 02 1 IASK At 1 3V ....(ASK BI....2 fQ.Zt> ; 2 M/ 2 <Q*2I> 1 ••••••(ASK •)••••••! 4V 49/ �24. 4IH CHIlo 3RD CHI ID 6TH CHIIO STH CHIIO Did any of your hioloqical brothers or sisters ever have a hirlh defeeL? (ASK A) t (SKIP TO p.25) 2 »«s 1 1 1 1 1 I I 1 1 1 Ho Don't know...(SKIP TO Q.2S) (Q.JI 1 *>/ <«.» 1 >!/ 9.) IQ.22I..-..I II 1 «/ A. (Q.2J1....2 S4/ (Q.22I....2 .k... .19.111 Mho had a defect, brothers, sisters, or both? J Brothers I Sisters 2 Both 55/ 3 FOB RACH SIBLING WITH A BIRTH DEFECT, ASKt What kind of birth defect did your (brother/sister) have? Has this sibling a half (brother/sister) or a full (brother/sister)? RECORD BELOW. Siblin9 I Slblinq 2 DEFECT! 56/ DEFECT t SB/ Half (brother/sister)...! Full (brother/sister)...} S7/ Half (brother/sister)..! Full (brother/sister) . .2 S9/ u> 01 Sibling ] Sibling 4 DEFECTi DEFECT i 62/ Half (brother/sister)...! Full (brother/sister)...I 25. 6O/ 61/ Half (brother/sister)..) Full (brother/sister)..2 63/ How I would like to ask you aoM questions about your biological parents. Did either your biological pother or biological father ever have • birth defect? US (GO TO 0.261 1 Mo ( S K I P TO 0.20) 64,' 2 Don't know..(SKIP TO O.2S)..B 26. Which parent had a birth defect? Mather only 1 Father o n l y . . . . . . . . . .2 Both parents 3 6S/ �DECKS 37-3B 27. t*1 !442ns<><--7 SF.CTIOtl 7l What kind of hirlh defect did your (PARENT) have? Hothen 66/ Father. HEALTH 67/ I. Now let's t a l k about health. Conpared to other people your aqe, would you say that your health is ... (READ CHOICES)? Now I have Bone different kind of questions. 20. Excellent Mas anyone near to you died In the last 12 Months? Good ........................ Fair ........................ 3 Poor ........................ 24/ 2 4 Yes..(ASK A AND B . . ). HAND CARD 1 Ho 2 68/ ................... » I A. What was the person's relationship to you? as apply. CODE ALL THAT APPLY. Please choose as Many 2. Since (DATE OF LAST INTERVIEW) have you had acne on your face, chest or back? A. Child Ol 69-70/ Ye B. Parent O2 71-72/ No c. Spouse/partner 03 7J-74/ D. Brother or sister...................................O4 7S-76/ E. Other near relative of you or your spouse/partner 3. F. Other (SPECIFY) ...06 79-BO/ BEQIH DECK 39 Ourinq what year, between (DATE OF LAST INTERVIEW) and now, did you last have acne on your face, chest or hack? 4. 26-27/ Think about the (first/next) tine you had acne on your face, chest or back between (DATE OF LAST INTERVIEW) and now. Mien did it start and until when did it last? (PROBE FOR ALL PERIODS OF TIME). IO-I1/ Third First Mist (was the date/Mrs the dates) of the death(s)? year? ENTE* MONTH AND YEARi 2 RECORD YEARi _O7 B. ( S K I P TO 0'8> 2V ...OS 77-7B/ Friend G. 1 I—I HO Mhst month and l_J_J I—I—LI—I Mo Yr I2-I5/ I EHTEM MONTH AND YEARi I ' LJ_J HO YR I6-I9/ L_l_l 20-J3/ MO YR I L-LJ—I—I Mo Ho Yr to to YR LJ_LJ_I ENTER MONTH AMD YEARi LJ_L_LJ Yr LJ_L_LJ U—l—LJ Mo Yr �IIKC KS JH- )•) 7. HAND CARD p. J A. 4 IHTPRVIFMRRi ASK A FOB FACH TIHR IN p. 4 What was Lhe name of the doctor or medical facility you consulted at the tine? Plea se show Me on this diaqran where the acne (is/was) located (the ( f i r s t / n e x t ) time). CIRCLE "ITS' OR -MOSECOND TIME Ho Yes FIRST TIHR Yes NO THIRD TIME Yes No First Nane Temples 1 2 52/ 1 2 «!/ 1 Eyes or eyelids 1 2 53/ 1 2 62/ 1 2 7»/ Ears t 2 54/ 1 2 6V 1 2 72/ Cheeks 1 2 55/ 1 2 64/ t 2 7V Nose 1 2 S6/ 1 2 65/ 1 2 7«/ 2 70/ OR Facility Nane A. What is the address of that (doctor/medical facility)? 14/ STRERT ADDRRSS Forehead 1 2 57/ 1 2 66/ 1 2 «/ Jaw, Chin other 1 2 58/ 1 2 «7/ 1 2 76/ Chest 1 2 59/ 1 2 6B/ 1 2 77/ 69/ 1 2 78/ Back 1 2 60/ 1 2 J I I (STATE) B. What Is your blood type? IS/ IF ANY "YES" TO TEMPLES, BYES, EYELIDS OH EARS IN A ABOVE. ASK 0.5 ALL OTHERS SKlr TO p. 8 S. Between (DATE or LAST INTERVIEW) and now, did you »»er consult a, doctor or Medical facility about tha acne on roar lt«ples/eyes/eyellds/ears)7 Yea Ho (SKIP TO p.») fi. 1 2 79/ What Month and year did you last consult a doctor about tha acne on your Itemples/eyes/eysllds/earsl? BEGIN DECK 39 LJLJLJLJ 10-iv A. 9. Is that positive or negative? During the last year, how often, on averaqe. would you say you use aspirin? Hare than 4 aspirin a day I 4 aspirin a day (2 dosss a day) 2 2 aspirin a day (1 dose a day) 3 6-e aspirin a week (I dose, 3-4 days/week)..4 4 aspirin a week or less S Hone 6 I 7/ �ASK OF ALL RESPONDENTS 10. IS. The next questions deal with your reaction to the sun without the uee of guntAn lotions. If your akin wa* exposed to the sunlight for the f i r s t tine in the snaMi-r for a period of 3O Minutes would you Alwaya burn, never tan I Usually burn, tan with difficulty LULU no 2 2J-26/ Tr IB/ 16. HAND I CARD H Nhat is the full name of the doctor who Made the diagnosis or the Medical facility where the dlaqnoels was Made? SoMetlsies burn Mildly, tan above avaraqe....! Rarely burn, tan with •••• 11. During whal Month and vear did a doctor first tell you that you had a peptic or RloMnrh ulcer? LAST 4 In the oiimmrr. once you have already been In the aun several tipiea. what reaction will your akin have the next tie* .you qo out la the sun for two or More hours on • brlqht day? Mould you say you qet A painful burn I A burn 2 SOMB redneas only 3 No Reaction FIRST I9/ 4 fACIUTl MMIE Ik. Hhat i« the addresa ot that (doctor/Medical f a c i l i t y ) ? 27/ STREET ADDRESS U) 00 12. After repeated sun exposures. Cor exaMple, a two wesk vacation outdoors, will your skin becos* . . . Only freckled or no suntan at all HAND I CARD N t Only Mildly tanned due to a tendency to peel 2 Moderately tanned 3 Very brown and deeply tanned.... 11. 14. l CITT , 2O/ Durlnq sny period in your life, did s doctor ever tell you that you had a peptic or stoMSch ulcer? fas (GO TO ff.15) 1 22/ .(SKIP TO Q.24) 2 I Da you have a peptic or atomach ulcer now? lea ...4 HAND SELr-Af*UNISTntn> FORM 2. He would like you to tell us sll the places youeve lived sines you ware born. Please list all the places you've lived for More than 12 Months starting with the flrat place 2I/ since birth, 'lease take your tine. It will probably take you to Minutes or so to f i l l out this fon. Please begin. Ho. 17« I (STATE) 28/ B.. o. IB. Mhat aionth and year did you laat conault a doctor for your peptic or atoMach ulcer? LJJLLJ Ho Tr 29-J2/ Have you ever during any period in your l i f e had a bleeding ulcer? 1 Tea No (SKIP TO p.21) 2 IV �2O. 24. During what nooth(a) and year(a) did you have a bleeding ulcer? FROM FROM FBOM Ir Tr TO Please indicate which of the following »e»lx!rs of your bioloqical f a n i l y have ever had a peptic or «to«ach ulcer? TO (to TO 2. Father Full Brother 4. Yr Mather 1. Ho I. ............... Half Brother ..O1 ................. O2 72-7I/ ........... O3 74-7S/ ........... O4 76-77/ OS 78- 79/ 21. 5. SB/ Tea. ............ Hone 8. Here you ever (during any period In your life) hospitalized for your peptic or atoMch ulcer? ............ Half Slater 7. Tr Full Slater 6. Ha 7U-7I/ Don't Knot. BEG III DECK 40 O6 1O-1I/ ................... O7 ............. 12-I3/ 98 14-1S/ Mo.. 2S. 22. Do you have or have you recently had aharp upper etomach pain? Tea Have you ever (during any period in your lifel had eurqery for your peptic or etoMdi ulcer? 26. ...... Tea 21. Are you currently taking any prescribed •edlclnee for your peptic or • toaiach ulcer? tee ......................... Ho ..... (SKIP TO 0.24) A. 1 ....... I6/ 2 Haa this pain relieved by food, Milk, or antacids? 7 UJ I ......................... MO ...... (SKIP TO Q. 28) 17/ Mo 60/ 27. Haa thle atoiuch pain awakened you fron aleep? 2 Tee IB/ Mo What are the naam ot the Medicine* you are taking? (PROBE I WHAT OTHEH87) 1) 6I-6V 2) 64-66X 3) 67-S9/ 20. Have you vonited blood recently? Tea. 19/ No.. 29. Have you recently experienced dark tar colored atoola or bowel •nvenanta? Tea. Ho.. 20/ �-76- Now I would like to ask you mome questions that deal only with the period of tl»e between (DATE OF LAST INTERVIEW) and now. rai ASK A rntouGB o ron EACH CONDITION CODED TES. A IIMK OF IASI • B»t.OMI IDAIt Of lltat l> tto lull IASI INKNVItH) tot ••>• »M titXttt ot MB, I* mat wxitk ••• y«*r «I4 • doctor tlr.t *•)! To. ttet rou tad <OJNDIIIUH)I > • doctor tol« (m ttat »ou ta< ... tt» ooclor «ho •»«• tt* «l*9no>l« or tta «.dle.l t.clllt, .l»r» t L_LJ LJU SKIP n PLEASE GO OH TO NEXT PAGE at City H«r»l« treklMl i t n/ SKIP nnciro 10 l_U * rir»t J_L_I Clf» It.t. �DECK 4O C 0 I Ara you currant ly taking any araccrlbad n»dlclaa» tor your ICOMDITIOHIt Oo you hava ICOHOIIIOH) oo.I Ta> 1 Ho Ta> Ho I 3V 1 2 SKIP TO F •kat ara tka aanaa ot aadlclnaa you ara taklaat tay otkaraf Nkan did you last concult a doctor lor (CONDIIION) batwaaa IIMTC OF 1 Alt INTCRV|[H) 0*4 ^o^f •> 1 1 11 1 1 SKIP TO F 3V t f Ha 21 31 Tr i Sine* tOA't Of l*SI M>at It tka full *a»a and addraM ot tka doctor or a*dlcal laclllty ««« |Bst conftaltad toM »<X1 ttot you ta4 .. 32. taamlat latt Mm Htot l> tte lull Mn* oiitf cd«r«ss of tte doctor •!» ~*« tlu) dlognosl* or tta Mdlol luclllry •teru tta *l I 4*/ I w H/ i sMI/ bt^on (DATE OF IASI INUmitV! «n« •c«. !• Btot «onth ••« y«ar «!• • doctor tlrtt toll you that you •»« CCONOIIIOHH iMnmitm «•• • «octor '^ 2 6V SKIP TO l_J_| LLJ M> Ir II/ »I-70/ Q.33 Flr«t Na-a 4£4*y OR Facility Hana Facility K«~ Straat Mara» Straat AddraM 1 City 1 I W 1 2 »!/ VIP TO F •1 1 1 II City 33. 1 1 A towt condltlo*! 2 SKIP TO F LJU 1 1 Stata SKIP (SPfCIFyi 21 ii-if/ 31 _ IV 10 0.34 «/ L_U N> Stat. LU IT 14-TI/ Flnt *wa H-aOV OR Facility Facility Ha»a Struot Addr*» Straat Mdraai 1 City 1 1 Stata J_LJ City �PFCIH nrx'K 41 MaiNOCCK 41 f Slaca (OAR OF LAS I INttmitX) kat a doctor tola you tfcat you ha« ... Do you hay* Ara you currant ly lading any arascrlbaa* (CONDIIION) aaalclnai for •Hat ara tka (CtfOIIICH) «<nt your (caomomi •aM si n4lclM« yew «r» Olilout k»«»M« IDATE Of lASf INKIt- lh« •actor or Mdlc*! iKlllty •tat oth»r>t »li«» w4 «o«l totum* IODK OF LAS1 IMICmlEK) >itd MM M* aaara» ol you l«t coo»»lt»< T.I No No 2 tO/ SKIP I 2 tut consult • voclor lor Wwt l» t»a lull •au. In >tat aoatk •»4 yuar ultf • uoctor tint tail you ttet you »M (COWIIKMII no«l» M» 34. 11/ Da, aa.larga« llnrl I TO F F 2 4J/ LL' LL' »r 4J-46/ SKIP 10 SKIP IO Hlu>t U thu lull •WMJ and •aorusf ol tta doctor ubo >m<le tl* diagnosis or tM «.0lc>l Uclllty • h»r« tto 4J/ O.M JI-JV Facility FKltlty I 01 ItrMt «»•»» -LLI U I 2 M/ SKIP TO F 1 2 IV SKIP TO F ,5«/ • LLJ LJLJ fc »r l«t 41/ 3?-4O/ St.t. City Cttr * w SKIP TO 9.1* LLI M> LLI «r 4H-JI/ Flrat I Facility Facility I Straat Adarasi Itraat M«ra» L.LJ City Stata .LLI City Stata �DECK 42 Do you kav* (CONDITION) Ar» you currently t.kliH) uny proscribed •uXIctiwi lor •out your ICONDIIIONIt to* Ho T.» Z M/ SKIP I Ho 2 3V SKIP TO 10 F F MMM did you iMt consult Sl.o. «MH Of l»S I d doctor lor Mwt or. IK. MM* ol •*4lclnu> you dr« tdklnit ICOKDIIION) B>tx« IOAH Of LAS I IM limit*) ..d now. In »h»t wMlh nnuu d«d •ddrots ol tatuuM (DATE OF lASf IMTER- Any otk»r*l Ik* doctor or •»dlc*l Uclllty you ld»t c •Ixt It Ite lull •»• y*v 4I« • doctor tlr«t toll you tk»t IHURVUHI kit • doctor tol« you that you KM ... Mwt I. tkn lull Ih. dl»gi>o>U or tl» ••dlul facility uh>r» tto dl««- you kM lOUNOiriunit ••«• MM] •dorusf ol tta doctor .to Mdu •osl* u»l pdd*? L_U •to Utt I Ir SKIP (to ID •!-*•/ Tr I*- It/ First I FKlllty I Fdclllty w Slrd*t Addroi Str..t Addrdts _LU 41 I JO/ SKIP TO F I 2 7I/ SKIP TO F Stdt* City LJU LJ_J »r City If. L_U l_U OrrkMll «l tk» llv«rt l..t SKIP 4-SIH-O-JIJ-l 10-IV n> Mo IT L»t n-n/ Flr.t I FKlllty I FdClllty ' Str«*t AddrMl Strxt Addruti _LU City J_LJ City Stutu �-84- Do you heve (CONDITION) eon! Mkee 414 yo* let! CeeMlt • 4octer lor Are you currently teklng eny ere»erlbe4 W>et ere tee Ho keteeee UMTf OF LAST INTEP> film eee «eet 2 26/ II 2 2V SKIP TO T». I btmii (OAK OF l«T IHIUWKm •«« K», I* «l»t •oom iw«l» ••• e***t 1 '»/ SKIP F 52-M/ tli* «l«9m>»l* or tto M4lc*l Uclllty •tor* tte *l*g- »o« lut co«>.lt*4 TO F •tat It tta lull •••• «B4 Mlar*»s ol <!• «octor »ho «M* e»4 y*er 414 • doctor flrtt t*ll you tt»t yea k>« (OWCIIIOMU rM> at LL.ILU ee4lcliiet lor yoer (CONDITION)! No Wxt U tlw t»ll (CONDITION) •em el e*4leleet yen ere teMefi Met etkeret Since (OAK UF IASI tHIunitli) h» e eoctor tele yen thet you he4 •• L_l_l LU He TO/ Ji-41/ Vr «0-*)/ O.S» Ftclllty N*« FKlllty I D3 Strwt M*-*u LJU I I 1 2 4V 1 2 «4/ SKIP TO F TO F I I SKIP City S»»*« Clff If. tell kIMMr ereblcuT 1 1 11 1 1 t..t H ~ M/ I 2 »V SKIP N> | | « TO 21 9.4* Jl I I I ( St.t. »»-*»/ Flr.t Hene FMlllty I Feclllty I Street Mereu Street A«er«i» City City _L_U �MM.KS •»4lcllM» lor your (CONOIIIOHX lM»t y*» mm 4I4 yo* Ho To» I Tl/ SKIP 10 No Wwt «•• tko ••Ml ol •*4lcl* you at* t*hl»«l «k*t otkM-ll F F iHHmum k» • «octor «kal l» tk* lull KONDITIOW MM «M •»*•>• of kotMM (DATE OF tAJI INTiMtlt»> M4 »o«» tk* •actor or •.4lol locltlty yo* lot comilt.4 LULU II SKIP TO SUo* IOAK OF I.HSI lut convlt • doctor for fr« you currootly taking My proscrlbof Do you bavo (CONDITION) 42-4) Ho » Tr AW» •!!« •«<»•••» of tl» ooctor >ho ••«• •no y«or 414 • 4octor tint toll you tftot tta <l«gno>l> or tM (Xim iKlllty you KM ICDMOIIIUMIl •tor* tr» «logM»|« ••• **4of A* otter (SHCIFyi • Itloul I 2 JV SKIP L_U L_U Ho Tr J6-H/ 13-ltV H«IN HOt 41 •l»t )• tM.lull M». Ill >k>t •ooth toU you tk>t you ta4 .. 40. la»t I ••turn* IDAIE Of 0.41 Flr.t I Facility Facility I T ttroot MOTIM Straot Mdro«» LLI LJU Cl ty City 2 It/ SKIP TO F I 2 IV LU LU II SKIP r ». 31 •• 23-»V 1*1 » Tr SKIP l*-32/ ID «.42 «•/ SI at* LJU LJU Ho Tr 42-4V Flr.t I Ficlllty Facility I Stroot M*r*l> -LU City Stoto City �BET.1H DECK Oo you navo I CONDI I ION) •ooJclm lor MOM o ooctor for 1COHDITIOH) •hot oro tko No 2 4I/ SKIP Of o«4lcl«0« Wwl It tko (oil »0t>000 IDATf O_ 1 TO F OF IAST INK* ! • • 9*4 ••*!••*• Ol !» L_U L.LJ I NOIIT tM ooctor •!» »o> tl« dlognoslt or tho •»•!€•! ffoclllty >N>ro tta AOffll ••• 42. tv X roiolrotory contltloa •tkor tkoo omi«io«l»l I 21 LLI 2 ID/ SKIP TO 1r 12-IV ».43 11/ 31 Ntat It tta lull A*M OA4 adtfross ol M0ve III »Mt nonth •M yo*r «l« • •octur Mrit toll you ttot you kM lOOHDIllUHII tko toctor or M4lcol Uclllty yov lost con«olto4 you wo tlklngt Wwt otkorol Voi euwnn (OAK Of LASI IHItKVIEKt •nd Sl»o> tDAH OF L«SI INHUXEII) MI • Doctor toN you thot you Iwd .. you yo«r (CGN0ltlON>t «o«J r«« Aro you corronfly toking ofty •roscrlboo' Flr.t Foclllty I Stroot Moro» Stroot MJron Stoto «J. My ofkor oojor coXltlont 1 1 2 oV SKIP TO f 1 1 «4/ SKIP TO F II f 1 II 1 1 31 M-tOf t SKIP re/ " St.t. City U I City " "' . K> 9.44 ^ U/ L,L_! L.L.I to If lf-22/ '•/ Flr.t Mow Foci 11 ty No»o FocllKy I Stroot Moroti Stroot M«ro» _LLJ City 44 City �0 Ar« you cur- •ton did you iMt comult • doctor lor •Mt It tb> lull Do you h*v«j rently taking t>ny pr«»erlb*i*} (CONDITION) (KdlclnM lor Vh«t «•• ti* your ICOMOIIION)! »«M* Ol M«JlcU Utnan «MIE W I AST INTtKmn> Md aoul •to., during «l»l Bonth tM doctor or •»dlc<l (Klllty »«• «nd M«r«ti ol tto doctor Mteo »«d« !<• dl»g«o>l> or •nd f««r did • doctor Mr it toll you «r»t CCONOIIIOm •onf tn> Mdlcal laclllty •fwro th» dlAgnosli you l«*t consHlto you *r« takliif t Wwt oth.»r«? «Mt l« tto (.11 TOO tod tOOMOIIIUNII 44. 2 24/ I i »/ SKIP SKIP 10 TO LU LU I) tr I. . .. »*»• you «.«r U toor •Mir. IM. kton tr..t.d f«r • »»«t«l or Motional I 2 SKIP ID I I I «—«—' "o 21 r ririt i SI facility FKlllty I Cd Str..t Addr*si Strxt Mdr«» _L_U CMy 2 40/ SKIP TO F 1 2 41/ LULU O »ir r ». SI Ha JV ir 0-4V flr.t Facility itoM StrMt _LLJ City J_U City �PFIGIN I1PCK 45 45. At any t(«e since (DATE OF LAST INTERVIEW) has a doctor told you that you had cancer? Ye* 1 IO/ 46. Did the itoclor tell you that thia wai a akin cancer or a ayatenic (body) cancer? No...(SKIP TO p.49) AT* you cor- Mm) 414 you last con»lt • 4octor lor Do you h«v* •CONDITION! r wit I f taking any pr««crlb«4 M4lclM< lor Kb** W« lOnOIHOW Mat l> tl» l»l I ••» m4 •44r«« of iN»I your (CONDITION)! • •••» of »«4lcla*s kolDMll (OAlt Of IkSI INliM- tin *oclor or a^lcol (acllltf TOW *re T«» No 2 6V SKIP TO F r«t I No 2 «V 91 IP llfcet OthMTSl < «Mt Skin cancer only..................................1 Syatemtc cancer only (SKIP TO p.48) BOTH SKIN MO SYSTEMIC CANCER ymi l«»t co«»lt«4 It/ 2 3 6KIH CANCER ONLY LULU I) 2 U.H 47. 2) rlcaae look at this chart and tall •» utiara each of your akin cancer* lla/waa) located. tt-W 5) INTERVIEWER! flrtt ! INDICATE THE AHATOttlCM. COOK FOR EACH SITE BEING REPORTED. SITE NUMBER RITE CODE 73-TT/ *-00 COOESi llr**t AMrm LU *OI) (14) 4O2) eye Lid (01) Ear Scalp or Forehead Aral or Hand, Not Otherwise Specified (15) Genitals 104) (05) (ICI (17) Le<j Foot (18) 1*9 or root. Not Otherwise Specified Skin. Not Otherwise Specified Upperlip, Not Otherwise Specified Lowerllp, Nat Otherwise Specified tip. Not Otherwise Specified Nose Head or Neck, Not Otherwise Specified (06) Cheek, chin or Jaw (07) M*dt or Supraclavlcular (08) Veraillion (09) Trunk, front (10) Trunk, Back (til Trunk, Not Otherwise Specified (12) Arm (11) Hand (19) (2O> (21) (22) �INTERVIEHERl DECKS 45-46 DECK 45 -93FOR EACH SITE REPORTED ASK A THROUGH F SKIM CANCER ONLY 47. SITE 1 A. SITE 2 In «hat Month and year wa* cancar or the (SITE) Heat dlaqnoaed? 1 1 I 1 1 A. In vhat aonth and year vaa cancer of the (SITE) flret dlagnoeed? A. L_l 1 L_LJ •*•»/ Ho. Tr. I ie-2./ D. In nhat Month and year nee cancer of the (SITE! firat diaonoeed? L_L_I L_LJ »>-»/ SITE 3 SITE 2 SITE I SITE 3 D. What treatMente or Medicine* (do/did) you take for cancer of the(SITE>7 CODE ALL THAT APPLY What treatments or Medlclnea (do/did) you take for cancer of thetSITEl? CODE ALL THAT APPLY Hhat treauwnta or Medlclnea (do/did) you take for cancer of the)SITE)7 CODE ALL THAT APPLY Tr. HAND CARD MAUD CARD B. Miat kind of akin cancer vaa thia7 Tr. HAND CARD » P B. Ho. Hhat kind of akin cancer vae thie7 B. 36/ Radiation 1 S6/ Radiation 2 37/ CheMOtherapy 2 57/ CheMOtherapy*....2 77/ 3 7B/ 79/ Surgery 3 Other.(SPECIFY)..4 Hhat kind of akin cancer Man thlal Baaal call SquaMoue cell SqiunotM cell 1 38/ Surgery.... X SB/ Surgery 39/ Other.(SPBCirr)..4 S9/ 76/ Other.(SPECIFY)..4 Baael cell SquaMoua cell Cancer Mtaetatlc to the akin 4 Cancar Metaatatlc to the akin 4 During vhat Month and year did you firat receive (EACH nCATHEHT COOED IN O) for cancer of the (SITE)7 Cancar Mataatatlc to the akin 4 C. what ! the full • naeie and addreaa of the doctor or the Medical facility where the dlegnoeie vae Made? Laat Maaw (that la the toll naM and addreaa of the doctor or live Medical facility vhere the dlaanoeia •aa Made? Uat Nee* IV rlrat Name C. "hat la the fall nee* and addreaa of the doctor or the Medical facility vhere the dlagnoeie vaa Made* nrat MM HO ULI YR 40-41/ L_L_I L.I-1 HO YR Facility HUM Facility NaM Street Addreaa Street Addraaa YR 48-S1/ OTHER L.L-1 HO L-L-I YR S2-SS/ ir SECOND SITE CODED IN Q.47 CO TO SITE 2 CUT city 1 1 1 1 1 State * 1 1 State * 1 1 State * HO L.L-1 L_ I-1 MO YR 60-63/ L_L_I L_l_l HO YR YR 10-13/ CHEMOTHERAPY | SURGERY | |_ | HO |_ L I YR I4-I7/ | | «4-67/ SURGERY |_LJ HO Street Addreaa City RADIA- 44-47/ OR raellltr Naaie E. O BEGIN DECK 46 During vhat Month and year did you f i r a t receive (EACH TREATMENT CODED IN D) for cancar of the (SITE)? TION Firat MM OR During vhat Month and year did you firat receive (EACH TREATMENT CODED IN O) for cancer of the (SITEI7 RADIA- I V OR t. RADIA- teat NaM W/ NODE NONE. HONE.. w C. t r Basal cell v£> Radiation CheMotherapy Ho. ir Q.46 CODED •!-, SKIP TO Q.49 L.L-1 YR •-I •7/ ««» LLI L-LJ HO YR | HO IF Q.46 CODED "1", SKIP TO 0.49 | 1S-2I/ OTHER L.LJ HO L.L-1 YR 2I-2S/ 72-7S/ IF THIRD SITE CODED IN Q.47 GO TO SITE 3 | YR IF Q.46 CODED SKIP TO Q.49 �DECK DEIK.S 46-47 47 48. (Continued) SYSTEMIC (HOOT) CANCER ONLY BODY PART I A. In what part of your body (is/was) cancer loca ted (RECORD VERBATIM) 26-4S/ BODY FART 2 in what part at your body (is/was) career located? 47 A* (RECORD VERBATIM) S1-70/ In what part of your body (Is/was) cancer located? (RECORD VERBATIM) IO-29/ Last Name Last Mas* Last "we B. OR Oil Oil B. Hhat king of cancer First Nane First first Nan* 4V 40/ 35/ Hhat kind of cancer was it? Hhat is the full na»e and address of the doctor or the •edlcal facility where the diagnosis was nade? Hhat is the full name and address of the doctor or the • •edlcal facility where the diagnosis was nade? D. What U the full name and address of the doctor or the •edtcsl facility where the diagnosis was *ada? BOOT PART 3 BODY PART 3 BODY PART 2 BODY PART I BEGIN DECK Hhat kind of cancer facility Facility Nane Street Address Street Address City •as It? Facility Ms Street Address was it? City City 30/ W C. In what Konth and year wac cancer of the (BODY PART) first dlaqnoaed? U_l LJU «>-*»/ Mr. Yr. In what Booth and year waa cancer of the (BOOT PART) first diagnosed? _l U_J I I_J LJ_I State In what awnth and year was cancer of the I BODY PART) first diagnosed? Hhat treatments or •edlclnes (do/did) you take for cancer of the (BODY PART)? CODE ALL THAT APPLY what treatments or •edlclnes (do/did) you take for cancer of the (BODY PART)? CODE ALL THAT APPLY Hhat treatments or medicines (do/did) you take for cancer of the (BODY PART)? CODE MX THAT APPLY LJU U_l Mo. Tr. State Radiation HAND CARD I 36/ Radiation t 4t/ Radiation I 46/ Chemotherapy 2 37/ Cbenother spy 2 42/ Chemotherapy 2 47/ Surgery 3 38/ Surgery 3 4V Surgery 3 48/ Other.(SPECIFY)..4 39, Other.(SPECIFY)..4 44/ Other.(SPECIFY)..4 49/ HOHE.. �4 . (Continued) 8 49. F. During what nonth and year did you first receive (EACH TREATMENT CODEO IN E) for cancer of the (BODY PART)? F. During what »onth and year did you flrat receive (EACH TREATMENT CODED IN E) for cancer of the (BODY PART)T RADIA- RAD1A- TIOH BODY PART 3 BODY PART 2 BODY PART 1 |_LJ I_U HO YR TION LJU I_LJ HO HO YR During what awnth and year did you flrat receive (EACH TREATMENT CODED IN E) for cancer of the (BODY PART)? RADIATION TR 47-70/ 50-53/ CHEMOTHERAPY LJLJ l_l_l F. CHEHO- ™ERA« LJ_J m 54-S7/ U_l Tt 7 i-;*/ |_J_J HO CHEHOT««APY LJ-J LJU 1R I5-I8/ M any lioe ainr* (DATP OF LAST INTERVIEW) has a doctor lolit you that you had leukevia? Yea CASK A-F) 1 32/ No. A. .(CO TO P.SU 2 Thinkinq about the period between (DATE OF (.AST INTERVIEW) and now, in what wonth and year was your leukeaia diaqnoaed? 33-S6/ LJU LJU Ho. Yr. B. Nhat la the name and addreaa of the doctor or the oedlcal facility where the diaqnoaia waa nade? LJU HO YR I9-22/ 37/ SURGERY LJU HO LJU- SURGERY LJU HO YR 58-6 1/ OTHE« LJU' LJU HO YR OTHER 62-6S/ C. IS THERE ANOTHER BODY PART AFFECTED? Yea.. (CO TO 48A-Body Part 2 . . l ).. LJLJ SURGERY LJU HO M 7S-7«/ BEGIN DECK 48 L J J U_l HO YR 10-IJ/ LJU YR rirat Haaie 2J-26/ OTHER LJ_J LJU HO YR 27-30/ Facility C. IS THERE ANOTHER BODY FART AFFECTED! Yea. . G TO (O 48A-Body Part 3 . . l ).. G. IS THERE ANOTHER BODY PART AFFECTED! Yea..(GO TO NEW QUO) street Mdreaa I I City No. .(SKIP TO Q 4q\ No..(SKIP TO Q.4f>... H . (SKIP TO o. } 66/ I4/ .2 Jl/ C. Mhat treatjnenta or Mediclnea have you taken for (DATE OF LAST INTERVIEW)? I) 21 I State leukevia aince 38-40/ 41-43/ 1) D. For the period between (DATE OF LAST INTERVIEW) and now, during what Booth and year did you flrat receive (EACH TREATMENT OR MEDICINE IN O? HJ. Yr. TREATMENT I TREATMENT 2 TREATMENT 3 LJU LJU LJU LJU l_l_l LJU 47-SO/ 51-54/ SS-58/ �INKKIIEMR: FOH IACH KS, ASK A IHUUGHB: A. 49. (Continued) ASK C-C FOR €ACH "IIS" Al 8. CARD E. What is Ihe none and nddreaa of the ftocLor or Medical facility you last consulted about your leukemia? J 90. 59/ Sl.c. WAIF. Of LAS1 IHUH»lfH) 6 , ,o. .. On .tat fffl at your body did ro. tev. (COHDIIIONII Any otter MTtl Laat Hade c. B. Did ,00 4llo>» ICOWITIQM) .It* • doctorl o. Mh*t ».s th. dl«gnosl«l Khar Is Ite docl-*•«* tot .*»..»! First I. Patctoi of row tkl. LJ_I Sit* CM* M-64/ raciittr I 2 M/ Flr«t N*«* Sit* Co*. LJ_J Street Mdreae Sit* Co*. F.clllf, A**r*ii 6».)0/ I J_J_J rttr F. w <J1 KJ Clt, » | I St.I. State During what Month and year did you laat conault (HJME IN El? I_U I_U Ho. Tr. 2. i*>l*r kTMlna el tk> (III* tkMi m**ll | || I 2 IV Sit* COM >)-?•/ i J io/ 11/ Sit. Co*. fsell It, Addr.o sit* CM* >*-ao/ at, Ist.t, I i CO fO n.M COHIIHUCU OODfSi (Oil Sc*l> ar For* I»M (»i« Er> tu l«>l CM(041 NOH <05I I*M or Mick, Hot OtterelM S»*el tl*« 10*1 Clmk. CM* or Jw <•!} Ntcft «r SDCTKlnloilcr 114) (13) (1*1 (III (Itl Off 174) 10*1 »mk. Fro«t 4101 **»k. IK* (2I> (III Oil (211 »»iMi. Hot Utter. IH SpoclllM »r- *-. or Mod, Mot Ottor.U. Steeltl*4 Qiill.ll l*« foot 1*9 or Foot, Hot Ottor.lt* S»*cltl*d fkl*. Not Utter.It* Sp.clll*4 UMorli*. Not Otter.lt. S»*clllM toiwll*. Ho* Otter.lM S»*cMU« llr. No' Otter.I w S»«clll.4 �99. ASK C-6 FUN EACH ICO.tl.Wdl •H5" AI 8. (WO CM» F. i . J Durl*f mint font* and y**r ••• tk* MM ***• «l tk* doctor tfc*t? (Mr I** (tot «o»tk ••*• y*or 414 *oa lo»t coiiMlt JO. Sl.c. WAR OF l«SI B. Ck> .k»t r*rt ol yoir ta» rou Ul« (On <H»cu»» txxV •!« you HIM you lost coa»vlt*d *ko»t (COMMIIOHIf tdMHIIUNI Nl»t ••» tta ICUMOIIIONIT otter »«rtt CH lit or aMIc*! facility .11 » • doctor? dl«gno>l«t >tet >•» L U L U Ha. LULU •>-,./ Tr. rir.t I •te. J. No Ski* tkot *M •«tr* Moiltlm | or too*** to tr. MM *iur*9il Sit* Cod* 1 da/ 1 Jl-JV kwrt for ** room! I I 3I/ | | | Sit* Cod* 14-1V F*ellltv I Addr*» Sit* Cod* J4-JI/ I Strwt «ddr*M City I I St.!. Clt, CUOiS FOR ».> LU Ha. LU »r. GO TO 0-50 COHIINUlO - LLI LU 22-2V flrtf i •a. Tr. COOtS t Soil or ror*k*M Ar» or Kn«. Hat OthM-.U* Onltcli Er» lid 10)1 Facility I 1141 MM OH 1011 IU» 1041 S»*cllU« UC MOM City t»»d tr 104) CIMk, cM« or J*. Faat U or Faat. Hot Otter. I >. | 1011 Mick or S**r*cl*vl«l«r llfl Skin. Hat Otteolw &p«cltl*d 101) IOt» Mr*4lla« k-Mik. fr*«t 1201 (JOMrllr. Hit otlw>l»* (101 »Mik. B*ck 121) low ll». Hat Otk>r>lu lilt »Mk. Hat Otter«lM Sp*clll*d 111! AT. II M MM H*ck. Mat Otlwuli* t»f III! (III Str**t Mdr**> 10)1 if*clll*d S**cltl*« III, Hat Otk*r«lw Sa*cltl*d �DECK 49 10. ICoo»ll»4l ASK A THROUGH J FOR iACH »IS F. I. *• Owing «tof iwM •tat yW Ml tMIt It tlM ••»• M4 III A>l«* fro* Injury. • Inc* tOHt OF IASI *••« »o« »" LU LU LL . J L U Na. Mo. *r. B. 10. During <HMt •44r*» of tlw doctor or n»4lc*l facility yo> lot eonult*4 •bout ICOMMTIONIt IT. Thinning about tko »•'•«"' «•»••« (OAli OF IAST IHIeWICKI nn4 no.. .k*n 414 >.. 4. LU LU I 2 »/ Ho. »r. JI-JV Fcclllty I M4TMI Clt» J_U In an, of Ooyouitlll h . (OMIITIDN) .. rM Mrtt notice (COOIIIONII *• * °; llblch Il«b> or matin w. •t«*cl*4l C. LU LU No. Tr. •*.«,*/ No �DF.I'KS 49-SI) JO. V). (Co«<lMw4) J. H. Fro» (DATE OF IASI Dlo you »oo • doctor lor wring >k>t Moth! ICONDI IIOHIt at Ml tko «l«g*o>l«t Owrlnij vh«t «onth and During >Mt ml y««r 4ld you l*tt consult •<lik-«» ol IK. doctor | T M M l •fto *mt* t •rid yo«rl»l us tk» (OOMDITIOH) -o»t _ Mo. Mol It th* MM m* ••*•«» al tko doct (NAME FROM III or th»BM-|c*l you Islt con*ult«tf •bout (COtOIIIOMK _rr, L U l_U KIP Ho. Flr.t H«U TO L.L-1 Vr. Jl/ •—'—' '—!—' 22-2V Ho. 1r. >€/ ,,-20 I—I—' I—I—' Mo. Tr. Vr. on ~U7 Fcclllty I Stroot M4r«< Street AMr«M U1 LLJ LLI "-•* Mo. Yr. TO K6IN OtOC M LULU "•» Mo. Yr. I 2 M/ _l L_U Kir Mo. »r. •"• FKlllly I FociIIty I ttroot M«r*» City J L L I City L U $t.t. JI-34/ �ASK A THROUGH I FOR EACH IE S A. e. Thinking .bout tk* period Utx»» (DATE SI. All** lro» Injury. >lK* (DATE or IAST IHIEIWIEMI tov* you "Mel. I Irti or *»>cl*s Mr* C. Do rou »tlll h*<* (COtOIIIOH) OF tASI yo« Mr»t nolle* <COMDI1KM>t I. P*r*l(l«it I_U LU •«n*nt lone In Mr el 2 IV Mo. Tr. J*-JV PLEASE GO OH TO HEXT PAGE 2. r*r«l.«»nt KkM •»« L.U LU I 2 «J/ «o. »r. 44-4W * 2 »0/ �DECKS B.IM*. (OAIE Of OK f ou M* • I I mtlmrlK) ••. U doctor lor »Mr*» ol t»» doctor •». «rlnf »t»t (OMJIIION)! »hoa»4» tto <!•(•<»I• Ma* *•• tta «l«jil»lU Mvt U •*• ••-• «H4 Dvrl*« -Ml ««>ll> •»« •o-tte ••* ,»«rl»l or tto •»«tc*l •MTMi ol tlx doctor y.«r 4U ro. l.tt co»».l> n tto (OmOIIION) l Ixllltf «h^. tto or M4lc»l (.cllllr INAK IKOM III you l«tt con«ult*4 .bout (CONDI IIOHIt IfOIHOCCK 51 L_U L_U Ho. Ir. 2 »/ SKIP IO U.t I 1 / 1 2 •to. TKT •I/ LL.I LULU __ Ir, Facility JU at, 17* City JLJU Hot* U l LJLJ L_L_I No. «r. 1 Ult I It/ II/ SKIP m L_LJ L.L.I •*-*" •to. II/ iMt I LULU LL« LLI •*.••/ Hr»t MM* Oil Ho. Ho. »r. >r. fed HIT J_U it*t« "Tut* Ir. SO-51 �51 109 A ASK * ItKOUOH ) fOH EACH US c. A. Thinking •tout IK *»!«• Iroa Injury, Kblck I!•*• or Do roll Mill r«rlod «>el« "»r« k»«« ICOOIIIOHI b«l>««« (DAIt Of l*SI IMIfmifNI •II*C««4I • IK* IDAIC Of IASI (on llr>) aotlc* (COMH1IOHII 2t-lt/ i LJ_J a xv Ho. PLEASE GO OH TO NEXT JACK J> 30 I Ir. 2 JO/ �-112- 92. DECK (Co.Ilnu.d) <Contlwi«dl H. fro. IDAIE Of 1*51 INTERVIEW md IK». during >k»t mtln •nd y.w(tl IMS Ik. (CONDI TION) noif l_LJ LU "-'^ 10 OH yo» » . . . doctor lor IOONDIIIOHII • t M« tk. dl«gnotl>t Mwl l> Ik. *tm» t*4 .ddr.il of tk* doctor •to *•** tk. dlogooiU or tk. wdlc.l laelllty .fcw. tk. 2 '. SKIP DurlM| .k.t Kntk wd >rar M> tk.II or Mdlc.l iKllltr you l.st con»«lt.d •bout ICONOIIIOHIt LL_l LU TO •hoi !• tk. nww >nd •ddr.n ol tk. docto „-«»/ y*»r did you l.tt consult •NAME FROM lit LU LLI 4,-,o/ •to. O.SJ l_U LU ricllltr w Durl«4 .k.t vontk and Str*.t JtMrxi cm -ww «•••«* ***-•« c»» -V.7.' *r. SI �DBUK SI 53. SECTION 81 (Maiden the prescribed medicines you 'told me about) are you currently taking any (other) prescribed medicines? Tea (ASK ' ft) S1 OECK 51 OFC:4«2BSec-fl HFALTH HABITS / The next net of questions refers to smoking habits* No...(SKIP TO SECTION 6 ) . . . . 2 I. A. For what conditions are the medicines? 1) 2) 3) What other condition*? ...__, . Have you ever smoked at least as many as 5 packs of cigarettes, that ls f IOO cigarettes, durlnq your entire life? . I Tea «/ No . (SKIP TO p.22) 5 J/ 2 _. 2. Do you now smoke cigarettes? 1 Yes No (SKIP TO p. 11) 54/ 2 CURRENT CIGARETTE SMOKE* SECTION 3. o^ o On average, how many cigarettes do you smoke a day? IHTEMVIENEMi ir « MiSHERS BT GIVINC NUHBEft OF PACKS OF CIGARETTES, RECORD VERBATIM. THEN MULTIPLY THE NtMBKIt OF PACKS BY 20 AND EHTEK THE M«BER OF CIGARETTES SMOKED. ENTER NUMBER OF CIGARETTES PER DAY I (IF HOT EVERY DATi) 11 | | | | PER MONTH | 55-5fi/ 57-SB/ OR (IF NOT EVERT DAYi ) 11 I j PER YEAR 59- 60/ �DECKS si-52 4. For how Many years have you been Moklnq (NUMBER IN p.3) (clqaretlea per day/per Month/per year)? Less than 2 yeara O1 2-5 yeara 7. 61-62/ Hhat type of ciqarettea are they? Are they . . . (READ EACH PAIR TOGETHER) CODE OHE A. Filter Lip or 1 O2 Non-filter tip?. CODE ONE 6-tO yeara*................02 II-1S year* O4 21-25 B. 03 16-2O year* Reqular size '. 1 7S/ OS yeara Kinq size o r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 100 H i l l i M t e r ? 8. 26-3O yeara e .•,..,,.,..... .06 JI-3S year* OS 36-40 year* O9 Nor* than 4O year* ..IO 3 Now I an qoinq to ehow you a dlaqran of d i f f e r e n t aize ciqarettea. Meaae look at the picture of the (KINO OF CIGARETTE IN p. 7A AND O.7B). Now, considering your atyle of snokinq--for exawple, how lonq you usually leave the ciqarette In an ashtray or just hold it in your hand — tell He the nmber which indicates now Much of the ciqarette you actually sanke. Section 1 ................... 1 76/ Section 2 5. 74/ Mhat brand of ciqarettea do you usually anok*T (IP HOUR THAN OHC BRNO OR NO REGULAR BRAND MENTIONED ASd Milch on* do you 'moke the Boat?) ................... 2 Section 3 ................... 3 Section 4 ................... 4 orrjce use ENTER BRAND 9e Mo reqular brand. .996 Durinq the period when you were Mokinit the nost heavily on a reqular basis* about how Many ciqarettes did you usually saoke in a day? ENTER NUMBER 1 6. For how lonq now have, yon been aanklna this particular brand? ENTER OATSi LJLJ (IF NOT EVERT DAT.) | I 70-71/ OR TEARS) 72-73/ (IF NOT EVERT OATl > 68-69/ OR MONTHSI 77-7H/ 79-80/ OR S6-67/ OR KEEKS t I PER DAY | | PER MONTH | | A. I | PEGIN DECK 52 ( | PER TF.AP 10-1 I/ When was that? FROM I_L_L_LJ Ho 12-1S/ Tr TO LJ_J_LJ 16-19/ �1O. when you saoke cigarettes, how deeply do you usually Inhale? sayi Would you As deeply Into the chest as possible. 13. Haw lonq hirl you been waking (NUMBER IN O. 12) (ciqarclles per day/per week/per Month)? 20/ Less than 2 years. As far back as the throat .Ol 2-5 years Only partly Into the chest .02 6-10 y e a r s . . . . . . . . .03 Hell back Into the mouth, or 4 11-15 years .04 Just puff and don't really draw It In at all 5 16-20 years OS DON'T UKM • 21-25 years O6 26-30 years O7 31-35 years OB 16-40 years O9 SKIP TO Q.22 FORMER CIGARETTE SMOKER SECTION II. How long haa It been since you smoked cigarettes fairly regularly (RECORD NUMBER) More than 4O y e a r s . . . . . . . . . I O ENTER DAYSt | |J _ OR WEEKS t || | 2J-24/ OR HONTHSi T || | 2S-26/ OR YEARSi || 2I-22/ j NEVER SMOKED REGULARLY. 12. >4. 27-2S/ (SKIP TO Q.22). 29/ ENTER NUMBER OP CIGARETTES PER DATl | l(_|_| PER MONTH ) | JO-31/ 32-3J/ OR (IF NOT EVERY DATl) f| | ,| PER YEAR I Yes IS. (SKIP TO P. 16) 38/ 2 Haw lonq did you stay off cigarettes at that tine? FNTCR DATS. | | 39-40/ | | | 4I-42/ OR HONTHSi I | | 43-44/ OR YEARSi 16. f OR WEEKS I IP R ANSWERS BT GIVING NUMBER OT PACKS OP CIGARETTES, RECORD VERBATIM. THEN MULTIPLY THE NUMBER OP PACKS BT 20 AND ENTER THE NUHBER OP CIGARETTES SMOKED. (IP NOT EVERT OAT.l You Benlioned that you have not sucked regularly for (TIME IN 0 . 1 1 ) . Did you ever stay off cigarettes for • longer period of time? No on the average, about how tuny cigarette* • day were you evoking at that time? INTERVIEWER! 36-37/ | | | 4S-46/ What brand of cigarettes did you usually smoke just before you stopped ssnklng cigarettes regularly? (IF MORE THAN ONE BRAND OR NO REGULAR BRAND MENTIONED, ASKi Which one did you smoke the most? OFFICE USE ENTER BRAND _ - 34-3S/ No regular brand ........................... 996 47-49/ �I > R « K 52 17. 2O. For how lonq did you smoke this particular brand? ENTER DAYSt I I I 50-St/ OR WEEKS 1 | | | 52-S3/ OR MONTHS I | | | S4-55/ OR YEARS 1 | | | During the period when you were smokint the most heavily on a regular basis, about how many cigarettes did you usually smoke in a day? 56- 57/ ENTER NUMBER I (IF NOT EVERY DAYi) | I 61-62/ | | PER DAY | 6J-64/ | | PER HONTH OR UF NOT EVERY DAYi) IB. A. Hhat type of cigarettes wore they? I | Filter tip or 1 Non-filter tip? SB/ 2 L_LJ_LJ Mo • CODE ONR B. Regular size 1 King site or... LJ_L_L_I Mo O UI 19. 7I-74/ Yr 3 21. CO 67-70/ Yr TO S9/ ..2 IOO Millimeter? | PER YEAR FROM CODE OMB A. | When was that? Now I an going to show you • diagram of different sire cigarettes. Please look at the picture of the (KIND or CIGARETTE IN p.ISA AMD O.IBB). How, considering yowr style of smoking—for example, how lonq you usually leave the cigarette In an ashtray or Just hold It In your hand—tell me the Busbar which indicates how such of the cigarette yon actually smoked. Mien you smoked cigarettes, how deeply did you usually inhale? you aayi Would As deeply Into the chest as possible Only partly Into the chest.. As far hack as the throat Hell back Into the mouth, or Section t I Section 2 . . 2 Just puff and don't really draw I t In at 3 DON'T KNOW Section 3 SO/ all. Section 4. CURRENT PIPE SMOKER SECTION 22. Doting your entire l i f e , have you smoked at least as many as SO plpefuls of tobacco? Yes Mo I (SKIP TO p.)S) 2 76/ �5} DECKS 52-53 2). 26. no you now seioke a I Yea (Continued) A. 77/ Mwn was that? FROM No (SKIP TO Q.26I 2 • I—LI—I—I No Yr 24. About how aany average elzed pipeful* of tobacco do you usually e»oke In a day? ENTER NUMBER OF PIPEFULS OF TOBACCO PER DAYi (IF HOT EVERY DAYlJ l| | | FBI MONTH 22-25/ TO LJ_U_I l__l__l 7B-79/ BB3IN DECK 53 tO-11/ Mo 27. OR Z6-29/ Yr Mien you exoke a pipe, how deeply do you usually inhale? Would you aayi As deeply into the cheat aa p o s a i b l e . . . . . . . . . . . . 1 3O/ 12-I3/ (IF NOT EVERY DAYl) Only partly into the cheat 25. Foe how many years have you been Mtokiit? (NUH0BR I* 0.34) lpipeful.1 per Aa far day/per wocvth/pec year)? .....2 back aa the throat 3 Hell back into the anuth, or 4 Juat puff and don't really draw I t in at a l l . . . . 5 DON'T KNOW cri (SKIP TO O.35) ...OJ 2^ R CURD FORMER PIPE SHOKEP SECTION ...OS 2B. Row long haa it NUMBER) ENTER DAYS i OR WEEKS i •tore than 40 year*. .»•.. been aince you emoked a pipe f a i r l y regularly? (RECORD U_l I_U 31-32/ S3-34/ OR MONTHSi 26. ENTER NUMBERI ( I F NOT EVERY D A Y i I I—l__l f | | rB * »¥ NOT EVERY OAYl) (SKIP TO 0.35). 39/ .1 I8-I9/ 29. OR (IF 37-3B/ NEVER SMOKED REGULARLY D | PER MONTH 35-36/ OR YEARS i During the period whan yon were aaoklng the mat heavily, about how •any pipafule of tobacco did you usually evoke in a day? 2O-2I/ On the average, about how «iany pipefula of aanklna at that tla«7 tobacco a day were you ENTER NUMBER OF PIPEFULS OF TOBACCO PER DAYi (IF NOT EVERY D A Y l J l| ( I F NOT EVER! O A Y l ) l| | | PER MONTH | | PER YEAR | | | 4O-41/ 42-43/ OR 44-4S/ �-12330. DECK SI DECT 53 14. For how Ions did you smoke (NUMBER IN p.29$ (plpetuls of tobacco per day/per week/per month)? , Ixtss than 2 years Ol 2-5 years 4S-47/ Mien you s«oked a pipe, how deeply did you usually Inhale? say. Mould you As deeply Into the chest as possible. O2 7I/ Only partly Into the chest 6-IO years.................03 HAHD CARD • T As far back as the throat 11-15 yesrs O4 16-2O yesrs OS Juat puff and don't really draw It In at all....S 21-25 years 06 DON'T KNOW 26-3O years O7 31-3S years O8 3»-4O years..... O9 4O or more years..... It. 12. IS. to 1 Yes (SKIP TO 0-311 | | | During your entire life have you smoked at least as »any as SO cigars? T 4B/ 3«. 2 > No....(SKIP TO O..S1) Do you now smoke clgsre? No ' 72/ 2 73/ TSS Haw long did you not smoke a pipe at that time? ENTER DATS t 8 CURRENT CIGAR SMOKER SECTION You mentioned that you have not smoked regularly for (TIME IN Q.3O). Old you ever not smoke a pipe for a longer period of time? Ho Hell back Into the muth, or (8K1F TO Q.42). 49-50/ OR WEEKS I .L_I_J SI-SV «« MONTHS. L_LJ S3-54/ OR YEARS I L_I_J SS-56/ 37. On average, about bow many cigars e day do you now smoke? ENTER NUMBER OF CIGARS PER DAIl ( I F NOT EVERT D»Tl » | | | »H_I *** MONTH 74-7S/ 76-77/ OR 3). During the period when you Mere smoking the most heavily on basis, about hew many plpefals of tobacco did you usually day? pl * °*T ENTER NUMBER I L_LJ (IT NOT EVERT DAYl ) •(_ |_ | PC* HONTH regular .• In a S7-58/ 59-60/ OR (IF HOT EVERT DAIl) K. t|_l_J PER TEM 61-62/ U_LU 61-66/ Mhen was that? Ho Tr TO I_L_U_I Ho Tr 67-70/ (IT NOT EVERT DMl> l| ( | PER TEAR 78-79/ �BEGIN OFCK 54 30. For how many years have you been anoking ( I in O.37) cigars per day/per month/per year)? Less than 2 years.. .Ol 2-5 yearn 11-15 years 16-2O years 21-25 Nhen you smoke cigars, how deeply do you u s u a l l y inhale? sav: As deeply into the chest as possible IO-II/ .02 6-IO years 4O. Hould you ............ 1 As far back as the throat ....................... .04 Hell back into the mouth, or .05 , ...................... .03 years Only partly into the chest 26/ Just puff and don't really draw it in at a l l . ...5 DON'T KNOW..... O6 26-JO years O9 Hare than 40 years 8 08 36-4O years ................................. 3 4 O7 31-35 years..- .................... 2 1O 41. What type of cigars do you usually smoke? Filter tip or... Non-filter tip?. 39. During the period when you war* smoking tike most heavily on a regular basis, about how nany cigars did you usually smoke In • day. am* NUMBER, (IF NOT EVERT DAI I ) | | ) pen Mr ij | | PKH MONTH A. 12-13/ Now I an going to show you a diaqran of different size cigars, please look at the picture of the (Kim or CIGAR IN 0 . 4 1 ) . Now considering your style of smoking—for example? how lonq you usually leave the cigar in an ashtray or just hold It In your hand—tell me the number which indicates how Much of the cigar you actually snoke. 14-IS/ Section I I Section 2 28/ 2 O« (I F HOT EVERT DM 11 'LJU *** TIUUI 16-17/ Section 1 A. 3 Section 4 Mien was that? 4 ntOM LJ_L_LJ SKIP TO Q.S1 1B-2I/ FORMER CIGAR SMOKER SECTION TO LjLLLJ Ho 22-2S/ 42. How long has it been since yousmoked cigars fairly regularly? Tr ENTER DATSi OR WEEKS. OR MONTHS OR TEARS L_I_J L_LJ L.LJ L_U NEVER SMOKER REGULARLY 29- JO/ 3I-32/ 33-J4/ 35- 36/ (SKIP TO p.5l)....t 37/ �47. 43. On the average, about how many ctqars a day were you smoking at that time? What type of ciqars did you usually smoke jusl tiefnrf you slopped smokinq ciqars reqularly? COOK ONE ENTER NUMBER OF CIGARS PER DAYi (IF HOT RVERY DAYi 1 l| | | 3B-39/ | F i l l e r lip or N o n - f i l l e r tip?. | | PER HOHTH 4O-4I/ | | PER YEAR 42-43/ OR (IF MOT EVERY D A Y i ) 44. l| 4fl. Pal how long did you smoke (HUMBFR PER DM IN O.43) cigars per oay7 Less than 2 years O1 2-5 years HAND I CARD R I * Section 1 16-20 years 21-2S 07 31-35 years OS 36-4O years O9 More than 4O years 4 During the period when you were smoking the most on a regular basis, about how many cigars did you usually smoke in a day? 1O ENTER NUMBER I ( I F NOT EVERY DAYi I a) a* 56/ O6 26-3O years 49. 3 Section 4..... OS yeara 2 Section 3 04 1 Section 2 44-45/ O3 II-IS years 5V Mow 1 am qoinq to show you a diagram of d i f f e r e n t nite ciqars. Please look at the picture of Ihe (KIND OP CIGAR IN O - 4 7 ) . Now, constderinq your a t y l e of aMokinq—for example, how long you usually leave the cigar in an aahlray or just hold it in your hand—tell ne the number which indicates how much of the cigar you actually moke. O2 6-10 year*.. 1 | | | PER DAY 11 S7-58/ | | PER MONTH 59-60/ | | PER YEAR 6I-62/ OR ( I F NOT EVERY DAYiJ A. 45. You Mentioned that you have not smoked regularly for (TIME IN Q.42). Did you ever stay off cigars for a longer period of time? 1 Yes l| When was that? FROM 46/ LJ_LJ_I Mo Mo {SKIP TO O.47J 63-6S/ Yr t TO 46. LJ_L_L_I How long did you stay oft cigars at that time? Mo ENTER DAYSI L_U 49-SO/ Oil WEEKS I OH MONTHSI OR YEARS I 47-4S/ L_U L_LJ SI-52/ S3-S4/ Yr 67-70/ �DECKS 54-5S -130- 5O. When you smoked a clqar, how deeply did you unually inhale? aayi Aa deeply into the cheat aa poaalble Hould you ............ 1 .............. 54. Approximately how many houra a week are you exposed to this smoke in your home? It/ 2 10 houra or less. Only partly Into the cheat ....... . 11 to 15 hours... to far back aa the throat ....................... 3 16 to 20 houra... Hell back Into the wroth, or .................... 4 21 to 25 hours... Just puff and don't really draw It in at A l l . . . . 5 26 or more houra. DON'T KNOH ................ . ..................... B 55. 51. INTERVIEWER > DORS R CURRENTLY HHVK It SPOUSE OR PMTMER? COOED IN SECTION 5, p'a 2.6c or Be 7 ) . YES .............. . (IS ANY "MO* .......... NO ..... (SKIP TO O.5JI ....... t for how many yeara have you bean expoaed to amoke in this way? (CHECK ONLY ONE) 72/ 1 to 2 S 52. Does your (epouae/partner) YES Cigarette* 01 BAND CARD HO. DON'T KHOM 1 2 • 16 to 21 to 73/ Clqars 1 2 8 1 2 a O7 Nor* 74/ Pipe OO 11 to M .a reqularly any of the following? O3 to 7S/ Don't BEGIN DECK 55 53. Approximately how Mich amoka la there In the air In your hone? A lot I A little 5*. 2 Hone (SKIP TO 0,56) 3 INTERVIEWER! DOES It WORK? (IS "YES* COOED AT SEC 2. JOB* CODED AT SEC 3. O.1F?) p.4 1 Yea Ho A. OR "CURRENT (SKIP TO O.62) IO/ 2 Approximately how such amok* i* there in the air In the transportation you take to and from work (For example, your car, the train, the bus, etc.)? A lot 1 A little 2 Hone....(SKIP TO p.59) 3 I1/ �-III- 57. . DECK Approximately how many hours a week are you exposed to this smoke? -132- 55 61. OfXT 55 for how nany years have you heen exposed to thia smoke at work? 10 hours or lesa Lesa than 1 year Ot 2 I to 4 y e a r s * • . .02 16 to 2O hours 3 S to IO years.. .03 21 to 25 hours 4 II to 15 years. .04 26 or store hours............5 16 to 2O years. .05 21 to 3O years..... .06 •tore than 3O years. 58. I 11 to IS hours .07 For how nany years have you been exposed to this smoke? Less than 1 year Ol 1 to 4 years.... O2 S to IO years HAND I CARD » I 1C to 20 years OS ...O6 O? Don't Know M Approximately how such eeoke Is there In the air where you work? A lot 1 A little None 60. (SKIP TO 0.621....J Approximately how many hours a week are you exposed to this smoker I tl to IS hours 2 1C to 20 hours Ihere are sope questions that are asked in survey research that are difficult to ask directly because Many people think they are too personal, while it is understandable that people feel this way, there Is a real ne.ed for the information for the population as a whole. He now have a way that Mkea it possible [or people to qive infcreation, without telling anyone about their own situation. Let aw show you how this worksi we will use the next question 1 have here as an example. HAND II CARD X. As you see. there are two questions on the card. One deals with the "real* question that the research is concerned with, the other Is completely unrelated. Both questions can be answered *yea* or *no.* n>e of the two questions is selected by chance and you anawer it. (I'll show you how thai works in a minute). I do not know which question you are answering. Mien all the questionnaires have been tallied, the researchers can tell how many people have smoked marijuana, but they have no way of knowing whether It was you or sny other person In particular who has smoked marijuana. please take the coin that you have been handed and flip it now. Don't tell me which side came up. If the coin shows heads, please answer only question t. If the coin shows tails, please answer only Question 2. I won't look to aee If the coin comes up heads or tailsi and you don't tell me which question you are answering. Just tell me if your answer is "yes" or "no". Te 1 IV 4 2C or cure hours........... ....S 16/ It Is very simple, as you will see. YOU will flip the coin, the question you will answer is selected by chance. In no way can a truthful answer prove harmful to you. There ia no identifying Information that can link you to your answers. 3 21 to 25 hours 17-1H/ HAND II COIN IS/ 2 10 hours or less HAND I CARD V I 62. HAND I CARD I 1 | O4 More than JO years 59. 13-14/ OJ it to II years 21 to 3O years ea 12/ Ho Don't Know. A. Now let's do that again, using the next question. HAND R CARD y. Flip the coin again. If the coin turns up heads, please anawer only question number 1. If the coin comes up tails, please anawar only question nwber 2. Don't tell me the question. la your answer "yea" or "no"? Tee No Don't Know. 20/ �S5 UBCK $5 63. 63. Have you ever been arrested tt)i * felony? Yes I Mo....(SKIP TO Q.64> (Continued) H. INTERVIEWER! HAS R EVER BEEN CONVICTED OF A THIRD FELONY? IN O.638 EQUAL TO 3 OR MORE?) 2 21 / Yes....(GO TO NEK OUEX) 1 No A. (IS I JB 2 Have you ever been convicted of a felony? 1 Ye* Bo (SKIP TO Q . 4 .6) 22/ 2 Next, I'd like SOM information about drinking alchollc beverages. B. How Many felonies have you been convicted of ? 64. ENTER HUMBERt C. | (_ | I3-24/ No 2S-2./ • (SKIP TO SECTION 9) JV 2 • 65. D. 1 Yes Hhat- Month and year were you convicted of this/your first) felony? l_L_LU Ho Yr Nave you had any alcoholic beveraqea. Including beer, wine, or liquor, since (DATE OF LAST INTERVIEW)? Since (DATE OF LAST INTERVZEN) have you had a drink of beer? 1 Yes On what charge were you convicted? No 29-JO/ 66. .(SKIP TO O.7II 40/ 2 Mov long has it been since your last drink of beer? ca t. INTERVIEHERi HAS ft BVER BEEH CONVICTED OT Ik SECOND FELONY? IN 0.66B EQUAL TO 2 OK NORC?) Yes Ho....(8Kir TO Q.64) r. I I 2 jl/ Mist mnth and year war* you convicted of this/your first) felony? U_l_U Ho G. (IS 32-J5/ Yr On what charge were you convicted? 36-S7/ 67. As you think back over the period of tlM between (DATE OF LAST INTERVIEW) and now, about how nany cans or hottlee -of beer would you drink on a typical day when you drank beer? ENTER NUMBER Of CANS OR BOTTLES I | | | 43-44/ �OFCS442BSOC-8 68. Main, thinking back over the period of tine between (DATE or LAST INTERVIEW) and now, about how regularly did you drink beer? PROBR IF HECESSARii it's BOBetUw* hard to reveaher. Jus I give me your beat quean. More often than once • day Every day 01 , 71. Since (DATE or LAST INTERIVEH) have you had a drink of wine? Yes 4S-46/ O2 72. 1 No. . . . ( S K I P TO p. 76) 2 How long has it been aince your laat drink of wine? 5 or 6 day* • week.........................•••...03 .02 8-14 day* ago .03 1S-3O day* ago .04 I Month aqo. . . . . . . . . . OS .01 1-7 day* ago O4 1 or 2 day* • week Today .05 2-3 Month* ago J or 4 day* • week.... .06 Lea* often than one* a week......................O4 IF CANNOT OBCIOEiDoa't know 69. 96 Haw large were the can* or bottle.* th*t you usually drank? 4-6 Month* ago....... 47/ 16 ox. (half quart) can* or bottle*. SI-S2/ .07 7-12 Months ago .OB More than 1 year ago. Standard I X ox. can* or bottle* 50/ .09 32 oc. (full quart) can* or bottle*. Le** than 12 ox. can* or bottlea.. More than 32 oc. can* or bottle* 70. 9 Don't drink can* or bottle* of beer 73. 4 A* you think back over the period of tiM between (DATE OF LAST INTERVIEW) and now, about how Many glasses of wine would you drink on a typical day when you drank wine? « ENTER NUMBER OP GLASSES: | During the laat 12 Month* that yon drank el nee (DATE OP LAST INTERVIEW), how often did yon have • or nor* can* of beer IB a *lngle day (3 quart* or •ore)? Every day or nearly every day O1 3-4 tlaws a week 74. 48-49/ Every day OS 3-6 tin** a year Once or twice • year.... Never 3 or 4 day* a w e e k . . . . . . . . . . . . . . . . . . . . 4 1 or 2 day* a week O6 .............O7 OB 2 S or 6 day* a w e e k . . . . . . . . . . . . . . . . . . . . 3 O4 ti«*s a year 53-S4/ Hore often than once a d a y . . . . . . . . . . . . 1 Once or twice a wa*k.............................O3 7-11 | Again, thinking back over the period of tin between (DATE OF LAST INTERVIEW) and now, about how regularly did you drink wine? PROBE If NECESSARY! It'* coMtlMea hard to reaenber. Just give Me your best one**. O2 1-3 tlM** a Month | 5 Leva often than once a week...........6 IP CANNOT DECIDE! Don't know 6 55 �OFCt442RSec-R 75. DECK 55 Durinq the last 12 months that you drank since (DATE or LAST INTERVIEW), how often did you have B or More qlasaea o( wine in a slnqle day (more than a f i f t h ) ? Every day or nearly every day Ol 3-4 times a week. DECK 55 79. 56-S7/ Again, thinking back over the period of time between (DATE OF LAST INTERVIEW) and now, about how reqularly did you drink hard liquor? PROBE IF NrCFSSARYi It'a sometimes hard to remember. Just qlve me your best queas. More often than once a day.. Once or twice a w e e k . . . . . . . . . . . . . . . . . 0 3 Every d y . . . . . . . . . a.......... 1-3 time* a month O4 S or 6 day* a week 7-11 OS J or 4 days a week O6 1 or 2 days a w e . . . . . ek..... once or twice, a year.................07 HAND CARD O2 Less often than once a week. time* a year CC 3-6 time* a year Never 76. 08 IF CANNOT DCCIDEl Since (DATE OF LAST INTERVIEH) have you had a drink containing liquor, such as whiskey, vodka, qln, brandy, etc.? Te* 1 No...(SHIP TO 0.81) 0O. 63/ Don't know. About how Many ounce* of hard liquor are there in the drinks that you usually drink? One ounce (one ahot) 2 58/ 1.5 ounce* (one jlqqer) 1 64/ 2 CO 2 ounces (2 shota) 3 3 ounce* (2 Jlqqer* or 3 shots) 4 4 ounce a (4 shots) 77. 5 How long ha* it been ainc* your la*t drink of hard liquor? Today Ol 1-7 day* aqo S9-6O/ .....O2 5 or aore ounces It or sure ]igqers)..6 8-14 day* aqo... 03 Don* t know. 1S-3O day* aqo I month aqo OS it O4 2-3 snath* aqo O6 4-6 month* aqo... 81. O7 Durinq Urn Isat 12 south* that you drank since (DATE or LAST IHTMIIVCH), how often did you have • or sore drinks of hard liquor in a • inqle day (a half pint or sore)? Hare than 1 year ago. 7«. 3-4 times a week 09 As you think back over the period of tiam between (DATE or LAST INTERVIEW) and now, about how many drinks of hard liquor would you drink on a typical day in which yon drank hard liquor? ENTER HIMMEK Of DRINKS I | ( | 61-62/ .02 Once or twice a week. OS .01 .03 1-3 tt«e» a Month 7-12 months aqo Every day or nearly every day. .04 7-M tines a year .05 3-6 times a y a . . . . . . . er....... .06 Once or twice a year .07 Never .OB 65-66/ �OFCi4428Sec-8 62. -MO- Have you had a drink of beer, wine or hard liquor Months? In the laat 12 8S. INTERVIEWER: ...... (SKIP 10 SECTION 9) ........... 56 HAS R WORKED THE PAST YEAR? 10/ YES...(ASK A THROUGH E). 67/ No BEGIN DPCK 2 NO...(SKIP TO 0.86) 81. About how often during the past 12 Month* did you drink enough to feel high — that la, happier or More carefree than usual, Maybe a little flushed or dl»y. but not drunk, for More than 24 hours In a row? Durinq the past yean No Yes S or More tinea ....... . .............. 01 4 tlana .............................. O2 HAND CARD 3 tiMss .............................. 03 EE 2 tlawa .............................. 04 Once ................................. OS 84. Have you atayed away froM work because of a hangover?..1 2 !!/ Have you qotten drunk when on the Job? 1 2 12/ drinking? < 2 IV D. Has drlnkinq led to your quitting a job? 1 2 147 E. Has drinking hurt your chances for promotion or .1 2 IS/ C. .07 Have you lost a Job, or nearly lost one, because of .06 Merer In My life. A. B. Never In the past year, but lMS before that 00 «8-69/ Now I would Ilk* to ask you scan questions about experiences that Many people have had with drinking. During the paat year. . . Tea U) A. raiaea or a better Job? Ho 2 7/ 0 •«. when you were growing up, do you think your father drank occasionally, drank frequently, bad a drinking probleM. or didn't he drink? B. Have you gotten Into a heated argument while drinking?..! 2 " / C. 2 T2/ 2 «/ 2 •>*/ 2 7*7 2 7«/ 2 tl/ Drank occasionally 1 2 ?/ • Drank frequently 2 Had a drinking probleM 3 Didn't drink 4 DON'T KHOH 8 D. Drank occasionally Drank frequently 1 Didn't drink 4 DON'T KNOW E, Her* you afraid you Might be an alcoholic or that 2 Had a drinking probleM Have you deliberately tried to cut down or quit 1 8 F. One* you * tar ted drinking, waa It difficult for you G. Have you awakened the next day not being able to reMenber thing* you had don* while drinking? H. Have you often taken a drink the ftret thing when I. Have your hand* abaken a lot the Morning after J. Have you eoMetlaoe gotten drunk when drinking by 1 2 K. •/ O When you were growing up, do you think your Mother drank occasionally, drank frequently, had a drinking probleM, or didn't ahe drink? 7»/ 2 •7. Have yon sonatinas kept on drinking after I7/ �OFC>442BSec-9 OFr>442BSrc-ll> SFCTIOM lOt SECTION 9i t. RECREATION, LEISURE, AND PHYSICAL ACTIVITIES Now we would like you to answer some questions about your leisure time activities. INTERVIEWER! HAND R SELF-ADMINISTERED FORM 1. Have any of the recreation, leisure, and/or physical actlvitlea you've participated In since (DATE OF LAST IHTrftVlEW) brought you in contact with any of the following substances . . . TE8 HO Listed below are a serlee of Leisure Time and Physical Actlvitlea. Related activities are grouped under general headings. I) Please read the list and circle number 1 for each activity you have performed In the laat 12 Months, and circle number 3 for those you have not. 2) Check which months you performed those activities. 3) Record the average number of days per month you performed those activities, and 4) Hecord how many hour* and Minutes you performed those activities, on am average day. 1 28/ -M2TOXIC SUBSTANCES FOR EACH SUBSTANCE COOED YES, ASK A THROUGH D. Since (DATE OF LAST INTERVIEW), In what Month and year did your recreation, leisure and/or physical activities first bring you In contact with (SUBSTANCE)? com MONTH Since (DATE OF LAST INTFJIVIEH), for how nany years did you continue to come In contact with (SUBSTANCE)? I I I years YEAR W-12/ 1. Have you participated three or MOT* tlmei 1 in (READ EACH ITEM),* Yee Scuba diving 1 Auto, boat, or Motorcycle racing t 2 industrial chemicals?.....! 1 IV IS/ ED ,..,. CD-CD MONTH YEAR 40-41/ 2O/ 2 2I/ 1 Plane racing or plane acrobatics, not Including flight training or any assignments for the Armed Forces 2 1 Hang gliding 2 22/ Skiing fast down a high mountain slop* 1 2 2 2S/ 1 2 2C/ 41-46/ I I 47-4B/ 24/ 2 I YEAR 42/ 23/ 1 Sailing long distance In email sailing craft . corn Insecticides or MONTH 1 Surf board riding INTERVIEHERl !/ • 1 Mountain climbing 2. 2 16- 19/ Skydiving W No degreaslng chemicals? 1 ' 49/ CD-CD MONTH years YEAR SO-Sl/ 54-S5/ HAS R ANSWERED "IBS* TO AMY QUESTIONS OM BELT-ADMINISTERED FORM 3 AMD TO 0.1 ABOVE? defoliants or I -! i MONTH MO years YEAR S7-60/ S6/ (SKIP TO 8BCTIOM I * . . . 2 D.... X-ray or nuclear . 61/ mm OD — 1 MONTH YEAR 61-63, 64-67/ 68-69 �I1FCKS 56-5 J OU'.ltiJHK&t - I I I (Continued} Since (DATE OF LAST IHTERVIFH), bow many days per year did you cone in contact with (SUBSTANCE)? On the days you came in contact with (SUBSTANCE) how often did you use protective clothing or qear or wash to remove (SMISTANCE)—all at the time, •one of the time, or never? Which of the following did you use? CODE ALL THAT APPLY. IV HED 75/ 77/ 78/ 70-72/ Self contained or auppli«d DATS SO/ BEGIN DECK M/ »3/ 57 is/ f«c^ vhleld All of UM tiM.{J.8K El«..1 3 IO-12/ 2O/ 2*a/ (ace •Hivld. .**•••*.. .**».] All of the tin. (ASK B>..*1 26/ 27/ 21-23/ -~J 29/ VJ1 Self contained or *upplled 3O/ 36/ J7/ 3V All of the ttM»(MK E)***1 39/ 32-J4/ Self contained or anpplted DAYS . 46/ All of the tine. (ASK B|..*1 ODD 41-45/ DAYS 47/ 40/ face ehield*.* ••.••»•••• ..3 «/ 5O/ 5I/ Seir contained or •applied 53/ se/ «7/ 59/ 6O/ 54-S6/ *2/ DATS Self contained or • applied air breathinq apperatne.. .6 NANO K SELF-AOrllNISTERFD FOKM 4 AND GENERAL PURPOSE ANSHER SHEET 16/ 17/ wachi.»q r«ciltti««... 5 Seir contained or aupplted 24/ INTER VIEWER I 63/ The Jenkins Activity Survey aaka questions about aspects of behavior that have been found helpful in Mdlcal diaqnosls. Each peraon la different, ao there are no "right* or 'wronq* answers. Por each question, choose the anawer that la true for you. on your answer sheet, fill 1» the circle below the letter of your answer. Use a black lead pencil, and Bake your Mrka heavy-and dark. Hark only one anawer for each question. If you change your Hind, eraae the old vark completely. �DECK 57 OFC:442HK«c-ll SECTION 1 1 1 2. INCOME Did you earn any income froei any Job during 1964? from retirement plans or pensions. Do not Include Income How I have acme questions about your incoM. Yes (ASK A) I 1. No (SKIP TO 0.3) 2 Please tell M which letter on this card best represents the total household income In 1984 before taxes or other deductions (or all people in your household, not including roomers. Itils Mount should include vages, net income llom business, interest, dividends, pensions, and any other money Income. Toll me the letter that COMS closest. A. In which of these groups did your earnings from jobs in 1984 f a l l — that la, before taxes or other deductions? Tell M the letter that come closest. (25,000 - f29,M9 OS »3O.OOO - (34,999 06 (35,000 - (39.999 07 (40,000 - (44,999 08 OS r. »JO,000 - (34,999 OC $3S,OOO - 139,999....... 07 H. (40,000 - 144,999 OB HAND CARD I. (45.000 - (49,999 09 rr J. (5O.OOO - $54,999 1O K. (55,000 - (59.999 11 L. $60,000 - $64,999 12 H. (65.000 - ««9,999 13 H. «70,000 - »74.999 14 " C. ea $25,000 - $29,999 04 B. .04 H. (20,000 - (24,999 $20,000 - $24,999. 03 D. .03 G. (15,000 - (19,999 $15,000 - (19,999. 02 C. 69-70/ .02 F. * 10.OOO - (14,9*9 $10,000 - $14.999. E. B. S6-67/ .01 D. Ol $5.0OO - (9,999... C. $5,OOO - (9,999 A. B. A. 68/ HAND (65,OOO - (69,999 13 $7O,OOO - (74,999 14 $75,000 - (79,999 15 P. JBO.OOO - (84,999 16 0. $85.000 - (89,999 17 R. S9O.OOO - (94,999 18 S. $95,OOO - $99,999... 19 T. $1OO,OOO or Bore 2O 17 $9O,OOO - (94,999 12 16 H. (60,000 - (64,999 IS $85,000 .- $89,999 11 O. Q. (55.OOO - (59,999 N. $80,000 - $84,999 IO H. P. O9 L. $75,OOO - |79,999 (4S.OOO - (49,999... (50,000 - (54.999 K. rr O. I. J. CAM) 18 8. $95,000 - (99.999 19 T. (tOO.OOO or sore 2O 3. INTERVIfMeRt 71-74/ AH RECORD TIME ENDED �OFCi4428REMKRK DECKS S7-5H INTERVIEWER REMARKS IHTERVIEHBRt OFCI4428RKHARK 6. Please complete these reurka as soon as you have finished the questionnaire. List the questions that confused, angered, or caused discovfort to the respondent or questions that you feel the respondent did not answer truthfully. EXPLAIN. NONE I. Lenqth of the interviawi (Section 1. p.1 throuqh section 12) 2O-21/ BBC IN DECK SB 2. C. Date ol the interviewi I_U L_LJ DM TKM 3. ?. ! Black 2 Other S2-34/ Describe Problem Race of Respondent! Nhlte 27-29/ JO-JI/ 7S-77/ 22-24X 2S-26/ L-LJLJ Himires List questions with skip errors, questions that were confusing to you, or questions that otherwise didn't work. EXPLAIN |6/ 1 NONE SECTION o. In general, -hat was the respondent's attitude toward the Interview? Friendly and interested I IT/ 39-41/ 42-43X 44-46/ C. 4. 37-38Y 47-487 49-5I/ Describe Problem S2/ Cooperative but not particularly interested ..........2 {•patient and restless.... Hostile 5. ] 4 0. 9. In general, was the respondent's understanding of the questions.... Good? 1 Fair? 2 IB/ Please record your interviewer ID li Please sign your nanei 5 3-SB/ �•one 4428 5/85 AIR FOBCB BKALTI BOBTXT roni Complete tale for* and enoloee m copy with eaoh oaaa Balled to Chicago. Int. 10 * Int. **m»i Date Halledi vounm It mot eaoloaad. a>plala ioa Bbvat Chlldraa** Record For* BuppleMBtary ChlldraK'e Record Fora Queetlonaalre Seir-AdBialetered Fora I ~~i Seir-ktelnlatered Fora * oo eeir-Aomlelexered For* J Medical Cooeeat Feral FOB OTFIC* on ont CIROU ' Date recelred la Chloafoi »1I required f«ra» vreeeati Tee �APPENDIX C* Physical Examination Methodology *0riginal forms were color coded; limited photocopy quality. �5) 0 'D 'D ® < 0 0 (D 0 '$ ® 0 0 ® 0;.l* 0 (D ® ••'*.' -I' VERSION 1.0 JRW:SCF 585 FAMILY HISTORY - PLEASE BLACKEN THE CIRCLE FOR ANY FAMILY MEMBER THAT HAS HAD ANY OF THE FOLLOWING: BLOOD RELATIVES -» NONE MOTHER FATHER SISTER BROTHER CHILD REVIEWER'S COMMENTS: O 0 0 DIABETES EPILEPSY CANCER HIGH BLOOD .PRESSURE O HEART DISEASE STROKE ALLERGY STOMACH TROUBLE O O O /— . f~\ NERVOUS TROUBLE w BLOOD DISEASE DEFORMITIES O O0 ARTHRITIS OTHER FAMILIAL DIS. '-» PLEASE UST HERE: Q © O O O O ;-> 6 c /""; O O O O O 0 O O O C'! O "••** 6 o '•-^' 0 .•'~v, 6 r-i '~^' o o O o o o o 0 o o o o o p) o o o '• 2 o • o •^^' o o o o o o c o o^ • o 0 0 o ; O • o -^ 0 Q •-. CURRENT FAMILY STATUS FATHER: LIVING-AGE CONDITION OF HEALTH? O EXCELLENT DEAD-AGE MOTHER: LIVING-AGE SISTERS: / LIVING NUMBER ARE YOU MARRIED? CHILDREN: QFAIR OPOOR CAUSE OF DEATH? CONDITION OF HEALTH? O EXCELLENT QGOOD OPOOR CAUSE OF DEATH? DEAD-AGE BROTHERS: / LIVING NUMBER^ D£AD QGOOD AGES CAUSES AGES CAUSES NO. OF YEARS WIFE'S I AGE HEALTH _ OF WIFE? O EXCELLENT IF WIFE IS DEAD, PLEASE GIVE AGE, YEAR, AND CAUSE OF DEATH: BOYSAGES GIRLS' AGES DO YOU HAVE ANY PHYSICAL OR NERVOUS COMPLAINTS? ALL HEALTHY? QGOOO ANY DEAD? ANY BIRTH DEFECTS? DO YOU HAVE ANY ALLERGIES OR SEVERE REACTIONS TO MEDICINES, FOODS, PLANTS. CHEMICALS. ETC.? OYES ONO EXPLAIN: PLEASE DESCRIBE: NCS Trans-Optic ® EP01-21182:321 C-l A8900 �PERSONAL HISTORY VES NO YES NO NO YES ® ~ ® HEPATITIS ACNE ® ® WORMS EXCESS HAtt GROWTH ® ® COUTIS ® ® ARTHRITIS < ® ® SCLERODERMA CIRCLE NEXT TO ANY OF ® ® RHEUMATIC FEVER OTHER SKIN TROUBLE ® ® CANCER OR TUMOR THESE CONDITIONS THAT YOU NOW HAVE OR HAVE HAD IN THE PAST, ® ® HEMORRHOIDS ® ® KIDNEY STONES . ® ® KIDNEY TROUBLE ®®VARieOSE VEINS ® ® PHLEBITIS BLACKEN THE © OTHERWISE BLACKEN ® . ; ® ® HERMA. (RUPTURE^ REVIEWER'S COMMENTS: ® ® ANEMIA ® ® CATARACTS @ ® PROSTATE TROUBLE ®®T!ONStLmS ® ® SINUSITIS ® ® GONORRHEA ®®HAY FEVER ®®FAWTING @® FITS OR CONVULSIONS ® ® MUMPS ®®MALARJA ®®GOUT ® ® DEPRESSION ® ® DIABETES ® ® NERVOUS BREAKDOWN ® ® MEASLES ® ® DYSENTERY 0® POUO © ® ASTHMA ® ® BRONCHITIS ® ® PARALYSIS ® ® PNEUMONIA ® ® MUSCLE PAIN ® ® MUSCLE WEAKNESS TUBERCULOSIS ® ® HEART TROUBLE ® ® STOMACH TROUBLE ® ® NUMBNESS ® ® LOSS OF SENSATION ® ® LOSS OF SEX DRIVE , ® ® RHEUMATOID ARTHRITIS ® ® SEVERE ARTHRITIS ® ® JAUNDICE g>®UVeR TROUBLE 6RYTHEMATOSIS LIST THE AVERAGE FOR EACH OF THE FOLLOWING DURING THE LAST 90 DAYS: FOR THE PAST 90 DAYS OR MORE: DID YOU TAKE REGULAR EXERCISE? PER DAY NUMBER CIGARETTES HOURS SLEEP PER NIGHT ALCOHOLIC DRINKS DAYS WORKED PER WEEK WHAT IS YOUR USUAL WEIGHT? LBS WHAT IS THE MOST YOU EVER WEIGHED? HOURS WORKED PER DAY LBS CUPS COFFEE DAYS VACATION PER YEAR AT WHAT AGE? HAVE YOU RECENTLY LOST OR GAINED WEIGHT? CHEWING TOBACCO IF SO, HOW MUCH? (+/-) SNUFF PAST HISTORY PLEASE LIST PREVIOUS OPERATIONS, INJURIES, AND SERIOUS ILLNESSES, INCLUDING THOSE INDICATED ABOVE YEAR LBS PLEASE BLACKEN THE CIRCLE IF YOU HAVE HAD REPEATED CASES OF ANY OF THE FOLLOWING IN THE PAST YEAR: YES NO DESCRIPTION OF OPERATION. INJURY, OR SERIOUS ILLNESS ® ® PNEUMONIA ® ® KIDNEY INFECTIONS ® ® SKIN BOILS DO NOT MARK IN THIS SPACE WHEN WAS YOUR LAST I PHYSICAL EXAM? I DID YOU TAKE ANY MEDICATIONS OR TREATMENT NOW OR OCCASIONALLY ANY ABNORMALITY FOUND? • YES O NO NO PLEASE DESCRIBE: NAME YOUR PERSONAL PHYSICIAN YES ® ® OTHER INFECTIONS SPECIFY STREET ADDRESS CITY, STATE, & ZIP C-2 ARE YOU UNDER ANY MEDICAL TREATMENT NOW? YES NO �NAME: DATE OF BIRTH: CASE NUMBER (0: 1 2 3 * 5 8 7 3 9 •: 1 2 3 #' 5 6 7 8 9 C '* 'i 2 3 4 5 6 7 3 9 0 'O, 1 f! 3 4 5 8 7 3 9 5 6: .'7 •'§. 9 f FEB APR WAV '.8' f- MH .9; MAR — ' I - 'f E 3 , i, MO TODAY'S DAY DATE YR BRQUP « fll (0. J . , i 301 20; 86, 90) REVIEW OF SYSTEMS FORM AFHS-2A 'Oi :T ''2; '••.-* < -i., .»', '•( '?; (8, '•) S•• 3') 4 . _V •—• • „. j ... -..- • v^- ''^ '-,V_ VERSION 1.0 JRW:SCF 585 INSTRUCTIONS IF YOU HAVE ANY OF THE FOLLOWING COMPLAINTS, BLACKEN THE CIRCLE IN, THE "YES" COLUMN, IF NOT, BLACKEN THE CIRCLE IN THE "NO" COLUMN. THE DOCTOR OR NURSE WILL ASK ABOUT THE DETAILS LATER. ANSWER ALL QUESTIONS. IF IN DOUBT, GUESS YES OR NO. QUESTIONNAIRE YES NO ••'£}•'$ ANY FOODS THAT TEND TO DISAGREE (WHICH ONES?) ••I*ITCH OR RASH (WHERE?) 0® SWELLING, LUMP, OR SORENESS ANYWHERE ON BODY (WHERE?) ® ® NUMBNESS OR TINGLING (WHERE?) 0 ® TWITCHING MUSCLES (WHERE?) 0 (NJ GET UP NIGHTS TO URINATE - HOW MANY TIMES A NIGHT? REVIEWER'S COMMENTS: YES NO ® ® 1. SEVERE HEADACHES OB HEAD PAINS • COMMENT YES NO 0 ® 31. WORRIED ABOUT YOUR HEART 0 (5) 2. ANY DISTURBANCE IN VISION 0 ® 32. BLOOD PRESSURE TOO HIGH ® ® 3. PAIN OR DISCOMFORT IN EYES 0 ® 33. BLOOD PRESSURE TOO LOW ® ® 4 . WEAR GLASSES (OR CONTACT LENSES?) 0 ® 34. PAINS IN HEART OR CHEST 0® 5. CONSTANT NOISE IN EARS 0 (N) 36. POUNDING OR SKIPPING OF HEART ® ® 8- HARD OF HEARING 0 ® 36. HEART STARTS RACING SUDDENLY 0 ® 7 ® ® 37. SHORTNESS OF BREATH OR WHEEZING - EAR ACHE WITH COLDS 0 (N,"' 8. EAR ACHE WITH PLANE FLIGHTS .Y, ..N,; 38. TROUBLE GETTING A DEEP BREATH 0 ® 0 ® 39. SWELLING ANKLES 0 9. CHRONIC RUNNING EARS ® 1°- CHRONIC STUFFY OR RUNNY NOSE ; 0 ® 11. NEED TO USE NOSE DROPS FREQUENTLY Y) 'S) 40. LEG CRAMPS IN BED OR SITTING STILL 0 ® 41. LEG CRAMPS WHILE WALKING 0 ® 12. BAD NOSE BLEEDS AT TIMES 0 ® 42. PAIN OR TROUBLE WITH SWALLOWING 0 ® 13. FREQUENT SEVERE COLDS OR SORE THROAT 0 ® 43. POOR APPETITE RECENTLY 0 ® 14. ANY KNOWN DENTAL PROBLEMS 0 (N) 44. POOR APPETITE ALWAYS 0 ® 16. SORENESS OR BLEEDING OF GUMS 0 ® 45. NAUSEA OR VOMITING 0 ® 16. MORE THAN A YEAR SINCE TEETH CHECKED 0 ® 46. VOMITING OF BLOOD 0 ® 17. SORE MOUTH OR TONGUE ® ® 47. BELCHING. BLOATING OR INDIGESTION 0 ® 18 - GOITER OR THYROID TROUBLE 0 ® 48. YELLOW SKIN OR EYES (JAUNDICE) ® ® 19. THYROID TEST TOO HIGH 0 ® 49. BURNING OR HUNGER PAINS IN STOMACH 0 ® 20. THYROID TEST TOO LOW 0 ® SO. USE ANTACIDS FOR STOMACH BURNING ® ® 21. FEELING OF LUMP IN THE THROAT 0 ® 51. SORENESS OR PAIN IN STOMACH. ABDOMEN 0 ® 22. NEED TO TAKE THYROID MEDICINE 0 (Sj) 23. HOARSENESS AT TIMES 0 ® 52. SUSPECT ULCERS OR STOMACH TROUBLE t 0 ® 53. CRAMPS IN STOMACH OR LOW DOWN 0 ® 24. RECENT OR CHRONIC COUGH 0 ® 54. LOOSE BOWELS OR DIARRHEA 0 ® 25. CHRONIC COUGHING UP OF SPUTUM ® ® 5B. BLACK OR TARRY STOOLS isowa MovEMtm) 0 'Mi 26. EVER COUGHING UP OF SPUTUM 0 ® 56. FRESH OR BRIGHT BLOOD WITH STOOLS 0 'S) 27. ACHE AU OVER 0 ® 57. MUCUS (SUME OR PHLEGM) IN STOOLS .Y' ..N; 28. HAVING CHILLS OR FEVER 0 ' N ) 53. CONSTIPATION 0 0 Cjjl 59. USE LAXATIVES FREQUENTLY *'> 29. SEVERE SOAKING NIGHT SWEATS Y"' N 30. LIVED WITH ANYONE HAVING T.B. ;?• .N, 60. USE ENEMAS FREQUENTLY NCS Trans-Optic ® EP01-21 161:321 C-3 A8900 �REVIEWER'S COMMENTS: YES NO YES NO ® ® 81. RECENT CHANGE IN BOWEL HABITS 0® 91. NAIL BITING 0 ® 62. RECTAL TROUBLE OR PAIN ® ® 9Z SLEEP WALKING © (N) 63. PAIN IN THE KIDNEY REGION ©® 91 BED WETTING AFTER AGE 12 ® ® 64. BLOOD OR PUS IN URINE 0® 94 ® ® 66. ALBUMN IN URINE ©® 95. IRREGULAR LIVING HABITS 0 ® 66. SUGAR IN URINE ©® 96. CANT GO TO SLEEP OR STAY ASLEEP 0 ® 67. SPELLS OF FREQUENT URINATION ® ® 97. NEARLY ALWAYS IN POOR HEALTH © ® 68. SEVERE BURNING OR PAIN ON URINATION ® ® 98. CONSIDERED TO BE A NERVOUS PERSON © ® 69. PAINS OVER BLADDER OR LOW DOWN ® ® 99. FROM SICKLY OR NERVOUS FAMILY © ® 70. TROUBLE STARTING URINE ® ® 100. TREMBLE AND SWEAT EASILY ® ® 71. URINARY STREAM HAS BECOME WEAK © ® 101. HAVE TROUBLE MAKING UP YOUR MIND - - CHRONICALLY TIRED OR OVERWORKED © ® 72. HARD TO EMPTY BLADDER COMPLETELY ® ® 10Z EASILY MIXED UP OR CONFUSED ©®71LOSE CONTROL Of PASSWGUR9« 0® © ® 74. PAINFUL OR SORE GENITALS (PRIVATES) 0® 104. FEEL SAD. LONELY OR DEPRESSED 103. CLUMSY OR HAVE FREQUENT ACCCENTS , "®© 1 * ON OFTEN 0 © ® 106. WISH I WERE DEAD 76. STIFFNESS OF MUSCLES OR JOINTS . SEVERE PAWS IN ARMS OR LEGS ® ® 107, WORRY COWTNUALLY 78. PAINFUL FEET © ® 108. UPSET BY LITTLE THINGS 0® 10ft A PERFECTIONIST ® ® 80. PAINS IN NECK 0® 110. SENSITIVE OR FEELINGS EASILY HURT ® ® St. EASYTOSUNBURN? ® 111. OFTEN MISUNDERSTOOD © ® 82. SUBJECT TO ACNE 112. OFTEN ACT ON SUDDEN IMPULSE ®® 83, SUBJECT TO SOLS Ofl INFECTION © ® 84. SUBJECT TO ATHLETE'S FOOT, SKIN FUNGUS (V)® 114. FREQUENTLY KEYED UP AND JITTERY ® ® 88. SUBJECT TO HIVES OR SKIM REACTIONS © ® 116. EASILY SCARED BY SUDDEN NOISE ® ® 86. EASY BLEEDING OR BRUISING ® ® 116. HAVE BAD DREAMS OR THOUGHTS © ® 87. MOLE OR SORE WHICH IS NOT HEALING ® ® 117. SUSPECT A SERIOUS DISEASE OR CANCER © ® 88. SEVERE DIZZINESS ® ® 118. HAVING TROUBLE GETTING ALONG WITH SOMEONE AT HOME OR AT WORK ® ® 88. GENERALIZED WEAKNESS ® ® 90. MUSCLE WEAKNESS EXPOSURE HISTORY HAVE YOU EVER BEEN EXPOSED TO ANY OF THE FOLLOWING SUBSTANCES OR TYPES OF RADIATION? EXPOSURE IS DEFINED AS SKIN OR RESPIRATORY CONTACT OF MORE THAN ONE DAY'S DURATION. FOR EACH "YES" RESPONSE, PLEASE COMPLETE ONE OF THE THREE BLOCKS ON FORM AFHS-2B. 0®, COMMENT REVIEWER'S COMMENTS: YES NO YES NO © ® CHLOROMETHYL ETHER 0® COAL TAR © ® CREOSOTE © ® ARSENIC ® ® ANTHRACENE 0®CHROMATES ® ® BENZENE ©©ASBESTOS 0® CUTTING OILS ® ® NAPHTHYLAMINE ® ® TRICHLOROETHYLENE ©<3) AMWODIPHENYL ® ® ULTRAVIOLET LIGHT (OTHER THAN SUN) ® ® MUSTARD GAS © ® X-RAYS (OTHER THAN ROUTINE) ® ® VINYL CHLORIDE © ® IONIZING RADIATION (OTHER THAN X-RAYS) REVIEWER'S NAME (PRINTED): C-4 �CASE NUMBER B NAME OF PARTICIPANT j PAGE OF FORM AFHS-2B EXPOSURE HISTORY DETAILS 0 FOR EACH "YES" EXPOSURE AT THE END OF FORM AFHS-2A, PLEASE FILL OUT ONE OF THE FOLLOWING BLOCKS. USE ADDITIONAL SHEETS IF NECESSARY. TYPE OF EXPOSURE (COAL TAR, ETC.) WAS EXPOSURE RECEIVED ON THE JOB? YES NO YES NO YES NO YES NO IF ON-THE-JOB EXPOSURE, JOB TITLE IF NOT ON-THE-JOB EXPOSURE, HOW EXPOSURE RECEIVED CHECK FREQUENCY OF EXPOSURE THAT BEST FITS YOUR EXPERIENCE DAILY WEEKLY MONTHLY YEARLY TYPE OF EXPOSURE (COAL TAR, ETC.) IN WHAT YEAR(S) WERE YOU EXPOSED? WAS EXPOSURE RECEfVED ON THE JOB? IF ON-THE-JOB EXPOSURE, JOB TITLE IF NOT ON-THE-JOB EXPOSURE, HOW EXPOSURE RECEIVED CHECK FREQUENCY OF EXPOSURE THAT BEST FITS YOUR EXPERIENCE DAILY WEEKLY MONTHLY YEARLY TYPE OF EXPOSURE (COAL TAR. ETC.) IN WHAT YEARIS) WERE YOU EXPOSED? WAS EXPOSURE RECEIVED ON THE JOB? IF ON-THE-JOB EXPOSURE, JOB TITLE IF NOT ON-THE-JOB EXPOSURE, HOW EXPOSURE RECEIVED CHECK FREQUENCY OF EXPOSURE THAT BEST FITS YOUR EXPERIENCE DAILY WEEKLY MONTHLY YEARLY TYPE OF EXPOSURE (COAL TAR, ETC.) IN WHAT YEARIS) WERE YOU EXPOSED? WAS EXPOSURE RECEIVED ON THE JOB? IF ON-THE-JOB EXPOSURE, JOB TITLE IF NOT ON-THE-JOB EXPOSURE, HOW EXPOSURE RECEIVED CHECK FREQUENCY OF EXPOSURE THAT BEST FITS YOUR EXPERIENCE DAILY WEEKLY MONTHLY YEARLY C-5 IN WHAT YEARIS) WERE YOU EXPOSED? �CASE NUMBER GROUP I NAME: DATE OF BIRTH: VERSION 1.0 JRW:SCF 585 GENERAL APPEARANCE • • • • • • ^ • • ^ • ^ • • • W B ^ ^ ^ M M ^ ^ M^ ^ ^ M ^ ^ ^ i ^ ^ ^ APPEARANCE VS STATED AGE HAIR DISTRIBUTION APPEARANCE OF ILLNESS OR DISTRESS WELL NOURISHED YOUNGER THAN YES NORMAL OBESE UNDER-NOURISHED OLDER THAN SAME AS NO ABNORMAL I v N COMMENT DESCRIBE ANY ABNORMAL HAIR DISTRIBUTION: NOTE: FILL IN VITAL SIGNS WITH "0"» IF REFUSED. I VITAL SIGNS HEIGHT CM ~ rtft r«N 00© ®®® CD CD® WEWHT (UNOMSSeO) KG ""(?! TIMniWTUM OHM. T '971 ''Oj (pi Q\ arrow BLOOD MIESSOBE SYSTOUC oV?i -^ 0® 0 «® 96) *.'f} (2) 7$® 'j) ftCs^ d) DIASTOtIC 'Oi • Oi : ^ X0 T) /2 , (2 t "11 '31 (Ji @'D® VPB* 2) -,3^ ?( PULSE RATE MINUTE •'S ' • f ''n'- NONOOMWAWT AflM, Hf-AIIT Lim '0' • ' O i •Ti rf) T: C*) (|; 'ii '.I^X '3. 1": vj) ' 2 ', :2i 4: '4~; •'•*) .:*• PER •'3^ • 3 , 0® 8®® '90® ••£ .?: (i) Js'i iia •'» :• T. 9 S 5 5 5 S 6 ei 8 '8 8 8 7 7 120 7 7 7 7 7 7 7 7 a a 9 9" 8 a a a a 8 9 9 9 N COMMENT 7 8'- Y' •«; 8 '»•: DESCRIBE ANY IRREGULARITIES: S. 5 "8; iob i'» I '5* -•. iib e~'i PULSE IS: .' REGULAR IRREGULAR 9; 9 '<£' T? 3 io> :«'.:.•«• i; 104 7 ' 7 a a" a9' a • ! ' »,' '4) '-t 9 FUNDUSCOPIC NORMAL YES NO YES EXTERNAL NO X' N. f LIGHT REFLEX v N HEMORRHAGES T- N! A-V NICKING v N) EXUDATES Y N ARTERIOLAR SPASM * N PAPILLEDEMA NO N; ARCUS SENILIS PRESENT N DISK PALLOR r; YES r: N. f CUPPING ABNORMAL REFUSED v Y N, ABNORMAL OCULAR PIGMENTATION COMMENT DESCRIBE VASCULAR LESIONS, HEMORRHAGES, EXUDATES, OR PAPILLEDEMA: NCS Trans-Optic ^ EP01-21157:321 C-6 A8900 �C NORMAL 0 ABNORMAL YES NO Y• N. Y N) Y a, RIGHT TYMPANIC MEMBRANE INTACT LEFT TYMPANIC MEMBRANE INTACT NASAL ULCERATIONS C REFUSED •"Y: :N) COMMENT DESCRIBE ABNORMALITY: C NORMAL ;Y- C THYROID GLAND -;MI PALPABLE r >!' NODULES LEFT •'." ".'. RIGHT ": : PAROTID GLAND ENLARGEMENT CAROTID PULSE ABSENT CAROTID BRUIT PRESENT Y N COMMENT ABNORMAL 'Y N .: REFUSED DESCRIBE ABNORMALITY: ENLARGED Y N TENDERNESS THORAX AND LUNGS O NORMAL O EXPMATOW ABNORMAL C CHEST CIRCUMFERENCE (CM) AT NIPPLE LEVEL REFUSED YES : Y• NO NI ASYMMETRICAL EXPANSION •Y) N. WHEEZES Y N 'D ' ,S> DULLNESS Y ; a"* ;'<»'> N) HYPERRESONANCE .Y; WSWIATION 0® Csi ft) ,7 7 RALES (NOTE LOCATION IN BOX BELOW) DESCRIBE ABNORMALITY: (Y,I : N) COMMENT HEART O NORMAL ,_ C ABNORMAL YES O . J HEART SOUNDS NORMAL DISPLACEMENT OF APICAL IMPULSE PRECORDIAL THRUST NO O ~ O ABNORMAL HEART SOUND(S) NO YES 'Si 'N: N; N' SI S2 S3 34 Y Y Y Y Y N COMMENT REFUSED MURMUR .. NO YES SYSTOLIC AORTIC Y N DIASTOLIC Y N PULMONIC Y N Y CHEST AREA APEX Y: N> N Y DESCRIBE ANY ENLARGEMENT, IRREGULARITY OF RATE, MURMUR, OR THRILL: C-7 N MITRAL Y, N' Y N �CASE NUMBER (o,, 'r '.*'• V •'".?• (o "V '2 ' 3 , 4. 5 (?: 'r f 3 4 ?,. 6 1 0,:<1; GROUP NAME: DATE OF BIRTH: 8 9 C" 8: 9 0 2 3; : 4) 5'. 8, 7 f: •>. I .2:.3j i-I) 'a.;.? • a .•.;(. F (a> '.!• '. MO TODAY'S DAY DATE YR FES MAR 30 201 101 38i 86! APR JUN M. Aue .9 JAM 'f 6 5) •98 87! 91' set i' «r» *i riov OCT 3' 94' LilJ 2 96) PHYSICAL EXAMINATION (CONTINUED) FORM AFHS-38 VERSION 1.0 JRW:SCF 585 CM YES NO O NORMAL O ABNORMAL C REFUSED ®<3> HEPATOMEGALY '£; '"N) SPLENOMEGALY Y N' TENDERNESS LIVER r; .r' TENDERNESS SPLEEN OTHER TENDERNESS Y N OTHER MASS WAIST MEASUREMENT 1!!^^^ (10) lj!0l -301 LIVER SPAN Y' N; COMMENT DESCRIBE ABNORMALITY: EXTREMITIES YES NO YES NO O NORMAL 0® PITTING EDEMA O ABNORMAL ®® CLUBBING OF NAILS ® ® VARICOSITIES ® ® LOSS OF HAIR ON TOES RIGHT ® ® LOSS OF HAIR ON TOES LEFT 'Y; •'$ NON-PITTING EDEMA •'»} ABSENCE (SPECIFY IN •NJ COMMENT BOX BELOW) DESCRIBE EDEMA. SIGNS OF VASCULAR INSUFFICIENCY, OR ABSENCE OF PART OR ALL OF EXTREMITY: O REFUSED PERIPHERAL PULSES RADIAL FEMORAL POPLITEAL DORSALIS PEDIS POSTERIOR TIBIAL COMMENTS: NORMAL .•~\ *—' DIMINISHED ABSENT REFUSED O O 0 "> 6 o 6 O O O O O 'o* 0 Q o o 6 ® (s) NCS Trans-Optic ® EP01-21156:321 C-8 A8900 �MUSCULOSKELETAL O NORMAL SPINE MUSCLES O ABNORMAL PRESENT ABSENT O O WEAKNESS O O TENDERNESS O O ATROPHY O O ABNORMAL CONSISTENCY PRESENT ABSENT O 0 SCOLIOSIS o o o o KYPHOSIS 0 PELVIC TILT 0 DECREASED RANGE OF MOTION 0 "0 O REFUSED STRAIGHT LEG RAISING O NORMAL O ABNORMAL TENDERNESS LEVEL © ® COMMENT o TENDERNESS O CERVICAL THORACIC O O LUMBAR SACRAL COMMENTS: GENITOURINARY/RECTAL/HERNIA YES NO VES NO ; ;O" NORMAL O; ABNORMAL. "®®KT INGUINAL HERNIA ®®LT INGUINAL HERNIA ®®VARICOCELE ® ® HEMORRHOIDS ® ® PROSTATE ENLARGEMENT ®®EPIDIDYMIS ®® RECTAL MASS ®®SCROTALMASS ®®®®®®®@®® O REFUSED 0® COMMENT TSSTE5 LEFT RIGHT NORM. O O ABS. ENGL. ATROPH. O O 0 o o o (DIAMETER-CM) COMMENTS: NON PALPABLE ENLARGED O NORMAL O ABNORMAL O REFUSED ' TENDER HARD CERVICAL O O O O OCCIPITAL SUPRACLAVICULAR AXILLARY EPTTROCHLEAR INGUINAL FEMORAL O O O O O O O O O O O O O O O O O O O O O O O O OTHER TESTS ORDERED ® YES ® NO DESCRIBE: ® ® COMMENT PRINTED NAME OF EXAMINING PHYSICIAN SIGNATURE C-9 oooooool o LYMPH NODES CONFLUENT O o o o o 0 o �• E EM E I ^ r B I • Z l S HJ ^ t B ™ I 1 2 3 * 5 8 7 3 0 1 2 3 4 5 8 » 8 9 C 0 ~ 0 9 * 1 * 3 4 5 6 7 8 9 0 0 1 . 4 3 4 9 8 7 8 9 NAME: DATE OF BIRTH: MO TODAY'S DAY DATE YR SROUP 1 JAN FEB MAR APR MAY JUN JUt AUG SEP OCT MOV 30 20 10 9 8 7 6 3 4 3 2 1 86 88 87 88 89. 90< 91 9Z 93) 94 95 98 <tt> G FORM AFHS-4 ^ DEC DERMATOLOGIC EXAMINATION AND BIOPSY VERSION 1.0 JRWtSCF 585 ^^^^^^^^^^^^^^^^^^^ BLACKEN CIRCLE IF LESION IS OBSERVED REFUSED EXAMINATION INDICATE TYPE AND LOCATION OF LESION ON ATTACHED ANATOMICAL CHART YES NO YES NO c I PALMAR KERATOSES COMEDONES ACTINIC KERATOSES ACNEFORM LESIONS . ACNEFORM SCARS PETECHIAE » DEPIGMENTATION C ECCHYMOSES INCLUSION CYSTS O CONJUNCT1VAL ABNORMALITY : CUTIS RHOMBOIDALIS ORAL MUCOSAL ABNORMALITY HYPERPIGMENTATION FINGER NAIL ABNORMALITY JAUNDICE TOE NAIL ABNORMALITY SPIDER ANGIOMATA O PALMAR ERYTHEMA 1" SUSPECTED MELANOMA DERMATOGRAPHIA SUSPECTED BASAL CELL CARCINOMA -.". SUSPECTED SQUAMOUS CELL CARCINOMA BIOPSY YES SKIN BIOPSY PERFORMED NO Y N BIOPSY LOCATION(S): YES CONSENT FORM OBTAINED Y NO N BIOPSY DIAGNOSIS: PHYSICAL FEATURES EYE COLOR BROWN HAZEL GREEN GREY BLUE HAIR COLOR SKIN COLOR ;T; •'*) -i: ;4.-..s; a" 7" KM) 10: .20! 30) '401 .BOI Wl fa S& 3$ «' i- •.? 'I! ® & •'•) 0rD ( PRINTED NAME OF EXAMINING PHYSICIAN SIGNATURE NCS Trans-Optic ! EP01-21160:321 C-10 A8900 �CASE NUMBER NAME: DATE OF BIRTH: MO TODAY'S DAY DATE YR BROUP I MAR APR WAV JUM -it* \Q '?) '?' (§) g) ^ §) |) l WOV '§) HE« >"C' 87J ($CT AU£ 101 sSl 36) FE8 20! JAM .961 .10 X- :|) • 4, .3, i;£) 1 ;£) X . ® 0 CD -X1 •'«) *'*><?} (i! ••£'0 ' i (f) O .(ji (D0 FORM AFHS-5 VERSION 1.0 JRW:SCF 585 HEAD AND NECK YES NO ® ® NORMAL TO PALPATION/INSPECTION 0 0 © NEUROLOGICAL EXAMINATION NECK RANGE OF MOTION O NORMAL O ABNORMAL O REFUSAL NO 0 ® ASYMMETRY ® DEPRESSION ® SCAR YES ® ® ® ® ,Y> 0 0 IMOTOR SYSTEM HANDEDNESS O LEFT I^^^^^^^^^^H ASSOCIATED MOVEMENT ARM SWING GAIT O NORMAL O ABNORIV AL O REFUSED O RIGHT ® 0 0 O ® BROAD BASED ® SMALL STEPPED (S)ATAXIC DECREASED LEFT DECREASED RIGHT DECREASED FORWARD DECREASED BACKWARD T O NORMAL O ABNORMAL LEFT O ABNORMAL RIGHT (SPECIFY) REFUSED MUSCLE STATUS BULK TONE: UPPER EXTREMITIES LOWER EXTREMITIES STRENGTH: DISTAL WRIST EXTENSORS ANKLE/TOE FLEXORS PROXIMAL DELTOIDS HIP FLEXORS O REFUSED ABNORMAL MOVEMENTS NORMAL ABNORMAL O O O O O O O O INCREASED LEFT RIGHT O O O O ONO OYES > 0®@0 TENDERNESS ONO OYES —> 0 ® ® ® TREMOR ONO OYES-I i LOWER UPPER LOWER UPPER LEFT LEFT RIGHT RIGHT O O O O O O OTHER —» O REFUSED EQUIUBRATORY (ROMBERG) FINGER - NOSE - FINGER HEEL -KNEE' -SHIN RAPIDLY ALTERNATING MOVEMENTS PRONATION/SUPINAT10N OF HANDS RAPID PATTING DEEP TENDON REFLEXES O O O ABNORMAL RIGHT BOTH O O O O O O O O O O O LEFT O O O O (0 - ABSENT, 1 - SLUGGISH, 2 - ACTIVE, 3 - VERY ACTIVE, 4 - THAN SIENT CLONUS, 3 - SUSTAINED CLONUS) EXTREMITY O O O O O O : COORDINATION NORMAL TICS, CHOREAS, FASCICULATIONS DECREASED LEFT RIGHT O O O O O O O O O O O O O O O O O O BICEPS TRICEPS PATELLAR ACHILLES BABINSKI LEFT ® © ® ® ® ® ® ® <I) ® ® © ® ® ® ® © ® ® ® O PRESENT O ABSENT O RESTING NORMAL O RESTING ABNORMAL O ESSENTIAL NORM. SKILLED ACTS O ESSENTIAL AB. O INTENTION NORM. SPEECH (ARTICULATION, APHASIA, AGNOSIA) O INTENTION AB. O NORMAL O GROSSLY ABNORMAL ® ® ® ® RIGHT (« Q) © © ® ® ® 0 (< ) 0 ® ® © ® £)PRESENT O ABSENT O DYSARTHRIA O APHASIA O AGNOSIA NCS Trans-Optic s> EP01-21169:321 c-ii A8900 �1 MENINGEAL IRRITATION AND SENSORY SYSTEM •^^^•^•^••^••^^^^^^^^^^^^^^^^^^^^^"^^^ ••" ' " — - ABNORMAL LEFT STRAIGHT LEG RAISING RIGHT O NORMAL 0 O UGHT TOUCH O PINPRICK VIBRATION AT ANKLE (128 HZ) O O O O POSITION (GREAT TOE) O o O O 0 0 0 BOTH 0 o . '"0 0 ' O O COMMENT NERVE STATUS (TENDERNESS, TUMORS, ETC.): It^^mm^^m CRANIAL NERVES • • • I B * H M M MM M ^ RIGHT LEFT O 0 o ABSENT 'scm "o o £'/' 0' ; 'b o am Io 0 f? B$I' ABNORMAL ^.asjc'wuttpR/ATfipw . EXUDATE f . -, Pa. -.;>0 "o •*i ,o o SENSORY NORMAL o o © puwts ?*<i" OROUAL • ., :..v : O o o o SS^jJBBSj? ®£K J& "d" :I '":-.0?'V. * >»vMTV ^ ' o" o - -O •' ''X3' -"~ ~ "0" (DBAW POSITION NYSTAGMUS VERTICAL ^-Ri«iP(K^;.FISSUlfeMpllli»a^^|^i^^? PALPEBRAL FISSURE ABNORMAL MOVEMENT DEVIATED i NYSTAGMUS HORIZONTAL o 0 FORM) • : ' PAUtfAfc'REFt^NORSiM^ PALATAL REFLEX ABNORMAL " i i i ON CONTINUATION PTOSIS PRESENT o • O'"; d" 0 • ' '^ •- ' 0 ^''TONGUE:PfKnlRUO£D»iMjaU^ ATROPHY OF TONGUE i '' • "„ . " . , • ' * , J• ' ' "" •.'" '"^"'": UGHT REACTION ABNORMAL '•-• 'Q. i NO YES PTOSIS ABNORMAL -.'-•0-' • : ^UOHrtReACTON NORMAL MENTAL STATUS (ALERT, CLEAR, COOPERATIVE, ETC.) EYEBALL POSmON NORMAL 'o SMILE ABNORMAL XII NYSTAGMUS ROTARY. (SROSS ABNORMALITIES? O NO SPECIFY : CORNEAL REFLEX ABNORMAL IX, XI PALATE AND UVULA 0 NYSTAGMUS ABSENT ROUND *,°r* ^^•'iSmjBB *'•&&;} ;;..;{*[• utHcR- 0 ; ... ay»£&'fiEn£$Mam^J;:i^*,'^ VII O O EYE MOVEMENT ABNORMAL SENSORY ABNORMAL >'^O»^' p::S*^;p9«p^!?0S|ftf|i^^^^^^^ ':'-': PUPIL SHAPE/POSITION O O OoNEQUAt D*f6RENC?(WMt;^®©®®^, EYE MOVEMENT NORMAL ; @@© •••":- III, IV, VI o o o : ; CLENCH JAW DEVIATED O NORMAL FIELDS ABNORMAL O Q 0 O O TQ • ' "O >Q-0 O 1 -f rw^tuMM^ ::^:- . ;::;5"^" ONORMAK:!-' •r-.'O"'•?.; O"" PI " r, fV' *£Qsl-, NORMAL HEMORRHAGE O NORMAL '??."""0!'/5: f?ifeDSiNORMAL (TO CONFRONTATION> !TOK O CLENCH JAW SYMMETRIC 0© FUNDU6 ;:.'' O o o PRESENT II ?!Cjtst RIGHT O SENSE OF SMELL o- V LEFT 1 , O YES 1 ( i 1 i i EYEBALL POSITION ABNORMAL i PRINTED NAME OF EXAMINING PHYSICIAN SIGNATURE I I I YES O NO O ADDITIONAL COMMENTS ENTERED ON FORM AFHS 7 C-12 �CASE NUMBER 0 i 2 3 4 5 8 7 3 9 0 1 2 3 4 3 8 7__ I 9 C 0 i 3 4 5 8 7 8 9 ^l 2 8 0 1 2 3 4 5 8 7 '8 9 0 1 2 3 4 5 6 7 8 £ 0 1 2 3 4 S 8' 7 3 f 0.'! VERSION 1.0 JRW SCF 385 TRAILS A TRAILS B 100 200 300 400 500 wo zoo 300 400 sw 10 20 30 40 50 60 70 80 90 to 20 30 40 50 60 70 30 90 f LOGICAL l '1 . IMMEDIATE 2 3 4 5 6 7 8 2 3 4 0 9 MEMORY 10 20 MEMORY 0 1 5 6 7 8 1 2 3 IMMEDIATE 5 3 7 3 9 10 20 F O R 9 4 0 1 2 3 4 5 6 7 8 9 6 7 3 9 DESIGNS 30' DELAY 30' DELAY 10 20 10 20 0 FIVE 1 3 2 3 4 5 9 'b 20 30 4a . 0 1 2 TOTAL 3 4 5 6 7 BO 10 SIX THREE 10 20 30 40 SEVEN 10 20 1 2 3 2 50) 4 5 6 7 3 9 0 1 2 3 4 5 6 7 8 5 7 0 9 4 «!! 10 20 30 40 0 1 30) 200 0 20. 10 20 30 40 TWO 6 80 10 20 30 40 7 90 FOUR 3 1 T P T WIN 9 °™ 0 1 2 3 4 5 6 7 3 9 S E C 0 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1S MEMORY 10 20 30 4O50 0 WIN SEC TOTAL 0 1 ° ' 2 3 2 3 . 2 3 4 5 8 DOMINANT 1 7 3 9 10 4 5 4 8 7 8 9 1 0 2 0 : 30 4 O 3 0 LOG 0 1 8 MIN NON DOMINANT SEASHORE RHYTHM * CORRECT SPEECH PERCEPTION 0 1 * & 30 * OF ERRORS 3 4 9 9 7 3 • 1 2 3 4 5 8 7 8 DOMINANT 9 10 » 30 40 50, 60 70 30 90 NQN DOMINANT 2 5 8 4 7 8 9 0 1 2 3 4 5 8 7 3 9 o RANK 9 ^ DOMINANT 2 3 4 PRINTED NAME OF EXAMINER 3 4 5 10 i 2 3 4 5 6 7 s 9 10 DOMINANT IS: 5 6 7 8 5~ 6 LEFT " RIGHT 9 ,0 20, 30 4« 50. 60. 70. 80 » 1 2 4 9 i 2 3 4 5 6 7 3 9 10 20 30 40 so so 70 80 90 1 3 3 10 11 12 13 14 1! TAPPING 10 20 30 40. so so ?o ao so 1 3 10 20 30 4» so 50 GRIP STRENGTH DOMINANT 2 2 7 7 I9 IMPAIRMENT INDEX .0 .1 1 .3 .4 .5 ii ,7 '.I' 9 1.0 SIGNATURE NCS Trans-Optic • EP01-21155:321 C-13 A3900 �CASE NUMBER B NAME OF PARTICIPANT LAST FIRST Ml CONT NUATION OF FORM AFHS-7 CONTINUATION FORM AFHS- • USE THIS FORM TO RECORD COMMENTS, OBSERVATIONS, OR PHYSICAL FINDINGS FOR WHICH THERE IS INADEQUATE SPACE ON THE OPTICAL MARK SENSE FORMS. PLEASE INDICATE THE ORGAN SYSTEM, IF APPROPRIATE, IN EACH COMMENT BLOCK. ORGAN SYSTEM: COMMENTS: ORGAN SYSTEM: COMMENTS: ORGAN SYSTEM: COMMENTS: ORGAN SYSTEM: COMMENTS: PRINTED NAME OF EXAMINING PHYSICIAN SIGNATURE C-14 DATE �CASE NUMBER 0 1 2 * 4 5 6 7 3 9 * 0 1 2 3 4 5 6 7 S * C 3 4* 0 1 2 GROUP * NAME: DATE OF BIRTH: 8 7 8 9 0 ,€} T; 2 3 4 5 3 7 " a ;£i ,'E. jHiX'T ?, * ? 3 7 'a sj 'p. ft (Tj 'I;' 5) 4) ''»'.- «... 7 8';• 9'; • 0 VERSION 1.0 JRW:SCF 585 INSTRUCTIONS FES 30 MAR 20 MO TODAY'S DAY DATE JUN JU(i AUQ 7 a S 90: APR MAY 3 10 91. 87 YR SEP OCT NOV DEC 3 4 93; 94i VIETNAM COMBAT INDEX FORM AFHS-8 INSTRUCTIONS ARE INCLUDED WITH EACH QUESTION. BELOW IS AN EXAMPLE OF THE CORRECT WAY TO ANSWER EACH QUESTION. EXAMPLE: DO YOU PLAN TO DO ANY OF THE FOLLOWING NEXT WEEK? (PLEASE BLACKEN EITHER "YES" OR "NO") YES NO • •«!' VISIT A RELATIVE ^••GO TO A MUSEUM TO A MOVIE (I WILL VISIT A RELATIVE AND GO TO A MOVIE NEXT WEEK) AIRCRAFT PLEASE INDICATE WHETHER YOU SERVED OR FLEW IN ANY OF THE FOLLOWING AIRCRAFT WHILE IN VIETNAM: (DO NOT INCLUDE TRANSPORTATION TO OR FROM VIETNAM) YES NO f ^_mm ®YES ®NO YES NO YES NO N)C-7 M. F-4 WERE YOU EVER A CREW MEMBER? ®C-54 ®C-118 ®C-123 ®C-130 ®F-105 ($8-52 ®B-66 ® ® HELICOPTER GUNSHIP 0® OTHER AIRCRAFT SPECIFY 3 EXPERIENCES SELOW IS A LIST OF DIFFERENT COMBAT ROLES AND FLYING EXPERIENCES THAT AIR FORCE PERSONNEL HAD DURING THE VIETNAM WAR. FOR EACH STATEMENT, PLEASE BLACKEN THE "YES" CIRCLE IF YOU HAD THAT EXPERIENCE DURING THE VIETNAM WAR OR THE "NO" CIRCLE IF YOU DID NOT. PLEASE BLACKEN EITHER "YES" OR "NO" FOR EACH EXPERIENCE. YES NO YES NO COMBAT PAY ®®FtEW IN AN AIRCRAFT THAT RECEIVED BATTLE DAMAGE '. :: / /: ® ® RECEIVED INCOMING ARTILLERY OR ROCKET FIRE AT HOME BASE OR CAMP ®®ENCOUNTERED MINES OR-BOOBY TRAPS :.;: • ® ® CRASH LANDED, BAILED OUT, OR SHOT DOWN f <3>® RECEIVED SNPER OR SAPPER FIRE IN. OR I ;"•'•;•'' AROUND BASE @ ® MOVED KILLED OR WOUNDED PERSONNEL ® ® KILLED VC OR NVA IN STRAFING OR BOMBING RUNS ®®WOUNDED: . ., ..-. "•-..:: •,": ;.;-: - ©® SERVED AS A FORWARD AIR CONTROLLER (FAC> ®®FLEW IN THE SAME AIRCRAFT WHEN FELLOW CREWMEMBER WAS WOUNDED OR KILLED ® ®FLEW IN THE SAME FORMATION OR ON 1• - - ; • • THE SAME SORTIE WHEN A FELLOW '•'•• CREWMEMBER WAS WOUNDED OR KILLED ®®HAD A CLOSE FRIEND KILLED IN ACTION VC OR IWA . ® ® CAPTURED BY THE ENEMY NCS Trent-Optic ® EP01-21134:321 C-15 AS900 �CASE NUMBER LAST FIRST NAME OF PARTICIPANT B Ml ATTACHMENT TO FORM FORM AFHS-9 ANATOMICAL CHART AFHS- RIGHT LEFT (OVER) C-16 �LEFT LEFT PRINTED NAME OF EXAMINING PHYSICIAN SIGNATURE C-17 DATE �CASE NUMBER B DATE ECG RECORDED NAME OF PARTICIPANT LAST FIRST Ml MO DAY YR FORM AFHS-10 ELECTROCARDIOGRAM REPORT ECG TECHNICIAN (INITIALS) t* 12- LEAD SCALAR ELECTROCARDIOGRAM NORMAL COMMENTS: ABNORMAL RBBB LBBB NON-SPECIFIC T-WAVE CHANGES TACHYCARDIA BRADYCARDIA ARRHYTHMIA PARTICIPANT COMPLIANCE WITH 4-HOUR ABSTINENCE: YES. No. RHYTHM STRIP INTERPRETATION OF ARRHYTHMIA: SIGNATURE PRINTED NAME OF CARDIOLOGIST C-18 DATE �CASE NUMBER B DATE OF X-RAY NAME OF PARTICIPANT MO DAY •V* FORM AFHS-11 PA CHEST X-RAY EXAMINATION X-RAY TECHNICIAN (INITIALS) RIGHT @ LEFT ANTERIOR PLEASE MARK THE LOCATION OF ANY SUSPECTED ABNORMALITY(IES) WITH AN ENCIRCLED NUMBER AND DESCRIBE BELOW. INTERPRETATION OF PA CHEST FILM NORMAL 1 ABNORMAL COMMENTS: SIGNATURE1 PRINTED NAME OF RADIOLOGIST C-19 DATE �CASE NUMBER B DATE OF ANTIGEN ADMINISTRATION LAST FIRST Ml NAME OF PARTICIPANT MO DAY YR TIME OF ANTIGEN ADMINISTRATION FORM AFHS-12 DELAYED SKIN TESTS © ADMINISTERED BY (INITIALS) RESULTS E - ERYTHEMA, MEASURED IN mm I - INDURATION, MEASURED IN mm ANTIGEN: 24-HOUR READING CANDIDA ALBICANS 1:1000 W/V MUMPS 2 CPU TRICHOPHYTON 1:1000 W/V STAPH-PHAGE-LYSATE 48-HOUR READING STAPH-6 TO 9 x 104 CPU PHAGE=0.5 TO 5 x 107 PFU TIME OF DAY THAT SKIN TESTS WERE READ SKIN TESTS READ BY (INITIALS) YES IS PARTICIPANT ON SYSTEMIC CORTICOSTEROIDS OR IMMUNOSUPPRESSANTS? NO DOSAGE: COMMENTS: PRINTED NAME OF REVIEWER SIGNATURE C-20 DATE �CASE NUMBER NAME OF PARTICIPANT B MO DAY DATE OF DIAGNOSIS YR FORM AFHS-16A OUTBRIEFIIMG CONDUCTED ICD9-CM CODE LAST YES DIAGNOSTIC SUMMARY (MEDICAL) NO CHECK ONE NEWLY PRE DIAGEXISTING NOSED DIAGNOSES BASED ON PHYSICAL EXAMS, ECG, DOPPLER. CHEST X-RAY, SKIN TESTS, AND LABORATORY STUDIES COMMENTS: PRINTED NAME OF DIAGNOSTICIAN SIGNATURE C-21 © �CASE NUMBER Ml NAME OF PARTICIPANT B MO DAY t-YR DATE OF DIAGNOSIS y OUTBRIEFINQ CONDUCTED ICD-9-CM CODE YES NO FORM AFH<! 1fiB FORM AFHS-16B CHECK ONE NEWLY DIAGNOSED PREEXISTING DIAGNOSTIC SUMMARY (PSYCHOMETRIC) DIAGNOSES BASED ON PSYCHOLOGICAL TESTING . COMMENTS: SIGNATURE PRINTED NAME OF DIAGNOSTICIAN r-?? m �APPENDIX D Study Selection and Participation Bias �TELEPHONE QUESTIONNAIRE 1. Compared to other people your age, would you say that your health is: Excellent Good Fair Poor 2. 1 2 3 4 Are you currently taking prescribed medicines for any illness or condition? Yes 1 (ANSWER A) No 2 (GO TO QUESTION 3) A. For what illness or condition are you taking prescribed medicines? 3. Within the last six months, did illness or injury keep you from going to work or from holding a job? Please do not count work around the house as a job. Yes 1 (ANSWER A and B) No 2 (GO TO QUESTION 4) A. How many days or weeks did you miss from work within the past six months? Days Missed OR Weeks Missed | | | | | | | | Not Able to Work at all 995 B. What illnesses or conditions caused you to miss work or prevented you from holding a job? D-l �Do you need the help of another person, or use of special equipment, to take care of personal needs such as eating, bathing, dressing, or getting up or around the house? [By "special equipment" we mean anything you use to help yourself, like a wheelchair, railing, walker, cane or other device, which is not usually used by most people.] Yes 1 (ANSWER A and B) No 2 (GO TO QUESTION 5) A. Is this help required because of a condition you've already mentioned, or because of some other health condition? Another condition 1 (ANSWER B) Condition 2 (GO TO already mentioned QUESTION 5) B. What illness or condition is that? 5. Now on to a different subject. Did you earn any income from any job in 1984? Yes 1 (ANSWER A) No 2 (DO NOT ANSWER A) A. What was your income in 1984? Less than $20,000 $20,000 to $40,000 More than $40,000 1 .. .2 3 NAME CASE ID I__|__|_J_|_J_| PHONEft( D-2 ) �TAHED-1. —„ ., . . -y as Efetemrined bv 1.154 Jfrriilarir <T^nB^T*! <WVQ — D B P £ V B C S K I N M M P I D B U N u S s G G P N U U C L V S E S EE E S E E M M E N R B B C —+ P U U —9 - G L L U M CC M E C G P L A T T I C Q N S S N C V V A A C H O 0 L A L K P P H 0 S C 0 O P P R R 0 A L L A T 4 T E S T T E D G T P TRTT.T .06 -.08 .01 .02 .02 .02 -.04 <.01 .02 .05 .02 .04 .01 -.10 -.01 -.01 .02 .08 -.01 -.06 -.07 .03 <.01 -.08 .14 EBP <.01 .03 .06 .02 -.04 .05 -.02 <.01 .03 <.01 .20 -.10 -.03 .03 .15 -.09 .23 .08 -.01 <.01 <.01 -.19 .12 .16 HBC .09 .07 -.04 -.09 .03 -.06 .02 .15 -.07 .12 -.05 .24 -.18 .04 -.07 .06 .15 .10 .10 .10 .05 .14 <.01 90N .06 -.01 .03 <.01 -.05 .05 .02 -.04 .06 -.04 .06 -.07 .01 -.06 <.01 .04 .10 .06 .04 -.02 .03 <.01 HMPID -.04 -.04 .02 -.04 .03 .04 .02 .12 -.03 -.01 -.IS .05 -.01 .10 .06 -.01 -.03 .03 -.01 .10 .03 BUN .22 .02 -.02 -.01 .01 <.01 .02 -.02 -.04 .07 -.03 -.01 .03 -.04 -.15 -.06 <0.1 -.09 .06 -.09 USG <.01 -.02 -.02 .06 .09 .U .03 -.08 <D.l -.05 -.01 -.07 <.01 -.12 .01 .02 -.11 -.02 -.07 PULSE <-01 .08 <.01 -.06 .11 -.10 .06 -.04 .05 -.06 .05 .07 -.08 -.02 .04 -.05 .08 -.01 NOTE -.06 .08 .12 -.06 .04 .05 .07 -.10 .30 .02 <.01 -.02 -.01 .02 .02 -.10 -.08 SEMEN -.02 -.10 .05 -.04 .04 <.10 -.02 .01 <.01 <.01 .03 -.08 <.01 <.01 .11 .02 BBC .07 .08 -.03 <.01 -.10 -.05 .10 .13 .07 -.11 .01 .20 .02 -.20 -.08 PULM .01 <.01 -.07 -.03 -.10 .14 <.01 -.04 -.04 .07 -.03 -.06 -.02 -.02 GLUC .07 .03 -.05 .06 -.10 .17 .12 -.05 -.02 .05 -.23 .B .09 BOG .01 .01 -.06 .03 -.02 -.07 .03 -.01 .03 .02 -.07 -.08 PLAT -.03 <.01 -.02 .07 .05 .09 -.05 .08 .02 .08 <.01 IQ -.06 -.07 -.04 -.08 .02 <.01 -.18 -.06 -.09 -.04 CMS -.13 .03 .04 .05 .01 -.03 .04 .12 .15 NCVA <.01 -.04 -.06 -.04 .07 .08 -.06 -.02 CHOL .03 <.01 -.04 .06 -.05 .18 .09 ALKPHOS -.03 -.01 .06 -.04 .13 .21 OOPRO .22 -.06 .08 .02 .05 ALA -.12 -.03 -.04 -.03 T4 .04 .07 .02 TEST -.13 <.01 SED .06 �TABLE D-2. Correlation Matrix of 20 Variables at First Follovup, as Determined by 1,293 Comparisons D B P TBILI DBP WBC SKIN MMPID ^ BUN USG PULSE RBC GLUC ECG PLAT CNS GGTP CHOL ALKPHOS COPRO T4 SED W B C .01 -.14 -.03 S K I N M M P I D B U N U S G P U L S E R B C G L U C E C G .01 .15 -.01 -.04 -.01 .06 .04 .05 01 .04 <.01 .05 <.01 01 .02 .07 .06 05 .14 .03 .03 .08 .09 -.11 <.01 02 <.01 -.01 .03 -.02 <.01 <.01 <.01 -.02 .03 .05 .03 .10 .09 05 02 -.01 <.01 .09 02 .01 .06 -.01 <.01 .05 .05 <.01 <.01 .10 P L A T -.08 <.01 .28 -.02 .07 -.06 -.03 -.01 .01 -.07 .01 C N S G G T P C H 0 L A L K P H 0 S .15 .05 .03 -.01 <.01 .01 <.01 <.01 .05 .03 .06 .17 -.03 <.01 -.01 -.04 .08 .03 .13 .02 -.05 .10 .06 -.05 .03 <.01 .03 .08 .01 .03 <.01 <.01 -.04 .14 .08 -.01 -.02 .22 .15 .07 .06 .08 .07 .08 .04 -.06 .05 .08 .16 .09 -.04 .06 .35 .03 C 0 P R 0 T 4 S E D .01 .03 .13 <.01 <.01 .15 -.05 .04 -.08 .10 <.01 <.01 <.01 .12 -.04 .04 .05 -.02 .13 <.01 .02 -.03 -.02 .01 .22 .01 .03 .04 .01 .03 .02 .08 <.01 .01 <.01 .18 <.01 .13 <.01 .03 .08 -.26 .16 .13 .12 .09 .17 .18 .27 .17 .05 T E S T <.01 !os .04 .01 -.08 -.17 -.05 -.05 .04 -.26 -.09 .04 .01 -.05 -.06 -.07 .04 .13 -.23 �APPENDIX E Statistical Methods �RUN: Initial screening models: Model A = all main effects Model B = Model A + all 2-way interactions Model C = Model B + all 3 way interactions involving exposure group (E) TEST: compare C with B for 3-way interactions: Is p-value > 0.05 for the global test of all 3-way interactions AND p value > 0.05 for each individual 3-way interaction? (i.e., can all 3-way interactions be dropped?) No Yes TEST: compare B with A for 2-way interactions: Is p-value > 0.05 for the global test of all 2-way interactions AND p value > 0.15 for each individual 2-way interaction? (i.e., can all 2-way interactions be dropped?) No Proceed stepwise eliminating individual effects with p-value > 0.05 (those with largest p-value first). "Initial Model" = Model A + all 2-way interactions with individual p-value <= 0.15 in Model B. I Yes •'Initial Model" = Model A. Proceed stepwise, eliminating individual effects with p-value > 0.05 (those with largest p-value first). Proceed stepwise eliminating individual effects with p-value > 0.05 (those with largest p-value first). Proceed stepwise, eliminating individual effects with p-value > 0.05 (those with largest p-value first). Yes Do all 2-way interactions drop out? I No Final model = "best model"; use for testing significance of E; use LSMEANS. "Initial Model" = Model B + all 3-way interactions with individual p-value <= 0.15 in Model D. Stratify and customize. Figure E-1. Modeling Strategy Yes Do all 3-way interactions drop out? No Stratify and customize. �Continuous Dependent Variables Use SAS® general linear models (GLM) procedure and follow the chart. Dichotomous Dependent Variable: Use BMDP®-LR (logistic regression) with MLR option, and follow the chart, If the number of covariates is huge, use ACE option for Models A, B, C. Polychotomous Dependent Variable: Use BMDP«-4F (log-linear model) adding "delta" =0.1 to each cell, and follow the chart. Use LAMB and COV options for parameter estimates when "best model" found. E-2 �APPENDIX F Exposure Index �TABLE F-l. Herbicide Orange Equivalent Gallons and Ranch Hand Manning by Month Mo./Yr. 10/61 11/61 12/61 01/62 02/62 03/62 04/62 05/62 06/62 07/62 08/62 09/62 10/62 11/62 12/62 01/63 02/63 03/63 04/63 05/63 06/63 07/63 08/63 09/63 10/63 11/63 12/63 01/64 02/64 03/64 04/64 05/64 06/64 07/64 08/64 09/64 10/64 11/64 12/64 01/65 02/65 03/65 Gallons Sprayed Pilot (Occ. 1) Navigator (Occ. 2) 0 0 0 191,426 324,216 191,426 0 0 0 0 0 334,126 334,126 0 90,879 0 0 0 0 0 174,024 259,150 0 0 339,588 377,172 942,630 121,454 363,758 755,312 56799 152,271 612,709 282,789 777,669 1,413,945 1,413,945 1,413,945 1,413,945 1,296,116 1,437,510 730,538 0 5 9 14 14 15 16 15 12 13 11 12 9 10 8 9 7 12 12 10 10 11 8 10 7 6 5 7 5 8 9 10 7 9 9 8 9 11 10 11 12 13 0 11 2 2 2 2 2 3 2 2 2 2 1 0 0 0 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 2 3 3 3 3 3 3 3 4 5 4 . F-l OfficerOther (Occ. 3) 0 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 Enlisted Flyer (Occ. 4) EnlistedOther (Occ. 5) 0 6 7 7 7 7 6 6 5 5 5 5 5 5 4 5 5 5 5 5 4 8 8 9 9 10 6 7 7 5 6 5 5 6 5 4 4 4 6 6 6 6 0 14 20 23 23 20 14 13 7 4 5 6 6 5 5 4 4 6 6 7 7 6 4 4 6 6 6 5 4 4 2 2 2 3 3 2 2 1 1 1 1 1 �TABLE F-l. (continued) Herbicide Orange Equivalent Gallons and Ranch Hand Manning by Month Mo./Yr. 04/65 05/65 06/65 07/65 08/65 09/65 10/65 11/65 12/65 01/66 02/66 03/66 04/66 05/66 06/66 07/66 08/66 09/66 10/66 11/66 12/66 01/67 02/67 03/67 04/67 05/67 06/67 07/67 08/67 09/67 10/67 11/67 12/67 01/68 02/68 03/68 04/68 05/68 06/68 Gallons Sprayed Pilot (Occ. 1) Navigator (Occ. 2) 659,841 1,767,431 0 942,630 26,500 44,650 78,850 106,900 148,525 152,450 129,150 135,600 141,050 183,900 191,830 112,300 192,050 213,970 122,040 164,800 212,100 202,360 363,830 285,400 208,300 251,320 335,860 253,884 162,895 298,615 265,335 372,425 383,605 333,595 27,450 48,200 307,740 336,300 226,325 14 15 16 19 19 22 23 24 23 21 22 21 22 21 20 21 26 28 34 41 45 49 59 51 50 53 55 51 63 60 55 55 58 54 65 69 72 75 77 3 4 4 4 4 4 4 6 5 6 6 4 5 6 6 8 8 9 8 8 9 9 13 13 14 15 13 15 13 18 19 17 18 19 19 20 20 18 18 OfficerOther (Occ. 3) 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 3 4 5 5 5 4 4 4 3 3 4 5 5 6 6 6 6 5 6 6 6 Enlisted Flyer (Occ. 4) 6 6 7 7 7 6 6 10 11 10 10 10 10 9 10 9 11 12 16 18 28 28 28 28 33 34 36 37 32 33 36 33 34 33 35 34 36 32 37 EnlistedOther (Occ. 5) 2 2 4 3 3 3 6 12 12 16 26 32 37 38 41 45 46 62 85 104 123 123 116 114 108 101 105 163 160 161 149 145 129 127 141 160 161 160 164 �TABLE F-l. (continued) Herbicide Orange Equivalent Gallons and Ranch Hand Manning by Month Mo./Yr. 07/68 08/68 09/68 10/68 11/68 12/68 01/69 02/69 03/69 04/69 05/69 06/69 07/69 08/69 09/69 10/69 11/69 12/69 01/70 02/70 03/70 04/70 05/70 06/70 07/70 08/70 09/70 10/70 11/70 12/70 01/71 02/71 03/71 04/71 05/71 06/71 07/71 08/71 09/71 10/71 Gallons Sprayed 258,100 289,160 216,300 72,250 189,100 218,750 264,450 197,450 356,500 339,800 353,800 383,533 287,425 299,100 206,800 181,000 205,100 276,900 186,350 152,100 153,730 45,700 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Pilot (Occ. 1) 84 91 89 89 101 94 98 91 90 94 93 88 91 85 83 83 90 76 66 58 59 54 51 47 44 40 40 34 30 25 23 23 23 23 23 28 29 29 29 29 Navigator (Occ. 2) 19 18 22 20 17 17 19 18 17 20 19 19 16 16 15 17 16 16 15 15 13 13 14 14 11 9 7 6 5 4 4 4 4 4 4 4 4 4 4 4 OfficerOther (Occ. 3) 7 9 8 8 7 8 7 5 5 6 6 7 6 6 6 6 6 5 5 5 5 5 5 3 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Source: Baseline Morbidity Report, 24 February 1984. Enlisted Flyer (Occ. 4) EnlistedOther (Occ. 5) 42 45 44 49 53 51 51 51 53 54 54 57 55 55 61 61 60 52 54 41 39 37 29 18 16 14 13 14 15 13 14 14 14 14 14 14 14 14 14 14 187 192 147 155 153 154 • 154 166 172 161 151 155 152 155 142 122 118 114 116 122 125 109 94 84 74 63 43 37 35 30 28 28 28 28 28 28 28 28 28 28 �APPENDIX G General Health �APPENDIX G: General Health Contents G-l Summary Statistics for General Health Covariates by Group G-l G-2 Unadjusted Analysis for Self-Perception of Health by Group (Original Comparisons Only) G-l G-3 Unadjusted Analysis for Appearance of Acute Illness or Distress by Group (Original Comparisons Only) G-2 G-4 Unadjusted Analysis for Appearance of Relative Age by Group (Original Comparisons Only) G-2 G-5 Adjusted Relative Risks of Appearance of Relative Age by Occupation (Original Comparisons Only) G-3 G-6 Unadjusted Analysis for Sedimentation Rate by Group (Original Comparisons Only) G-3 G-7 Unadjusted Analysis for Percent Body Fat by Group (Original Comparisons Only) G-4 G-i �TABLE G-l. Summary Statistics for General Health Covariates by Group Group Ranch Hand Comparison Covariate Covariate Category Percent Percent Race Black Nonblack 5.9 94.1 6.4 93.6 0.673 Occupation Officer Enlisted Flyer Enlisted Groundcrew 37.4 17.4 45.2 37.4 16.2 46.3 0.727 p-Value Mean±SE Age (at Baseline) Personality Score 43.94±0.24 — 43.85±0.22 0.789 -1.18±0.29 — Meant SE -1.56±0.25 0.332 — Covariate not categorized for these results. TABLE G-2. Unadjusted Analysis for Self-Perception of Health by Group (Original Comparisons Only) Self-Perception of Health Excellent Group Good Pair Poor Number Percent Number Percent Number Percent Number Percent Total Ranch Hand 490 48.2 434 42.7 74 7.3 18 1.8 1,016 Original Comparison 490 51.3 403 . 42.2 55 5.8 7 0.7 0.7 955 p=0.075 G-l �TABLE G-3. Unadjusted Analysis for Appearance of Acute Illness or Distress by Group (Original Comparisons Only) Acute Illness or Distress Yes Group No Number Percent Number 4 0.4 1,010 Ranch Hand Percent Total p-Value* 1,014 99.6 0.340 Original Comparison 6 0.6 949 955 99.4 *Fisher's exact test, 1-sided. TABLE G-4. Unadjusted Analysis for Appearance of Relative Age by Group (Original Comparisons Only) Appearance of Relative Age Younger Group Ranch Hand Same Older Number Percent Number Percent Number Percent Total 16 1.6 957 94.3 42 4.1 p-Value 1,015 0.454 Original Comparison 9 0.9 906 94.9 G-2 40 4.2 955 �TABLE G-5. Adjusted Relative Risks of Appearance of Relative Age by Occupation (Original Comparisons Only) Adjusted Relative Risk 95% C.I. p-Value Officer 3.27 (0.67,15.8) 0.142 Enlisted Flyer 0.47 (0.20,1-12) 0.089 Enlisted Groundcrew 1.17 (0.65,2.08) 0.603 Occupation TABLE G-6. Unadjusted Analysis for Sedimentation Rate by Group (Original Comparisons Only) Sedimentation Rate Normal <20 mm/hr Group Ranch Hand Abnormal >20 mm/hr Number Percent Number Percent Total 957 94.2 59 5.8 p-Value 1,016 0.059 Original Comparison 918 96.1 37 G-3 3.9 955 �TABLE G-7. Unadjusted Analysis for Percent Body Fat by Group (Original Comparisons Only) Percent Body Fat Lean/Normal <25% Group Ranch Hand Number Percent 831 81.8 Obese >25% Number Percent 185 18.2 Total p-Value 1,016 0.230 Original Comparison 759 79.6 195 G-4 20.4 954 �APPENDIX H Malignancy �APPENDIX H: Malignancy Contents Table H-l H-2 H-3 H-4 H-5 H-6 H-7 H-8 H-9 Unadjusted Analyses of Followup Participants with Verified and Suspected Neoplasms in the Baseline-Followup Interval by Group (Nonblacks and Blacks, Original Comparisons Only).. H-l Unadjusted Analyses of Nonblack Followup Participants with Verified and Suspected Malignant Skin Neoplasms in the Baseline-Followup Interval by Cell Type and Group (Original Comparisons Only) H-3 Association Between Baseline-Followup Interval Basal Cell Carcinoma Incidence and the Covariates for Combined Followup Ranch Hand and Comparison Nonblack Participants H-4 Association Between Baseline-Followup Interval Sun Exposure-Related Skin Malignancy Incidence and the Covariates for Combined Followup Ranch Hand and Comparison Nonblack Participants H-9 Summary Results of Main Effects Models for Selection of Covariates for Basal Cell Carcinomas in the Baseline-Followup Interval , H-14 Adjusted Analyses of Nonblack Followup Participants for Malignant Skin Neoplasm Incidence During Baseline-Followup Interval (Original Comparisons Only) H-15 Summary of Group-by-Covariate Interactions for Malignant Skin Neoplasms in the Baseline-Followup Interval (Nonblacks Only) H-16 Summary of Group-by-Covariate Interactions for Malignant Skin Neoplasms in the Baseline-Followup Interval (Original Comparisons Only) H-18 Followup Participants with Verified Malignant Systemic Neoplasms in Baseline-Followup Interval by Group (Original Comparisons Only) H-22 H-10 Unadjusted Analyses of Followup Participants with Verified and Suspected Malignant Systemic Neoplasms in the Baseline-Followup Interval by Group (Original Comparisons Only) H-23 H-ll Association Between Baseline-Followup Interval Incidence of All Malignant Systemic Neoplasms Combined and the Covariates for the Combined Followup Ranch Hand and Comparison Groups H-24 H-i �APPENDIX H: Malignancy Contents (continued) Table H-12 Adjusted Analyses of Followup Participants for the Incidence of All Malignant Systemic Neoplasms During the Baseline-Followup Interval (Original Comparisons Only)...... H-28 H-13 Summary of Group-by-Occupation Interaction for All Malignant Systemic Neoplasms (Verified Plus Suspected) During the Baseline-Followup Interval H-29 H-14 Summary of Group-by-occupation Interaction for All Malignant Systemic Neoplasms (Verified Plus Suspected) During the Baseline-Followup Interval (Original Comparisons Only) , H-30 H-15 Unadjusted Analyses of Followup Participants with Lifetime Occurrence of Verified and Suspected Lifetime Neoplasms by Group (Nonblacks and Blacks, Original Comparisons Only) . H-31 H-16 Unadjusted Analyses of Nonblack Followup Participants with Lifetime Occurrence of Verified and Suspected Lifetime Malignant Skin Neoplasms by Cell Type and Group (Original Comparisons Only) H-33 H-17 Association Between Lifetime Incidence of Sun-Exposure Related Skin Malignancies and the Covariates for Combined Followup Ranch Hand and Comparison Nonblack Participants.... H-34 H-18 Summary of Group-by-Covariate Interactions for Lifetime Malignant Skin Neoplasm Incidence (Nonblacks Only) H-39 H-19 Adjusted Analyses of Nonblack Followup Participants for Lifetime Malignant Skin Neoplasm Incidence (Original Comparisons Only) H-40 H-20 Summary of Group-by-Covariate Interactions for Lifetime Malignant Skin Neoplasm Incidence (Original Comparisons Only*) H-21 Summary of Followup Participants with Lifetime Incidence of Verified Malignant Systemic Neoplasms by Group (Original Comparisons Only) H-22 Unadjusted Analyses of the Lifetime Incidence Rates of All Malignant Systemic Neoplasms Combined, by Group (Original Comparisons Only) H-23 Summary of Group-by-occupation Interactions for Lifetime Incidence of All Malignant Systemic Neoplasms Combined H-ii H-41 H-45 H-46 H-47 �APPENDIX H: Malignancy Contents (continued) Table H-24 Adjusted Analyses for Lifetime Incidence of All Malignant Systemic Neoplasms Combined (Original Comparisons Only) H-25 Summary of Group-by-Occupation Interactions for Lifetime Incidence of All Malignant Systemic Neoplasms Combined (Original Comparisons Only) H-49 H-50 H-26 Unadjusted Exposure Index Analyses for Followup Participants for Occurrence of Malignant Neoplasms in the Baseline-Followup Interval H-52 H-27 Unadjusted Exposure Index Analysis for Followup Participants for Lifetime Occurrence of Malignant Neoplasms ". H-55 H-iii �TABLE H-l. Unadjusted Analyses of Followup Participants with Verified and Suspected Neoplasms in the Baseline-Follovup Interval by Group (Nonblacks and Blacks, Original Comparisons Only) Group* Ranch Hand Site Skin Neoplasm Behavior and Status Original Comparison Number** Percent Number** Percent Total** p-Value*** 37 47 3.6 4.6 31 45 3.2 4.7 68 92 0.711 0.999 Benign Verified Verified and Suspected 76 78 7.5 7.7 65 68 6.8 7.1 141 146 0.600 0.667 Uncertain Behavior and Unspecified Nature: Verified Verified and Suspected 1 1 0.1 0.1 1 1 0.1 0.1 2 2 0.999 0.999 Any Skin Neoplasm* Verified Verified and Suspected Systemic Malignant Verified Verified and Suspected 114 124 11.2 12.2 96 113 10.1 11.8 210 237 0.422 0.835 Malignant Verified Verified and Suspected 8 12 0.8 1.2 6 11 0.6 1.2 14 23 0.791 0.999 Benign Verified Verified and Suspected 42 48 4.1 4.7 41 49 4.3 5.1 83 97 0.911 Uncertain Behavior and Unspecified Nature: Verified Verified and Suspected 6 6 0.6 0.6 6 9 0.6 0.9 12 15 0.999 0.442 Any Systemic Neoplasm1* Verified Verified and Suspected 55 65 5.4 6.4 51 66 5.3 6.9 106 131 0.999 0.652 H-l 0.679 �TABLE H-l. (continued) Unadjusted Analyses of Followup Participants with Verified and Suspected Neoplasms in the Baseline-Followup Interval by Group (Nonblacks and Blacks, Original Comparisons Only) Group* Ranch Hand Site All Neoplasms Neoplasm Behavior and Status Malignant, Benign, Uncertain Behavior. Unspecified Nature Verified Verified and Suspected Original Comparison Number** Percent Number** Percent 161 177 15.9 17.4 140 169 14.7 17.7 Total** p-Value*** 301 346 *Sample sizes: 1,016 Ranch Hands, 955 Original Comparisons. **Number of participants. ***Fisher's exact test. "Participant has one or more malignant, benign, or unspecified skin neoplasms. b Participant has one or more malignant, or benign skin or systemic neoplasms. c Participant has one or more malignant, benign, or unspecified systemic neoplasms. H-2 0.491 0.906 �TABLE B-2. Unadjusted Analyses of Nonblack Follovup Participants with Verified and Suspected Malignant Skin Neoplasms in the Baseline-Followup Interval by Cell Type and Group (Original Comparisons Only) 4 Group** Cell Type Status Statistic* Ranch Hand Original Comparison Est. Relative Risk (95Z C.I.) p-Value Verified Number/Z Total Neoplasms 29 42 3.0Z 22 31 2.5Z 1.24 (0.71,2 .18) 0.480 Verified & Suspected Number/Z Total Neoplasms 36 53 3.8Z 36 46 4.0Z 0.94 ( . 8 1 . 0 05, 5) 0.811 Verified Number/Z Total Neoplasms 4 6 0.4Z 3 3 0.3Z 1.25 ( . 8 5 .61) 02, 0.999 Verified & Suspected Number/Z Total Neoplasms 4 6 0.4% 3 3 0.3Z 1.25 (0.28,5 .61) 0.999 Verified Number/Z Total Neoplasms 1 2 0.1Z 3 3 0.3Z 00, 0.31 ( . 3 3 .01) 0.359 Verified & Suspected Number/Z Total Neoplasms 1 2 0.1Z 4 4 0.4Z 0.23 (0.03,2 .10) All Malignant Verified Skin Neoplasms Verified & Suspected Number/Z Total Neoplasms 37 56 3.9Z 31 41 3.5Z 1.13 (0.69,1 .83) 0. 711 Number/Z Total Neoplasms 47 70 4.9Z 45 58 5.0Z 0.98 ( . 4 1 .49) 0.999 06, Sun-Exposure Related Malignant Neoplasms' Verified Number/Z Total Neoplasms 32 47 3.4Z 25 33 2.8Z 0) 1.21 (0.71,2 . 6 Verified & Suspected Number/Z Total Neoplasms 39 58 4.1Z 39 50 4.4Z 0.94 (0.59,1 .47) 0.817 Basal Cell Carcinoma Squamous Cell Carcinoma Melanoma *Huraber and percent of participants; total number of malignant neoplasms of specified cell type. **Number of participants — 956 Ranch Hands, 897 Original Comparisons. *Basal cell carcinoma, melanoma, and malignant epithelial neoplasms NOS. 0. 204 0..504 �TABLE H-3. Association Between Baseline-Follovup Interval Basal Cell Carcinoma Incidence and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Age Born >1942 Born 1923-41 Born <1922 11 43 5 1.3 3.6 5.8 0.001 14 61 9 1.6 5.1 10.3 <0.001 Occupation Officer Enlisted Flyer Enlisted Groundcrew 31 11 17 3.7 3.1 1.8 0.047 47 14 23 5.5 3.9 2.4 0.003 Total Lifetime Smoking (Pack- Years) 0 >0-20 >20-40 >40 16 21 12 10 2.6 2.1 3.1 6.4 0.023 21 28 24 11 3.4 2.8 6.1 7.0 0.005 Total Lifetime Alcohol Consumption (Drink- Years) 0 >0-5 >5-30 >30-100 >100 4 20 18 13 4 2.8 2.8 2.8 2.7 3.9 0.980 5 27 24 24 4 3.6 3.8 3.7 5.0 3.9 0.800 Ethnic Background3 A B C D 49 8 1 0 3.1 1.9 1.6 0.0 0.333 72 9 1 0 4.6 2.1 1.6 0.0 0.046 Skin Color Dark Medium Pale Dark Peach Pale Peach 0 1 3 43 12 0.0 1.4 1.0 3.4 2.3 0.156 0 2 5 56 21 0.0 2.7 1.6 4.4 4.0 0.232 �TABLE H-3. (continued) Association Between Baseline-Followup Interval Basal Cell Carcinoma Incidence and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Eye Color Brown Hazel Green Grey Blue 18 17 2 2 20 2.8 3.7 1.7 2.2 2.4 0.590 24 18 5 4 33 3.7 4.0 4.2 4.3 3.9 0.998 Hair Color Black Dark Brown Light Brown Red Blond 11 25 18 2 3 2.5 2.4 3.2 12.5 2.8 0.152 16 37 24 3 4 3.6 3.6 4.3 18.8 3.7 000 .4 Residential History (Average Latitude) >37° <37° 21 38 1.9 3.7 008 .0 31 53 2.7 5.2 0.004 Skin Reaction to First 30 Min. of Sun Exposure Burns Usually Burns Burns Mildly Rarely Burns 7 24 17 11 2.8 5.6 2.1 1.6 0.001 11 32 22 19 4.5 7.5 2.7 2.8 <0.001 Skin Reaction to >2 Hrs. of Sun After First Exposure Burns Painfully Burns Becomes Red No Reaction 2 21 22 14 1.7 6.2 2.1 2.1 <0.001 4 23 35 22 3.3 6.8 3.4 3.3 0.027 Skin Reaction After Repeated Exposure to Sun Freckles, No Tan Tans Mildly Tans Moderately Tans Deep Brown 0 20 22 16 0 6,4 2.2 2.0 <0.001 1 25 33 24 2.2 8.0 3.2 3.1 0.001 PC I Cn �TABLE H-3. (continued) Association Betveen Baseline-Pollowup Interval Basal Cell Carcinoma Incidence and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Sun Reaction Index Tends to Burn. Mild Reaction Tends to Tan Exposures to Carcinogens Number* Percent Verified and Suspected p-Value Number* Percent p-Value 1.4 6.0 1.9 <0.001 4 32 47 2.8 7.1 3.0 <0.001 Asbestos Yes 10 No 49 2.2 2.9 0.519 17 67 3.7 3.9 0.999 Nonmedical x rays Yes 17 No 42 3.4 2.5 0.272 26 58 5.3 3.5 0.084 Industrial Chemicals Yes 28 No 31 2.5 3.0 0.511 40 44 3.6 4.2 0.437 Herbicides Yes 40 No 19 3.2 2.1 0.142 57 27 4.5 3.0 0.072 Insecticides Yes 40 No 19 3.1 2.2 0.281 55 29 4.2 3.4 0.365 Degreasing Chemicals Composite Carcinogen Exposure 2 27 29 Yes 35 No 24 2.8 2.7 0.894 47 37 3.7 4.1 0.735 Yes 15 No 42 3.1 2.5 0.523 18 64 3.7 3.9 0.999 �TABLE H-3. (continued) Association Between Baseline-Follovup Interval Basal Cell Carcinoma Incidence and the Covariates for Combined Pollovup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value ~~ Yes No 0 59 0.0 2.7 0.999 0 84 0.0 3.9 0.999 Yes 2 No 56 4.9 2.6 0.301 3 80 7.3 3.8 0.206 Benzene PC iJ Anthracene Arsenic Exposure to Individual Carcinogens Yes No 4 55 5.4 2.6 0.140 5 79 6.8 3.8 0.207 Benzidine Yes 1 No 58 9.1 2.7 0.263 1 83 9.1 3.9 0.354 Chroma tes Yes No 3 54 3.6 2.6 0.484 4 78 4.8 3.8 0.558 Coal Tar 1 Yes No 58 1.5 2.8 0.999 2 82 2.9 3.9 0.999 Creosote Yes 6 No 53 3.8 2.6 0.441 7 77 4.4 318 0.669 Aminodiphenyl Yes No 0 58 0.0 2.7 0.999 0 83 0.0 3.9 0.999 Chloromethyl Ether 1 Yes No 58 4.8 2.7 0.442 1 83 4.8 3.9 0.566 Mustard Gas Yes No 0 59 0.0 2. 7 0.999 0 84 0.0 3.9 0.,999 * �TABLE H-3. (continued) Association Between Baseline-Follovup Interval Basal Cell Carcinoma Incidence and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Naphthylamine Yes No 3 55 5.8 2.6 0.161 4 79 7.7 3.7 0.136 Cutting Oils Yes 9 No 50 4.0 2.6 0.200 12 72 5.3 3.7 0.271 Trichloroethylene Yes 4 No 55 2.2 2.8 0.814 6 78 3.3 3.9 0.842 Ultraviolet Light Yes 2 No 57 4.6 2.7 0.339 2 82 4.6 3.9 0.688 Vinyl Chloride Exposure to Individual Carcinogens (continued) Yes 0 No 59 0.0 2.8 0.999 1 83 3.2 3.9 0.999 sa CO *Number of participants with basal cell carcinoma. a Ethnic Background: A B C D - English, Welsh, Scottish, Irish, Scandinavian, German, Polish, Russian, other Slavic, Jewish, French. Spanish, Italian, Greek. Mexican, American Indian, Asian. �TABLE H-4. Association Between Baseline-Foliovup Interval Sun Exposure-Related Skin Malignancy Incidence and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Age 14 46 5 1.6 3.8 5.8 0.004 19 64 9 2.2 5.4 10.3 <0.001 Occupation 33 I Born >1942 Born 1923-41 Born <L922 Officer Enlisted Flyer Enlisted Groundcrew 33 12 20 3.9 3.3 2.1 0.077 50 15 27 5.9 4.2 2.8 0.006 Total Lifetime Smoking (Pack- Years) 0 >0-20 >20-40 >40 19 22 13 11 3.1 2.2 3.3 7.0 0.012 24 31 25 12 3.9 3.1 6.4 7.6 0.007 Total Lifetime Alcohol Consumption (Drink-Years) 0 >0-5 >5-30 > 30-100 >100 4 22 18 15 5 2.8 3.1 2.8 3.1 4.8 0.858 5 29 26 26 5 3.6 4.0 4.0 5.4 4.8 0.736 Ethnic Background A B C D 54 8 1 0 3.4 1.9 1.6 0.0 0.213 78 10 1 0 4.9 2.4 1.6 0.0 0.036 Skin Color Dark Medium Pale Dark Peach Pale Peach 0 1 4 48 12 0.0 1.4 1.3 3.8 2.3 0.116 0 2 6 62 22 0.0 2.7 2.0 4.9 4.2 0.213 �TABLE H-4. (continued) Association Between Baseline-Follovup Interval Sun Exposure-Related Skin Malignancy Incidence and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Eye Color Brown Hazel Green Grey Blue 20 19 2 2 22 3.1 4.2 1.7 2.2 2.6 0.461 27 20 5 4 36 4.2 4.4 4.2 4.3 4.2 0.999 Hair Color Black Dark Brown Light Brown Red Blond 12 27 20 2 4 2.7 2.6 3.6 12.5 3.7 0.171 17 40 26 3 6 3.9 3.9 4.6 18.8 5.6 0.051 Residential History (Average Latitude) >37° <37° 24 41 2.1 4.0 0.011 34 58 3.0 5.7 0.003 Skin Reaction to First 30 Min. of Sun Exposure Burns Usually Burns Burns Mildly Rarely Burns 8 26 20 11 3.2 6.1 2.5 1.6 <0.001 13 34 26 19 5.3 7.9 3.2 2.8 <0.001 Skin Reaction to >2 Hrs. of Sun After First Exposure Burns Painfully Burns Becomes Red No Reaction 4 22 24 15 3.3 6.5 2.3 2.3 0.001 6 25 38 23 5.0 7.4 3.6 3.5 0.016 Skin Reaction After Repeated Exposure to Sun Freckles, No Tan Tans Mildly Tans Moderately Tans Deep Brown 0 22 23 19 0 7.0 2.3 2.4 <0.001 1 27 35 28 2.2 8.6 3.4 3.6 <0.001 EC 4 �TABLE H-4. (continued) Association Between Baseline-Followup Interval Sun Exposure-Related Skin Malignancy Incidence and the Covariates for Combined Followup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Sun Reaction Index Tends to Burn Mild Reaction Tends to Tan Exposures to Carcinogens Number* Percent Verified and Suspected p-Value Number* Percent p-Value <0.001 6 34 51 4.1 7.5 3.3 <0.001 Asbestos Yes 10 No 55 2.2 3.2 0.283 17 75 3.7 4.4 0.602 Yes 18 No 47 3.6 2.8 0.367 27 65 5.5 3.9 0.129 Industrial Chemicals Yes No 33 32 2.9 3.1 0.900 45 47 4.0 4.5 0.594 Herbicides Yes No 43 22 3.4 2.4 0.203 61 31 4.8 3.4 0.130 Insecticides Yes 42 No 23 3.2 2.7 0.523 58 34 4.4 4.0 0.664 Degreasing Chemicals Composite Carcinogen Exposure 2.8 6.2 2.1 Nonmedical X Rays as 4 28 32 Yes No 37 28 2.9 3.1 0.899 50 42 4.0 4.6 0.451 Yes 15 No 48 3.1 2.9 0.879 18 72 3.7 4.4 0.608 �TABLE H-4. (continued) Association Between Baseline-Follovup Interval Sun Exposure-Related Skin Malignancy Incidence and the Covariates for Combined Followup Ranch Hand and Comparison Nonblack Participants Verified Covariate Exposure to Individual Carcinogens Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Anthracene Yes No 0 65 0.0 3.0 0.999 0 92 0.0 4.3 0.999 Arsenic Yes No 2 62 4.9 2.9 0.343 3 88 7.3 4.1 0.246 Benzene Yes No 4 61 5.4 2.9 0.280 5 87 6.8 4. 2 0.241 Benzidine 1 Yes No 64 9.1 3.0 0.285 1 91 9. 1 4.2 0.380 Chromates Yes No 3 60 3.6 2.9 0.734 4 86 4.8 4. 1 0.777 Coal Tar 1 Yes No 64 1.5 3.1 0.721 2 90 2.9 4.3 0.999 Creosote Yes 6 No 59 3.8 2.9 0.473 7 85 4.4 4.2 0.839 Aminodiphenyl Yes No 0 64 0.0 3.0 0.999 0 91 0.0 4.2 0.999 Chloromethyl Ether Yes No 1 64 4.8 3.0 0. ,475 1 91 4.8 4.3 0.,600 Mustard Gas Yes 0 No 65 0.0 3.0 0.,999 0 92 0.0 4.3 0.999 �•' TABLE H-4. (continued) Association Between Baseline-Followup Interval Sun Exposure-Related Skin Malignancy Incidence and the Covariates for Combined Followup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value 4 Exposure to Individual Carcinogens (continued) Naphthylamine Yes 3 No 61 5.8 2.9 0.197 4 87 7.7 4.1 0.277 Cutting Oils Yes No 9 56 4.0 0.406 12 80 5.3 4.1 0.384 Yes 4 No 61 2.2 0.653 6 86 3.3 4.4 0.701 Yes No 2 63 4.6 0.384 2 90 4.6 4.3 0.711 Yes No 0 65 0.0 0.999 1 91 3.2 4.3 0.999 Trichloroethylene so Ultraviolet Light Vinyl Chloride 2.9 3.1 3.0 3.1 *Number of participants with sun exposure-related skin malignancies. "Ethnic Background: A B C D - English, Welsh, Scottish, Irish. Scandinavian, German, Polish, Russian, other Slavic, Jewish, French. Spanish, Italian, Greek. Mexican, American Indian, Asian. �TABLE H-5. Summary Results of Main Effects Models for Selection of Covariates for Basal Cell Carcinomas in the Baseline-Followup Interval Basal Cell Carcinoma in Baseline-Follovup Interval p-Value Covariate Model 1* Category Model 2b Group Ranch Hand, Comparison 0.264 0.259 Age Born XL942, 1923-1941, <1922 0.001 0.002 Occupation Officer, Enlisted Flyer, Enlisted Groundcrew 0.683 0.609 Lifetime Smoking 0-20, >20-40, >40 Pack-years 0.091 0.096 Lifetime Alcohol 0-30, >30 0.743 0.819 Ethnic Background0 A, Not A 0.391 0.425 Skin Color Peach, Not Peach 0.084 0.055 Hair Color Blond or Red, Other 0.880 0.761 Reaction of Skin to 30 Minutes Sun Burns, Usually Burns, Other 0.195 Reaction of Skin to >2 Hours of Sun Burns Painfully, Burns, Other 0.007 Reaction of Skin to Repeated Sun Exposure Freckles With no Tan, Tans Easily, Other Sun-Reaction Index Residential History <0.001 Burns With Freckles, Intermediate Reaction, Tans Easily — <37° N latitude, >37° N latitude 0.006 *Model includes three reactions of skin to sun variables. Model includes sun-reaction index. °Ethnic background: A - English, Welsh, Scottish, Irish. H-14 _ <0.001 0.008 �TABLE H-6. Adjusted Analyses of Nonblack Followup Participants for Malignant Skin Neoplasm Incidence During Baseline-Followup Interval (Original Comparisons Only) Variable Status Adj. Relative Risk (95% C.I.) p-Value Basal Cell Carcinoma Verified 1.46 (0.82,2.61) 0.198 AGE (p<0.001) LAT (p=0.015) SUNREAC (p=0.001) GRP*SUNREAC (marginal: p=0.051) **** **** GRP*SUNREAC (p=0.007) AGE (p<0.001) LAT (p=0.006) **** **** GRP*SKIN (p=0.036) AGE (p<0.001) SUNREAC (p=0.003) LAT (p=0.015) **** **** GRP*SUNREAC (p=0.030) SKIN*SUNREAC (p=0.042) AGE (p<0.001) LAT (p=0.008) Verified & Suspected Verified Sun-Exposure Skin Neoplasms Verified & Suspected Covariate Remarks* *Abbreviations: LAT: average lifetime residential latitude SUNREAC: sun reaction index GRP: group OCC: occupation SKIN: skin color ****Group-by-covariate interactions—adjusted relative, confidence interval, and p-value are not presented. H-15 �TABLE H-7. Summary of Group-by-Covariate Interactions for Malignant Skin Neoplasms in the Baseline-Followup Interval (Nonblacks Only) Group Ranch Hand re Interaction Basal Cell Group-byCarcinoma Occupation (Verified Only) Percent Number Percent Adj . Relative Risk (95% C.I.) p- Value Stratification Statistic Officer Variable Number Comparison n Yes No 372 14 358 3.8 96.2 475 16 459 3.4 96.6 1.14 (0.54,2.40) 0.726 Enlisted Flyer n Yes No 166 9 157 5.4 94.6 193 2 191 1.0 99.0 6.50 (1.36,31.01) 0.019 Enlisted Groundcretf n Yes No 416 6 410 1.4 98.6 539 11 528 2.0 98.0 0.81 (0.29,2.23) 0.680 �TABLE H-7. (continued) Summary of Group-by-Covariate Interactions for Malignant Skin Neoplasms in the Baseline-Followup Interval (Nonblacks Only) Group Ranch Hand Variable Interaction Stratification Statistic Burns Easily Intermediate Basal Cell Group-byCarcinoma Sun Reaction Reaction (Verified Index Plus Suspected) Tans Easily a Number Percent Adj. Relative Number Percent Risk (95% C.I.) p- Value n Yes No 68 0 68 0 100 77 4 73 5.2 94.8 n Yes No 192 17 175 8.9 91.1 262 15 247 5.7 94.3 1.76 (0.84,3.69) 0.134 n Yes No 693 19 674 2.7 97.3 867 28 839 3.2 96.8 0.90 (0.50,1.64) 0.731 Adjusted relative risk not calculated because of a zero cell count. Fisher's exact test. Comparison 0.121b �TABLE H-3. Summary of Group-by-Covariate Interactions for Malignant Skin Neoplasms in the Baseline-Follovup Interval (Original Comparisons Only*) Group Ranch Hand Variable Stratification Statistic Burns Easily 93 I t-» 00 Interaction Basal Cell Group-byIntermediate Carcinoma Sun Reaction (Verified Index Only) Tans Easily Original Comparison Number Percent Adj. Relative Risk (95% C.I.) p-Value Number Percent n Yes No 68 0 68 0 100 56 2 54 3.6 96.4 0.202b n Yes No 192 15 177 7.8 92.2 186 6 180 3.2 96.8 2.81 (1.05,7.55) 0.040 n Yes No 694 14 680 2.0 98.0 653 13 640 2.0 98.0 1.13 (0.52,2.43) 0.758 �TABLE H-8. (continued) Summary of Group-by-Covariate Interactions for Malignant Skin Neoplasms in the Baseline-Follovup Interval (Original Comparisons Only*) Group Ranch Hand Variable CE l-> vO Interaction Basal Cell Intermediate Carcinoma Group-by(Verified Sun Reaction Plus Index Suspected) Tans Easily Adj. Relative Percent Risk (95% C.I.) p-Value Number Percent Number n Yes No 68 0 68 0 100 55 4 51 7.3 92.7 —a n Yes No 192 17 175 8.9 91.1 185 8 177 4.3 95.7 2.38 (0.98,5.76) 0.055 n Yes No 694 19 675 2.7 97.3 646 23. 623 3.6 96.4 0.86 (0.46,1.60) 0.627 Stratification Statistic Burns Easily Original Comparison 0.038b �TABLE H-8. (continued) Summary of Group-by-Covariate Interactions for Malignant Skin Neoplasms in the Baseline-Follovup Interval (Original Comparisons Only*) Group Ranch Hand Number Percent Original Comparison Number Q3 NS 0 Percent Adj. Relative Risk (95% C . I . ) p-Value Variable Interaction Stratification Statistic Sun Exposure Related Group-byNeoplasms Skin Color (Verified Only) Peach n Yes No 768 31 737 4.0 96.0 744 20 724 2.7 97.3 0.20 (0.02,1.80) 0.150 Not Peach n Yes No 185 1 184 0.5 99.5 150 4 146 2.7 97.3 1.70 (0.95,3.04) 0.073 �TABLE H-8. (continued) Summary of Group-by-Covariate Interactions for Malignant Skin Neoplasms in the Baseline-Pollovup Interval (Original Comparisons Only*) Group Ranch Hand Original Comparison EC Interaction Statistic n Yes No 68 2 66 2.9 97.1 56 4 52 7.1 92.9 0.47 (0.08,2.74) 0.401 n Yes No 192 .18 174 9.4 90.6 186 8 178 4.3 95.7 2.74 (1.14,6.63) 0.025 n Yes No 693 19 674 2.7 97.3 652 26 626 4.0 96.0 0.75 (0.41,1.37) 0.343 Sun Exposure Related Neoplasms Group-byIntermediate (Verified Sun Reaction Plus Index Suspected) Tans Easily Number Percent *Blacks excluded. "Adjusted relative risk not calculated because of a zero cell count. b Fisher's exact test. Number Adj . Relative Risk (95% C.I.) p-Value Stratification Burns Easily Variable Percent �TABLE H-9. Followup Participants With Verified Malignant Systemic Neoplasms in Baseline-Followup Interval by Group (Original Comparisons Only) Group Site Ranch Hand Original Comparison Total Oral Cavity and Pharynx 2a ' b 0 2 Thyroid Gland 0 1 1 Bronchus and Lung 1 0 1 Colon 0 3C ' d 3 Kidney and Bladder 2 1 3 Prostate 1 1 2 Testicles 1 0 1 _1 _0 J. 8 6 14 Connective and Other Soft Tissue Total "includes one Ranch Hand with separate malignancies of tongue and epiglottis and also malignant neoplasm of bone. Includes one Ranch Hand with separate malignant neoplasms of tongue and oropharynx and secondary malignant neoplasm of other site. c Incudes one comparison with secondary malignant neoplasms of liver and bone and bone marrow. Includes one comparison with secondary malignant neoplasm of liver. H-22 �TABLE B-10. Unadjusted Analyses of Followup Participants with Verified and Suspected Malignant Systemic Neoplasms in the Baseline-^ollowp Interval by Group (Original Comarisons Only) Group Statistic Original Ranch Hand Comparison Est. Relative Risk ( 5 C.I.) 9% p-Value Verified Number/% Total Neoplasms 8 08 6 06 .% .% 12 9 1.26 ( . 3 3 6 ) 04,.3 071 .9 Verified & Suspected Number/% Total Neoplasms 12 1 2 11 1 2 1.03 ( . 5 2 3 ) .% .% 04,.4 23 14 099 .9 Status H-23 �TABLE H-ll. Association Between Baseline-Follovup Interval Incidence of All Malignant Systemic Neoplasms Combined and the Covariates for the Combined Followup Ranch Hand and Comparison Groups Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Age l' 10 4 0.1 0.8 4.6 <0.001 4 16 4 0.4 1.3 4.6 0.001 Race en ro Born XL942 Born 1923-41 Born <1922 Nonblack Black 15 0 0.7 0.0 0.999 23 1 1.1 0.7 0.999 Occupation Officer Enlisted Flyer Enlisted Groundcrew 10 2 3 1.2 0.5 0.3- 0.056 13 4 7 1.5 1.0 0.7 0.193 Total Lifetime Smoking (Pack- Years) 0 >0-20 >20-40 >40 2 6 5 2 0.3 0.6 1.2 1.3 0.220 4 11 7 2 0.6 1.0 1.7 1.3 0.374 Total Lifetime Alcohol Consumption (Drink- Years) 0 >0-5 >5-30 >30-100 >100 0 2 3 7 1 0 0.3 0.4 1.4 0.9 0.082 1 5 5 9 2 0.7 0.7 0.7 1.8 1.9 0.233 Exposures to Carcinogens Asbestos Yes 3 No 12 0.6 0.7 0.999 5 19 1.0 1.1 0.999 Nonmedical X Rays Yes 5 No 10 0.9 0.6 0.364 10 14 1.9 0.8 0.049 �TABLE H-ll. (continued) Association Between Baseline-Follovup Interval Incidence of All Malignant Systemic Neoplasms Combined and the Covariates for the Combined Followup Ranch Hand and Comparison Groups Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Industrial Chemicals Yes No 5 10 0.4 0.9 0.196 11 13 0.9 1.2 0.682 Herbicides Yes No 9 6 0.7 0.6 0.999 14 10 1.1 1.0 0.999 Insecticides Yes No 4 11 0.3 1.2 0.014 10 14 0.7 1.5 0.091 Degreasing Chemicals Exposures to Carcinogens (continued) Yes No 6 9 0.5 0.9 0.191 14 10 1.0 1.0 0.999 Yes No 4 11 0.8 0.6 0.757 5 18 1.0 1.0 0.999 Anthracene Yes No 0 15 0.0 0.7 0.999 0 24 0.0 1.0 0.999 Arsenic Yes No 0 15 0 0.7 0.999 2 22 4.8 1.0 0.069 Benzene Yes No 1 14 1.2 0.6 0.424 1 23 1.2 1.0 0.587 Yes No 1 14 7.1 0.6 0.088 1 22 7.1 1.0 0.131 EC N> Composite Carcinogen Exposure Exposure to Individual Carcinogens Benzidine „ �TABLE H-ll= (continued) Association Between Baseline-Follovup Interval Incidence of All Malignant Systemic Neoplasms Combined and the Covariates for the Combined Follovup Ranch Hand and Comparison Groups Verified Covariate Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Chroma tes Yes No 2 13 2.3 0.6 0.110 2 22 2.3 1.0 0.232 Coal Tar Yes No 2 13 2.7 0.6 0.079 2 22 2.7 1.0 0.175 Creosote Exposure to Individual Carcinogens (continued) Yes No 1 14 0.6 0.7 0.999 3 21 1.8 1.0 0.240 Aminodiphenyl Yes No 0 15 0 0.7 0.999 1 23 16.7 1.0 0.061 Chloromethyl Ether Yes No I 14 4.4 0.6 0.140 2 22 8.7 1.0 0.023 Mustard Gas Yes 0 No 15 0 0.7 0.999 1 23 11.1 0.090 Naphthylamine Yes No 2 13 3.6 0.6 0.050 3 21 5.4 0.9 0.019 Cutting Oils Yes No 3 12 1.2 0.6 0.204 5 19 2.1 0.9 0.100 Trichloroethylene Yes No 3 12 1.5 0.6 0.135 4 20 2.0 1.0 0.149 CO i OS 1.0 �TABLE H-ll. (continued) Association Between Baseline-Follovup Interval Incidence of All Malignant Systemic Neoplasms Combined and the Covariates for the Combined Followup Ranch Hand and Comparison Groups Verified Covariate Exposure to Individual Carcinogens (continued) BC to Covariate Category Number* Percent Verified and Suspected p-Value Number* Percent p-Value Ultraviolet Light Yes No 1 14 2.0 0.6 0.286 I 23 2.0 1.0 0.417 Vinyl Chloride Yes No 0 15 0.0 0.7 0.999 1 23 3.0 1.0 0.294 *Number of participants with malignant systemic neoplasms . �TABLE H-12. Adjusted Analyses of Follovup Participants for the Incidence of All Malignant Systemic Neoplasms During the Baseline-Followup Interval (Original Comparisons Only) Variable Systemic Neoplasms (Verified) Systemic Neoplasms (Verified & Suspected) Adj. Relative Risk (95X C.I.) 1.47 (0.50,4.33) p-Value Covariate Remarks AGE (p<0.001) **** **** 0.481 GRP*OCC (p=0.028) AGE (p<0.001) RACE*PACKYR (p=0.032) ****Group-by-covariate interaction—adjusted relative risk, confidence interval, and p-value not presented. H-28 �TABLE H-13. Summary of Group-by-Occupation Interaction for All Malignant Systemic Neoplasms (Verified Plus Suspected) During the Baseline-Follovup Interval Group Ranch Hand Variable Percent Number 380 5 375 1. 3 98.7 484 8 476 1.7 98.3 N5 VD n Yes No 175 4 171 2. 3 97.7 209 0 209 0 100 n Yes No 457 3 454 0.7 99.3 598 4 594 0.7 99.3 Statistic n Yes No Enlisted Flyer Enlisted Groundcrev Systemic Cancer (Verified Group-byoccupation Plus Suspected) Stratification Officer . CO Interaction Number 'Adjusted relative risk not calculated because of a zero cell. 'Fisher's exact test. Comparison Percent Adj. Relative Risk (95% C.I.) p-Value 0.79 (0.25,2.44) 0.679 —a 0.042b 0.93 (0.20,4.26) 0.923 �TABLE H-14. Summary of Group-by-Occupation Interaction for All Malignant Systemic Neoplasms (Verified Plus Suspected) During the Baseline-Follovup Interval (Original Comparisons Only) Group • Ranch Hand Variable Number CO O Statistic n Yes No 380 5 375 1.3 98.7 350 7 343 2.0 98.0 Enlisted Flyers n Yes No 175 4 171 2.9 97.1 174 0 174 0.0 100.0 Enlisted Groundcrew All Systemic Cancer Group-by(Verified Occupation Plus Suspected) Stratification Officers ffi Interaction Original Comparison n Yes No 457 3 454 0.7 99.3 430 4 426 0.9 99.1 Percent 'Adjusted relative risk not calculated because of a zero cell count. 'Fisher's exact test. Number Percent Adj. Relative Risk (95% C.I.) p-Value 0.72 (0.22,2.30) 0.577 _ —a u. 0.123b 0.71 (0.16,3.28) 0.666 �TABLE H-15. Unadjusted Analyses of Followup Participants with Lifetime Occurrence of Verified and Suspected Lifetime Neoplasms by Group (Nonblacks and Blacks, Original Comparisons Only) Group Ranch Hand Site Skin Neoplasm Behavior and Status Comparison Number** Percent Number** Percent Total** p-Value*** 66 75 6.5 7.4 48 62 5.0 6.5 114 137 0.177 0.479 Benign Verified Verified and Suspected 84 86 8.3 8.5 67 70 7.0 7.3 151 156 0.310 0.360 Uncertain Behavior and Unspecified Nature: Verified Verified and Suspected 1 1 0.1 0.1 1 1 0.1 0.1 2 2 0.999 0.999 Any Skin Neoplasm* Verified Verified and Suspected Systemic Malignant Verified Verified and Suspected 150 159 14.8 15.7 111 128 11.6 13.4 261 287 0.046 0.160 Malignant Verified Verified and Suspected 17 21 1.7 2.1 14 19 1.5 2.0 31 40 0.722 0.999 Benign Verified Verified and Suspected 51 57 5.0 ' 5.6 52 60 5.5 6.3 103 117 0.687 0.568 Uncertain Behavior and Unspecified Nature: Verified Verified and Suspected 15 15 1.5 1.5 11 14 1.2 1.5 26 29 0.560 0.999 Any Systemic Neoplasm13 Verified Verified and Suspected 81 91 8.0 9.0 71 86 7.4 9.0 152 177 0.673 0.999 H-31 �TABLE H-15. (continued) Unadjusted Analyses of Follovup Participants vith Lifetime Occurrence of Verified and Suspected Lifetime Neoplasms by Group (Nonblacks and Blacks, Original Comparisons Only) Group Ranch Hand Site Neoplasm Behavior and Status Comparison Number** Percent Number** Percent Total** p-Value*** All Malignant, Benign, Neoplasms Uncertain Behavior, Unspecified Nature Verified 216 Verified and Suspected 231 21.3 22.7 167 196 17.5 20.5 383 427 *Sample sizes: 1,016 Ranch Hands, 955 Original Comparisons. **Number of participants. ***Fisher's exact test. "Participant has one or more malignant, benign, or unspecified skin neoplasms. b Participant has one or more malignant, benign, or unspecified systemic neoplasms. Participant has one or more malignant or benign skin or systemic neoplasms. H-32 0.035 0.251 �TABLE H-16. Unadjusted Analyses of Nonblack Follovup Participants with Lifetime Occurrence of Verified and Suspected Lifetime Malignant Skin Neoplasms by Cell Type and Group (Original Comparisons Only) Group Cell Type Basal Cell Carcinoma Status Verified Verified & Suspected Squamous Cell Carcinoma Verified Verified & Suspected as i Melanoma Verified Verified & Suspected All Malignant Verified Skin Neoplasms Verified & Suspected Sun-Exposure Verified Related Malignant Neoplasms9 Verified & Suspected Statistic Ranch Hand Number /% Total Neoplasms 53 77 5.5% Number /% Total Neoplasms 59 88 6.2% Number/% Total Neoplasms 4 6 0.4% Number /% Total Neoplasms 4 6 0.4% Number /% Total Neoplasms 5 6 0.5% Number/% Total Neoplasms 5 6 0.5% Number/% Total Neoplasms 66 103 6.9% Number /% Total Neoplasms 75 117 7.9% Number /% Total Neoplasms 59 90 6.2%. Number /% Total Neoplasms 65 101 6.8% Original Comparison 34 Est. Relative Risk ( 5 C.I.) p-Value 9% 3.8% 1.49 (0.96,2 .32) 0.079 5.4% 1.16 ( . 9 1 .72) 07, 0.486 0.6% 0.75 ( . 0 2. 0 02, 8) 0.747 0.6% 0.75 8) (0.20,2 . 0 0.747 0.6% 0.94 (0.27,3 .25) 0.999 0.7% 0.78 (0.24,2 .57) 0.768 5.4% 1.31 ( . 9 1 .93) 08, 0.176 6.9% 1.15 (0.81,1 .63) 0.478 4.4% 1.45 (0.96,2 .19) 006 .9 5.9% 1.16 (0.80,1 .69) 0.448 52 48 68 5 6 5 6 5 6 6 7 48 73 62 90 39 58 53 75 *Number and percent of participants; total number of malignant neoplasms of specified cell type. **Number of participants — 956 Ranch Hands and 897 Original Comparisons. a Basal cell carcinoma, melanoma, and malignant epithelial neoplasms NOS. �TABIEH-17. Association Between Lifetime Incidence of Sun-Exposure Belated Skin Malignancies and the Covariates for Gcnbined Followup Ranch Band and Gcnparison Nonblack Participants Verified Covariate Covariate Category Number* Verified and Suspected Percent p-Value Number* Percent p-Value Age Bom >1942 Bom 1923-41 Born <L922 8 83 23 9.2 6.9 2.6 <0.001 28 99 12 3.2 8.3 13.8 <0.001 Occupation Officer fiilisted Flyer Enlisted Groundcrev 60 17 37 7.1 4.7 3.9 009 .0 75 20 44 8.8 5.6 46 . 001 .0 Total Lifetime Smoking (Pack-Years) 0 X)-20 >20-40 >40 35 40 24 15 5.7 40 . 6.1 96 . 0.021 40 49 34 16 6.5 4.9 8.7 10.2 002 .1 Total Lifetime Alcohol Consumption (Drink-Years) 0 X)-5 >5-30 >30-100 >100 8 39 29 24 11 5.7 5.4 4.4 5.0 10.6 0.140 9 45 36 35 11 6.4 6.3 5.5 7.3 10.6 0.341 Ethnic Background3 A B C D 95 16 1 0 60 . 3.8 1.6 00 . 005 .5 118 17 1 0 7.5 40 . 1.6 00 . 007 .0 Skin Color Dark Medium Pale Dark Peach Pale Peach 0 2 10 76 26 00 . 2.7 3.3 60 . 5.0 025 .8 0 2 12 90 35 00 . 2.7 3.9 7.1 6.7 010 .9 �TABLE B-17. (continued) Association Between Lifetime Incidence of SUV-Exposure Related Skin Malignancies and the Covariates for Combined BoUowup Ranch Bend and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Verified and Suspected Percent p-Value Number* Percent p-Value Eye. Color Brown Hazel Green Grey Blue 34 31 4 5 40 5.3 6.8 3.4 5.4 4.7 046 .6 40 32 7 7 53 6.2 7.0 5.9 7.5 6.2 090 .6 Hair Color Black Dark Brown Light Brown Red Blond 22 46 36 2 8 5.0 44 . 6.4 12.5 7.4 0.230 26 58 42 3 10 5.9 5.6 7.5 18.8 9.3 006 .9 Residential History (Average Latitude) >37° <37° 48 66 4.2 6.5 006 .2 56 83 4.9 8.1 003 .0 Skin Reaction to First 30 Min. of Sun Exposure Burns Usually Bums Bums Mildly Rarely Bums 23 39 34 17 9.3 9.1 4.2 2.5 O01 .0 28 47 38 25 11.3 11.0 4.7 3.7 O01 .0 Skin Reaction to >2 Hrs of Sun After First Exposure Bums Painfully Bums Becomes Red No Reaction 13 32 46 23 10.8 9.5 4.4 3.5 O01 .0 15 35 59 30 12.5 10.4 5.7 4.5 O01 .0 Skin Reaction After Repeated Exposure to Sun Freckles, No Tan Tans Mildly Tans Moderately Tans Deep Brown 5 33 40 35 11.1 10.5 3.9 4.5 <D.001 6 38 51 43 13.3 12.1 5.0 5.5 001 .0 �TABLE B-17. (continued) and the Covariates for Grinned FoUowup Ranch Eland and Comparison Nonblack Participants Verified Covariate Covariate Category Sun Reaction Index Tends to Burn Mild Reaction Tends to Tan Exposures to Carcinogens Number* Verified and Suspected Percent p-Value Number* Percent p-Value 9.7 9.3 3.7 O01 .0 16 48 74 11.0 10.6 4.7 <0.001 Asbestos Yes 19 No 95 4.2 5.6 0.288 26 113 5.7 6.6 0.520 Nonmedical X Rays Yes 31 No 83 6.3 5.0 0.252 39 100 7.9 6.0 0.143 Industrial Chemicals Yes 54 No 60 4.8 5.8 0.336 65 74 5.8 7.1 0.220 Herbicides Yes 74 No 40 5.9 4.4 014 .4 91 48 7.2 5.3 006 .7 Insecticides Yes 76 No 38 5.8 4.5 0.200 90 49 6.9 5.7 0.324 Degreasing Chemicals Composite Carcinogen Exposure 14 42 57 Yes 64 No 50 5.1 5.5 067 .9 77 62 6.1 6.9 0.534 Yes 21 No 91 4.3 5.5 0.355 24 113 4.9 6.8 0.141 �TABLE H-17. Continual) ad the Covariates for Conbined Followup Ranch Band and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Verified and Suspected Percent p-Value Number* Percent p-Value Vs Anthracene Yes 0 No 114 00 . 5.3 099 .9 0 139 00 . 6.4 099 .9 Arsenic Exposure to Individual Carcinogens Yes 4 No 109 9.8 5.1 0.163 5 133 12.2 6.3 0.181 Benzene Yes 6 No 108 8.1 5.2 0.281 7 132 9.5 6.3 0.327 Benzidine Yes 1 No 113 9.1 5.3 049 .4 1 138 9.1 6.4 0.519 Chromates Yes 4 No 108 4.8 5.2 099 .9 5 132 6.0 6.4 099 .9 Coal Tar Yes 2 No 112 2.9 5.3 050 .8 3 136 4.4 6.5 079 .9 Creosote Yes 9 No 105 5.7 5.2 0.715 10 129 6.3 6.4 099 .9 Aminodiphenyl Yes 0 No 113 00 . 5.2 099 .9 0 138 00 . 6.4 099 .9 Chloromethyl Ether Yes 2 No 112 9.5 5.2 0.305 2 137 9.5 6.4 0.641 Mustard Gas Yes 0 No 114 00 . 5.3 099 .9 0 139 00 . 6.4 099 .9 �TAHE H-17. (continued) and tiie Govariates for Combined Followup Ranch Band and Comparison NonUack Participants Verified Covariate Category Verified and Suspected Number* 5.8 5.2 0.751 4 134 7.7 6.3 050 .7 Yes 12 No 102 5.3 5.3 099 .9 15 124 6.6 6.4 086 .8 Yes 5 No 109 2.7 5.5 010 .2 7 132 3.8 6.7 0.156 Ultraviolet Light Yes 5 No 109 11.4 5.1 008 .7 5 134 11.4 6.3 020 .0 Vinyl Chloride Exposure to Individual Carcinogens (continued) p-Value Trichloroethylene Covariate Yes 0 No 114 0.0 5.3 046 .0 1 138 3.2 6.5 0.718 Naphtnylamine , Cutting Oils £ Number* Percent Yes 3 No 110 *Number of participants with sun-exposure related skin malignancies. a • Ethnic Background• , Welsh, Scottish, Irish. B—Scandinavian, German, Polish, Russian, other Slavic, Jewish, French. C—Spanish, Italian, Greek. D—Mexican, American Indian, Asian. Percent p-Value �TABLE H-18. Summary of Group-by-Covariate Interactions for Lifetime Malignant Skin Neoplasm Incidence (Nonblacks Only) Group Ranch Hand Number Percent 03 I M3 Interaction Stratification Statistic Burns Easily Variable n Yes No 68 3 65 n Yes No 192 25 167 n Yes No 693 31 662 Basal Cell Group-byIntermediate Carcinoma Sun Reaction Reaction (Verified Index Plus Tans Easily Suspected) Original Comparison Number Percent Adj. Relative Risk ( 5 C.I.) p-Value 9% 11.7 88.3 0.29 (0.07,1.28) 0.102 95.6 77 9 68 13.0 87.0 262 21 241 8.0 1.97 (1.04,3.73) 0.038 92.0 867 36 831 95.8 4.4 4.5 95.5 4.2 1.15 (0.70,1.88) 0.591 �TABLE H-19. Adjusted Analyses of Nonblack Follovup Participants for Lifetime Malignant Skin Neoplasm Incidence (Original Comparisons Only) Adj. Relative Risk (95% C.I.) Variable Status Basal Cell Carcinoma Verified **** **** GRP*SUNREAC (p=0.010) SKIN*SUNREAC (p<0.001) OCC*AGE (p=0.010) OCC*SUNREAC (p=0.007) LAT (p=0.006) Verified & Suspected **** **** GRP*SUNREAC (p=0.003) AGE (p<0.001) LAT (p=0.002) SUNREAC*OCC (p=0.043) SKIN*SUNREAC (p<0.001) Verified **** **** GRP*SUNREAC (p=0.045) SKIN*LAT (p=0.028) SKIN*SUNREAC (p<0.001) OCC*AGE (p=0.023) Verified & Suspected **** **** GRP*SUNREAC (p=0.016) AGE (p<0,001) SUNREAC*LAT (p=0.017) SKIN*SUNREAC (p<0.001) Sun-Exposure Skin Neoplasms p-Value Covariate Remarks ****Group-by-covariate interactions—adjusted relative risk, confidence interval, and p-value not presented. H-40 �TABLE H-20. Summary of Group-by-Covariate Interactions for Lifetime Malignant Skin Neoplasm Incidence (Original Comparisons Only*) Group Ranch Hand Number Percent EC Interaction Stratification Statistic Burns Easily Variable n Yes No 68 3 65 n Yes No 192 23 169 n Yes No 693 27 666 Basal Cell Carcinoma Group-byIntermediate (Verified Sun Reaction Only) Index , Tans Easily Original Comparison Number Percent 95.6 56 6 50 10.7 89.3 12.0 88.0 186 8 178 95.7 652 19 633 97.1 Adj. Relative Risk ( 5 C . . p-Value 9% 1) 4.4 3.9 96.1 4.3 2.9 0.17 ( . 2 1.67) 0.128 00, 3.76 (1.60,8.86) 0.002 1.45 (0.79,2.67) 0.226 �TABLE H-20. (continued) Sunmary of Group-by-Covariate Interactions for Lifetime Malignant Skin Neoplasm Incidence (Original Comparisons Only*) Group Ranch Hand Number Percent Original Comparison Number Percent 83 to Interaction Adj. Relative Risk (95% C.I O p-Value Stratification Statistic Burns Easily Variable n Yes No 68 3 65 4.4 95.6 56 8 48 14.3 85.7 0.19 (0.03,1.15) 0.070 n Yes No 192 25 167 13.0 87.0 186 10 176 5.4 94.6 3.25 (1.47,7.18) 0.003 n Yes No 693 31 662 4.5 95.5 652 29 623 4.4 95.6 1.10 (0.65,1.86) 0.719 Basal Cell Intermediate Carcinoma Group-by(Verified Sun Reaction Plus Index Suspected) Tans Easily �TABLE H-20. (continued) Summary of Group-by-Covariate Interactions for Lifetime Halignant Skin Neoplasm Incidence (Original Comparisons Only*) Group Original Variable Comparison Number Percent Adj. Relative Risk ( 5 C.I . ) p-Value 9% Stratification Statistic Burns Easily at i u> Interaction Ranch Hand Number Percent n Yes No 68 7 61 10.3 89.7 56 8 48 14.3 85.7 1.20 ( . 4 4 22) 0.777 03,. n Yes No 192 26 166 13.5 86.5 186 10 176 5.4 3.98 (1.70,9.33) 0.002 94.6 n Yes No 693 32 661 652 34 618 94.8 Sun Intermediate Exposure Group-byRelated Sun Reaction Cancers Index (Verified Tans Easily Plus Suspected) 4.6 95.4 5.2 1.17 (0.31,4.41) 0.819 �TABLE H-2Q. (continued) Summary of Group-by-Covariate Interactions for Lifetime Malignant Skin Neoplasm Incidence (Original Comparisons Only*) Group Variable Original Comparison Number Percent Adj. Relative Risk (95% C.I.) p-Value w Stratification Statistic Burns Easily Sun Exposure Related Cancers (Verified Only) Interaction Ranch Hand Number Percent n Yes No 68 7 61 10.3 89.7 56 6 50 10.7 89.3 0.80 (0.25,2.53) 0.700 n Yes No 192 24 168 12.5 87.5 186 8 178 4.3 95.7 3.42 (1.56,7.51) 0.002 n Yes No 693 28 665 4.0 96.0 652 24 628 3.7 96.3 0.97 (0.59,1.61) 0.911 Group-byIntermediate Sun Reaction Index *Blacks excluded. Tans Easily �TABLE H-21. Summary of Followup Participants With Lifetime Incidence of Verified Malignant Systemic Neoplasms by Group (Original Comparisons Only) Group Ranch Hand Original Comparison Total Eye I 0 1 Oral Cavity and Pharynx -a , b 0 3 Larynx 0 1 1 Thyroid Gland 0 2 2 Bronchus and Lung 2 0 2 Colon 0 Kidney and Bladder 4 3 7 Prostate 2 2 4 Testicles 3 0 3 Connective and Other Soft Tissue 1 1 2 J.e _l f _2 17 14 31 Site Ill-Defined Sites Total 4c,d 4 a Includes one Ranch Hand with separate malignancies of tongue and epiglottis and also malignant neoplasm of bone. Includes one Ranch Hand with separate malignant neoplasms of tongue and oropharynx and secondary malignant neoplasm of other site. c lncudes one Comparison with secondary malignant neoplasms of liver and bone and bone marrow. d lncludes one Comparison with secondary malignant neoplasm of liver. 'Malignant neoplasm of thorax. Malignant neoplasm of face, head, or neck. H-45 �TABLE H-22. Unadjusted Analyses of the Lifetime Incidence Bates of Ml Malignant Systemic Neoplasms Combined, by Group (Original Comparisons Only) Statistic Group Original Ranch Hand Comparison Est. Relative Risk ( 5 C.I.) 9% p-Value Verified Number/% Total Neoplasms 17 1 7 14 1.5% .% 25 19 1.14 ( . 6 2 3 ) 05,.3 0.722 Verified & Suspected Number/% Total Neoplasms 21 2.1% 36 1 0 (.619) . 4 05,.5 Status IM6 19 2 0 .% 24 099 .9 �TABLE H-23. Summary of Group-by-Occupation Interactions for Lifetime Incidence of All Malignant Systenic Neoplasms Combined Group Variable Interaction Stratification Statistic Ranch Hand Number Percent Comparison Number Percent Officer Group-by- . Occupation a 380 8 372 2.1 97.9 484 12 472 2.5 97.5 Enlisted Flyer n Yes No 177 5 172 2.8 97.2 210 0 210 0 100 Enlisted Groundcrew Malignant Systemic Neoplasm (Verified Only) n Yes No n Yes No 459 4 455 0.9 99.1 599 5 594 0.8 99.2 Adjusted relative risk not calculated because of a zero cell. Fisher's exact test. Adj. Relative Risk (95% C.I.) p-Value 0.84 (0.34,2.08) 0.698 _ w. 0.019b 1.07 (0.28,4.03) 0.922 �TABLE H-23. (continued) Summary of Group-by-Occupatioh Interactions for Lifetime Incidence of All Malignant Systemic Neoplasms Combined Group Ranch Hand Number Percent ta i 00 Interaction Comparison Number Percent Malignant Systemic Group-byNeoplasm Occupation (Verified plus Suspected) a Stratification Statistic Officer Variable n Yes No 380 9 371 2.4 97.6 484 14 470 2.9 97.1 Enlisted Flyer n Yes No 175 7 168 4.0 96.0 209 0 209 0 100 Enlisted Groundcrew n Yes No 457 5 452 1.1 98.9 598 8 590 1.3 98.7 Adjusted relative risk not calculated because of a zero cell. Fisher's exact test. Adj. Relative Risk (95% C.I.) p-Value 0.80 (0.34,1.89) 0.616 0.004b 0.79 (0.26,2.47) 0.690 �TABLE H-24. Adjusted Analyses for Lifetime Incidence of All Malignant Systemic Neoplasms Combined (Original Comparisons Only) Variable Malignant Systemic Neoplasms (Verified) Malignant Systemic Neoplasms (Verified & Suspected) Adj. Relative Risk (95% C.I.) p-Value Covariate Remarks **** **** GRP*OCC (p-0.034) AGE (p<0.001) **** **** GRP*OCC (p=0.003) AGE (p<0.001) RACE*PACKYR (p=0.033) ****Group-by-covariate interaction—adjusted relative risk, confidence interval, and p-value not presented. H-49 �TABLE H-25. Summary of Group-by-Occupation Interactions for Lifetime Incidence of All Malignant Systemic Neoplasms Combined (Original Comparisons Only) Group Variable Interaction Stratification Statistic Ranch Hand Number Percent Original Comparison Number Percent Officer EC Malignant Systemic Group-byNeo plasms Occupation (Verified Only) n Yes No 380 8 372 2.1 97.9 350 9 341 2.6 97.4 Enlisted Flyer n Yes No 177 5 172 2.8 97.2 174 0 174 0 100 Enlisted Groundcrev n Yes No 459 4 455 0.9 99.1 431 5 426 1.2 98.8 Adj. Relative Risk (95% C.I.) p-Value 0.90 (0.34,2.38) 0.834 0.061b 0.81 (0.21,3.07) 0.757 �TABLE H-25. (continued) Summary of Group-by-Occupation Interactions for Lifetime Incidence of All Malignant Systemic Neoplasms Combined (Original Comparisons Only) Group Ranch Hand Number Percent Original Comparison Number Percent CC Un a Statistic n Yes No 380 9 371 2.4 97.6 350 11 339 3.1 96.9 Enlisted Flyer n Yes No 175 7 168 4.0 96.0 174 0 174 0 100 Enlisted Groundcrew Malignant Systemic Group-byNeoplasms Occupation (Verified Plus Suspected) Stratification Officer Variable Interaction n Yes No 457 5 452 1.1 98.9 430 8 422 1.9 98.1 Adjusted relative risk not calculated because of zero cell. b Fisher's exact test. Adj. Relative Risk (95% C.I.) p-Value 0.81 (0.33,1.98) 0.638 0.015a 0.59 (0.19,1-85) 0.369 �TABLE H-26. unadjusted Exposure Index Analyses for FdUxwup Participants in the Bas^iTv>_Rn llnuin Tnfprual Exposure Index3 Low Variable Medium High Contrast Est. Relative Risk (95%C.I.) p-Value 2.5 97.5 Overall M vs. L H vs. L 018 .7 2.73 ( . 1 1 . 5 0.216 07,05) 1.03 ( . 1 5 2 ) 0 9 9 02,.3 .9 52 1 51 1.9 98.1 Overall M vs. L H vs. L 0.073 0.27 ( . 5 1 4 ) 0.145 00,.1 0 1 ( . 2 1 3 ) 0.113 .6 00,.5 1.3 98.7 129 1 128 08 . 99.2 Overall 0.627 01,.2 M vs. L 0.61 ( . 0 3 7 ) 0.674 00,.2 H vs. L 0.35 ( . 4 3 4 ) 0.623 127 12 115 9.4 90.6 121 4 121 3.3 96.7 Overall 002 .4 M vs. L 3.16 ( . 9 1 . 7 0 0 8 09,00) .6 H vs. L 1.03 ( . 5 4 2 ) 0 9 9 02,.3 .9 11.1 88.9 61 2 59 3.3 96.7 52 1 51 1.9 98.1 Overall 003 .7 00,.1 M vs. L 0.27 ( . 5 1 4 ) 0.145 H vs. L 0.16 ( . 2 1 3 ) 0.113 00,.5 2.9 97.1 149 2 147 1.3 98.7 129 1 128 08 . 99.2 Overall M vs. L H vs. L Nunber Percent Number Percent Number Percent Occupation Statistic Officer n Abnormal Normal • 125 3 122 2.4 97.6 127 8 . 119 6.3 93.7 121 3 118 n Abnormal Normal 54 6 48 11.1 8. 89 61 2 59 3.3 96.7 n Abnormal Normal 138 3 135 2.2 97.8 149 2 147 Officer n Abnormal Normal 125 4 121 3.3 96.8 Enlisted Flyer n Abnormal Normal 54 6 48 Enlisted Groundcrew n Abnormal Normal 138 4 134 Pagal Cell Enlisted Flyer Carcinoma (Verified Only)b Enlisted Groundcrew Basal Cell Carcinoma (Verified and Suspected) 0.372 0 4 ( . 8 2 5 ) 0.432 .6 00,.3 ,0.26 ( . 3 2 3 ) 0.372 00,.7 �TABLE B-26. (continued) Unadjusted Exposure Index Analyses for PoLLovup Participants for Occurrence of Malignant Neoplasms in the Baseline-Followp Interval Exposure Indexa Medium Low High Contrast Est. Relative Risk ( 5 C I ) p-Value 9% .. 2.5 97.5 Overall M vs. L H vs. L 006 .9 3.10 ( . 2 1 . 4 0 1 6 08,17) . 3 1.03 ( . 1 5 2 ) 0 9 9 02,.3 .9 52 1 51 1.9 98.1 Overall M vs. L H vs. L 0.073 0.27 ( . 5 1 4 ) 0.145 00,.1 0.16 ( . 2 1 3 ) 0.113 00,.5 u> 2.0 9. 80 129 2 127 1.6 98.4 Overall M vs. L H vs. L 0.929 0.93 ( . 8 4 6 ) 0.999 01,.6 0.71 ( . 2 4 3 ) 0 9 9 01,.1 .9 127. 13 114 10.2 8. 98 121 4 117 3.3 96.7 Overall M vs. L H vs. L 0.021 3.45 ( . 9 1 . 9 0 0 2 10,08) .4 1.03 ( . 5 4 2 ) 0 9 9 02,.3 .9 Sun-Exposure Belated Malignancies (Verified and Suspected) 11.1 88.9 61 2 59 3.3 96.7 52 1 51 1.9 98.1 Overall M vs. L H vs. L 0.073 0.27 ( . 5 1 4 ) 0.145 00,.1 0.16 ( . 2 1 3 ) 0.113 00,.5 2.9 97.1 149 3 146 2.0 9. 80 129 2 127 1.6 98.4 Overall M vs. L M vs. L 0.742 0.69 ( . 5 3 1 ) 0.714 01,.3 0.53 ( . 0 2 9 ) 0.685 01,.3 n Abnormal Normal 125 3 122 2.4 97.6 127 9 118 7.1 92.9 121 3 118 Bilisted Flyer n Abnormal Normal 54 6 43 11.1 88.9 61 2 59 3.3 96.7 138 3 135 2.2 97.8 149 3 146 Officer Ul Statistic n Bilisted Abnormal Groundcrev Normal Sun-Exposure Belated Malignancies (Verified Cnly)b Occupation Officer Variable Number Percent Number Percent Number Percent n Abnormal Normal 125 4 121 3.2 96.8 Enlisted Flyer n Abnormal Normal 54 6 48 n Bilisted Abnormal Groundcrev Normal 138 4 134 �TABLE H-26. (continued) Unadjusted Exposure Index Analyses for Followjp Participants for Occurrence of Malignant Neople in the Baseline-Follouup Interval Exposure Index8 Medium Low Variable Number Percent High Number Percent Number Percent Contrast Est. Relative Risk ( 5 C I ) p-Value 9% . . Occupation Statistic Officer n Abnormal Normal 127 1 126 0.8 99.2 130 3 127 2.3 97.7 123 0 123 00 . 100 0. Overall 016 .8 M vs. L 2.98 ( . 1 2 . 0 0.622 03,90) H vs. L 0.34 ( . 1 8 4 ) 0 9 9 00,.6* . 9 n Abnormal Normal 55 1 54 1.8 98.2 65 1 64 1.5 98.5 57 0 57 00 . 100 0. Overall 0.612 M vs. L 0 8 ( . 5 1 . 1 0 9 9 .4 00,38) .9 H vs. L 0.32 ( . 1 7 9 ) 0 4 1 00,.2* . 9 Enlisted Groundcrev n Abnormal Normal 154 0 154 00 . 100 0. 163 1162 0.6 99.4 142 1 141 0.7 99.3 Overall 0.598 M vs. L 2.85 ( . 2 7 . 5 * 0 9 9 01,05) .9 H vs. L 3.28 ( . 3 8 . 6 * 0 4 0 01,10) . 8 Officer n Abnormal Normal 127 1 126 08 . 99.2 130 4 126 3.1 96.9 123 0 123 00 . 100 0. Overall H vs. L H vs. L 001 .3 40 (.43.9 030 .0 04,62) .7 0.34 ( . 1 8 4 ) 0 9 9 00,.6* . 9 Enlisted Flyer n Abnormal Normal 55 2 53 3.6 96.4 65 1 64 1.5 98.5 57 1 '6 5 1.8 98.2 Overall M vs. L H vs. L 0 4 (.446) . 1 00,.9 04 (.453) .7 00,.7 Enlisted Groundcrev n Abnormal Normal 154 0 154 00 . 100 0. 163 1 162 06 . 99.4 142 2 140 1.4 98.6 Overall M vs. L H vs. L 0.323 2.85 ( . 2 7 . 5 * 0 9 9 01,05) .9 5.50 ( . 6 1 5 5 ) 0.229 02,1.1* Enlisted Systemic Malignancies Flyer (Verified Only)c Systemic Malignancies (Verified and Suspected)0 a Nunber and percent b Nonblacks only. c of total participants. Blacks and nonblacks. *0.5 added to each cell to calculate relative risk and confidence interval, due to zero cell frequency. 078 .0 0.593 0.615 �TABLE B-27. unadjusted Exposure Index Analysis for FaUowp Participants for Lifef-imp Occurrence of Malignant Neoplasms Low Number Percent Exposure Index3 Medium High Number Percent Number Percent Contrast Est. Relative Risk ( 5 C.I.) p-Value 9% 7.4 92.6 Overall M vs. L H vs. L 061 .4 1.60 ( . 0 4 2 ) 0 4 4 06,.7 .6 1.36 ( . 9 3 7 ) 0.613 04,.6 52 1 51 1.9 98.1 Overall M vs. L H vs. L 000 .3 0.23 ( . 5 1 1 ) 0.081 00,.5 0.13 ( . 2 1 1 ) 0 0 0 00,.1 .6 4.0 9. 60 129 5 124 3.9 96.1 Overall M vs. L H vs. L 094 .8 1.12 ( . 3 3 7 ) 0 9 9 03,.4 .9 1.07 ( . 0 3 8 ) 0 9 9 03,.0 .9 127 14 113 11.0 89.0 121 10 111 8.3 91.7 Overall M vs. L H vs. L 049 .1 1.81 ( . 3 4 4 ) 0.265 07,.9 1.32 ( . 0 3 4 ) 0 6 0 05,.6 .3 13.0 8. 70 61 2 59 3.3 96.7 52 1 51 1.9 98.1 Overall 000 .3 M vs. L 0.23 ( . 5 1 1 ) 0 0 1 00,.5 .8 00,.1 H vs. L 0.13 ( . 2 1 1 ) 0 0 0 .6 4.3 95.7 149 6 143 40 . 96.0 129 5 124 3.9 96.1 Overall M vs. L H vs. L Occupation Statistic Officer n Abnormal Normal 125 7 118 5.6 94.4 127 11 116 8.7 91.3 121 9 112 n Abnormal Normal 54 7 47 13.0 87.0 61 2 59 3.3 96.7 Enlisted Groundcrew n Abnormal Normal 138 5 133 3.6 96.4 149 6 143 Officer Variable n Abnormal Normal 125 8 117 6.4 93.6 Enlisted Flyer n Abnormal Normal 54 7 47 Enlisted Groundcrew n Abnormal Normal 138 6 132 Basal Cell Enlisted Carcinoma Flyer (Verified Only)b Basal Cell Carcinoma (Verified and Suspected) 090 .8 0.92 ( . 9 2 9 ) 0 9 9 02,.3 .9 0.89 ( . 6 2 9 ) 0 9 9 02,.8 .9 �TABLE H-27. (ctntinued) Unadjusted Exposure Index Analysis for Followp Participants for Lifetime Occurrence of Malignant Neoplasms Exposure Index3 Contrast Est. Relative Risk (95%C.I.) p-Value 7.4 92,6 Overall M vs. L H vs. L 0.625 1.47 ( . 1 3 5 ) 0 5 4 06,.7 .0 1.04 ( . 0 2 7 ) 0 9 9 04,.0 .9 52 1 51 1.9 98.1 Overall M vs. L H vs. L 008 .5 0.35 ( . 9 1 4 ) 0.186 00,.2 0.13 ( . 2 1 1 ) 0 0 0 00,.1 .6 4.7 95.3 129 5 124 3.9 96.1 Overall M vs. L H vs. L 0.910 1.31 ( . 1 4 2 ) 0.772 04,.3 1.07 ( . 0 3 8 ) 0 9 9 03,.0 .9 127 16 111 12.6 87.4 121 10 111 8.3 91.7 Overall M vs. L H vs. L 0.382 1.66 ( . 2 3 8 ) 0.301 07,.1 1.04 ( . 2 2 5 ) 0 9 9 04,.9 .9 13.0 87.0 61 3 58 49 . 95.1 52 1 51 1.9 98.1 Overall M vs. L H vs. L 008 .5 0.35 ( . 9 1 4 ) 0.186 00,.2 0.13 ( . 2 1 1 ) 0 0 0 00,.1 .6 4.3 95.7 149 7 142 4.7 95.3 129 5 124 3.9 96.1 Overall M vs. L M vs. L 0.945 1.09 ( . 6 3 3 ) 0 9 9 03,.1 .9 0.89 ( . 6 2.98) 0 9 9 02, .9 Low Number Percent Medium High Number Percent Number Percent Sun-Exposure Related Malignancies (Verified and Suspected) n Abnormal Normal 125 9 116 7.2 92.8 127 13 114 10.2 89.8 121 9 112 Enlisted Flyer n Abnormal Normal 54 7 47 13.0 87.0 61 3 58 4.9 95.1 138 5 133 3.6 96.4 149 7 142 Officer 1° Statistic n Enlisted Abnormal Groundcrew Normal Sun-Exposure Belated Malignancies (Verified Only) Occupation Officer Variable n Abnormal Normal 125 10 115 80 . 92.0 Enlisted Flyer Abnormal Normal 54 7 47 n Enlisted Abnormal Groundcrew Normal 138 6 132 �TABIE H-27. (continued) Unadjusted Exposure Index Analysis for EoUowup Participants for T.iferimp Occurrence of Malignant Naoplc Low Number Percent Exposure Index3 Medium High Number Percent Number Percent Contrast Est. Relative Risk (95ZC.I.) p-Value 2.4 97.6 Overall M vs. L H vs. L 1.48 ( . 4 8 9 ) 02,-9 1.56 ( . 6 9 5 ) 02,.2 086 .7 099 .9 060 .8 57 1 56 1.8 98.2 Overall M vs. L H vs. L 08 (.261) .4 01,.8 0.47 ( . 4 5 3 ) 00,.7 0.825 099 .9 0.615 1.8 98.2 142 1 141 0.7 99.3 Overall M vs. L H vs. L 025 .0 6.74 ( . 5 1 1 5 ) 0 2 8 03,3.3* .4 3.28 ( . 3 8 . 6 * 0 4 0 01,10) . 8 130 4 126 3.1 96.9 123 3 120 2.4 97.6 Overall M vs. L H vs. L 0.730 1.98 ( . 6 1 . 3 0 6 4 03,10) .8 1.56 ( . 6 9 5 ) 0 6 0 02,.2 .8 5.5 94.5 65 2 63 3.1 96.9 57 2 55 3.5 96.5 Overall M vs. L H vs. L 0.55 ( . 9 3 4 ) 00,.2 0.63 ( . 0 3 9 ) 01,.3 0.0 100.0 163 3 160 1.8 98.2 142 2 140 1.4 98.6 Overall M vs. L H vs. L 0.261 6.74 ( . 5 1 1 5 ) 0.248 03,3.3* 5.50 ( . 6 1 5 5 ) 0 2 9 02,1.1* .2 Occupation Statistic Officer n Abnormal Normal 127 2 125 1.6 98.4 130 3 127 2.3 97.7 123 3 120 n Abnormal Normal 55 2 53 3.6 96.4 65 2 63 3.1 96.9 Enlisted Groundcrev n Abnormal Normal 154 0 154 00 . 100.0 163 3 160 Officer n Abnormal Normal 127 2 125 1.6 98.4 Enlisted Flyer n Abnormal Normal 55 3 52 Enlisted Groundcrew Variable n Abnormal Normal 154 0 154 Enlisted Systemic Malignancies Flyer (Verified Only)c Systemic Malignancies (Verified and Suspected)0 a Number and percent b Nonblacks only. c of total participants. Blacks and nonblacks. *0.5 added to each cell to calculate relative risk and confidence interval due to zero cell. 074 .8 060 .6 066 .7 �APPENDIX I Neurological Assessment �APPENDIX I: Neurological Assessment Contents Table 1-1 Unadjusted Exposure Index Analyses for Neurological Variables by Occupation 1-1 1-2 Interaction Summaries of Adjusted Exposure Index Analyses for Selected Neurological Variables 1-16 1-3 Exclusions and Missing Data for Neurological Assessment by Group (Original Comparisons Only) 1-19 1-4 Unadjusted Analyses for Verified Neurological Diseases by Group— 1982-1985 (Original Comparisons Only) 1-20 1-5 Unadjusted Analyses for Verified Neurological Diseases by Group— Baseline and First Followup Studies Combined (Original Comparisons Only) 1-21 1-6 Unadjusted Analyses for Cranial Nerve Function by Group (Original Comparisons Only) 1-22 1-7 Adjusted Analyses for Selected Variables of Cranial Nerve Function by Group (Original Comparisons Only) 1-25 1-8 Unadjusted Analyses for Peripheral Nerve Function by Group (Original Comparisons Only) 1-26 1-9 Adjusted Analyses for Selected Variables of Peripheral Nerve Function by Group (Original Comparisons Only) 1-27 1-10 Summary of Group-by-Diabetic Class Interaction for Pin Prick (Original Comparisons Only) 1-28 1-11 Unadjusted Analyses for CNS Coordination Variables by Group (Original Comparisons Only) 1-29 1-12 Adjusted Analyses for Selected Variables of CNS Coordination by Group (Original Comparisons Only) 1-30 1-13 Longitudinal Analyses of Romberg Sign and Babinski Reflex: A Contrast of Baseline and First Followup Examination Abnormalities (Original Comparisons Only) 1-31 I-i �TftBtEI-1. Variable Exposure Index High Medium Number Percent Number Percent Contrast Est. Relative 'Risk ( 5 C I ) phValue 9% .. 08 . . 9. 92 Overall M vs. L H vs. L 0.361 0 1 ( . 1 4 0 ) 0.243 . 9 00,.5* 00,.2 0.51 ( . 5 5 7 ) 0 9 9 .9 57 2 55 3.5 96.5 Overall M vs. L H vs. L 0.739 08 (.51.1 099 .4 00,38) .9 01,23) .9 1.96 ( . 7 2 . 0 0 9 9 06 . 99.4 142 1 141 0.7 99.3 Overall M vs. L H vs. L 095 .9 09 (.61.3 099 .4 00,52) .9 10 (.71.1 099 .9 00,75) .9 130 0 130 00 . 100 0. 123 1 122 08 . 99.2 Overall 0.593 M vs. L 0.32 ( . 1 8 0 ) 0 4 4 00,.1* . 9 H vs. L 1.03 ( . 6 1 . 0 0 9 9 00,67) .9 1.8 98.2 65 0 65 00 . 100 0. 57 0 57 00 . 100 0. Overall M vs. L H vs. L 0.328 00,.5* 0.28 ( . 1 6 9 ) 0.458 0.32 ( . 1 7 9 ) 0 4 1 00,.2* . 9 1.3 98.7 163 0 163 00 . 100 0. 142 1 141 0.7 99.3 Overall M vs. L H vs. L 0.356 0 1 ( . 1 3 9 ) 0.235 . 9 00,.2* 00,.1 .9 0.54 ( . 5 6 0 ) 0 9 9 Low Number Percent n Abnormal Normal 127 2 125 1.6 98.4 130 0 130 00 . 100 0. 123 1 122 Enlisted Flyer n Abnormal Normal 55 1 54 1.8 98.2 65 1 64 1.5 98.5 Enlisted Groundcrew n Abnormal Normal 154 1 153 0.7 99.3 163 1 162 Officer Visual Fields Statistic Officer Smell Occupation n Abnormal Normal 127 1 126 08 . 99.2 Enlisted Flyer n Abnormal Normal 55 1 54 Enlisted Groundcrew n Abnormal Normal 154 2 152 �TfiBLE 1-1. (continued) Unadjusted Exposure Index Analyses for Neurological Variables by Occupation Exposure Index Variable Occupation Statistic n Officer * Abnormal Normal n Light Reaction Enlisted Flyer ^ Enlisted Groundcrew Abnormal Normal n Abnormal Normal n Officer Abnormal Normal n Ocular. Movements Enlisted Flyer Normal Enlisted Groundcrew n Abnormal Normal Abnormal Low Number Percent Medium High Number Percent Number •Percent Contrast Est. Relative Risk (95% C.I.) p-Value 1.6 98.4 130 0 130 0.0 100.0 123 0 123 0.0 100.0 Overall M vs. L H vs. L 0.133 0.19 (0.01,4.01)* 0.241 0.20 (0.01,4.24)* 0.498 55 1 54 1.8 98.2 65 0 65 0.0 100.0 57 0 57 0.0 100.0 Overall M vs. L H vs. L 0.328 0.28 (0.01,6.95)* 0.458 0.32 (0.01,7.92)* 0.491 154 2 152 1.3 98.7 163 2 161 1.2 98.8 142 1 141 0.7 99.3 Overall M vs. L H vs. L 0.94 (0.13,6.79) 0.54 (0.05,6.01) 127 1 126 0.8 99.2 130 0 130 0.0 100.0 123 0 123 0.0 100.0 Overall M vs. L H vs. L 0.368 0.32 (0.01,8.01)* 0.494 0.34 (0.01,8.46)* 0.999 55 1 54 1.8 98.2 65 1 64 1.5 98.5 57 0 57 0.0 100.0 Overall M vs. L H vs. L 0.612 0.84 (0.05, 13.81) 0.999 0.32 (0.01,7.92)* 0.491 154 1 153 0.7 99.3 163 1 162 0.6 99.4 142 1 141 0.7 99.3 Overall M vs. L H vs. L 0.995 0.94 (0.06,15.23) 0.999 1.09 (0.07,17.51) 0.999 126 2 124 0.866 0.999 0.999 �_», Unadjusted Exposure Index Anal\*ses for 1fenmlncriral \fariahle*: fw (VnnvtHm ^ —x ^ Variable Occupation Statistic Officer Abnormal Normal Enlisted Flyer Abnormal Normal Enlisted Groundcrew Abnormal Normal n n Facial Sensation n n Officer Abnormal Normal n Comeal Reflex Enlisted Flyer Abnormal Normal n Enlisted Groundcrew Abnormal Normal Low Number Percent Exposure Index Medium High Number Percent Number Percent Contrast Est. Relative Risk (95%C.I.) p-Value 126 1 125 0.8 99.2 130 1 129 0.8 99.2 123 0 123 0.0 100.0 Overall M vs. L H vs. L 0.617 0.97 (0.06,15.66) 0.999 0.34 (0.01,8.40)* 0.999 55 I 54 1.8 98.2 65 0 65 0.0 100.0 57 0 57 0.0 100.0 Overall M vs. L H vs. L 0.328 0.28 (0.01,6.95)* 0.458 0.32 (0.01,7.92)* 0.491 154 1 153 0.7 99.3 163 0 163 0.0 100.0 141 0 141 0.0 100.0 Overall M vs. L H vs. L 0.372 0.31 (0.01,7.74)* 0.486 0.36 (0.02,8.95)* 0.999 125 0 125 0.0 100.0 130 0 130 0.0 100.0 123 0 123 0.0 100.0 Overall M vs. L H vs. L 55 0 55 0.0 100.0 65 0 65 0.0 100.0 57 0 57 0.0 100.0 Overall M vs. L H vs. L — — — — 152 0 152 0.0 100.0 161 0 161 0.0 100.0 142 0 142 0.0 100.0 Overall M vs. L H vs. L — — — — — — — — ___ �T6HE T-l. (continued) Unadjusted Exposure Index Analyses for Neurological ^ariahles by Occupation Exposure Index Variable Low Number Percent Medium High Number Percent Number Percent Statistic Officer n Abnormal Normal 127 1 126 08 . 99.2 130 0 130 00 . 100 0. 123 0 123 00 . 100 0. Overall 0.368 M vs. L 0.32 ( . 1 8 0 ) 0 4 4 00,.1* . 9 00,.6* . 9 H vs. L 0.34 ( . 1 8 4 ) 0 9 9 Enlisted Flyer n Abnormal Normal 55 1 54 1.8 98.2 65 0 65 00 . 100 0. 57 0 57 00 . 100 0. Overall 0.328 00,.5* M vs. L 0.28 ( . 1 6 9 ) 0.458 00,.2* H vs. L 0.32 ( . 1 7 9 ) 0.491 Enlisted Groundcrew n Abnormal Normal 154 0 154 00 . 100 0. 163 0 163 00 . 100 0. 142 0 142 00 . 100 0. Overall M vs. L H vs. L Officer n Abnormal Normal 127 1 126 08 . 99.2 130 1 129 08 . 99.2 123 0 123 00 . 100 0. Overall 0.618 M vs. L 0 9 ( . 6 1 . 9 0 9 9 .8 00,57) .9 H vs. L 0.34 ( . 1 8 4 ) 0 9 9 00,.6* . 9 Enlisted Flyer n Abnormal Normal 55 1 54 1.8 98.2 65 1 64 1.5 98.5 57 0 57 00 . 100 0. Overall 0.612 M vs. L 0 8 ( . 5 1 . 1 0 9 9 .4 00,38) .9 00,.2* H vs. L 0.32 ( . 1 7 9 ) 0.491 Enlisted Groundcrew n Abnormal Normal 154 0 154 00 . 100 0. 163 1 162 0.6 9. 94 142 2 140 1.4 9. 86 Overall 0.323 01,05) M vs. L 2.85 ( . 2 7 . 5 * 0.999 H vs. L 5.50 ( . 6 1 5 5 * 0.229 02,1.) * Jaw Clench Snile Contrast Est. Relative 9% . . Risk ( 5 C I ) p-Value Occupation — — — — �t • uhadjusted Exposure Index Anal]rses for 1feiirrtlrvHipal \fariaKl«ag by Occupation Variable Low Number Percent n Abnormal Normal 127 1 126 0.8 99.2 Enlisted " Flyer n Abnormal Normal 55 0 55 0.0 100.0 Enlisted Groundcrev n Abnormal Normal 154 0 154 Officer Balance Statistic Officer Palpebral Fissures Occupation n Abnormal Normal Enlisted Flyer Enlisted Groundcrev Exposure Index Medium High Number Percent Number Percent Contrast Est. Relative Risk (952C.I.) p-Value 1.5 98.5 122 1 121 0.8 99.2 Overall M vs. L H vs. L 0.801 1.97 (0.18, 21.99) 0.999 1.04 (0.06, 16.84) 0.999 65 1 64 1.5 98.5 57 0 57 0.0 100.0 Overall M vs. L H vs. L 0.0 100.0 163 1 162 0.6 99.4 142 1 141 0.7 99.3 Overall M vs. L H vs. L 0.420 2.58 (0.10,64.65)* 0.999 — — 0.598 2.85 (0.12, 70.55)* 0.999 3.28 (0.13,81.06)* 0.480 127 0 127 0.0 100.0 130 0 130 0.0 100.0 123 0 123 0.0 100.0 Overall M vs. L H vs. L n Abnormal Normal 55 0 55 0.0 100.0 64 0 64 0.0 100.0 57 0 57 0.0 100.0 Overall M vs. L H vs. L n Abnormal Normal 154 1 153 0.7 99.3 163 0 163 0.0 100.0 142 1 141 0.7 99.3 Overall M vs. L H vs. L 130 2 128 — — — — — — — — 0.574 0.31 (0.01,7.74)* 0.486 1.09 (0.07,,17.51)* 0.999 �unadjusted Exposure Index Analyses for 1 Variable Low Number Percent cal %ruibles by Occupation Exposure Index High Medium Number Percent Number Percent Contrast Est. Relative Risk ( 5 C I ) p-Value 9% .. n Abnormal Normal 126 0 126 00 . 100.0 130 0 130 00 . 100 0. 123 0 123 00 . 100 0. Overall M vs. L H vs. L Enlisted Flyer n Abnormal Normal 55 1 54 1.8 98.2 65 0 65 00 . 100 0. 57 0 57 00 . 100 0. Overall 0.328 00,.5* . 5 M vs. L 0.28 ( . 1 6 9 ) 0 4 8 H vs. L 0.32 ( . 1 7 9 ) 0 4 1 00,.2* . 9 Enlisted Groundcrew n Abnormal Normal 153 0 153 00 . 100 0. 163 0 163 00 . 100 0. 142 0 142 00 . 100 0. Overall M vs. L H vs. L Officer Speech Statistic Officer Gag Reflex Occupation n Abnormal Normal 127 0 127 00 . 100 0. 130 0 130 00 . 100 0. 123 0 123 00 . 100 0. Overall M vs. L Hvs. L Enlisted Flyer n Abnormal Normal 55 1 54 1.8 98.2 65 0 65 00 . 100 0. 57 0 57 00 . 100 0. Overall M vs. L H vs. L Enlisted Groundcrew n Abnormal Normal 154 1 153 0.7 99.3 163 0 163 00 . 100 0. 142 1 141 07 . 9. 93 038 .2 0.28 ( . 1 6 9 ) 0 4 8 00,.5* . 5 0.32 ( . 1 7 9 ) 0 4 1 00,.2* . 9 Overall 0.574 M vs. L 0.31 ( . 1 7 7 ) 0 4 6 00,.4* . 8 00,75) .9 H vs. L 1.09 ( . 7 1 . 1 0 9 9 �T^HIE 1-1. (continued) Unadjusted Exposure Index Analyses for Neurological Variables by Occupation Variable Low Nunber Percent Exposure Index Medium High Number Percent Number Percent Contrast Est. Relative Risk ( 5 C I ) p-Value 9% .. 00 . 100 0. Overall M vs. L H vs. L 0.368 0.32 ( . 1 8 0 ) 0 4 4 00,.1* . 9 0.34 ( . 1 8 4 ) 0 9 9 00,.6* . 9 57 0 57 00 . 100 0. 0.328 Overall M vs. L 0.28 ( . 1 6 9 ) 0.458 00,.5* H vs. L 0.32 ( . 1 7 9 ) 0 4 1 00,.2* . 9 06 . 9. 94 142 0 142 00 . 100 0. Overall 044 .0 M vs. L 2.83 ( . 2 7 . 9 * 0 9 9 01,00) . 9 H vs. L — 130 0 130 00 . 100 0. 123 0 123 00 . 100 0. Overall 0.365 Mvs. L 0.32 ( . 1 7 9 ) 0 4 2 00,.4* . 9 H vs. L 0.34 ( . 1 8 4 ) 0 9 9 00,.0* . 9 1.8 98.2 65 0 65 00 . 100 0. 57 0 57 00 . 100 0. Overall 0.328 M vs. L 0.28 ( . 1 6 9 ) 0 4 8 00,.5* . 5 H vs. L 0.32 ( . 1 7 9 ) 0 4 1 00,.2* . 9 00 . 100 0. 163 0 163 00 . 100 0. 142 0 142 00 . 100 0. Overall M vs. L H vs. L 127 1 126 08 . 99.2 130 0 130 00 . 100 0. 123 0 123 Enlisted Flyer n Abnormal Normal 55 1 54 1.8 98.2 65 0 65 00 . 100 0. n Abnormal Normal 153 0 153 00 . 100 0. 163 1 162 Officer Palate and Uvula Movement n Abnormal Normal Enlisted Groundcrev M Statistic Officer Tongue Position Relative to Midline Occupation n Abnormal Normal 126 1 125 08 . 99.2 Misted Flyer n Abnormal Normal 55 1 54 Enlisted Groundcrev n Abnormal Normal 153 0 153 — — — — — �Unadjusted Exposure Index Analyses for Neurological \foriables by Occupation Variable Low Number Percent Exposure Index Medium High Number Percent Number Percent Statistic Officer n Abnormal Normal 127 9 118 7.1 92.9 130 9 121 6.9 93.1 123 10 IB 8.1 91.9 Overall 094 .2 M vs. L 0 9 ( . 7 2 5 ) 0 9 9 .9 .8 03,.4 04,.) . 1 H vs. L 1.16 ( . 6 2 % 0 8 4 Enlisted Flyer n Abnormal Normal 55 5 50 9.1 9. 09 65 6 59 9.2 9. 08 57 4 53 7.0 93.0 081 .9 Overall M vs. L 1.02 ( . 9 3 5 ) 0 9 9 .9 02,.3 H vs. L 0.76 ( . 9 2 9 ) 0 7 0 01,.7 .4 Enlisted Groundcrev n Abnormal Normal 154 9 145 5.8 94.2 163 6 157 3.7 96.3 142 3 139 2.1 97.9 020 .5 Overall 02,.7 M vs. L 0.62 ( . 1 1 7 ) 0.433 H vs. L 0.35 ( . 9 1 3 ) 0.142 00,.1 Officer Neck Range of Motion Contrast Est. Relative Risk ( 5 C I ) p-Value 92.. Occupation n Abnormal Normal 122 15 107 12.3 87.7 130 12 118 9.2 9. 08 123 13 109 10.7 89.3 074 .3 Overall M vs. L 0.73 ( . 3 1 6 ) 0.542 03,.2 .4 H vs. L 0 8 ( . 9 1 8 ) 0 8 1 .5 03,.7 Enlisted Flyer n Abnormal Normal 55 7 48 12.7 87.3 64 8 56 12.5 87.5 57 5 52 88 . 91.2 n Enlisted Abnormal Grounder®/ Normal 151 15 136 9.9 90.1 161 12 149 7.5 92.5 141 9 132 6.4 93.6 £ Cranial Nerve Function Index Overall M vs. L 0 9 .8 H vs. L 0 6 .6 , Overall M vs. L 0.73 H vs. L 0.62 0.754 (.329) 099 03,.0 .9 ( . 0 2 2 ) 0.554 02,.2 0.512 (.316) 056 03,.2 .4 ( . 6 1 4 ) 0.290 02,.6 �— —— — x ' • "r Unadjusted Exposure Index Analyses for Neurological \fariables by Occupation Variable Low Number Percent Exposure Index Medium Higfc Number Percent Number Percent n Abnormal Normal 122 8 114 6.6 93.4 130 3 127 2.3 97.7 122 3 119 2.5 97.5 Overall 0.136 M vs. L 0.34 ( . 9 1 3 ) 0.127 00,.0 H vs. L 0.36 ( . 9 1 3 ) 0.216 00,-9 Enlisted Flyer n Abnormal Normal 55 3 52 5.5 94.5 64 2 62 3.1 96.9 57 2 55 3.5 96.5 Overall M vs. L H vs. L Enlisted Groundcrew n Abnormal Normal 151 8 143 5.3 94.7 161 6 155 3.7 96.3 141. 7 134 5.0 95.0 Overall 075 .8 M vs. L 0 6 ( . 3 2 0 ) 0.590 .9 02,.4 H vs. L 0.93 ( . 3 2 6 ) 0 9 9 03,.5 .9 Officer Pin Prick Statistic Officer Cranial Nerve Function Index (Neck Range of Motion Excluded) Contrast Est. Relative Risk ( 5 C I ) p-Value 9% .. Occupation n Abnormal Normal 126 9 117 7.1 9. 29 127 6 121 4.7 95.3 122 7 115 5.7 9. 43 Overall 0.713 02,.7 M vs. L 0.65 ( . 2 1 8 ) 0 4 0 .4 H vs. L 0 7 ( . 9 2 2 ) 0 7 7 . 9 02,.0 .9 Enlisted Flyer n Abnormal Normal 55 5 50 9.1 9. 09 64 3 61 4.7 95.3 57 5 52 88 . 91.2 Overall 0.584 M vs. L 0 4 ( . 1 2 1 ) 0 4 9 .9 01,.6 .6 . 02,.2 H vs. L 0 % ( . 6 3 5 ) 0 9 9 .9 Enlisted Groundcrew n Abnormal Normal 152 13 139 8.6 91.4 162 8 154 49 . 95.1 138 3 135 2.2 97.8 0.052 Overall M vs. L 0 5 ( . 2 1 3 ) 0.259 .6 02,.8 H vs. L 0.24 ( . 7 0 8 ) 0 0 0 00,.5 .2 0.791 0 5 ( . 9 3 4 ) 0.661 .6 00,.7 0.63 ( . 0 3 9 ) 0 6 6 01,.3 .7 �TABLE 1-1. (cmtinued) Unadjusted Exposure Index Analyses for Neurological variables by Occupation Exposure Index Variable Low Number Percent Medium High Number Percent Number Percent Officer 126 9 117 7.1 92.9 127 4 123 3.2 96.8 3 119 2.5 97.5 0.141 Overall M vs. L 0.42 ( . 3 1 4 ) 0.167 01,.1 H vs. L 0.33 ( . 9 1 2 ) 0.137 00,-4 Bilisted ' Flyer n Abnormal Normal 55 3 52 5.5 9. 45 64 2 62 3.1 9. 69 57 4 53 7.0 9. 30 Overall 069 .1 M vs. L 0 5 ( . 9 3 4 ) 0 6 1 . 6 00,.7 .6 H vs. L 1.31 ( . 8 6 1 ) 0 9 9 02,.3 .9 n Abnormal Normal 152 5 147 3.3 96.7 162 5 157 3.1 96.9 138 3 135 2.2 97.8 Overall 084 .3 M vs. L 0 9 ( . 7 3 3 ) 0 9 9 .4 02,.0 .9 H vs. L 0 6 ( . 5 2 7 ) 0.725 .5 01,.9 Officer Muscle Status n Abnormal Normal Enlisted Groundcrev Light Touch Statistic n Abnormal Normal 127 5 122 3.9 96.1 130 2 128 1.5 98.5 123 5 118 4.1 95.9 Overall 048 .2 M vs. L 0 3 ( . 7 2 0 ) 0 2 8 . 8 00,.0 .7 H vs. L 1.03 ( . 9 3 6 ) 0 9 9 02,.6 .9 Enlisted Flyer n Abnormal Normal 55 1 54 1.8 98.2 65 1 64 1.5 98.5 57 1 56 1.8 98.2 Overall M vs. L H vs. L 092 .9 0 8 (.5 1 . 1 0 9 9 . 4 00, 3 8 ) . 9 09 (.61.1 099 .6 00,58) .9 Enlisted Groundcrew n Abnormal Normal 154 3 151 2.0 9. 80 163 5 158 3.1 96.9 142 3 139 2.1 97.9 Overall M vs. L H vs. L 071 .8 1.59 ( . 7 6 7 ) 0.724 03,.8 1.09 ( . 2 5 4 ) 0 9 9 02,.7 .9 199 Contrast Est. Eelative Risk ( 5 C I ) p-Value 9% . . Occupation �Unadjusted Exposure Index Analyses for Neurological Variables by Occupation Variable Exposure Index Medium High Number Percent Number Percent Contrast Est. Relative Risk ( 5 C I ) p-Value 9%.. 2.5 97.5 Overall M vs. L H vs. L 052 .4 0.33 ( . 3 3 1 ) 0.370 00,-7 1.03 ( . 1 5 2 ) 0 9 9 02,.2 .9 57 1 56 1.8 98.2 Overall 068 .2 Mvs. L 2.62 ( . 1 6 . 8 * 0 9 9 01,56) . 9 H vs. L 2.95 ( . 2 7 . 0 * 0 9 9 01,39) .9 1.2 98.8 138 0 138 00 . 100 0. 130 1 129 08 . 99.2 123 3 120 2.4 9. 76 Overall 0.156 M vs. L 2.95 ( . 2 7 . 8 * 0 9 9 01,31) .9 H vs. L 7.41 ( . 8 1 4 9 ) 0 1 8 03,4.0* . 1 1.8 98.2 65 1 64 1.5 98.5 57 1 56 1.8 98.2 092 .9 Overall M vs. L 0 8 ( . 5 1 . 1 0 9 9 .4 00,38) .9 H vs. L 0 % ( . 6 1 . 1 0 9 9 . 00,58) .9 2.0 9. 80 163 0 163 00 . 100.0 142 1 141 0.7 99.3 Overall M vs. L H vs. L Low Number Percent n Abnormal Normal 126 3 123 2.4 97.6 127 1 126 08 . 99.2 122 3 119 Enlisted Flyer n Abnormal Normal 55 0 55 00 . 100 0. 64 1 63 1.6 9. 84 Enlisted Groundcrew n Abnormal Normal 152 0 152 00 . 100 0. 162 2 160 Officer Patellar Reflex Statistic Officer Vibratory Sensation Occupation n Abnormal Normal 127 0 127 00 . 100 0. Enlisted Flyer n Abnormal Normal 55 1 54 Enlisted Groundcrew n Abnormal Normal 154 3 151 Overall Mvs. L H vs. L 016 .6 4.75 ( . 3 9 . 6 * 0 4 9 02,97) .9 0.170 0.13 ( . 1 2 5 ) 0 1 4 00,.8* . 1 0.36 ( . 4 3 4 ) 0 6 4 00,.7 .2 �TABIE 1-1. (continued) Ihadjusted Exposure Index Analyses for Neurological Variables by Occupation Variable Low Nunber Percent Exposure Index Medium High Number Percent Number Percent Est. Relative Risk ( 5 C I ) p-Value 9% . . n Abnormal Normal 124 8 116 6.5 93.5 129 6 123 4.7 95.3 123 8 115 6.5 93.5 074 .7 Overall M vs. L 0.71 ( . 4 2 1 ) 0 5 1 02,.0 .9 03,.8 H vs. L 1.01 ( . 7 2 7 ) 0 9 9 .9 Enlisted Flyer n Abnormal Normal 55 5 50 9.1 90.9 64 0 64 00 . 100 0. 57 4 53 70 . 93.0 Overall M vs. L H vs. L Enlisted Groundcrew n Abnormal Normal 153 10 143 6.5 93.5 163 9 154 5.5 9. 45 141 8 133 5.7 94.3 Overall 090 .2 M vs. L 0 8 ( . 3 2 1 ) 0 8 4 .4 03,.2 .1 H vs. L 0 8 ( . 3 2 2 ) 0 8 2 .6 03,.5 .1 Officer Biceps Reflex Statistic Officer • Achilles Reflex Occupation n Abnormal Normal 127 1 126 0.8 99.2 130 0 130 00 . 100 0. 123 2 121 1.6 98.4 Overall M vs. L H vs. L 034 .4 0.32 ( . 1 8 0 ) 0 4 4 00,.1* . 9 2.08 ( . 9 2 . 7 0 6 8 01,32) .1 Enlisted Flyer n Abnormal Normal 55 0 55 00 . 100.0 65 1 64 1.5 98.5 57 0 57 00 . 100 0. Overall M vs. L H vs. L 040 .2 2.58 ( . 0 6 . 5 0 9 9 01,46) .9 — — Enlisted Groundcrew n Abnormal Normal 154 1 153 0.7 99.3 163 3 160 1.8 98.2 142 1 141 0.7 99.3 Contrast 009 .5 0 0 (.0,.2* 0 0 9 . 7 00413) . 1 0.76 ( . 9 2 9 ) 0 7 0 01,.7 .4 Overall 0.515 03,78) M vs. L 2.87 ( . 0 2 . 8 0.623 00,75) .9 H vs. L 1.09 ( . 7 1 . 1 0 9 9 �TAKE 1-1. (continued) Ibadjusted Exposure Index Analyses for Neurological Variables by Occupation Exposure Index Variable Low Number Percent n Abnormal Normal 127 0 127 Enlisted Flyer n Abnormal Normal Enlisted Groundcrew Officer Tremor Statistic Officer Babinski Reflex Occupation High. Medium Nunber Percent Number Percent 00 . 100 0. 128 0 128 00 . 100 0. 122 0 122 55 0 55 00 . 100.0. 65 0 65 n Abnormal Normal 153 2 151 1.3 98.7 n Abnormal Normal 127 5 Enlisted Flyer Enlisted Groundcrew Contrast Est. Relative Risk ( 5 C I ) p-Value 9% . . 00 . 100 0. Overall M vs. L H vs. L 00 . 100 0. 57 0 57 00 . 100 0. Overall M vs. L H vs. L 162 0 162 00 . 100 0. 142 2 140 1.4 9. 86 Overall M vs. L H vs. L 199 3.9 96.1 130 1 129 08 . 99.2 123 3 120 2.4 97.6 Overall 0.248 M vs. L 0.19 ( . 2 1 6 ) 0.117 00,.4 H vs. L 0.61 ( . 4 2 6 ) 0.722 01,.1 n Abnormal Normal 55 1 54 1.8 98.2 65 3 62 46 . 95.4 57 3 54 5.3 94.7 Overall M vs. L H vs. L n Abnormal Normal 154 5 149 3.3 96.7 163 4 159 2.5 97.5 142 1 141 07 . 99.3 0.312 Overall M vs. L 0 7 ( . 0 2 8 ) 0 7 4 . 5 02,.5 .4 H vs. L 0.21 ( . 2 1 8 ) 0.216 00,.3 — — — — __ — — — — 0.329 0.19 ( . 1 3 9 ) 0.235 00,.2* 10 (.577) 099 .8 01,.6 .9 069 .0 2.61 ( . 6 2 . 6 0 6 4 02,58) . 2 3.00 ( . 0 2 . 6 0.618 03,97) �TABLE 1-1. (continued) \y Occupation Unadjusted Exposure Index Analyses for Neurological Variable Occupation Statistic Low Number Percent Exposure Index Medium High Number Percent Number Percent Contrast Est. Relative Risk (95XC.I.) p-Value . Officer n Abnormal Normal 127 0 127 00 . 100 0. 130 1 129 08 . 99.2 123 1 122 08 . 99.2 Overall 063 .0 01, 31) . 9 M vs. L 2.95 ( . 2 7 . 8 * 0 9 9 01, 73) .9 H vs. L 3.12 ( . 3 7 . 9 * 0 4 2 Coordination Enlisted • Flyer n Abnormal Normal 55 0 55 00 . 100.0 64 1 63 1.6 98.4 57 0 57 00 . 100.0 0.415 Overall M vs. L 2.62 ( . 1 6 . 8 * 0 9 9 01, 56) . 9 H vs. L — — Enlisted Groundcrew n Abnormal Normal 154 2 152 1.3 98.7 163 1 162 06 . 99.4 142 3 139 2.1 97.9 Overall M vs. L H vs. L Officer n Abnormal Normal 127 0 127 00 . 100.0 130 0 130 00 . 100.0 123 0 123 00 . 100.0 Overall M vs. L H vs. L — — — — Enlisted Flyer n Abnormal Normal 55 0 55 00 . 100.0 64 0 64 00 . 100.0 57 0 57 00 . 100.0 Overall M vs. L H vs. L — — — — Enlisted Groundcrew n Abnormal Normal 154 1 153 0.7 99.3 163 0 163 00 . 100.0 142 1 141 07 . 99.3 Overall M vs. L H vs. L Romberg Sign 0.516 0.47 ( . 4 5.23) 0.613 00, 1.64 ( . 7 9 % 0 6 4 02, . ) . 7 054 .7 0.31 ( . 1 7 7 ) 0 4 6 00,.4 .8 1.09 ( . 7 1 . 1 0 9 9 00,75) .9 �— — —- X ' " f Unadjusted Exposure Index Anal}ses for Neurological \feriahles by Occupation Variable Exposure Index Medium High Number Percent Mmber Percent Contrast Est. Relative Risk ( 5 C I ) p-Value 9%.. 1.6 98.4 Overall M vs. L H vs. L 051 .9 0.32 ( . 3 3 1 ) 0.366 00,.2 0.68 ( . 1 4 1 ) 0 9 9 01,.6 .9 57 0 57 00 . 100 0. Overall M vs. L H vs. L 0.370 08 (.261) 099 .4 01,.8 .9 00,.7* . 3 0.19 ( . 1 3 9 ) 0 2 9 1.2 9. 88 142 4 138 2.8 97.2 Overall M vs. L H vs. L 050 .8 04 (.625) 047 .7 00,.8 .3 1.09 ( . 7 4 4 ) 0 9 9 02,.3 .9 130 2 128 1.5 9. 85 123 4 119 3.3 9. 67 Overall M vs. L H vs. L 0.123 02 (.511) 008 .3 00,.2 .5 0 5 ( . 5 1 7 ) 0.377 . 0 01,.1 5.5 94.5 64 5 59 7.8 92.2 57 3 54 5.3 9. 47 Overall M vs. L H vs. L 080 .1 03,.5 1.47 ( . 4 6 4 ) 0.724 .9 0 % (.949) 099 . 01,.9 6.5 93.5 163 7 156 43 . 95.7 142 6 136 42 . 95.8 0.585 Overall . 5 02,.4 .5 M vs. L 0 6 ( . 4 1 7 ) 0 4 9 H vs. L 0 6 ( . 3 1 8 ) 0 4 8 . 4 02,.0 .4 Lew Number Percent n Abnormal Normal 127 3 124 2.4 97.6 130 1 129 08 . 99.2 123 2 121 Enlisted Flyer n Abnormal Normal 55 2 53 3.6 96.4 65 2 63 3.1 9. 69 Enlisted Groundcrew n Abnormal Normal 154 4 150 2.6 97.4 163 2 161 Officer n Abnormal Normal 127 8 119 6.3 93.7 Enlisted Flyer n Abnormal Normal 55 3 52 Enlisted Groundcrew CMS Sunnary Index Statistic Officer Gait Occupation n Abnormal Normal 154 10 144 *Estimated relative risk and confidence interval calculated after adding 0.5 to each cell. —No abnormals present in contrast; estimated relative risk, confidence interval, and p-value not calculated. �TABLE 1-2. Interaction Sumaries of Adjusted Exposure variable Interaction (Occupation) Stratification Diabetic . Cranial Nerve Function Index Light Touch . Exposure Index-byDiabetic Class (Enlisted Groundcrew) Exposure Index-byInsecticide Exposure (Enlisted Flyer) Exposure Index Low Adj. Relative Medium High 9Z . . Statistic Number Percent Number Percent Number Percent Contrast Risk ( 5 C I ) n Abnormal Normal n Impaired Abnormal Normal p-Value 7 0 7 00 . 100 0. 12 3 9 25.0 75.0 14 2 12 14.3 85.7 5.53 ( . 5 1 4 4 ) * 02,2.0* 30 (.37.1* .0 01,13)* 0.339* 0.263* 0.533* 11 0 11 00 . 100.0 14 2 12 14.3 85.7 18 0 18 Overall . 0 02,0.0* 0 0 M vs. L 4 6 ( . 0 1 6 3 ) * . — 1 0 0 H vs. L 0. 0.114* 047 .8* — Overall M vs. L H vs. L Normal* n Abnormal Normal 133 15 118 11.3 88.7 135 7 128 5.2 9. 48 109 7 102 64 . 93.6 Overall M vs. L 0 5 ( . 1 1 4 ) .5 02,.4 H vs. L 0 5 ( . 2 1 4 ) .6 02,.4 036 .3 0.225 028 .2 Yesb n Abnormal Normal 38 1 37 2.6 97.4 45 2 43 44 . 9. 56 40 4 36 10.0 9. 00 Overall M vs. L 0 4 ( . 2 8 8 ) . 2 00,.1 H vs. L 5 4 ( . 1 7 . 4 . 8 04,15) 006 .8 053 .7 014 .9 No n Abnormal Normal 17 2 15 11.8 88.2 19 0 19 00 . 100 0. 17 0 17 Overall 0 0 M vs. L 0.16 ( . 1 3 5 ) * . 00,.6* 1 0 0 H vs. L 0.18 ( . 1 3 9 ) * 0. 00,.8* 0.111* 026 .1* 045 .8* �TABLE 1-2. (continued) DatezacticD ^••ai-iog of AHjigt**j Exposure Variable Interaction (Occupation) Stratification Exposure Index High Low Medium Adj. Relative Statistic Number Percent Number Percent Number Percent Contrast Risk ( 5 C I ) 9% .. p-Value Born >1942 AchiUes Reflex Exposure Index-byAge (Enlisted Groundcrew) 1.1 98.9 129 2 127 1.6 93.4 82 0 82 00 . 100 0. Overall M vs. L 1.35 ( . 2 1 . 7 * 01,51) H vs. L 0.35 ( . 1 8 7 ) * 00,.0* 059 .3* 099 .9* 099 .9* Bom 1923-1941 n Abnormal Normal 63 1 62 1.6 9. 84 33 2 31 6.1 93.9 54 3 51 5.6 94.4 Overall M vs. L 4 0 ( . 5 4 . 4 * . 0 03,58) H vs. L 3.65 ( . 7 3 . 3 * 03,61) 046 .3* 021 .7* 034 .3* n Abnormal Normal 4 1 3 2. 50 75.0 1 1 0 100 0. 00 . 6 0 6 Bom £L942b Exposure Ihdex-byAge (Enlisted Groundcrew) 87 1 86 Born <L922 Muscle Status n Abnormal Normal n Abnormal Normal 86 2 84 2.3 97.7 129 6 123 4.7 95.3 81 1 80 1.2 98.8 Overall M vs. L 2.18 ( . 3 1 . 2 04,11) H vs. L 0.57 ( . 5 6 3 ) 00,.7 039 .0 036 .4 064 .4 n Bom 1923-1941" Abnormal Normal 63 5 58 7.9 92.1 33 2 31 6.1 93.9 54 6 48 11.1 88.9 Overall M vs. L 0 6 ( . 1 3 7 ) .4 01,.5 H vs. L 1.40 ( . 8 5 1 ) 03,.9 069 .4 064 .2 066 .1 n Abnormal Normal 4 3 1 75.0 2. 50 1 1 0 100.0 0;0 6 1 5 16.7 83.3 Overall M vs. L 1.29 ( . 3 5 . 1 * 00,35)* H vs. L 0 0 ( . 0 , . 1 * . 7 00315) 010 .0* 099 .9* 010 .9* Born <L922 Overall 0 0 M vs. L 7.00 ( . 7 2 1 3 ) * . 01,9.4* 1 0 0 H vs. L 0.18 ( . 1 5 6 ) * 0. 00,.8* 001 .5* 040 .0* 040 .0* �TABLE 1-2. (continued) Index AralV9PS I™* Selected tinimlniriral UaHahilM Variable Central Nervous System Index Interaction (Occupation) Exposure Index-byInsecticide Exposure (Enlisted Groundcrew) Stratification Exposure Index High Low Medium Adj. Relative Statistic Nuriber Percent Number Percent Nunber Percent Contrast Risk (95Z C.I.) p-Value Yes" n Abnormal Normal 102 4 98 3.9 96.1 113 7 106 6.2 93.8 95 3 92 3.2 96.8 Overall M vs. L H vs. L 1.68 (0.48,5.95) 0.87 (0.19,3.99) 0.560 0.419 0.853 No n Abnormal Normal 52 6 46 11.5 88.5 50 0 50 0.0 100.0 47 3 44 6.4 93.6 Overall M vs. L H vs. L 0.07 (0.004,1.29)** 0.38 (0.10,1.54)* 0.050* 0.027* 0.209* *Estimated relative risk or p-value, based on stratified tables. **Estimated relative risk and confidence interval calculated after adding 0.5 to each cell. —No abnormal participants present in contrast; adjusted relative risk, p-value, and confidence interval not presented. 'Results adjusted for age. b Results adjusted for diabetic class. Note: Small sample sizes may affect validity of overall p-value. �TABLE 1-3. Exclusions and Missing Data for Neurological Assessment by Group (Original Comparisons Only) Group Data Category Ranch Hand Original Comparison Total Lifetime Alcohol History (Drink-years); Missing Data 39 27 66 Peripheral Edema (Exclusion Category for Pin Prick, Light Touch, and Ankle Vibration) 13 12 25 Diabetic Class (Missing Data) 0 3 3 Positive Syphilis Serology (RPR and FTA); Exclusion Category 0 0 0 1-19 �TABLE 1-4. Unadjusted Analyses for Verified Neurological Diseases by Group*—1982-1985 (Original Comparisons Only) Group Abnormalities Original Comparison Ranch Hand Disease Category Number Percent Number Percent Total p-Value** Inflammatory Diseases 0 0.0 0 ' 0.0 0 Hereditary and Degenerative Diseases 2 0.2 0 0.0 2 0.500 Peripheral Disorders 18 1.8 20 2.1 38 0.626 Disorders of the Eye 5 0.5 5 0.5 10 0.999 Disorders of the Ear 6 0.6 6 0.6 12 0.999 Other Disorders 8 0.8 3 0.3 11 0.288 *Based on 1,016 Ranch Hands and 955 Original Comparisons; some participants may be classified in more than one category. **Fisher's exact test. 1-20 �TABLE 1-5. Unadjusted Analyses for Verified Neurological Diseases by Group*—Baseline and First Followup Studies Combined (Original Comparisons Only) Group Abnormalities Original Comparison Ranch Hand Disease Category Number Percent Number Percent Total p- Value**'* Inflammatory Diseases 3 0.3 2 0.2 5 0.999 Hereditary and Degenerative Diseases 2 0.2 3 0.3 5 0.678 Peripheral Disorders 23 2.3 27 2.8 50 0.475 Disorders of the Eye 16 1.6 16 1.7 32 0.861 Disorders of the Ear 24 2.4 26 2.7 50 0.668 Other Disorders 15 1.5 12 1.3 27 0.703 *Based on 1,016 Ranch Hands and 955 Original Comparisons; some participants may be classified in more than one category. **Fisher's exact test. *Group significance for patterns of disease: 1-21 p=0.702. �TABLE 1-6. Unadjusted Analyses for Cranial Nerve nncn.cn iay uroup vunginajL \ ^jflnjujri^ riR uniy /j ••—** **•*«/ Gn3UP Original Comparison Number Percent Est. Relative Risk (95% C.I.) Variable Cranial Nerve Statistic Ranch Hand Number Percent Smell I Olfactory n Abnormal Normal 1,016 10 1,006 1.0 99.0 955 9 946 0.9 99.1 1.05 (0.42,2.58) 0.999 n Abnormal Normal 1,016 6 1,010 0.6 99.4 955 4 951 0.4 99.6 1.41 (0.40,5.02) 0.754 n Abnormal Normal 1,015 8 1,007 0.8 99.2 954 7 947 0.7 99.3 1.08 (0.39,2.98) 0.999 Oculomotor IV Trochlear VI Abducens n Abnormal Normal 1,016 6 1,010 0.6 99.4 955 5 950 0.5 99.5 1.13 (0.34,3.71) 0.999 Facial Sensation V Trigeminal n Abnormal Normal 1,014 4 1,010 0.4 99.6 953 2 951 0,2 99,8 1.88 (0.34,10.31) 0.688 Jaw Clench V Trigeminal n Abnormal Normal 1,016 2 1,014 0.2 99.8 955 1 954 0.1 99.9 1.88 (0.17,20.79) 0.999 n Abnormal Normal 1,016 7 1,009 0.7 99.3 955 2 953 0,2 99.8 3.31 (0.69,15.95) 0.181 n Abnormal Normal 1,015 7 1,008 0.7 99.3 955 5 950 0.5 99.5 1.32 (0.42,4.17) 0.775 Visual Fields Light Reaction n Optic m Oculomotor p-Value m Ocular Movements Smile vn Facial Palpebral Fissures vn Facial 1-22 �TABLE 1-6. (ocntiiuad) UHU justea UK uyses tor Lianiai. Nerve Function by Group (Original Comparisons Only) Group Variable Balance Cranial Nerve vm Acoustic Gag Reflex IX Glossopharyngeal Speech X Vagus Original Comparison Nunber Percent Statistic Ranch Hand Number Percent n Abnormal Normal 1,015 2 1,013 0.2 99.8 955 1 954 0.1 99.9 1.88 ( . 7 2 . 1 0.999 01,08) n Abnormal Normal 1,014 1 1,013 0.1 99.9 955 1 954 0.1 99.9 0.94 ( . 6 1 . 8 00,50) n Abnormal Normal 1,016 3 1,013 0.3 99.7 954 0 954 0.0 100.0 6.59 ( . 4 1 7 8 ) 0.250 03,2.0* n Abnormal Normal 1,015 3 1,012 0.3 99.7 955 0 955 0.0 100.0 6.61 ( . 4 1 8 0 ) 0.250 03,2.6* Est. Relative Risk (95%C.I.) p-Value 0.999 Tongue Position Relative to Midline X Vagus Palate and Uvula Movement H Spinal Accessory n Abnormal Normal 1,014 2 1,012 0.2 99.8 955 1 954 0.1 99.9 1.89 ( . 7 2 . 3 0.999 01,08) Neck Range of Motion xn Hypoglossal n Abnormal Normal 1,016 61 955 6.0 94.0 955 64 891 6.7 93.3 0.89 ( . 2 1 2 ) 06,.8 0.579 n Abnormal Normal 1,003 . 96 907 9.6 90.4 941 89 852 9.5 90.5 1.01 ( . 5 1 3 ) 07,.7 0.939 Cranial Nerve Function Index 1-23 �TABLE 1-6. (continued) Unadjusted Analyses for Cranial Nerve junction ay uroup tunguEu umpansons uuy; Group Variable Cranial Nerve Cranial Nerve Function Index (Neck. Range of Motion Excluded) Statistic Ranch Hand Number Percent n Abnormal Norral 1,003 42 961 4.2 9. 58 Original Comparison Number Percent 941 29 912 3.1 9. 69 Est. Relative Risk ( 5 C.I.) p-Value 9* 1.34 ( . 5 2 2 ) 0.226 08,.3 ^Estimated relative risk and confidence interval calculated after adding 0.5 to each cell. 1-24 �TABLE 1-7. Adjusted Analyses for Selected Variables of Cranial Nerve Function by Group (Original Caparisons Only) Group Variable Statistic Ranch Hand Member Percent Original Comparison Nurber Percent Adj. Relative Risk ( 5 C.I.) p-Value 9% Covariate Remarks* n Neck Range of Abnormal Motion Normal 1,016 61 955 6.0 94.0 955 64 891 6.7 93.3 0.92 ( . 4 1 5 ) 0.647 06,.7 ACE(p<0.001) Cranial n Nerve Function Abnormal Index Normal 1,003 % 907 9.6 90.4 941 89 852 9.5 90.5 1.04 ( . 6 1 4 ) 0.812 07,.1 AGE(p<0.001) 964 38 926 3.9 96.1 911 28 883 3.1 96.9 1.27 ( . 7 2 1 ) 0.350 07,.2 DIAB*INS(p=0.032) OCC*OKm(p=0.016) OCC*DIAB(p=0.032) Cranial Nerve Function Index (Neck Range of Motion Excluded) n Abnormal Nornel Alternative Model—Includes Missing Drink-Year Participants* 'b n Abnormal Normal 1,003 42 961 4.2 95.8 938 28 910 3.0 97.0 1.41 ( . 6 2 3 ) 0.170 08,.2 DIAB*INS(i>O.022) OCC*DIAB(p=0.027) ^Abbreviations; DIAB: diabetic class INS: insecticide exposure OCC: occupation CROR: drink-years 'lifetime alcohol consumption (drink-years) not used as a covariate. b 66 missing drink-year participants: 4/39 Ranch Hands abnormal; 0/27 Original Comparisons abnormal. 1-25 �TABLE 1-8. Unadjusted Analyses for Peripheral Nerve Function by Group (Original Comparisons Only) Group Variable Statistic Ranch Hand Number Percent Original Comparison Number Percent Est,. Relative Risk (95% C.I.) p-Value n Pin Prick Abnormal Normal 1,003 59 944 5.9 94.1 943 62 881 6.6 93.4 0.89 (0.62,1 .28) 0.573 Light Touch n Abnormal Normal 1,003 38 965 3.8 96.2 943 33 910 3.5 96.5 1.09 (0.68,1 .75) 0.809 Muscle Status n Abnormal Normal 1,016 26 990 2.6 97.4 955 26 929 2.7 97.3 0.94 (0.54,1 .63) 0.888 n Vibratory Abnormal Sensation Normal 1,003 11 992 1.1 98.9 943 7 936 0.7 99.3 1.48 (0.57,3 .84) 0.482 Patellar Reflex n Abnormal Normal 1,016 11 1,005 1.1 98.9 954 14 940 1.5 98.5 0.74 (0.33,1 .63) 0.547 Achilles Reflex n Abnormal Normal 1,009 58 951 5.7 94.3 950 62 888 6.5 93.5 0.87 (0.60,1 .26) 0.510 Biceps Reflex n Abnormal Normal 1,016 9 1,007 0.9 99.1 955 10 945 1.0 99.0 0.85 (0.34,2 .09) 0.819 Babinski Reflex n Abnormal Normal 1,011 4 1,007 0.4 99.6 951 4 947 0.4 99.6 0.94 (0.24,3 .77) 0.999 1-26 �TABLE 1-9. Adjusted Analyses for Selected Variables of Peripheral Nerve Function by Group (Original Comparisons Only) Group Variable Statistic n Pin Prick Abnormal Nonnal Ranch Band Number Percent Original Comparison Number Percent 1,003 59 944 5.9 94.1 941 61 880 6.5 93.5 Adj. Relative Risk (95XC.I.) p-Value **** **** Covariate Remarks* GBP*DIAB(p=0.001) AGE(p<0.001) Light Touch Abnormal Normal 964 37 927 3.8 96.2 914 31 883 3.4 96.6 1.12 ( . 9 1 8 ) 0.652 06,.2 AGE(p=0.049) Muscle Status n Abnormal Normal 977 25 952 2.6 97.4 925 24 901 2.6 97.4 1.04 ( . 7 1 8 ) 0.893 05,.9 EHOR*AGE(D=O.Q37) DIAB*INS(p"0.028) Achilles Reflex n Abnormal Normal 1,009 58 951 5.7 94.3 947 61 886 6.4 93.6 0.92 ( . 3 1 3 ) 0.677 06,.5 AGE(p<0.001) DIAB(p<0.001) n *Additional Abbreviation: GBP: group ***><toup-by-<»variate interaction—adjusted relative risk, confidence interval, and p-value are not presented. 1-27 �TABLE 1-10. Summary of Group-by-Diabetic Class Interaction for Pin Prick (Original Comparisons Only) Group Variable Interaction Stratification Statistic Ranch Hand Number Percent Original Comparison Number Percent Adj. Relative Risk (95% C.I.) p-Value Abnormal Group-byPin Prick Diabetic Class n Abnormal Normal 76 13 63 17.1 82.9 74 7 67 9.5 90.5 2.03 (0.76,5.43) 0.161 Impaired n Abnormal Normal 105 1 104 1.0 99.0 135 14 121 10.4 89.6 0.08 ( . 1 0 6 ) 0.015 00,.1 Normal n Abnormal Normal 822 45 777 5.5 94.5 732 40 692 5.5 94.5 1.01 (0.65,1.57) 0.951 M 1 to 00 �TABLE 1-11. lhadjusted Analyses for CMS Coordination Variables by Group (Original Comparisons Only) Group Ranch Hand Number Percent Original Comparison Est. Relative Number Percent Risk ( 5 C.I.) p-Value 9% Variable Statistic Tremor n Abnormal Normal 106 ,1 26 990 2.6 97.4 955 14 941 1.5 1.77 ( . 2 3 4 ) 0.110 09,.0 9. 85 n Coordination Abnormal Normal 1,015 9 106 ,0 0.9 99.1 955 7 948 0.7 1.21 ( . 5 3 2 ) 0 8 4 04,.7 .0 99.3 Romberg Sign n Abnormal Normal 1,015 2 1,013 0.2 9. 98 955 1 954 0.1 1 8 ( . 7 2 . 1 0 9 9 .8 01,08) .9 9. 99 Gait n Abnormal Normal 106 ,1 20 996 2.0 9. 80 954 11 943 1.2 1.72 ( . 2 3 6 ) 0.153 08,.1 9. 88 n Abnormal Normal 1,015 48 967 4.7 95.3 954 29 925 3.0 1.58 ( . 9 2 5 ) 0.062 09,.3 9. 70 CNS Sunmary Index 1-29 �TABLE 1-12. Adjusted Analyses for Selected Variables of CMS Coordination by Group (Original Comparisons Only) Group Original Comparison Number Percent Variable Statistic Ranch Hand Number Percent Adj. Relative Risk ( 5 C.I.) p-Value 9% Covariate Remarks Tremor n Abnormal Normal 1,016 26 990 2.6 97.4 952 14 938 1.5 98.5 1.68 ( . 7 3 2 ) 0.112 08,.5 DIAB(p=0.002) Gait n Abnormal Normal 1,016 20 996 2.0 98.0 951 10 941 1.1 98.9 1.85 ( . 6 3 9 ) 0.105 08,.9 DIAB(p=0.035) n Abnormal Normal 1,015 48 967 '4.7 95.3 951 28 923 2.9 97.1 1.58 ( . 8 2 5 ) 0.055 09,.5 DIAB(p=0.009) M I CNS Summary Index �TABLE 1-13. Longitudinal Analysis of Romberg Sign and Babinski Reflex: A Contrast of Baseline and Follovup Examination Abnormalities (Original Comparisons Only) 1985 1982 Followup Baseline Exam Odds p-Value Exam Abnormal Normal Ratio (OR)* (ORt>u vs ORO C n^) Rn Variable Group 2 0 188 777 0 0 1 180 690 0.006 Ranch Hand Abnormal Normal 1 3 7 953 0.43 Original Abnormal Comparison Normal Babinski Reflex Abnormal Normal Abnormal Original Comparison Normal Romberg Sign 0 4 1 863 Ranch Hand 0.31 0.07 4.00 Number Normal Baseline, Abnormal Followup *OHH<! Rat In; - .... . Number Abnormal Baseline, Normal Follovup 1-31 �APPENDIX J Psychological Assessment �APPENDIX J: Psychological Assessment Contents Table Page J-l Summary of Kolmogorov-Smirnov Tests on MMPI for Officers J-l J-2 Summary of Kolmogorov-Smirnov Tests on MMPI for Enlisted Flyers J-2 J-3 Summary of Kolmogorov-Smirnov Tests on MMPI for Enlisted Groundcrev J-3 J-4 Summary of Group-by-Covariate Interactions for Psychological Variables J-4 J-5 Summary of Kolmogorov-Smirnov Tests on CMI J-7 J-6 Unadjusted Exposure Index Analyses for Psychological Variables by Occupation J-8 J-7 Interaction Summaries of Adjusted Exposure Index Analyses for Psychological Variables J-17 J-8 Unadjusted Analyses for Reported Psychological Illnesses by Group; Baseline and First Followup Studies Combined (Original Comparisons Only) J-27 J-9 Summary of Kolmogorov-Smirnov Tests on MMPI for Officers (Original Comparisons Only) J-28 J-10 Summary of Kolmogorov-Smirnov Tests on MMPI for EnlistedFlyers (Original Comparisons Only) J-29 J-ll Summary of Kolmogorov-Smirnov Tests on MMPI for Enlisted Groundcrew (Original Comparisons Only) J-30 J-12 Summary of Kolmogorov-Smirnov Tests on CMI (Original Comparisons Only) J-31 J-13 Unadjusted Analyses for MMPI by Group (Original Comparisons Only) J-32 J-14 Adjusted Analyses for MMPI by Group (Original Comparisons Only) J-34 J-15 Unadjusted Analyses for the Cornell Medical Index (CMI) by Group J-36 J-16 Adjusted Analyses for CMI Variables by Group (Original 'Comparisons Only) J-37 J-i �APPENDIX J: Psychological Assessment Contents Table J-17 Summary Results for the Halstead-Reitan Battery Impairment Index Unadjusted and Adjusted Analyses (Original Comparisons Only) J-38 J-18 Summary of Group-by-Covariate Interactions for Psychological Variables (Original Comparisons Only) J-39 J-ii �TABLE J-l. Summary of Kolmogorov-Smirnov Tests on MMPI for Officers Parameter* Group Mean Score Standard Deviation KolmogorovSmirnov p-Value Anxiety Ranch Hand Comparison 52.82 53.64 7.38 7.77 0.281 Consistency Ranch Hand Comparison 49.23 49.23 6.92 6.62 0.995 Defensiveness Ranch Hand Comparison 49.64 49.61 6.64 6.97 0.990 Denial Ranch Hand Comparison 59.47 59.81 7.31 7.46 0.877 Depression Ranch Hand Comparison 55.11 55.27 10.17 10.28 0.999 Hypochondria Ranch Hand Comparison 55.60 55.02 8.85 8.76 0.970 Hysteria Ranch Hand Comparison 60.45 59.97 7.00 7.47 0.627 Mania/Hypomania Ranch Hand Comparison 54.36 53.39 8.94 9.12 0.092 Masculinity/ Femininity Ranch Hand Comparison 58.83 58.10 8.64 9.07 0.366 Paranoia Ranch Hand Comparison 53.71 53.91 7.34 7.34 0.975 Psychopathic/ Deviate Ranch Hand Comparison 56.82 57.08 8.87 10.02 0.837 Schizophrenia Ranch Hand Comparison 54.80 55.14 8.48 9.02 0.942 Social Introversion Ranch Hand Comparison 46.56 47.19 7.43 7.77 0.594 Validity Ranch Hand Comparison 0.93 0.89 3.11 3.20 0.776 *n=380 Ranch Hands; n=480 Comparisons (except for validity, where n=483 Comparisons). J-l �TABLE J-2. Summary of Kolmogorov-Smirnov Tests on MMPI for Enlisted Flyers Group Mean Score Standard Deviation KolmogorovSmirnov p-Value Anxiety Ranch Hand Comparison 54.78 55.09 9.73 10.09 0.678 Consistency Ranch Hand Comparison 52.24 51.83 10.30 8.22 0.995 Defensiveness Ranch Hand Comparison 53.05 52.00 7.96 8.31 0.270 Denial Ranch Hand Comparison 55.95 56.95 8.73 8.50 0.807 Depression Ranch Hand Comparison 58.21 58.32 11.79 11.86 0.744 Hypochondria Ranch Hand Comparison 58.48 59.25 12.16 12.85 0.496 Hysteria Ranch Hand Comparison 60.00 61.32 8.94 9.47 0.366 Mania/Hypomania Ranch Hand Comparison 55.05 54.99 10.00 9.87 0.496 Masculinity/ Femininity Ranch Hand Comparison 55.79 56.57 7.85 8.46 0.776 Paranoia Ranch Hand Comparison 52.77 51.92 8.43 7.83 0.695 Psychopathic/ Deviate Ranch Hand Comparison 57.48 58.95 11.40 10.79 0.088 Schizophrenia Ranch Hand Comparison 56.56 57.06 13.03 11.49 0.178 Social Introversion Ranch Hand Comparison 50.35 49.48 8.89 8.14 0.678 Validity Ranch Hand Comparison 3.93 6.31 42.48 54.96 0.864 Parameter* *n=176 Ranch Hands; n=208 Comparisons (except for validity, where n=177 Ranch Hands and n=210 Comparisons). J-2 �TABLE J-3. Summary of Kolmogorov-Srairnov Tests on HMPI for Enlisted Groundcrew Group Mean Score Standard Deviation KolmogorovSmirnov p-Value Anxiety Ranch Hand Comparison 55.86 55.61 10.96 9.58 0.993 Consistency Ranch Hand Comparison 53.05 52.03 10.64 8.43 0.877 Defensiveness Ranch Hand Comparison 52.42 52.60 8.27 8.35 0.970 Denial Ranch Hand Comparison 55.53 56.52 8.57 8.88 0.118 Depression Ranch Hand Comparison 58.79 57.66 12.98 10.78 0.594 Hypochondria Ranch Hand Comparison 58.44 57.13 13.22 11.12 0.142 Hysteria Ranch Hand Comparison 60.36 59.78 9.80 8.99 0.577 Mania/Hypomania Ranch Hand Comparison 55.26 56.33 9.78 10.17 0.450 Masculinity/ Femininity Ranch Hand Comparison 56.18 57.18 8.24 8.82 0.292 Paranoia Ranch Hand Comparison 53.12 53.73 9.05 8.47 0.393 Psychopathic/ Deviate Ranch Hand Comparison 59.01 59.26 11.29 10.66 0.728 Schizophrenia Ranch Hand Comparison 58.20 57.36 14.22 11.12 0.837 Social Introversion Ranch Hand Comparison 51.86 50.43 9.82 8.71 0.107 Validity Ranch Hand Comparison 1.97 0.89 26.48 4.80 0.999 Parameter* *n=458 Ranch Hands; n=598 Comparisons (except for validity, where n=459 Ranch Hands and n=i599 Comparisons). J-3 �TABLE J-4. SuBBary of Group-by-Covariate Interactions for Psychological Variables Group Variable Interaction Anxiety Group-byEducation Statistic High School n Number/* Abnormal Normal 564 n Number/* Abnormal Normal 448 n Number /Z Abnormal Normal 537 •n Number/* Abnormal Normal 437 n Number /Z Abnormal Normal 192 n Number/* Abnormal Normal 368 n Number/* Abnormal Normal 414 College Consistency Group-byEducation High School College Depression Group-byCombat Index Ranch Hand Stratification Lov Medium High 57 507 16 432 30 507 6 431 28 164 37 331 43 371 Comparison Adj. Relative Risk (95Z C.I.) p-Value 685 10. 1Z 89. 9Z 49 636 7.2Z 92. 8Z 1.39 (0.92,2.08) 0.114 5.0Z 95. OZ 0.68 (0.36,1.28) 0.233 2.9Z 97. 5Z 1.81 (1.00,3.28) 0.051 2.6Z 97. 4Z 0.46 (0.17,1.19) 0.110 8.2Z 91.8* 1.73 (1.03,2.91) 0.039 9.1Z 90. 9Z 1.09 (0.67,1.75) 0.657 13.9* 86.1* 0.74 (0.48,1.16) 0.159 600 3.6Z 96. 4Z 30 570 661 5.6Z 94. 4Z 19 642 585 1.4Z 98.6* 15 570 490 14.6* 85.4* 40 450 417 10. 1Z 89. 9Z 38 379 339 10.4* 89.6* 47 292, �TABLE J-4. (continued) Summary of Group-by-Covariate Interactions for Psychological Variables Group Variable Paranoia Interaction Group-by Age Statistic Born >1942 n Number/Z Abnormal Normal 410 n Number/Z Abnormal Normal 602 n Number/Z Abnormal Normal 537 n Number/Z Abnormal Normal 439 n Number/Z Abnormal Normal 198 n Number/Z Abnormal Normal 384 n Number/* Abnormal Normal 430 Born <1942 Schizophrenia Group-byEducation High School «-. I College Social Introversion Group-byCombat Index Ranch Hand Stratification Low Hedium High 15 395 16 586 72 465 18 421 11 187 6 378 9 421 Adj. Relative Comparison Risk (95Z C.I.) p-Value 3.5Z 96. 5Z 0.88 (0.43,1.77) 0.712 1.2Z 98.8* 2.63 (1.11,6.20) 0.027 9.5Z 90. 5Z 1.51 (1.05,2.16) 0.033 6.3Z 93. 7Z 0.63 (0.35,1.12) 0.119 1.2Z 98.8* 4.86 (1.77,13.36) 0.002 0.9* 99. 1Z 1.59 (0.44,5.68) 0.478 2.6* 97.4* 0.82 (0.32,2.09) 0.677 545 3.7Z 96.3* 19 526 740 2.7Z 97. 3Z 9 731 665 13.4* 8.* 66 63 602 585 4.1Z 95. 9Z 37 548 509 5.6Z 94. 4Z 6 503 427 1.6* 98.4* 4 423 349 2.1Z 97. 9Z 9 340 �TABLE J-4. (continued) Summary of Group-by-Covariate Interactions for Psychological Variables Group Variable Interaction Validity Group-by- Stratification Nonblack Statistic Ranch Hand n 208 746 Risk ( 5 C.I.) 9% p-Value 1,206 954 Number/% >0 0 Race Adj. Relative Comparison 235 971 1.20 (0.97,1.49) 0.095 42.2% 57.8% 0.46 ( . 2 0 9 ) 02,.6 0.038 31.43 95% C.I.* ( 0 2 ,45.65) (25 . 6 3 . 6 3.5 4,87) Black 19.5% 80.5% (-1 Total CMI I Group-byEducation High School — <0.001 574 n 433 Number/% Adj. Mean* 23.72 24.22 95% C.I.* (19.10,29.41) (19 .57,29.91) — 0.657 n 15 45 83 60 Number/% >0 0 21.8% 78.2% n Number/% 25.0% 75.0% 35 48 655 529 Adj. Mean* 37.18 College M-R Subscore Group-byEducation High School College n Number/% 0 (Low) l-lO(Medium) >10 (High) n Number/% 675 555 246 267 42 44.3% 48.1% 7.6% 0.030 352 283 40 52.1% 41.9% 5.9% Medium vs. Low 1.37 (1.07,1.75) High vs. Low 1.33 (0.82,2.15) 0.014 63.2% 33.9% 2.9% Overall: Medium vs. Low 0.91 (0.70,1.18) High vs. Low 0 8 (0.35,1.84) .0 590 443 0 (Low) 292 l-lO(Medium) 141 >10 (High) 10 Overall: 65.9% 31.8% 2.3% 373 200 17 transformed from log(X+l) scale, where X was the number of questions answered "yes." —No relative risk given for Total CMI, which was analyzed as a continuous variable. 0.254 0.699 0.465 0.597 �TABLE J-5. Summary of Kolmogorov-Smirnov Tests on CHI Group Mean Score Ranch Hand Comparison 402 543 15.88 13.07 15.21 12.40 <0.001 Born 1923-1941 Ranch Hand Comparison 562 674 15.49 14.64 12.99 12.52 0.281 Born <1922 Ranch Hand Comparison 36 51 18.56 16.57 » 15.94 9.27 0.970 941 Ranch Hand Comparison 1,187 15.63 13.94 13.83 12.00 0.003 Ranch Hand Comparison 59 81 17.81 15.68 16.91 17.01 0.481 Ranch Hand Comparison 556 677 18.88 15.77 15.80 13.50 <0.001 Ranch Hand Comparison 444 591 11.85 12.08 10.18 10.63 0.999 Current Alcohol Use Drinker 851 Ranch Hand Comparison 1,092 15.16 13.72 13.12 12.00 0.024 Stratification Age Born XL942 Race Nonblack Black Education High School College Standard Deviation KolmogorovSmirnov p-Value Sample Size Ranch Hand Comparison 148 176 18.84 16.06 17.71 14.38 0.149 Ranch Hand Comparison 377 476 11.23 10.88 9.01 9.09 0.807 Enlisted Flyer Ranch Hand Comparison 174 204 17.72 15.63 14.84 13.53 0.220 Enlisted Groundcrev Ranch Hand Comparison 449 588 18.80 16.06 16.06 13.69 0.007 Nondrinker Occupation Officer J-7 �TABLE J-6. Unadjusted Exposure Index Analyses for Psychological Variables by Occupation Exposure Index Lov Medium n Number/Z Abnormal Normal 127 n Number/Z Abnormal Normal 54 65 Enlisted Flyer 8 14.82 46 85 .2Z 4 6.2Z 61 93. 8Z n Number/Z Abnormal Normal 154 163 Enlisted Groundcrev 19 12 .3Z 135 87 .7Z Officer Anxiety Contrast p-Value Statistic n Number/Z Abnormal Normal n Number/Z Abnormal Normal 54 65 57 Overall 0.476 Enlisted Flyer 4 7.4Z 50 92 .6Z 2 3.1Z 63 96.92 2 3.5Z 55 96. 5Z M vs. L H vs. L 0.40 (0.07,2. 26) 0.409 0.46 ( . 8 2 59) 0.430 00,. n Number/Z Abnormal Normal 154 163 Overall 0.813 Enlisted Groundcrev 9 5.82 145 94 .2Z 9 5.5Z 6 4.3Z 154 94. 5Z 135 95. 7Z H vs. L H vs. L 0.94 (0.36,2.44) 0.999 0.72 (0.25,2.07) 0.603 130 High Est. Relative Risk (95Z C.I .) Occupation Officer Variable Overall 0.876 M vs. L H vs. L 1.48 (0.24,8.99) 0.999 1.56 (0.26,9.52) 0 6 0 .8 57 Overall 0.211 4 7.0Z 53 93. OZ H vs. L H vs. L 0.38 (0.11,1.33) 0.137 0.43 (0.12,1.54) 0.230 Overall 0.560 18 11. OZ 12 8.5Z 145 89. OZ 129 91. 5Z M vs. L B vs. L 0.88 (0.44,1. 75) 0.730 0.66 (0.31,1.42) 0.343 127 130 Overall 0.380 2 1. Z 6 125 98 .4Z 0 O.OZ M vs. L 2 1.5Z 128 98.52 123 100. OZ H vs. L 0.98 (0.14,7.04) 0.999 0.20 (0.01,4. 28)* 0.498 123 2 1 .62 3 2.3Z 3 2.4Z 125 98 • 4Z 127 97. 7Z 120 97. 6Z 4 I 00 Consistency 141 123 141 �TABLE J-6. (continued) Unadjusted Exposure Index Analyses for Psychological Variables by Occupation Low Exposure Index Medium High Est. Relative Risk (95% C.I.) Occupation Statistic n Number/% Abnormal Normal 127 Officer n Number/% Abnormal Normal 54 n Number /% Abnormal Normal 154 Enlisted Groundcrew n Number/% Abnormal Normal 127 Officer Variable n Number/% Abnormal Normal 54 65 Overall 0.366 Enlisted Flyer 2 3.7% 52 96.3% 2 3.1% 0 0.0% M vs. L 63 96.9% 57 100.0% H vs. L 0.83 ( .11,6.06) 0.999 0 0.18 ( .01,3.89)* 0.234 0 n Number/% Abnormal Normal 154 Enlisted Groundcrew Defensiveness Enlisted Flyer Denial 130 123 Contrast p-Value Overall 0.603 0 0.0% 1 0.8% 1 0.8% M vs. L 127 100.0% 129 99.2% 122 99.2% H vs. L 2.95 (0.12,73.18)* 0.999 3.12 (0.13,77.39)* 0.492 65 57 Overall 0.991 1 1.9% 1 1.5% 1 1.8% M vs. L 53 98.1% 64 98.5% 56 98.2% H vs. L 0 0.83 ( .05,13.56) 0.999 0.95 ( .06,15.52) 0.999 0 Overall 0.697 163 141 5 3.1% 6 3.9% 7 5.0% M vs. L 148 96.1% 158 96.9% 134 95.0% H vs. L 130 123 Overall 1 0.8% M vs. L 6 4.7% 2 1.5% 121 95.3% 128 98.5% 122 99.2% H vs. L 57 141 0.78 ( .23,2.61) 0 0 1.29 ( .42,3.93) 0.765 0.779 0.094 0.32 ( .06,1.59) 0 0.17 ( .02,1.39) 0 0.169 0.120 Overall 0.326 2 1.3% 0 0.0% 2 1.4% M vs. L 152 98.7% 163 100.0% 139 98.6% H vs. L 0.19 ( .01,3.92)* 0.235 0 0 1.09 ( . 15,7.87) 0.999 163 �TABLE J-6. (continued) Unadjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Low Exposure Index Medium High c_ ! Statistic n Number/% Abnormal Normal 127 130 Officer 8 6.3% 119 93.7% 6 4.6% 10 8. 1% 124 95.4% 113 91.9% n Number/% Abnormal Normal 54 65 Enlisted Flyer 11 20.4% 43 79.6% 7 10.8% 58 89.2% n Number/% Abnormal Normal 154 Enlisted Groundcrev Officer Depression Occupation Hypochondria Enlisted Flyer Enlisted Groundcrev 123 Est. Relative Contrast Overall Risk (95% C.I •) p-Value 0.517 M vs. L H vs. L 0.72 (0.24,2.14) 0.593 1.32 (0.50,3.45) 0.631 57 Overall 0.223 6 10.5% 51 89.5% M vs. L H vs. L 0.47 (0.17,1.32) 0.199 0.46 (0.16,1.35) 0.191 Overall 0.695 25 16.2% 21 12.9% 20 14.2% 129 83.8% 142 87.1% 121 85.8% M vs. L H vs. L 0.76 (0.41,1.43) 0.428 0.85 (0.45,1.62) 0.746 n Number/% Abnormal Normal 127 Overall 0.530 M vs. L H vs. L 0.57 (0.20,1.61) 0.312 0.92 (0.36,2. 36) 0.999 n Number/% Abnormal Normal 54 65 57 Overall 0.059 11 20.4% 43 79.6% 4 6.2% 61 93.8% 10 17.5% 47 82.5% H vs. L H vs. L 0.26 ( . 8 0 86) 0.026 00,. 0.83 (0.32,2.15) 0.810 n Number/% Abnormal Normal 154 163 25 16.2% 129 83.8% 18 11.0% 26 18.4% 145 89.0% 115 81.6% 163 130 141 123 9 7.3% 10 7.9% 6 4.6% 117 92.1% 124 95.4% 114 92.7% 141 Overall M vs. L H vs. L 0.176 0.64 (0.33,1.23) 0.192 1.17 (0.64,2. 13) 0.646 �TABLE J-6. (continued) Unadjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Low Exposure Index Medium High 130 7 5.5% 120 94.5% 9 6.9% 14 11.4% 121 93.1% 109 88.6% n Number/% Abnormal Normal 54 65 57 Overall Enlisted Flyer 11 20.4% 43 79.6% 6 9.2% 59 90.8% 7 12.3% 50 87.7% M vs. L H vs. L 0.40 (0.14,1.16) 0.115 0.55 (0.20,1.54) 0.307 n Number/% Abnormal Normal 154 163 Overall 0.065 Enlisted Groundcrew 28 18.2% 126 81.8% 16 9.8% 25 17.7% 147 90.2% 116 82.3% M vs. L H vs. L 0.49 (0.25,0.95) 0.035 0.97 (0.54,1.76) 0.999 n Number/% Abnormal Normal 127 130 Overall 0.585 Officer Mania/ Hypomania n Number/% Abnormal Normal 127 Officer Hysteria Statistic 8 6.3% 119 93.7% 8 6.5% 5 3.9% 125 96.1% 115 93.5% M vs. L H vs. L 0.60 (0.19,1.87) 0.407 1.04 (0.38,2.85) 0.999 n Number /% Abnormal Normal 54 65 57 Overall 0.623 Enlisted Flyer 3 5.6% 51 94.4% 6 9.2% 59 90.8% 6 10.5% 51 89.5% M vs. L H vs. L 1.73 (0.41,7.27) 0.509 2.00 (0.47,8.44) 0.491 n Number/% Abnormal Normal 154 163 Overall 0.568 Enlisted Groundcrew 11 7.1% 143 92.9% 6 4.3% 10 6.1% 153 93.9% 135 95.7% M vs. L H vs. L 0.85 (0.35,2.06) 0.823 0.58 (0.21,1.61) 0.326 123 141 123 141 Contrast Est. Relative Risk (95% C.I . ) Occupation p-Value Overall 0.200 M vs. L H vs. L 1.28 (0.46,3.53) 0.797 2.20 (0.86,5.66) 0.113 0.198 �TABLE J-6. (continued) Unadjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Statistic Officer n Number/% Abnormal Normal n Masculinity/ Femininity Enlisted Flyer Low Exposure Index Medium High 127 130 7 5.5% 120 94.5% 115 88.5% Est. Relative Contrast Risk (95% C • I.) p-Value 0.142 123 15 11.5% Overall 15 12.2% 108 87.8% M vs. L B vs. L 08, 2.24 ( . 8 5.68) 0.118 2.38 ( . 4 6.06) 0.075 09, Overall 0.478 M vs. L H vs. L 3.48 (0.38,32.06) 0.375 1.93 (0.17,21.89) 0.999 54 65 57 1 1.9% 53 98.1% 4 6.2% 61 93.8% 55 96.5% Number/% Abnormal Normal 2 3.5% 154 163 141 Overall 0.360 Normal 10 6.5% 144 93.5% 5 3.1% 158 96.9% 7 5.0% 134 95.0% M vs. L H vs. L 0.46 (0.15,1.37) 0.189 0.75 (0.28,2.03) 0.625 n Number/% 127 130 123 Overall 0.194 Abnormal 1 0.8% 126 99.2% 1 0.8% 129 99.2% 4 3.3% 119 96.7% M vs. L H vs. L 0.98 ( . 6 15.79) 0.999 00, 4.24 (0.47,38.44) 0.208 Overall 0.447 M vs. L H vs. L 0.27 (0.03,2.63) 0.328 0.94 (0.18,4.90) 0.999 141 Overall 0.735 7 5.0% 134 95.0% M vs. L H vs. L 1.14 (0.34,3.81) 0.999 1.56 ( . 8 5.02) 0.560 04, n Enlisted Groundcrev Officer Number/% Abnormal Normal Paranoia Enlisted Flyer n Number/% Abnormal Normal n Enlisted Abnormal 65 3 5.6% 51 94.4% 1.5% 3 5.3% 64 98.5% 54 94.7% 1 154 163 5 3.3% 149 96.7% 157 96.3% Number/% Groundcrev 57 54 Normal 6 3.7% �TABLE J-6. (continued) Unadjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Low Exposure Index Medium High Est. Relative Occupation Statistic n Number/% Abnormal Normal 127 130 Officer 9 7.1% 118 92.9% 7 5.4% 11 8.9% 123 9 . % 112 91.1% 46 n Number/% Abnormal Normal 54 65 8 14.8% 46 85.2% 11 16.9% 6 10.5% 54 83.1% 51 89.5% n Number/% Abnormal Normal 154 163 Enlisted Groundcrev 25 16.2% 129 83.8% Officer n Number/% Abnormal Normal n Number/% Abnormal Normal 54 65 57 Overall 8 14.8% 46 85.2% 6 9.2% 59 9 . % 08 5 8.8% 52 91.2% M vs. L H vs. L 0.59 (0.19,1 . 0 0 4 0 8) .0 0.55 (0.17,1 .81) 0.385 n Number/% Abnormal Normal 154 163 141 Overall 0.600 18 11.7% 136 88.3% 25 15.3% 138 84.7% 21 14.9% 120 85.1% M vs. L H vs. L 1.37 (0.71,2 .62) 0.413 1.32 (0.67,2 . 0 0.492 6) Psychopathic/ Enlisted Deviate Flyer - Schizophrenia Enlisted Flyer Enlisted Groundcrev 123 Contrast Risk (95% C.I.) p-Value Overall 0.545 M vs. L H vs. L 0.75 (0.27,2 . 7 0.614 0) 05, .4 1.29 ( . 1 3 .23) 0 6 6 Overall 0.593 M vs. L H vs. L 1.17 (0.43,3 .16) 0 8 6 .0 0.68 (0.22,2 .10) 0.574 Overall 0.220 18 11.0% 25 17.7% 145 8 . % 116 82.3% 90 M vs. L H vs. L 0.64 (0.33,1 .23) 0.192 1.11 (0.61,2 . 4 0.758 0) 127 130 Overall 0.546 5 3.9% 122 96.1% 2 1.6% 4 3.1% 126 96.9% 121 98.4% M vs. L H vs. L 0.78 (0.20,2 • 95) 0.747 0.40 ( . 8 2 .12) 0.447 00, 57 141 123 0.519 �TABLE J-6. (continued) Unadjusted Exposure Index Analyses for Psychological Variables by Occupation Exposure Index Medium High p-Value Statistic n Number/Z Abnormal Normal 127 n Number/Z Abnormal Normal 54 65 57 Overall 0.696 2 3.7Z 52 96. 3Z 1 1.5Z 64 98.5% 1 1.8Z 56 98. 2% M vs. L H vs. L 0.41 ( . 4 4 .61) 0.590 00, 0.46 ( . 4 5 .27) 0.611 00, n Number/Z Abnormal Normal 154 Overall 0.635 Enlisted Groundcrew 6 4.3Z 8 5.2Z 5 3.1Z 146 94. 8Z 158 96. 9Z 135 95. 7Z M vs. L H vs. L 0.58 (0.19,1 .81) 0.403 4) 0.81 (0.27,2 . 0 0.788 n Number/Z >0 0 127 Overall 0.052 Officer 32 25. 2Z 95 74. 8Z 34 26. 2Z 18 14. 6Z 96 73. 8Z 105 85.4% H vs. L H vs. L 1.05 ( . 0 1 .84) 0.887 06, 0.51 (0.27,0 .97) 0.041 n Number/Z Abnormal Normal 55 65 57 Overall 0.310 Enlisted Flyer 10 18. 2Z 45 81. 8Z 8 12. 3% 57 87. 7Z 13 22. 8Z 44 77.2% M vs. L H vs. L 0.63 (0.23,1 .73) 0.445 1.33 (0.53,3 .35) 0.642 n Number/Z >0 0 154 Overall 0.802 Enlisted Groundcrev M vs. L H vs. L 1.11 (0.66,1 .87) 0.790 1.20 (0.70,2 .05) 0.584 Enlisted Social Introversion Flyer c_ i _ Validity Low Contrast Est. Relative Risk (95% C.I.) Occupation Officer Variable Overall 0.385 0 O.OZ 1 0.8% M vs. L 2 1.5Z 126 99.2% 128 98. 5Z 123 100. OZ H vs. L 1.97 (0.18,21.99) 0.999 4) 0.34 (0.01,8 . 6 * 0.999 130 163 130 163 123 141 123 142 34 22. 1Z 39 23. 9% 36 25. 4Z 120 77. 9Z 124 76. 1Z 106 74. 6Z �TABLE J-6. (continued) Unadjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Exposure Index Medium High Contrast Cn Statistic Officer Enlisted Flyer n 53 Mean** 13.69 95% C.I.** (10.63, 17.56) 65 13.13 (11.16, 15.42) 56 13.96 (11.51, 16.88) Overall M vs. L H vs. L n 153 Mean** 14.46 95% C.I.** (12.64, 16.52) 159 13.03 (11.57, 14.65) 137 14.92 (13.05, 17.05) Overall M vs. L M vs. L n Number/% >0 0 125 Officer 41 32.8% 84 67.2% 37 28.9% 91 71.1% n Number/% Abnormal Normal 53 22 41.5% 31 58.5% n Number/% >0 0 154 Est. Relative Risk (95% C.I.) 128 123 n 126 Overall 8.55 Mean** 10.04 M vs. L 7.99 95* C.I.** (6.84,9.31) (7.52,9.69) (8.99,11.20) H vs. L Enlisted Groundcrev Total Cornell Medical Index Occupation M-R Subscore Enlisted Flyer Enlisted Groundcrev Low 91 59.1% 63 40.9% 121 — — " p-Value 0.049 0.506 0.018 0.906 0.784 0.906 0.304 0.252 0.745 Overall 0.247 47 38.8% 74 61.2% M vs. L H vs. L 0.83 (0.49,1.42) 0.586 1.30 (0.77,2.19) 0.353 65 56 Overall 0.400 35 53.8% 30 46.2% 26 46.4% 30 53.6% H vs. L H vs. L 1.64 (0.79,3.42) 0.200 1.22 (0.57,2.61) 0.700 Overall 0.592 M vs. L H vs. L 0.86 (0.55,1.35) 0.567 0.79 (0.50,1.25) 0.346 128 157 87 55.4% 70 44.6% 139 74 53.2% 65 46.8% �TABLE J-6. (continued) Unadjusted Exposure Index Analyses for Psychological Variables by Occupation Est. Relative Exposure Index Medium c_ I Statistic n Number/% >0 0 113 57 50.4% 56 49.6% 56 50.0% 56 50.0% n Number/2 >0 0 50 Enlisted Flyer 33 66.0% 17 34.0% n Number/% >0 0 136 Enlisted Groundcrev A-H Area Subscore Occupation Officer Variable n Number/% Abnormal Normal 126 n Number/% Abnormal Normal Officer HRB Impairment Index Enlisted Flyer n Enlisted Groundcrev Number/% Abnormal ' Normal Low 112 High 114 Contrast Risk (95% C.I.) p-Value Overall 090 .8 56 49.1% 58 50.9% M vs. L H vs. L 0.98 ( . 8 1 .66) 0 9 9 05, .9 0.95 (0.56,1 . 0 0.895 6) 60 50 Overall 0.946 39 65.0% 21 35.0% 34 6 . % 80 16 32.0% H vs. L H vs. L 0.96 (0.43,2 .11) 04, 1.10 ( . 8 2 .52) 0.999 0.999 Overall 0.557 M vs. L H vs. L 0.82 ( . 0 1 •34) 0.455 05, 1.07 (0.63,1 . 0 0 8 4 8) .9 Overall 0.401 29 23.0% 97 77.0% 21 17.4% M vs. L 31 23.9% 99 76.1% 100 82.6% H vs. L 1.05 ( . 9 1 .87) 0.884 05, 0.70 (0.38,1 .32) 0.342 52 65 57 Overall 0.253 20 38.5% 32 61.5% 35 53.9% 30 46.1% 27 47.4% H vs. L 30 52.6% H vs. L 1.87 ( . 9 3 .92) 0.136 08, 1.44 (0.67,3 . 9 0.439 0) Overall 0.272 93 68.4% 43 31.6% 153 64 41.8% 89 58.2% 150 96 6 . % 40 54 36.0% 130 163 72 44.2% 91 55.8% 129 90 69.8% 39 30.2% 121 139 49 35.3% 90 64.7% M vs. L H vs. L 1.10 (0.71,1 .72) 0.76 (0.47,1 .22) ^Estimated relative risk and confidence interval calculated after adding 0.5 to each cell. **Converted from log (X+l) scale, where X was the number of questions answered "yes." —No relative risk given for Total Cornell Medical Index, which was analyzed as a continuous variable. 0.733 0.280 �TffilE J-7. Variable Interaction (Occupation) Anxiety Exposure Index-byAge (Officer) High Adj. Relative Contrast Risk (95% C I ) .. p-Value 138 n Number/% Abnormal Normal 16 14 5 31.3% 11 68.7% 2 12 n Number/% Abnormal Normal 43 22 11 2 4.7% 41 95.3% 0 00 .% 22 1 0 0 0.% 00,.2* 0 0.0% M vs. L 0.37 ( . 2 8 0 ) * 0.545* 0.% 00,61)* . 9 * 11 1 0 0 H vs. L 0.72 ( . 3 1 . 2 * 0 9 9 n Number/% Abnormal Normal 79 93 Born 192314() 91a 3 3.8% 76 96.2% 1 92 n Number/% Abnormal Normal 5 15 Born <=1922 . 1 20.0% 4 80.0% 1 14 Black Born >=1942 Denial Exposure Index Medium n Number/% Abnormal Normal Nonblack Exposure Index-byRace (Enlisted Groundcrew) Low Stratification Statistic 14 124 149 10.1% 89.9% 16 133 128 1.% 07 89.3% 12 9.4% M vs. L 1.07 ( . 0 2 2 ) 05,.7* 116 90.6% H vs. L 0.92 ( . 1 2 0 ) 04,.6* 13 14.3% 85.7% Overall Overall 1 0.9% M vs. L 0.27 ( . 3 2 6 ) 00,.1 108 99.1% H vs. L 0.25 ( . 3 2 4 ) 00,.9 3 6.7% 93.3% Overall 092 .3* 099 .9* 099 .9* 004 .7* 0 0.0% M vs. L 0.37 ( . 6 2 2 ) 00,.9* 0 3 9 .9* 0.% 13 1 0 0 H vs. L 0 0 ( . 0 , . 6 * 0 0 8 . 8 00415)* . 4 * 109 1.1% 9.% 89 Overall Overall 0.455* 0.337 0.255 0.238 0.558* .4* 0 0.0% M vs. L 0 2 ( . 1 5 6 ) . 9 00,.6* 0 4 7 3 1 0 0 H vs. L 0 4 ( . 1 1 . 8 * 0 9 9 0.% . 3 00,40)* . 9 * �— -- X —- — • • •f Interaction Summaries of Adjusted Exposure Index Analyses for Psychological Variables "Variable Interaction (Occupation) Stratification Statistic Nonblack Depression Exposure Index-byRace (Enlisted Groundcrew) Black High. School Exposure Hypochondria Index-byEducation (Officer) College Nonblack Exposure Index-byIfypochondria Race (Enlisted Groundcrew) Low Exposure Index Medium n Number/% Abnormal Normal 138 n Number/% Abnormal Normal 16 14 4 25.0% 12 75.0% 1 13 n Number/% Abnormal Normal 11 17 n Number/% Abnormal Normal 116 IB 6 5.2% 110 9 . % 48 3 110 n Number/% Abnormal Normal 134 146 n Number/% Abnormal Normal 14 21 117 149 15.2% 8.% 48 4 3.% 64 7 63.6% 17 117 12.7% 87.3% 20 129 3 14 15 131 High 128 13.4% 86.6% 086 .5* 077 .3* 099 .9* 002 .9* Overall 001 .1* .% . 8 00,.6* 0 3 1 0 0 0 M vs. L 0 3 ( . 7 2 1 ) .8* 24 1 0 0 H vs. L 0.03 ( . 0 , . 1 * 0 0 6 0.% 00207)* . 0 * 99 2.7% 97.3% Overall p-Value 0 0 0 M vs. L 0.23 ( . 2 2 3 ) .% 0 0 , . 7 * 0.336* 0.% 00,.1* 13 1 0 0 H vs. L 0.10 ( . 1 2 1 ) * 0 1 7 .0* 24 17.7% 82.3% Overall 04,.7* 20 15.6% M vs. L 0.86 ( . 5 1 6 ) 44 108 8 . % H vs. L 1.03 ( . 3 2 0 ) 05,.1* 13 7.1% 92.9% Adj. Relative Contrast Risk (95%C.I.) Overall 9 9.1% M vs. L 0 5 ( . 2 2 0 ) . 0 01,.5* 90 90.9% H vs. L 1.83 ( . 3 5 3 ) 06,.4* 010 .2* 049 .9* 023 .9* 119 10.3% 8.% 97 .2 03,.4 25 21.0% M vs. L 0 8 ( . 9 1 7 ) 94 79.0% H vs. L 1.67 ( . 3 3 3 ) 08,.3 066 .0 018 .4 o Black 6 42.9% 8 57.1% 14 3 11 12 21.4% 78.6% 1 8.3% M vs. L 0.31 ( . 6 1 7 ) 00,.0 11 91.7% H vs. L 0 1 ( . 1 1 0 ) . 0 00,.2 0.179 002 .5 �TABLE J-7. (continued) U — —-^ - TntorarHm &BBari«>>3 nf A/HnstpH Rnwam> Tnrlear AnalLvaes Variable Interaction (Occupation) Stratification Statistic High School Hysteria Exposure Index-byEducation (Officer) College<b) Nonblack <r Hysteria Exposure Ihdex-byRace (Enlisted Groundcrev) Black 0 Mania/ Hypomania Exposure Index-byDrink-years (Officer) 0-50 >50 Low Exposure Index High Medium n Number/% Abnormal Normal 11 n Number/% Abnormal Normal 116 n Number/% Abnormal Normal 138 n Number/% Abnormal Normal 16 n Number/% Abnormal Normal 5 n Number/% Abnormal Normal 105 n Number/% Abnormal Normal 15 18 0 0.0% 15 100.0% 0 18 2 9 5 11 1 23 115 5 11 17 18.2% 81.8% 11.8% 88.2% 4.3% 95.7% 7 106 149 16.7% 83.3% 13 136 14 31.2% 68.8% 3 11 21.4% 78.6% 25 19.5% H vs. L 103 80.5% H vs. L 0.029* 0.48 (0.23,0.99)* 1.21 (0.65,2.27)* 0.050* 0.633* 0.094* 0 0.0% M vs. L 13 100.0% H vs. L 0.60 (0.11,3.15)* 0.689* 0.08 (0.004,1.56)** 0.048* Overall — — — — 0.207* 0.58 (0.18,1.84)* 0.27 (0.06,1.30)* 0.402* 0.108* Overall 2 2.2% M vs. L 90 97.8% H vs. L 22 0.0% 100.0% 0.795 0.024 Overall 92 4.6% 95.4% 1.17 (0.35,3.92) 3.49 (1.17,10.32) Overall 0 0.0% M vs. L 6 100.0% H vs. L 109 0.025 Overall 14 14.1% M vs. L 85 85.9% H vs. L 6 — — 5 104 0.60 (0.07,5.03)* 0.999* 0.08 (0.003,1.77)** 0.092* 13 0 7.6% 92.4% 0 0.0% M vs. L 24 100.0% H vs. L 128 8.7% 91.3% p-Value 0.129* 99 6.2% 93.8% Adj. Relative Contrast Risk (95ZC.I.) Overall 24 113 0 0.0% 5 100.0% 8 97 2 15 — for Psvchnlnoricail feriahles Overall 0.016* 5 22.7% M vs. L 17 77.3% H vs. L — — 9.74 (0.50,190.8)** 0.067* �TABLE J-7. Variable Interaction (Occupation) Stratification Statistic High School Exposure Masculinity/ Index-byFemininity Education (Officer) College Born >=1942 Paranoia Exposure Index-byAge (Officer) Born <L942 Born >1942 Paranoia Exposure Ihdex-byAge (Enlisted Flyer) Born <1942 (continued) Low Exposure Index Medium High Adj. Relative .. Contrast Risk (95% C I ) p-Value n Number/% Abnormal Normal 11 16 1 9.1% 10 9 . % 09 4 12 n Number/% Abnormal Normal 114 n Number/% Abnormal Normal 43 22 0 00 .% 43 1 0 0 0.% 1 21 n Number/% Abnormal Normal n Number/% Abnormal Normal 11 18 0 00 .% 11 1 0 0 0.% 1 17 n Number/% Abnormal Normal 43 47 48 3 7.0% 40 93.0% 0 00 .% 47 1 0 0 0.% 3 6.3% M vs. L 0.12 ( . 1 2 4 ) * 00,.3* 45 93.7% H vs. L 0 8 ( . 7 4 6 ) .9 01,.6* 6 5.3% 108 94.7% 24 25.0% 75.0% 0.150 047 .6 % 110 11 99 1 4.2% M vs. L 5 % ( . 3 6 . 9 . 05,75) 00,.5 23 95.8% H vs. L 0.33 ( . 2 6 5 ) 1.% 00 9.% 00 13 13.5% M vs. L 3.03 ( . 1 9 0 ) 10,.8 83 86.5% H vs. L 4 4 ( . 9 1 . 4 .8 14,34) 11 Overall 008 .4 008 .0 028 .8* 4.5% 95.5% 84 108 02,5.0* 0 0.0% M vs. L 6.07 ( . 4 1 5 3 ) * 0.338* 11 1 0 0 H vs. L 0.% — — 112 Overall 016 .0* 1 1.2% 83 9 . % 88 0 00 .% 108 1 0 0 0.% 4 3.6% M vs. L 0 2 ( . 1 6 3 ) * 0 4 7 . 6 00,.8* .3* 108 96.4% H vs. L 3.07 ( . 4 2 . 2 * 0.394* 03,80) 9 5.6% 9.% 44 Overall 055 .6* 0 0.0% M vs. L 1.97 ( . 7 5 . 1 * 0 9 9 00,27)* . 9 * — 0.% 9 1 0 0 H vs. L — Overall 0.195* 0.105* 099 .9* �———_ V V- X" * — • • « - f Interaction Sunmarles of Adjusted Exposure Index Analyses for Psychological Variables Variable Interaction (Occupation) Stratification Statistic Nonblack Paranoia Exposure Index-byRace (Enlisted Groundcrew) Black High School Psychopathic/ Deviate Exposure Jndex-byEducation (Enlisted Groundcrew) College Low Exposure Index Medium High Adj. Relative 9%.. Contrast Risk ( 5 C I ) Overall 038 .4* 7 5.5% M vs. L 121 94.5% H vs. L 1.56 ( . 7 6 6 ) 0 3 , . 6 * 0.724 .0* 2.60 ( . 6 1 . 0 * 0 2 4 06,03) Overall 042 .2* n Number/% Abnormal Normal 138 149 3 2.2% 135 97.8% 5 144 n Number/% Abnormal Normal 16 14 2 12.3* 14 87.5% 1 13 n . 110 Number/% Abnormal 16 14.5% Normal 94 85.5% 131 . 15 116 11.5% 8.% 85 n Number/% Abnormal Normal 38 3 26 3.4% 9.% 66 13 7.12 92.9% 29 8 21.1% 30 78.9% 128 .% 0 0 0 M vs. L 0 5 ( . 4 6 6 ) . 4 00,.7* 0.% 13 1 0 0 H vs. L 0.22 ( . 1 4 8 ) * 00,.9* 108 10.3% 8.% 97 099 .9* 048 .3* Overall 03,.4 24 22.2% M vs. L 0.75 ( . 5 1 6 ) 84 77.8% H vs. L 1.50 ( . 2 3 0 ) 07,.9 23 p-Value 045 .7 0.278 Overall 1 4.3% M vs. L 0.52 ( . 2 2 2 ) 01,.3 22 95.7% H vs. L 0.12 ( . 1 1 1 ) 00,.1 037 .7 002 .6 �TABLE J-7. Variable Interaction (Occupation) Stratification Statistic Low (continued) Exposure Index Medium High Adj. Relative Contrast Risk ( 5 C I ) p-Value 9% .. () c () c _ 0.228 M vs. L H vs. L () c () c 094 .6 008 .5 M vs. L H vs. L () c () c 007 .0 000 .8 0 Total Off Exposure Index-byDrink-years (Officer Enlisted) n Adj. Mean 95% C I .. 5 0 40 .1 (.7 80) — 17, .6 — M vs. L 6 M vs. L 6.87 (.7 35, (25) 1.3 0-50 n Adj. Mean 95%C.I. 104 6.95 (.0 51, 9.37) 107 6.98 9.39) (5.13; 92 84 .0 (.9 61, 11.29) 15 13.74 (.1 89, 2.2 09) 17 7.17 (.8 46, 1.5 07) 22 9.27 (.4 63, 13.36) >50 n Adj. Mean 95% C.I. �_• . - X-""-— ""t Interaction Sunmaries of Adjusted Exposure Index Analyses for Psychological Variables Variable Interaction (Occupation) Stratification Statistic Low Exposure Index Medium High Adj. Relative Contrast Risk ( 5 C I ) 9% . . p-Value High School 0 n Adj. Mean 95% C.I. 2 4 4.00 17.79 (.11.8 (.83.9 10,14) 86,54) 4 11.32 (.1 53, 2.9 30) M vs. L M vs. L () c () c 0.018 0.106 High School 0-50 'n Adj. Mean 95%C.I. 25 36 27 16.06 10.87 12.68 (12,28) (.71.3 (.2 1.32.1 92,72) 78, 1.7 49) M vs. L M vs. L () c () c 0.191 005 .4 Exposure Ihdex-byEducation High School >50 n Adj. Mean 95% C I .. 14 11 M vs. L 10 16.32 18.27 M vs. L 19.64 (05,48) (23,08) (16, 1.92.7 1.83.5 1.8 28.27) () c () c 059 .0 060 .8 Exposure Index-byDrink-Years (Enlisted Flyer) College 0 n Adj. Mean 95% C.I. 1 0.84 (,.6 056) College 0-50 n Adj. Mean 95% C.I. College >50 n Adj. Mean 95% C.I. Total CMT 1 0 10.17 (.33.6 — 21,88) — 6 8 10 11.92 6.43 11.84 ( . 5 1 . 7 ( . 3 1 . 1 (.6 3 2 , 1 9 ) 7 3 , 8 8 ) 70 19.72) M vs. L M vs. L () c () c 009 .4 — M vs. L M vs. L () c () c 0.100 0.112 1 20.55 (.4 50, 7.4 58) M vs. L M vs. L () c () c — — 0 — — 0 — — �——— — * - x f Interaction Smmarles of Adjusted Exposure Index Analyses for Psychological Variables Variable Interaction (Occupation) Stratification Statistic 0 M-R Subscore Exposure Index-byDrink-years (Enlisted Flyer 05() -0a >50(a) n Number/% X) 0 Low Exposure Index Medium 3 4 1 4 5 .% 0 00 3 100 0.% n Number/% X) 0 31 n Number/% X) 0 14 12 19 8.% 00 20.0% 7 50.0% 7 50.0% 25 21 p-Value 000 .6* 3 7 . % M vs. L 50 1 25.0% H vs. L 21.00 ( . 4 , 9 . ) * 0 1 3 06 6 0 0 * . 4 * 16.33 ( . 8 , 5 . ) * 0.143* 04 5 5 6 * Overall 004 .5 50.0% 50.0% 10 28.6% M vs. L 1.98 ( . 7 5 0 ) 07, . 9 25 71.4% H vs. L 0.65 ( . 3 1.83) 02, 0.158 040 .1 12 54.3% 45.7% 10 5 5 Adj. Relative Contrast Risk ( 5 C. . 9% 1) Overall 35 46 38.7% 61.3% High 040 .4 Overall 9 75.0% M vs. L 1.00 ( . 0 5.07) 02, 3 25.0% H vs. L 2.67 ( . 9 1 . 6 04, 4 4 ) 1.000 0.255 �TABLE J-7. (continued) Variable Interaction (Occupation) Stratification Statistic Low Exposure Index Medium High Adj. Relative Contrast Risk ( 5 C.I.) 9% p-Value 2 4 4 Normal 0 00 . 2 100 0. 4 100 0. 0 00 . 4 1 0 0 M vs. L 4 . 0 ( . 7 3 4 ) * 0 0 7 0. 5 0 06,03* .6* 0 0 0 H vs. L 45.00 ( . 7 3 4 ) * 0 0 7 . 06,03* .6* n Number/% 22 33 22 Overall 005 .3* Abnormal 18 8 . 18 4 18.2 18 15 10 45.5 12 54.5 M vs. L 0 2 ( . 7 0 % * . 7 00,.) Y vs. L 0 1 ( . 5 0 7 ) . 9 00,.3* 007 .4* 007 .2* 14 9 11 Overall 0.583* 10 71.4 4 28.6 8 1 8. 89 11.1 9 81.8 2 18.2 M vs. L 3.20 ( . 0 3 . 9 * 0.611* 03,45) 02,23) .6* H vs. L 1.80 ( . 6 1 . 0 * 0 6 1 n 0 0 .— 0 — 6 1 0 00 . 1 100 0. 1 0 00 . 1 100 0. M vs. L H vs. L 10 7 Overall 1 16.7 5 83.3 6 4 5 71.4 2 28.6 M vs. L 7.50 ( . 2 9 . 5 * 0.145* 06,06) H vs. L 12.50 ( . 4 1 6 3 ) 0.103* 08,8.0* 0 0 1 Overall 0 — 0 — 0 0 1100 0. 0 00 . M vs. L H vs. L High School n 0 Drink-years Nanber/% Ahnprmal High School 0-50 Drinkyears Normal Exposure Index-byEducation ft A-fl Area Subscore Index-byDrink-years (Enlisted Flyer) High School >50 Drinkyears n Number/% Abnormal Normal Exposure College 0 Drink-years Abnormal Normal College 0-50 Drinkyears n Njrober/% Abnormal Normal College >50 Drinkyears n Number/% Abnormal Normal 54.5 45.5 6. 00 4. 00 — — Overall 007 .0* Overall — — — — — 0.115* — — — — — �TftBLE J-7. (continued) Variable Interaction (Occupation) Nonblack HRB Impairment Index Exposure Ihdex-byRace (Enlisted Groundcrew) Low Stratification Statistic Black Exposure Index Medium n Number/% Abnormal Normal 132 n Number/% Abnormal Normal 13 14 11 84.6% 2 15.4% 10 4 48 84 59 85 Adj. Relative 9% . . Contrast Risk ( 5 C I ) p-Value 115 144 36.4% 63.6% High 4.% 10 59.0% 10,.5 40 3 . % M vs. L 1.88 ( . 9 3 2 ) 48 75 6 . % H vs. L 0 7 ( . 3 1 4 ) 52 .7 04,-0 004 .2 040 .0 12 71.4% 28.6% 3 2 . % M vs. L 0 3 ( . 5 2 7 ) 50 .9 00,.2 9 7 . % H vs. L 0 0 ( . 1 0 4 ) 50 .6 00,.4 030 .4 006 .0 ^Unadjusted estimate of relative risk and confidence interval, or p-value, based on stratified tables. **unadjusted estimate of relative risk and confidence interval calculated after adding 0.5 to each cell. (a)Results have been adjusted for education. (b)Results have been adjusted for age. —Zero counts in cells do not allow for calculation of percent, relative risk, confidence interval, or p-value. (c)No relative risk given for total Off, which was analyzed continuously. Note: Snail sample sizes may affect validity of overall p-value. Note: Results without (*) or ( * are adjusted for all other main effects in model (age, race, education, drink-years, and combat index *) [for MMPI variables]), unless otherwise noted. Note: Participants judged to have PTSD are eliminated from analysis of Total .Off, Section M-R Subscore, Section A-H Area Subscore, and HRB Impairment Index. �TABLE J-8. Unadjusted Analyses for Reported Psychological Illnesses by Group: Baseline and First Followup Studies Combined* (Original Comparisons Only) Group Abnormalities Type of Illness Original Ranch Hand Comparison Number Percent Number Percent Total p-Value** 14 1.4 8 0.8 22 0.289 Alcohol Dependence 9 0.9 4 0.4 13 0.268 Anxiety 7 0.7 8 0.8 15 0.798 72 7.1 46 4.8 118 0.037 Psychoses Other Neuroses ^Analyses based on 1,016 Ranch Hands and 955 Original Comparisons; some participants had more than one illness. **Fisher's exact test. J-27 �TABLE J-9. Summary of Kolmogorov-Smirnov Tests on MMPI for Officers (Original Comparisons Only) Parameter* Mean Score Group Standard Deviation KolmogorovSmirnov p-Value Anxiety Ranch Hand Original Comparison 52.82 53.22 7.38 7.06 0.328 Consistency Ranch Hand Original Comparison 49.23 49.02 6.92 5.50 0.995 Defensiveness Ranch Hand Original Comparison 49.64 49.69 6.64 7.14 0.822 Denial Ranch Hand Original Comparison 59.47 59.46 7.31 7.40 0.901 Depression Ranch Hand Original Comparison 55.11 54.86 10.17 9.38 0.975 Hypochondria Ranch Hand Original Comparison 55.60 54.96 8.85 8.09 0.990 Hysteria Ranch Hand Original Comparison 60.45 59.79 7.00 6.81 0.711 Mania/Hypomania Ranch Hand Original Comparison 54.36 53.68 8.94 9.10 0.436 Masculinity/ Femininity Ranch Hand Original Comparison 58.83 58,14 8.64 8.94 0.577 Paranoia Ranch Hand Original Comparison 53.71 53.57 7.34 6.86 0.995 Psychopathic/ Deviate Ranch Hand Original Comparison 56.82 56.61 8.87 9.65 0.644 Schizophrenia Ranch Hand Original Comparison 54.80 54.66 8.48 7.89 0.661 Social Introversion Ranch Hand Original Comparison 46.56 47.13 7.43 7.60 0.627 Validity Ranch Hand Original Comparison 0.93 0.80 3.11 2.86 0.594 *n=380 Ranch Hands; n=350 Original Comparisons. J-28 �TABLE J-10. Summary of Kolmogorov-Smirnov Tests on MMPI for Enlisted Flyers (Original Comparisons Only) Group Mean Score Standard Deviation KolmogorovSmirnov p-Value Anxiety Ranch Hand Original Comparison 54.78 54.16 9.73 8.78 0.993 Consistency Ranch Hand Original Comparison 52.24 51.41 10.30 8.15 0.923 Defensiveness Ranch Hand Original Comparison 53.05 52.09 7.96 8.34 0.353 Denial Ranch Hand Original Comparison 55.95 56.81 8.73 8.44 0.760 Depression Ranch Hand Original Comparison 58.21 57.27 11.79 10.35 0.760 Hypochondria Ranch Hand Original Comparison 58-48 58.44 12.16 11.71 0.577 Hysteria Ranch Hand Original Comparison 60.00 60.70 8.94 8.77 0.465 Mania/Hypomania Ranch Hand Original Comparison 55.05 55.50 10.00 9.42 0.155 Masculinity/ Femininity Ranch Hand Original Comparison 55.79 56.22 7.85 8.39 0.877 Paranoia Ranch Hand Original Comparison 52.77 51.72 8.43 7.68 0.877 Psychopathic/ Deviate Ranch Hand Original Comparison 57.48 57.85 11.40 9.72 0.163 Schizophrenia Ranch Hand Original Comparison 56.56 56.27 13.03 10.59 0.496 Social Introversion Ranch Hand Original Comparison 50.35 49.00 8.89 7.94 0.194 Validity Ranch Hand Original Comparison 3.93 7.18 42.48 60.31 0.864 Parameter* *n=176 Ranch Hands; n=172 Original Comparisons (except for validity, where n=177 Ranch Hands and n^!74 Original Comparisons). J-29 �TABLE J-ll. Summary of Kolmogorov-Smirnov Tests on MMPI for Enlisted Groundcrew (Original Comparisons Only) Group Mean Score Standard Deviation KolmogorovSmirnov p-Value Anxiety Ranch Hand Original Comparison 55.86 55.72 10.96 9.34 0.877 Consistency Ranch Hand Original Comparison 53.05 51.78 10.64 8.19 0.728 Defensiveness Ranch Hand Original Comparison 52.42 52.69 8.27 8.51 0.923 Denial Ranch Hand Original Comparison 55.53 56.80 8.57 8.72 0.112 Depression Ranch Hand Original Comparison 58.79 57.85 12.98 10.63 0.661 Hypochondria Ranch Hand Original Comparison 58.44 57.50 13.22 11.30 0.577 Hysteria Ranch Hand Original Comparison 60.36 59.75 9.80 8.84 0.644 Mania/Hypomania Ranch Hand Original Comparison 55.26 55.69 9.78 9.72 0.822 Masculinity/ Femininity Ranch Hand Original Comparison 56.18 57.24 8.24 8.49 0.304 Paranoia Ranch Hand Original Comparison 53.12 53.73 9.05 8.48 0.340 Psychopathic/ Deviate Ranch Hand Original Comparison 59.01 59.34 11.29 10.70 0.644 Schizophrenia Ranch Hand Original Comparison 58.20 57.58 14.22 10.64 0.328 Social Introversion Ranch Hand Original Comparison 51.86 50.99 9.82 8.63 0.353 Validity Ranch Hand Original Comparison 1.97 0.97 26.48 5.41 0.999 Parameter* *n=458 Ranch Hands; n=430 Original Comparisons (except for validity, where n=459 Ranch Hands and n=431 Original Comparisons). J-30 �TABLE J-12. Summary of Rolmogorov-Smirnov Tests on CHI (Original Comparisons Only) KolmogorovSmirnov Standard Deviation p-Value Sample Size Mean Score Ranch Hand Original Comparison 402 366 15.88 12.95 15.21 11.57 <0.001 Born 1923-1941 Ranch Hand Original Comparison 562 528 15.49 14.55 12.99 11.88 0.776 Ranch Hand Original Comparison 36 45 18.56 16.47 15.94 9.28 0.980 Ranch Hand Original Comparison 941 882 15.63 13.96 13.83 11.47 0.038 Ranch Hand Original Comparison 59 57 17.81 14.82 16.91 14.62 0.561 Ranch Hand Original Comparison 556 524 18.88 15.62 15.80 13.12 <0.001 Ranch Hand Original Comparison 444 415 11.85 11.99 10.18 9.16 0.577 Current Alcohol Use Drinker Ranch Hand Original Comparison 851 812 15.16 13.77 13.12 11.32 0.170 Ranch Hand Original Comparison 148 127 18.84 ' 15.61 17.71 13.69 0.162 Ranch Hand Original Comparison 377 345 11.23 10.99 9.01 7.53 0.970 Enlisted Flyer Ranch Hand Original Comparison 174 170 17.72 15.31 14,84 12.51 0.270 Enlisted Groundcrew 449 424 18.80 15.96 16.06 13.48 0.003 Stratification Age Born XL942 Born <1922 Race Nonblack Black Education High School College Nondr inker Occupation Officer Group Ranch Hand Original Comparison J-31 �TABLE J-13. Unadjusted Analyses for HMFI by Group (Original Comparisons Only) Group Variable Anxiety Consistency Defensiveness i Ranch Hand Statistic Number Percent Original Comparison Number Percent Est. Relative Risk ( 5 C.I.) 9% p-Value n Abnormal Normal 1,014 73 941 7.2 92.8 952 49 903 5.1 94.9 1.43 (0.98,2.08) 0.062 n Abnormal Normal 1,014 36 978 3.6 96.4 952 22 930 2.3 97.7 1.56 (0.91,2.67) 0.111 n Abnormal Normal 1,014 23 991 2.3 97.7 952 30 922 3.2 96.8 0.71 (0.41,1.24) 0.265 n Abnormal Normal 1,014 17 997 1.7 98.3 952 37 915 3.9 96.1 0.42 (0.24,0.75) 0.003 n Abnormal Normal 1,014 114 900 11.2 88.8 952 87 865 9.1 90.9 1.26 (0.94,1.69) 0.136 n Abnormal Normal 1,014 119 895 11.7 88.3 952 91 861 9.6 90.4 1.26 (0.94,1.68) 0.125 n Abnormal Normal 1,014 123 891 12.1 87.9 952 89 863 9.4 90.6 1.34 (1.00,1.79) 0.050 ro Denial Depression Hypochondria Hysteria �TABLE J-13. (continued) Unadjusted Analyses for MMFI by Group (Original Comparisons Only) Group Variable Original Ranch Hand Comparison Statistic Number Percent Number Percent Est. Relative Risk (95% C.I.) p-Value Mania/Hypomania n Abnormal Normal 1,014 63 951 6.2 93.8 952 56 896 5.9 94.1 1.06 (0.73,1.54) 0.777 Masculinity/ Femininity n Abnormal Normal 1,014 66 948 6.5 93.5 952 83 869 8.7 91.3 0.73 (0.52,1.02) 0.073 Paranoia n Abnormal Normal 1,014 31 983 3.1 96.9 952 17 935 1.8 98.2 1.73 (0.95,3.16) 0.079 Psychopathic/ Deviate n Abnormal Normal 1,014 120 894 11.8 88.2 952 104 848 10.9 89.1 1.09 (0.83,1.45) 0.570 Schizophrenia n Abnormal Normal 1,014 94 920 9.3 90.7 952 71 881 7.5 92.5 1.27 (0.92,1.75) 0.166 Social Introversion n Abnormal Normal 1,014 26 988 2.6 97.4 952 14 938 1.5 98.5 1.76 ( . 2 3 4 ) 09,.0 0.109 Validity n >0 0 1,016 224 792 22.0 78.0 955 198 757 20.7 79.3 1.08 (0.87,1.34) 0.510 CH I �TABLE J-14. Adjusted Analyses for MMPI by Group (Original Comparisons Only) Group Variable Ranch Hand Total Original Comparison Total 1,012 950 Consistency 976 926 Defensiveness 976 926 Denial 974 Depression Hypochondria Anxiety Hysteria Mania/Hypomania Adj. Relative Risk (95% C.I.) 1.39 (0.95,2.05) 0.092 Covariate Remarks* EDUC (p<0.001) AGE*CI (p=0.010) **** AGE (p=0.025) RACE*DRKYR (p=0.024) GRP*EDUC (p=0.034) 0.68 (0.38,1.19) 0.175 EDUC (p<0.001) AGE*DRKYR (p=0.043) 924 0.43 (0.23,0.80) 006 .0 AGE*DRKYR (p=0.026) EDUC*CI (p=0.040) 1,014 952 1.26 ( . 4 1.70) 09, 0.120 EDUC (p<0.001) 1,014 . **** p-Value 952 1.29 (0.96,1.72) 0.091 AGE (p=0.005) EDUC (p<0.001) 1,014 952 1.37 (1.02,1.83) 0.033 AGE (p=0.001) EDUC (p<0.001) 974 924 0.97 (0.66,1.43) 0.882 AGE (p=0.048) CI (p=0.010) DRKYR (p=0.016) �TABLE J-14. (continued) Adjusted Analyses for MMFI by Group (Original Comparisons Only) Group Variable Ranch Hand Total Original Comparison Total Adj. Relative Risk ( 5 C.I.) 9% p-Value Covariate Remarks* Masculinity/ Femininity 952 Paranoia (-1 I 1,014 1,012 950 Psychopathic/ Deviate 974 924 1.13 (0.84,1.52) 0.433 EDUC (p=0.010) AGE*CI (pXLOOl) RACE*DRKYR (p=0.001) Schizophrenia 976 926 1.27 (0.91,1.77) 0.153 EDUC (p<0.001) RACE*DRKYR (p=0.003) Social Introversion 1,014 952 1.72 (0.89,3.31) 0.099 Validity 1,014 953 0.71 (0.51,1.00) **** 0.047 **** Ul **** **** EDUC (p<0.001) RACE* AGE (p=0.006) AGE*CI (p=0.004) GRP*AGE (p=0.040) AGE (p=0.005) AGE*CI (p=0.034) GRP*RACE (p=0.016) GRP*CI (p=0.030) *Abbreviations : EDUC: education CI: combat index DRKYR: drink-years GRP: group ****Group-by-covariate interaction: presented. adjusted relative risk, confidence interval, and p-value are not �TABLE J-15. Unadjusted Analyses for the Cornell Medical Index (CHI) by Group (Original Comparisons Only) Group Variable Statistic Ranch Hand Total CMI n Mean* 95% C.I.* 1,000 11.74 (11.17,12.35) M-R Subscore Subscore C«i i n Number/% 0 (Low) 1-10 (Medium) (High) 998 538 408 52 Original Comparison 939 10.55 (10.02,11.11) 531 370 35 n Number/% 0 (Low) 1-3 (Medium) 4-8 (High) 1,148 914 360 449 105 39 .4% 49 .1% 11 .5% 375 394 82 0.004 44 .1% 46 .3% 9.6% 0.198 Medium vs. Low 1.09 (0.91,1.31) High vs. Low 1.47 (0.99,2.29) 0.371 Overall 56 .7% 39 .5% 3.8% u> A-H Area Subscore — p-Value Overall 936 53 .9% 40 .9% 5.2% Est. Relative Risk (95% C.I.) 0.106 Medium vs. Low 1.19 (0.97,1.45) High vs. Low 1.33 (0.97,1.84) 0.090 0.095 0.086 transformed from log (X+l) scale, where X was the number of questions answered "yes." —No relative risk given for Total CMI, which was analyzed as a continuous variable. �TABLE J-16. Adjusted Analyses for CMI Variables by Group (Original Comparisons Only) Variable Statistic Total CMI LO -vl n Adj. Mean 95% C.I. M-R Subscore A-H Area Subscore Ranch Hand Group Original Comparison 962 **** **** 998 935 n 914 850 ****Group-by-covariate not presented. p-Value Covariate Remarks* **** 913 **** **** n Adj. Relative Risk (95% C.I.) **** Overall Medium vs. Low: 1.20 ( . 9 1 4 ) 09,.7 High vs. Low: 1.32 (0.92,1.89) PTSD (p<0.001) RACE*DRKYR (p=0.027) AGE*EDUC (p=0.010) GRP*EDUC (p=0.002) **** AGE (p<0.001) PTSD (p<0.001) GRP*EDUC (p=0.006) 0.108 0.063 0.132 AGE (p<0.001) EDUC (p<0.001) PTSD (p<0.001) interaction — adjusted mean/relative risk, confidence interval, and p-value are —No relative risk given for total CMI, which was analyzed as a continuous variable. *Additional Abbreviations: PTSD: post-traumatic stress disorder �TABLE J-17. Summary Results for the Halstead-Reitan Battery Impairment Index Unadjusted and Adjusted Analyses (Original Comparisons Only) Group Analysis Unadjusted Analysis 00 n Abnormal Normal 1,006 348 658 Adjusted Analysis u> Ranch Hand Statistic Number Percent n 1,006 34.6 65.4 Original Comparison Number Percent 947 319 628 947 33.7 66.3 Est. Relative Risk (95% C.I). p-Value 1.04 (0.86,1.26) 0.703 1.10 (0.90,1.34) 0.359 Covariate Remarks AGE (p<0.001) RACE (p<0.001) EDUC (p<0.001) �TfiBlE J-18. for Psychological Variables (Original Comparisons Only) Group Variable Consistency Interaction Group-byEducation Statistic High School n Number/% Abnormal Normal 537 n Number/% Abnormal Normal 437 n Number/% Abnormal Normal 410 n Number/% Abnormal Normal 602 College V ^> Paranoia Group-byAge Ranch Hand Stratification Bom XL942 Born <1942 30 507 6 431 15 395 16 586 Original Comparison Adj. Relative Risk ( 5 C I ) 9% . . p-Value 2.5% 97.5% 2.31 ( . 8 4 5 ) 11,.0 0.014 2.6% 97.4% 0 6 (.217) . 1 02,.4 0.359 3.5% 96.5% 0 % (.421) . 04,.3 0.927 1.2% 98.8% 3 8 (.61.4 . 5 13,09) 0.011 516 5.6% 94.4% 13 503 585 1.4% 98.6% 15 570 545 3.7% 96.3% 19 526 740 2.7% 97.3% 9 731 �TABLE J-18. (continued) Summary of Group-by-Cbvariate Interactions (Original Comparisons Only) Group Variable Interaction Validity Group-byRace Ranch Hand Stratification Statistic Nonblack CX=Low n Nunfcer/% X) 0 178 n Number/% X) 0 367 n Number/% X) 0 409 n Number/% X) 0 20 n Number/% X) 0 18 n Number/% X) 0 22 Nonblack CkMediun Nonblack CI=Bigh Group-by Combat Index Black CT=low Black Q=Medium Black d=fligii 47 131 78 289 83 326 6 14 6 12 3 19 Original Conparison Adj. Relative Risk (95% C.I.) p-Value 20.7% 79.3% 1.49 (0.98,2.26) 0.059 14.6% 85.4% 1.60 (1.07,2.38) 0.021 23.1% 76.9% 0.83 (0.57,1.20) 0.319 37.5% 62.5% 0.55 (0.24,1.29) 0.170 41.7% 58.3% 0.59 (0.25,1.41) 0.239 50.0% 50.0% 0.31 (0.13,0.71) 0.006 353 26.4% 73.6% 73 280 295 21.3% 78.7% 43 252 247 20.3% 79.7% 57 190 24 30.0% 70.0% 9 15 12 33.3% 66.7% 5 7 22 13.6% 86.4% 11 11 �TABLE J-38. (ccntuued) for Psychological Variables (Original CoDfEorisons Only) Group Interaction Total Off Group-byEducation M-R Subscore Group-byEducation Stratification Statistic Ranch Hand Original Comparison High School n Adj. Mean* 95% C I * .. 529 35.50 (80,47) 2.94.9 511 29.65 (32,77) 2.73.0 — <0.001 College Variable n Adj. Mean* 95% C I * .. 433 22.70 (77,89) 1.62.5 402 23.49 (83,99) 1.72.7 — 048 .9 High School n 555 Number/% 0 (Low) 246 1-10 (Med) 267 >10 (High) 42 0.015 281 213 27 53.9% 40.9% 5.2% Overall Med vs. Low 1.42 ( . 1 1 8 ) 11,.2 High vs. Low 1.51 ( . 4 2 7 ) 08,.0 414 Overall Med vs. Low 60.4% 0.77 ( . 8 1 0 ) 05,.2 37.7% High vs. Low 1.9% 0.99 ( . 4 2 8 ) 03,.7 01 .% College 443 n Nunber/% 0 (Low) 292 1-10 (Med) 141 >10 (High) 10 521 44.3% 48.1% 7.6% 65.9% 31.8% 2.3% 250 156 8 ^Converted from log (X+l) scale, where X was the number of questions answered "yes". —No relative risk given for Total Off, which was analyzed as a continuous variable. Adj. Relative Risk ( 5 C I ) 9% .. p-Value 0.005 0.167 0.071 0.991 �APPENDIX K Gastrointestinal Assessment �APPENDIX K: Gastrointestinal Assessment Contents Table K-l Page Summary of Group-by-Covariate Interactions for Hepatic Function Variables and Porphyrin Determinations Analyzed Continuously K-l K-2 Summary of Group-by-Covariate Interactions for Hepatic Function Variables and Porphyrin Determinations Analyzed Categorically K-4 K-3 Unadjusted Categorical Exposure Index Analyses for Hepatic Function Variables by Occupation K-5 K-4 Unadjusted Continuous Exposure Index Analyses for Hepatic Function Variables and Tvo Porphyrin Determinations by Occupation K-8 K-5 Summary of Adjusted Exposure Index, Analyses Involving Interactions with a Continuous Covariate for Hepatic Function Variables K-13 K-6 Summary of Adjusted Exposure Index, Analyses Involving Interactions with a Categorical Covariate for Hepatic Function Variables K-15 K-7 Unadjusted Analyses for Baseline and Interval History of Liver Disease by Group (Verified by Medical Record Review) (Original Comparisons Only) K-16 K-8 Unadjusted Analyses of Enlarged Livers Diagnosed at Physical Examination by Group (Original Comparisons Only) K-17 K-9 Unadjusted Continuous and Categorical Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Group (Original Comparisons Only) K-18 K-10 Adjusted Continuous and Categorical Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Group (Original Comparisons Only) K-21 K-ll Summary of Group-by-Covariate Interactions for Hepatic Function Variables (Original Comparisons Only) K-27 K-12 Summary of Group-by-Covariate Interactions for Variables and Porphyrin Determinations Analyzed Categorically (Original Comparisons Only) K-29 K-i �APPENDIX K: Gastrointestinal Assessment (continued) Contents Table Page K-13 Unadjusted Analysis for Interval History of Skin Bruises, Skin Patches, or Skin Sensitivity by Group (Original Comparisons Only) K-30 K-14 Unadjusted Analyses for Porphyrin Abnormalities by Group and Skin Patch, Bruise, or Sensitivity Reported at Followup Questionnaire (Original Comparisons Only) K-31 K-15 Unadjusted Analyses for Uroporphyrin Abnormalities by Skin Patch, Bruise, or Sensitivity Reported at Followup Questionnaire and by Group (Original Comparisons Only) K-32 K-16 Longitudinal Results for SCOT, SGPT, and GGTP: A Contrast of Baseline and First Followup Examination Test Means (Original Comparisons Only) K-ii K-33 �TABLE K-l. v Summary of Group-by-Covariate Interactions for Hepatic Function Variables and Porphyrin Determinations Analyzed Continuously Group Ranch Hand Comparison n Mean Adj . Mean 95% C.I. 652 32.2 33.4 (32.4,34.4) 843 32.3 33.4 (32.5,34.4) 1-4 n Mean Adj . Mean 95% C.I. 275 35.3 36.5 (35.1,38.0) 335 33.4 34.4 (33.2,35.7) 0.010 >4 Mean Adj . Mean 95% C.I. 76 37.5 38.2 (35.7,40.9) 108 38.1 39.4 (37.2,41.8) 0.477 <1 Group-bycur rent Alcohol Use (Drinks/day) Statistic n SCOT Interaction Stratification <1 Variable n Mean Adj . Mean 95% C.I. 652 20.5 19.8 (18.6,21.2) 843 21.9 21.1 (19.8,22.5) 0.011 1-4 n Mean Adj . Mean 95% C.I. 275 23.4 22.2 (20.2,24.3) 335 22.6 21.5 (19.7,23.5) >4 n Mean Adj . Mean 95% C.I. 76 23.9 23.3 (19.7,27.5) 108 27.0 26.7 (22.8,31.3) • SGPT Group-bycur rent Alcohol Use (Drinks/day) p-Value 0.889 0.428 0.058 �TABLE R-l. (continued) Summary of Group-by-Covariate Interactions for Hepatic Function Variables and Porphyrin Determinations Analyzed Continuously Group Statistic Ranch Hand n Mean Ad j . Mean 95% C.I. 504 681 94.8 89.9 102.4 97.2 <0.001 (97.9,107.2) (93.1,101.5) Not Exposed to Industrial Chemicals n Mean Adj . Mean 95% C.I. 499 88.9 88.5 (84.9,92.1) 604 88.7 88.5 (85.0,92.1) Born XL942 Interaction Stratification Exposed to Industrial Chemicals Variable n Mean Adj . Mean 95% C.I. 408 122.2 124.8 (121.9,127.7) 547 121.2 0.425 123.6 (121.0,126.3) Born 1923-1941 n Mean Adj . Mean 95% C.I. 565 123.8 127.0 (124.5,129.7) 691 125.3 0.217 128.6 (126.1,131.1) Born <1922 n Mean Adj . Mean 95% C.I. 36 135.1 138.9 (130.8,147.5) 51 135.5 139.1 0.966 (132.2,146.5) Comparison p-Value Alkaline Group-byPhosphatase Industrial Chemicals 0.973 to LDH Group-by-Age �TABLE K.-1. (continued) Summary of Group-by-Covariate Interactions for Hepatic Function Variables and Porphyrin Determinations Analyzed Continuously Group Variable Ranch Hand Comparison n Mean Ad j . Mean 95% C.I. 406 107.1 95.6 (88.0,103.9) 546 110.2 99.7 0.321 (92.5,107.4) Born 1923-1941 n Mean Adj . Mean 95% C.I. 689 561 126.0 124.5 0.784 116.8 115.6 (108.6,125.5) (108.0,123.8) n Mean Adj. Mean 95% C.I. 51 36 139.0 105.6 130.7 . 98.3 0.039 (105.4,161.9) (81.9,118.0) BUN <L4 Uroporphyrins Statistic Born <1922 Group-by-Age Stratification Born >1942 Triglycerides Interaction n Mean Adj . Mean 95% C.I. 547 16.2 16.3 (15.4,17.4) 701 18.4 18.6 (17.6,19.7) <0.001 BUN >14 n Mean Adj . Mean 95% C.I. 453 17.8 17.9 (16.8,19.2) 582 17.4 17.6 (16.6,18.7) 0.686 p-Value Group-byBlood Urea Nitrogen (BUN) �TABLE K-2. Sunnary of Group-by-Cbvariate Interactions for Hepatic Function Variables Group Ranch Hand Variable Interaction Stratification Exposed to Industrial Chemicals Direct Bilirubin Statistic n High Normal Number 507 27 480 Percent 5.3 94.7 Comparison Number 683 20 663 Percent Adj. Relative Risk(95%C.I.) p-Value 2.9 97.1 1.89 ( . 5 3 4 ) 0 0 5 10,.2 .3 * Group-byIndustrial Chemicals 4 n High Normal 4 % 11 485 2.2 97.8 603 23 580 3.8 96.2 0.58 ( . 8 1 2 ) 0 1 3 02,.0 . 4 Officer n High Normal 376 34 342 90 . 91.0 484 26 458 5.4 94.6 1.77 ( . 4 3 0 ) 0 0 5 10,.1 . 3 Enlisted flyer n High Normal 177 15 162 8.5 91.5 209 14 195 6.7 93.3 1.27 ( . 9 2 7 ) 0 5 9 05,.1 .3 Enlisted Groundcrev n High Normal 456 19 437 4.2 95.8 5% 39 557 6.5 93.5 0.63 ( . 6 1 1 ) 0 1 4 03,.0 .0 Not Exposed to Industrial Chemicals Triglycerides Group-byOccupation �TABLE K-3. unadjusted Categorical Exposure Index Analyses for Hepatic Function Variables by Occupation Exposure Index Low Variable Medium High Contrast Est. Relative Risk ( 5 C I ) p-Value 9% . . 7.5 92.5 Overall M vs. L H vs. L 059 .1 1.57 ( . 5 3 7 ) 0.384 06,.7 1.05 ( . 0 2 7 ) 0 9 9 .9 04,.5 57 4 53 7.0 93.0 0.234 Overall M vs. L 5.49 ( . 4 4 . 9 0.123 06,70) H vs. L 4 0 ( . 4 3 . 7 0.364 .8 04,76) 8.1 91.9 142 17 125 12.0 8. 80 005 .6 Overall .4 M vs. L 1.86 ( . 2 4 7 ) 0 2 9 07,.9 H vs. L 2.86 ( . 5 7 1 ) 0.031 11,.1 130 20 110 15.4 8. 46 120 16 104 13.3 86.7 069 .6 Overall M vs. L 0 8 ( . 4 1 6 ) 0.736 .6 04,.7 H vs. L 0 7 ( . 6 1 4 ) 0 3 4 .8 . 3 03,.6 3.6 96.4 65 10 55 15.4 8. 46 57 6 51 10.5 89.5 0.105 Overall M vs. L 4.82 ( . 1 2 . 3 0.037 10,30) H vs. L 3.12 ( . 0 1 . 7 0.272 06,61) 11.0 8. 90 160 27 133 16.9 83.1 142 17 125 12.0 88.0 Overall M vs. L H vs. L 126 9 117 7.1 92.9 130 14 116 10.8 89.2 120 9 111 Enlisted Flyer n Abnormal Normal 55 1 54 1.8 98.2 65 6 59 9.2 90.8 n Abnormal Normal 154 7 147 4.5 95.5 160 13 147 Officer SGPT n Abnormal Normal Enlisted Groundcrev E Statistic Officer SOOT Occupation Number Percent Number Percent Number Percent n Abnormal Normal 126 22 104 17.5 82.5 Enlisted Flyer n Abnormal Normal 55 2 53 Enlisted Groundcrew n Abnormal Normal 154 17 137 025 .6 1.64 ( . 5 3 1 ) 0.147 08,.4 1.10 ( . 4 2 2 ) 0 8 6 05,.4 .5 �TABIE K-3. (continued) unadjusted Categorical Exposure Index Analyses for Hepatic Function Variables by Occupation Exposure Index Low Medium n Abnormal Normal 126 11 115 8.7 91.3 130 9 121 6.9 93.1 120 9 111 7.5 92.5 089 .5 Overall M vs. L 0 7 ( . 1 1 9 ) 0 6 6 .8 03,.5 .4 H vs. L 0.73 ( . 6 1 4 ) 0.384 03,.6 Enlisted Flyer n Abnormal Normal 55 5 50 9.1 9. 09 65 7 58 1. 08 89.2 57 6 51 1. 05 89.5 Overall 0.950 M vs. L 1.21 ( . 6 4 0 ) 0 9 9 03,.4 .9 H vs. L 1.18 ( . 4 4 1 ) 0 9 9 03,.0 .9 n Abnormal Normal 154 16 138 1. 04 89.6 160 13 147 8.1 91.9 142 14 128 99 . 90.1 Overall 0.773 M vs. L 0.76 ( . 5 1 6 ) 0.561 03,.4 H vs. L 0 9 ( . 4 2 0 ) 0 9 9 . 4 04,.1 .9 Officer Alkaline Phosphatase Statistic Enlisted Groundcrew GGEP Number Percent Number Percent Est. Relative Risk ( 5 C I ) p-Value 9%.. Occupation Officer Variable Number Percent High n Abnormal Normal 126 4 122 3.2 96.8 130 7 123 5.4 94.6 120 3 117 2.5 97.5 Overall 048 .4 M vs. L 1.74 ( . 0 6 0 ) 0.540 05,.8 H vs. L 0 7 ( . 7 3 5 ) 0 9 9 .8 01,.7 .9 Enlisted Flyer n Abnormal Normal 55 1 54 1.8 98.2 65 5 60 7.7 92.3 57 5 52 88 . 91.2 0.258 Overall M vs. L 4.50 ( . 1 3 . 4 0.217 05,97) H vs. L 5.19 ( . 9 4 . ) 0 2 6 05,5% .0 Enlisted Groundcrew n Abnormal Normal 154 8 146 5.2 9. 48 160 10 150 6.3 93.8 142 13 129 9.2 9. 08 0.378 Overall M vs. L 1.22 ( . 7 3 1 ) 0.810 04,.7 H vs. L 1.84 ( . 4 4 5 ) 0.257 07,.8 Contrast �T&BLEK-3. (continued) 4 Unadjusted Categorical Exposure Index Analyses for Hepatic Ruction \fariables by Occupation Exposure Index Medium Low n Abnormal Normal 126 3 123 2.4 97.6 130 2 128 1.5 98.5 120 3 117 2.5 97.5 Overall 086 .4 M vs. L 0 6 ( . 1 3 9 ) 0 6 0 . 4 01,.0 .8 H vs. L 1.05 ( . 1 5 3 ) 0 9 9 02,.1 .9 Enlisted Flyer n Abnormal Normal 55 0 55 00 . 100.0 65 1 64 1.5 9. 85 57 2 55 3.5 96.5 Overall 0.353 M vs. L 21.58 ( . 0 6 . 5 0 9 9 01,46) .9 H vs. L 5.00 ( . 4 1 6 5 ) 0 4 02,0.4 .% n Abnormal Normal 154 7 147 4.5 95.5 160 3 157 1.9 98.1 142 6 136 4.2 95.8 Overall M vs. L H vs. L Officer Direct Bilirubin Statistic Enlisted Groundcrew Total Bilirubin Number Percent Number Percent Contrast Est. Relative Risk ( 5 C I ) p-Value 9%.. Occupation Officer Variable Number Percent High n Abnormal Normal 126 2 124 1.6 98.4 130 5 125 3.8 96.2 120 6 114 5.0 95.0 Overall 0.327 04,30) .4 M vs. L 2.48 ( . 7 1 . 2 0 4 7 H vs. L 3.26 ( . 5 1 . 0 0.164 06,65) Enlisted Flyer n Abnormal Normal 55 2 53 3.6 9. 64 65 5 60 7.7 92.3 57 3 54 5.3 94.7 Overall 0.624 M vs. L 2.21 ( . 1 1 . 6 0 4 1 04,18) .5 H vs. L 1.47 ( . 4 9 1 ) 0 9 9 02,.7 .9 Enlisted Groundcrew n Abnormal Normal 154 4 150 2.6 97.4 160 6 154 3.8 96.3 142 5 137 3.5 96.5 Overall 0.834 M vs. L 1.46 ( . 0 5 2 ) 0 7 0 04,.8 .5 H vs. L 1.37 ( . 6 5 2 ) 0 7 2 .4 03,.0 034 .7 0 4 ( . 0 1 5 ) 0.211 .0 01,.8 0.93 ( . 0 2 8 ) 0 9 9 03,.3 .9 �TABLE K-4. Unadjusted Continuous Exposure Index Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index Occupation Statistic Low Officer n Mean 95% C.I. 126 33.3 (31.5,35.3) Enlisted Flyer n Mean 95% C.I. Enlisted Groundcrew Variable High Contrast p-Value 130 34.5 (32.6,36.5) 120 33.8 (31.9,35.8) Overall M vs. L H vs. L 0.695 0.400 0.757 55 30.9 (28.5,33.4) 65 33.3 (31.0,35.8) 57 33.2 (30.7,35.8) Overall M vs. L H vs. L 0.289 0.151 0.192 n Mean 95% C.I. 154 32.9 (31.4,34.5) 160 33.2 (31.7,34.8) 142 34.4 (32.8,36.1) Overall M vs. L H vs. L 0.418 0.801 0.211 Officer SCOT Medium n Mean 95% C.I. 126 22.5 (20.5,24.6) 130 22.1 (20.2,24.2) 120 21.3 (19.4,23.39) Overall M vs. L H vs. L 0.712 0.804 0.421 Enlisted Flyer n Mean 95% C.I. 55 18.5 (16.4,20.8) 65 21.9 (19.7,24.5) 57 22.0 (19.6,24.7) Overall M vs. L H vs. L 0.061 0.037 0.041 Enlisted Groundcrew n Mean 95% C.I. 154 20.9 (19.3,22.6) 160 22.3 (20.7,24.1) 142 21.5 (19.9,23.4) Overall M vs. L H vs. L 0.495 0.237 0.595 00 SGPT �TABLE K-4. (continued) Unadjusted Continuous Exposure Index Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index Occupation Statistic Low Officer n Mean 95% C.I. 126 31.4 (27.8,35.5) Enlisted Flyer n Mean 95% C.I. Enlisted Groundcrew GGTP Alkaline Phosphatase High Contrast p-Value 130 32.7 (29.0,36.9) 120 32.6 (28.7,36.9) Overall M vs. L H vs. L 0.885 0.653 0.688 55 29.7 (24.6,35.9) 65 34.9 (29.3,41.5) 57 37.1 (30.8,44.7) Overall M vs. L H vs. L 0.243 0.225 0.104 n Mean 95% C.I. 154 34.7 (31.1,38.7) 160 33.0 (29.6,36.7) 142 31.1 (27.8,34.9) Overall M vs. L H vs. L 0.412 0.527 0.183 Officer Variable Medium n Mean 95% C.I. 126 87.2 (83.5,91.1) 130 86.8 (83.2,90.6) 120 88.9 (85.0,92.9) Overall M vs. L H vs. L 0.735 0.893 0.546 Enlisted Flyer n Mean 95% C.I. 55 88.9 (83.1,95.0) 65 96.3 (90.5,102.4) 57 95.2 (89.1,101.6) Overall M vs. L H vs. L 0.189 0.086 0.153 Enlisted Groundcrew n Mean 95% C.I. 154 93.2 (89.7,96.9) 160 95.1 (91.5,98.8) 142 96.2 (92.4,100.2) Overall M vs. L H vs. L 0.535 0.484 0.270 �TABLE K-4. (continued) Unadjusted Continuous Exposure Index Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index Occupation Statistic Low Officer n Mean 95% C.I. 126 0.76 (.208) 07,.0 Enlisted Flyer n Mean 95% C.I. Enlisted Groundcrew Total Bilirubin High . Contrast p-Value 130 0.73 (.907) 06,.7 120 0.77 (.308) 07,.1 Overall M vs. L H vs. L 0.382 0.324 0.722 55 0.66 (0.61,0.72) 65 0.73 (0.67,0.78) 57 0.76 (0.70,0.83) Overall M vs. L H vs. L 0.045 0.092 0.014 n Mean 95% C.I. 126 0.74 (.007) 07,.8 130 0.75 (.107) 07,.8 120 0.78 (.408) 07,.2 Overall M vs. L H vs. L 0.373 0.783 0.182 Officer Variable n Mean 95% C.I. 126 0.18 (0.17,0.21) 130 0.17 (0.16,0.19) 120 0.19 (0.17,0.21) Overall M vs. L H vs. L 0.606 0.484 0.784 Enlisted Flyer n Mean 95% C.I. 55 0.16 (.402) 01,.0 65 0.18 (0.15,0.21) 57 0.18 (0.15,0.21) Overall M vs. L H vs. L 0.684 0.455 0.440 Enlisted Groundcrew n Mean 95% C.I. 126 0.17 (0.15,0.19) 130 0.15 (0.17,0.20) 120 0.19 (0.16,0.19) Overall M vs. L H vs. L 0.541 0.277 0.728 i o Direct Bilirubin Medium �TABLE K~4. (continued) Unadjusted Continuous Exposure Index Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index Occupation Statistic Low Officer Variable n Mean 95% C.I. Enlisted Flyer High p-Value 126 130 120 125.1 121.8 124.2 (121.2,129.1) ( 2 . , 2 . ) (117.9,125.8) 104181 Overall M vs. L H vs. L 0.489 0.742 0.246 n Mean 95% C.I. 55 65 57 120.2 126.2 117.6 (114.4,126.4) (112.4,123.3) (120.2,132.5) Overall M vs. L H vs. L 0.118 0.533 0.175 n Mean 95% C.I. 154 160 142 123.1 127.9 122.3 (120.1,126.2) (119.4,125.3) (124.6,131.2) Overall M vs. L H vs. L 0.031 0.710 0.037 Officer Cholesterol Contrast Enlisted Groundcrew LDH Medium n Mean 95% C.I. 126 130 120 220.7 210.7 211.8 (213.3,228.3) (204.8,219.0) (203.5,218.2) Overall M vs. L H vs, L 0.122 0.093 0.063 Enlisted Flyer n Mean 95% C.I. 55 65 57 221.1 227.7 213.9 (210.2,232.6) (204.2,224.2) (216.6,239.3) Overall M vs. L H vs. L 0.203 0.351 0.418 Enlisted Groundcrew n Mean 95% C.I. 154 160 142 211.4 213.3 211.1 (205.4,217.6) (205.2,217.2) (207.0,219.8) Overall M vs. L H vs. L 0.873 0.984 0.677 �TABLE K-4. (continued) Unadjusted Continuous Exposure Index Analyses for Hepatic Function Variables and Tvo Porphyrin Determinations by Occupation Exposure Index Triglycerides Occupation Statistic Low Officer Variable n Mean 95% C.I. Enlisted Flyer Enlisted Groundcrew Medium High Contrast p-Value 126 130 120 114.2 117.9 125.1 (100.2,130.1) (103.8,134.1) (109.5,142.9) Overall M vs. L H vs. L 0.624 0.728 0.338 n Mean 95% C.I. 65 57 55 123.8 119.9 123.1 (105.2,145.7) (103.2,139.3) (104.9,144.5) Overall M vs. L H vs. L 0.953 0.775 0.811 n Mean 95% C.I. 142 154 160 118.7 113.8 117.9 (108.1,130.3) (103.9,124.7) (107.0,129.9) Overall M vs. L H vs. 'L 0.796 0.528 0.919 �TABLE K-5. Summary of Adjusted Exposure Index, Analyses Involving Interactions with a Continuous Covariate for Hepatic Function Variables p-Value Interaction Exposure Index Level Slope p-Value on Test of Slope Against Null Hypothesis of Zero Slope Variable Occupation Interaction Covariate SCOT Enlisted Flyer Current Alcohol Use <0.001 Low Medium High -0.012 0.068 0.029 0.266 004 .0 0.138 SCOT Enlisted Groundcrew Current Alcohol Use <0.001 Low Medium High 0.004 007 .4 0.062 0.628 <0.001 <0.00i SGPT Enlisted Groundcrew Current Alcohol Use 0.006 Low Medium High -0.026 0.027 0.071 0.097 0.189 0.013 GGTP Enlisted Flyer Current Alcohol Use 0.003 Low Medium High 0.022 0.183 0.154 0.455 <0.001 0.006 GGTP Enlisted Groundcrew Current Alcohol Use 009 .0 Low Medium High 0.035 0.091 0.170 0.129 <0.001 <0.001 Direct Bilirubin Enlisted Flyer Current Alcohol Use 0.017 Low Medium High 0.013 0.097 0.065 0.445 <0.001 0.090 Cholesterol Enlisted Groundcrew Current Alcohol Use 0.006 Low Medium High 0.003 -0.008 0.033 0.591 0.285 0.002 �TABLE K-5. (continued) Summary of Adjusted Exposure Index, Analyses Involving Interactions vith a Continuous Covariate for Hepatic Function Variables Exposure Index Level Slope Variable I Enlisted Groundcrew Current Alcohol Use 0.025 Low Medium High -0.005 0.016 0.003 0.322 0.020 0.728 Triglycerides *> Occupation Interaction Covariate LDH l-» p-Value Interaction p-Value on Test of Slope Against Null Hypothesis of Zero Slope Enlisted Flyer Age 0.012 Low Medium High 0.014 0.015 0.038 0.286 0.247 0.032 �TABLE K-6. Summary of Adjusted Exposure Index, Analyses Involving Interactions vith a Categorical Covariate for Hepatic Function Variables Variable Occupation Total Bilirubin Enlisted Groundcrew Interaction Covariate Stratification Exposure Index Level Adjusted Mean (n) l-» Ul Black Low Medium High 0.63 (14) 0.90 (13) 0.70 (13) Nonblack Race Low Medium High 0.75 (138) 0.73 (147) 0.80 (127) Interaction p- Value 0.007 �TABLE K-7. Unadjusted Analyses for Baseline and Interval History of Liver Disease by Group (Verified by Medical Record Review) (Original Comparisons Only) Group Comparison Ranch Hand Disease Statistic Number Percent Est. Relative Number Percent Risk (95% C.I.) p-Value Hepatitis (Viral and Alcoholic) n Yes No 1,016 37 979 3.6 96.4 955 34 3.6 921 96.4 1.02 (0.64,1.65) 0.999 Jaundice n Yes No 1,016 20 996 2.0 98.0 955 21 2.2 934 97.8 0.89 (0.48,1.66) 0.754 n Yes No 1,016 3 1,013 0.5 99.5 955 1 0.1 954 99.9 2.83 (0.29,27.21) 0.625 Enlarged Liver n Yes No 1,016 17 999 1.7 98.3 955 15 1.6 940 98.4 1.07 (0.53,2.15) 0.999 Miscellaneous Liver Disorders n Yes No 1,016 17 999 1.7 98.3 955 7 0.7 948 99.3 2.31 (0.95,5.58) 0.065 Cirrhosis K-16 �TABLE K-8. Unadjusted Analyses of Enlarged Livers Diagnosed at Physical Examination by Group* (Original Comparisons Only) Enlarged Liver Yes Group Number No Percent Number Percent Total p-Value 0.227 Ranch Hand 8 0.8 1,002 99.2 1,010 Original Comparison 3 0.3 950 99.7 953 *Excludes participants with positive HBgAg. K-17 �TAHEK-9. and Two Borphyrin Determinations by Group (Original Comparisons Only) Group Variable Statistic Ranch Hand Original Comparison Est. Relative Risk (95% C.I.) SOOT n Mean 95% C.I. Nunber/% Normal High 1,009 33.5 (32.9,34.1) 952 33.4 (32.8,34.0) — 929 92.3% 80 7.9% 872 91.6% 80 8.4% n Mean 95% C.I. »jmber/% Normal High 1,C09 21.6 (21.0,22.3) 952 ' 22.5 (21.8,23.2) 872 86.4% 137 13.6% 816 85.7% 136 14.3% n Mean 95% C.I. Nunber/% Normal High 1,009 32.8 (31.4,34.3) 952 32.7 (31.3,34.2) 919 91.1% 90 8.9% 862 90.6% 90 9.4% n Mean 95% C.I. Number/% Normal High 1,009 91.8 (90.4,93.3) 952 89.7 (88.3,91.2) 953 94.5% 56 5.5% 911 95.7% 41 4.3% SOT QGTP Alkaline Phosphatase 0.94 (0.68,1.30) — 0.94 (0.73,1.22) — 0.94 (0.69, 1.28) — 1.31 (0.87,1.97) p-Value 0.903 0.701 0.060 0.651 0.921 0.682 0.044 0.205 �TABIEK-9. (ccntinuad) and Two Borphyrin Determinations by Group (Original Comparisons Cbly) Group • Variable Statistic Ranch Hand Original Comparison Total Bilirubin n Mean 95% C I .. Number/% Normal High 1,009 0.74 (.307) 07,.6 952 07 .5 (.307) 07,.6 982 9 . % 73 27 2 7 .% 919 9 . % 65 33 3 5 .% n Mean 95% C I .. Number/% Normal High 1,009 02 .8 (.702) 02,.8 952 02 .8 (.702) 02,.8 971 9 . % 62 38 38 .% 920 9 . % 66 32 3 4 .% n Mean 95% C I .. Number/% Normal High 1,039 123.5 (2.,2.) 122148 952 124.3 (2.,2.) 128157 999 9 . % 90 10 1 0 .% 941 11 n Mean 93% C I .. Number/% Normal High 109 ,0 214.3 (1.,1.) 218268 952 216.7 (1.,1.) 240293 863 8 . % 55 146 1 . % 45 778 8 . % 17 174 IS.3% - Direct Bilirubin LDH Cholesterol . 98.8% 12 .% Est. Relative Risk ( 5 C I ) 9%.. — 0.77 ( . 6 1 2 ) 04, . 8 — 1.13 ( . 0 1 8 ) 07, . 2 — 08 (.620) .6 03,.2 — 0.76 ( . 0 0 % 06,.) p-Value 0.772 030 .1 075 .9 069 .2 049 .4 0.724 0.205 003 .2 �TABLE K-9. (continued) and Two Parphyrin Determinations by (koup (Original < Xflll|Xil. 'SOPS 'fll'yj Grcwp Variable Statistic Ranch Band Original Comparison Triglycerides n Mean 95%C.I. Nunber/% Normal High 1,009 118.5 (113.8,123.3) 952 118.1 (113.4,122.9) 941 93.3% 68 6.7% 891 93.6% 61 6.4% n Mean 95% C.I. 1,006 16.9 (16.2,17.7) 949 17.7 (16.9,18.4) — 0.176 n Mean 95% C.I. 1,008 119.1 (116.2,122.0) 950 115.0 (112.0,118.0) — 0.053 Uroporphyrin Coproporphyrin — No relative ride or confidence interval given for continuous analyses. Est. Relative Risk (95% C.I.) — 1.06 (0.74,1.51) p-Value 0.908 0.767 �TABLE K-10. and Two Borphyrin Determinations by Group (Original Comparisons Oily) Group Ranch Hand Original Comparison n Adj. Mean 95% C.I. 1,003 34.69 (33.66,35.75) 949 34.64 (33.61,35.70) 0.908 n Adj. Mean 95% C.I. 1,003 35.97 (34.82,37.17) 949 35.99 (34.83,37.18) 0.974 n 1,003 949 CC SGPT Statistic DD SOOT Analysis OC Variable n Adj. Mean 95% C.I. 1,003 21.52 (20.89,22.17) 949 22.53 (21.86,23.23) 0.035 ALC*DC (p=0.002) AGE*ALC (p=0.006) n Adj. Mean 95% C.I. 1,003 21.49 (20.21,22.86) 949 22.52 (21.18,23.%) 0.032 ALODC (p=O.043) AGE*ALC (p=0.008) n 1,003 949 CD CD DD Adj. Relative Risk (95%C.I.) 0.93 (0.66,1.29) 0.91 (0.70,1.18) p-Value 0.655 0.467 Covariate Remarks* ALC*DC (p<0.001) AGE*ALC (p=0.033) RACE (p=0.003) RACE (p=0.004) AIC (pO.OOl) 1C (p=0.032) DC (p=0.044) AGE*ALC (p<D.OOi) CCC*ALC (p<0.001) RACE (p=0.026) AGE (p=0.008) ALC (p^O.003) �TABLE K-10. (continued) and T\» Pdrphyrin Determinations by Group (Original Comparisons Only) Group Adj. Relative Risk (95%C.I.) Govariate Reniarks* Ranch Hand Original Comparison n Adj. Mean 95% C.I. 1,003 37.47 (34.93,40.20) 949 37.39 (34.84,40.12) — 0.941 ALC*DC (pO.OOl) AG£*DC (p=0.029) RACE*IC (p=0.005) n Adj. Mean 952C.I. 1,003 42.35 (38.17,46.98) 949 42.24 (38.08,46.86) — 0.932 AG£*ALC (p=0.006) RACE (pO.OOl) n 1,003 949 OC Alkaline Fhosphatase Statistic ED QGTP Analysis OC Variable n Adj. Mean 9SSC.I. 1,003 91.5 (89.1,93.9) 950 89.1 (86.8, 91.5) n Adj. Mean 95% C.I. 1,003 950 -I t i t , ffTCKff AAXA AA AA AAAA n 1,003 950 CD CD DD 0.94 (0.68,1.28) p-Value 0.680 — 0.020 _ **** 1.39 (0.91,2.12) 0.121 AGE*ALC (p=0.005) RACE (p=0.049) RACE*IC (p=0.002) WINE (pO.OOl) AGE (p<0.001) OCC (p<0.001) GEP*IC (p=0.006) AGE*IC (p=0.040) RACE*IC ()M).002) WINE (pO.OOl) OCC (pO.OOl) WINE*DC (p=0.010) AGE*IC (p=0.006) RACE*IC (p=O.C04) OCC*IC (p=0.011) GRP*IC (marginal) (p=0.052) �TfiBlEK-lO. (continued) and Two Barphyrin Determinations by Qxup (Original Comparisons Only) Group Total Bilirubin U) Direct Bilirubin Statistic Ranch Hand Original Comparison OC n Adj. Mean 95* C.I. 1,003 0.78 (0.75,0.81) 949 0.78 (0.75,0.82) 0.714 CCC*RACE (p4D.002) ALC*RACE (p=O.008) AGE*DC (p=0.043) n Adj. Mean 95* C.I. 1,003 0.84 (0.80,0.88) 949 0.84 (0.80,0.88) 0.774 OCC*RACE (p=0.003) OCOALC (p4).047) RACE*ALC (p=0.002) n 1,009 952 0.314 RACE (p<O.C01) OC - Analysis ED Variable n Adj. Mean 95% C.I. 1,003 (0.278) (0.271,0.285) 949 0.276 (0.269,0.283) 0.746 n Adj. Mean 95%C.I. 1,003 0.308 (0.291,0.325) 949 0.305 (0.288,0.323) 0.673 1,003 949 **** CD CD ED Adj. Relative Risk (95% C.I.) 0.77(0.46,1.29) p-Value Covariate Remarks* ALC (p<0.001) IC*DC (p=0.014) ALC*DC (p=0.012) ALC««ACE (p=0.030) ALC*CCC (p=0.007) GEP*IC (p=0.010) RACE (p=0.014) ALC (p=0.016) �TABLE K-10. (continued) and fao Borphyrin Determinations by Group (original Canparisons Only) Group Ranch Band Original Comparison n Adj. Mean 95% C.I. 1,009 952 "fCfCffX "X7CK7C TfifcxX "JfXffJf n Adj. Mean 95% C.I. 1,009 129.4 (126.5,B2.2) 952 130.0 (127.2,132.9) n 1,003 949 CC Cholesterol Statistic ED U3H Analysis OC Variable n Adj. Mean 95% C.I. 1,003 219.3 (214.1,224.6) 949 221.3 (216.0,226.6) — 0.301 n Adj. Mean 95% C.I. 1,003 223.7 (217.3,230.3) 949 226.0 (219.6,232.5) — 0.243 n 1,003 949 CD CD ED Adj. Relative Risk (95% C.I.) p-Value Covariate Renarks* GRP*A2E (p=0.038) IOBACE (jMD.029) IC*OGC (^0.049) — **** — 0.555 M3S, (pO.OOl) RACE (p=O.006) 0.706 ALC (p=0.046) 0.85 (0.36,2.01) 0.74 (0.58,0.94) 0.015 OOO«ACE (p=0.003) DC*RACE (p=0.006) AGE (jxD.COl) AI£ (pOD.OOl) 1C (j^O.016) OCC««ACE (p=0.029) ALC (p<0.001) AGE (ixO.OOl) RACE*ALC (p=0.009) AGE (p=0.020) OCC (p=0.037) �TfiBLEK-10. (continued) and Two Borphyrin Daterminatians by Group (Original Comparisons Only) Group 'feriable Analysis CC Statistic Ranch Hand Original Comparison n 1,003 949 95% C.I. n 1,003 112.7 (103.4,122.9) 949 112.3 (103.1,122.4) n 1,003 949 Uroporphyrin CC — 0.904 946 Adj. Mean ED **** AAAA 95% C.I. CD p-Value _ Adj. Mean Triglycerides Adj. Relative Risk (95% C.I.) n 1,000 Adj. Mean AAA A 95% C.I. "XTfTCX 1.06 (0.74,1.51) — AAA A' 0.770 **** Covariate Remarks* GBP*AGE (p=0.012) ALC (p<0.001) RACE (pO.COl) OCC (pO.OOl) OCC (p<0.001) RACE (p<0.001) AGE (p<0.001) ALC (p=0.007) AGE*ALC (p=0.023) RACE (p=0.044) GRP*BUN (p4).014) OCO«CC (p=0.006) ALC (p=0.048) �T&BLEK-10. (continued) Adjusted and Categorical Analyses for Hepatic Ruction Variables and Tuo Porphyrin Determinatians by Group (Original Confiarisons Ckily) . Group Variable Analysis Statistic Ranch Hand Original Comparison Coproporphyrin CC n Adj. Mean 95% C I .. 1,002 119.1 (1.,2.) 162119 947 115.2 (1.,1.) 124182 Adj. Relative Risk ( 5 C I ) 9%.. — p-Value 006 .6 *Abbreviations; GRP: group OCC: occupation ALC: current alcohol use VINE: vine consumption DC: exposure to degreasing chemicals 1C: exposure to industrial chemicals BUN: blood urea nitrogen — No relative risk or confidence interval given for continuous analyses. ***^Group-by-covariate interaction—adjusted mean/relative risk, confidence interval, and p-value not presented. Covariate Remarks* ALC ( < . 0 ) p001 BUI ( < . 0 ) p001 �TABLE K-U. LY) (Origii Group Variable Interaction Stratification Statistic Ranch Hand Original Comparison Alkaline Phosphatase Grcqp-by-Industrial Chemicals Exposed to Industrial Chemicals n Mean Adj. .Mean 95% C.I. 504 94.8 96.9 (93.5,100.6) 506 90.0 92.1 (88.8,95.5) Not Exposed to Industrial Chemicals n Mean Adj. Mean 95% C.I. 499 88.9 92.5 (89.3,95.8) 444 89.5 93.3 (90.0,96.7) Bom >1942 n Mean Adj. Mean 95% C.I. 408 •122.2 124.1 (121.1,127.1) 369 121.7 123.6 (120.6,126.7) n Mean Adj. Mean 95% C.I. 565 123.8 126.4 (123.7,129.1) 538 125.2 127.8 (125.1,130.5) n Mean Adj. Mean 95% C.I. 36 135.1 138.0 (129.9,146.5) 45 134.6 137.4 130.1,145.1) Lffi Group-by-rAge Born 1923-1941 Bom <1922 p-Value 0.001 0.608 0.756 0.295 0.913 �TABLE K-ll. x« • • -1 — . * » i v (QciginaJ- firmpangmg Only) Group Variable Interaction Stratification Statistic Ranch Band Original Comparison Triglycerides Group-by-rAge Born XL942 n Mean Adj. Mean 95% C.I. 406 107.14 95.55 (87.62,104.20) 368 110.24 98.49 (90.12,107.63) n Mean Adj. Mean 95%C.I. 561 125.98 116.13 (107.59,125.35) 536 124.23 115.14 106.64,124.31) n Mean Adj. Mean 95%C.I. 36 139.00 129.40 (104.04,160.94) 45 106.00 98.40 (80.80,119.83) n Mean Adj. Mean 95% C.I. 547 16.23 16.39 (15.41,17.44) 509 18.18 18.45 (17.33,19.65) n Mean Adj. Mean 952C.I. 453 17.79 18.02 (16.85,19.27) 547 17.15 17.45 (16.30,18.67) Bom 1923-1941 • Bom <L922 Uroporphyrin Group-by-BIN BIW <L4 BUM >14 p-Value 0.055 0.824 0.510 0.005 0.479 �TABLE K-12. Sunnary of Group-by-Covariate Interactions for Variables and (Original Comparisons Only) Group Ranch Hand Variable Interaction Stratification Exposed to Industrial Chemicals Direct Bilirubin Statistic Number Percent Comparison Number Percent Adj. Relative Risk ( 5 C I ) p-Value 9% .. n High Normal 507 27 480 5.3 9. 47 506 14 492 2.8 97.2 1.96 ( . 2 3 8 ) 0 0 5 10,.0 .4 n High Normal 4% 11 485 2.2 97.8 443 18 425 4.1 95.9 0.53 ( . 5 1 1 ) 0 1 7 02,.5 .0 Group-byIndustrial Chemicals Not Exposed to Industrial Chemicals �TABLE K-13. Unadjusted Analysis for Interval History of Skin Bruises, Skin Patches, or Skin Sensitivity by Group (Original Comparisons Only) Bruises, Patches, or Sensitivity Yes Group No Number Percent Number Percent Ranch Hand 265 26.2 746 73.8 1,011 Original Comparison 195 20.5 758 79.5 953 K-30 Total p-Value 0.003 �B/796I/disk 11/document 11 TABLE K-14. Unadjusted Analyses for Porphyrin Abnormalities by Group and Skin Patch, Bruise, or Sensitivity Reported at Followup Questionnaire (Original Comparisons Only) Abnormal Porphyrin Findings for a Participant Group Skin Patch, Bruise, or Reported Sensitivity 0 Number 1 Percent Number 2 Percent Number Percent Total p-Value* Both Groups Yes No 412 1,361 89.8 91.0 45 129 9.8 8.6 2 6 0.4 0.4 459 1496 0.754 Ranch Hand Yes No 239 670 90.5 90.3 24 70 9.1 9.4 1 2 0.4 0.3 264 742 0.950 Original Comparison Yes No 173 691 88.7 91.6 21 59 10.8 7.8 1 4 0.5 0.5 195 754 0.419 *Chi-square test, 2 d . f . �TABLE K-15. Unadjusted Analyses for Uroporphyrin Abnormalities by Skin Patch, Bruise, or Sensitivity Reported at Follovup Questionnaire and by Group (Original Comparisons Only) Group Ranch Hand Original Comparison Stratification Statistic Skin Patch, Bruise, or Sensitivity Reported Variable n Abnormal Normal 264 12 252 4.5 95.5 195 9 186 Skin Patch, Bruise, or Sensitivity Not Reported n Abnormal Normal 742 42 700 5.7 94.3 754 40 714 Number Percent Number Est* Relative Risk (95Z C.I.) p-Value 4.6 95.4 0.98 ( . 1 2 3 ) 04,.8 0.999 5.3 94.7 1.07 ( . 9 1 6 ) 06,.7 0.821 Percent Uroporphyrins �TABLE K-16. Longitudinal Results for SCOT, SGPT, and GGTP: A Contrast of Baseline and First Follovup Examination Test Means (Original Comparisons Only) Means Variable SCOT SGPT GGTP Group Total 1982 Baseline 1985 Followup Ranch Hand Original Comparison 971 32.91 33.73 872 32.93 33.69 971 20.08 21.82 872 20.51 22.36 971 39.26 33.16 872 38.76 32.56 Ranch HandOriginal Comparison Ranch Hand Original Comparison *Analyzed in log units. K-33 p-Value* (Equality of Difference) 0.90 0.86 0.78 �APPENDIX L Dermatological Evaluation �APPENDIX L: Dermatological Evaluation Contents Table Page L-l Unadjusted Exposure Index Analyses for Dermatological Variables by Occupation L-l L-2 Interaction Summaries of Adjusted Exposure Index Analyses of Dermatological Variables L-5 L-3 Unadjusted Analysis for Reported Historical Occurrence of Acne by Group (Original Comparisons Only) L-8 L-4 Unadjusted Analysis for Reported Historical Occurrence Relative to 1961 by Group (Original Comparisons Only) L-8 L-5 Unadjusted Analysis for Reported Historical Occurrence of Acne Relative to SEA Tour of Duty for Post-1961 Acne by Group (Original Comparisons Only) L-9 L-6 Adjusted Analysis for Duration of Acne (in Years) for Post-1961 Acne by Group L-9 L-7 Results Discussed Narratively in Questionnaire Data Section of Chapter (Original Comparisons Only) L-10 L-8 Unadjusted Analyses for Dermatological Variables by Group (Original Comparisons Only) L-ll L-9 Adjusted Analyses for Dermatological Variables by Group (Original Comparisons Only) L-13 L-10 Summary of Group-by-Presence of Pre-SEA Acne Interaction for the Dermatology Index (Original Comparisons Only) L-15 L-ll Longitudinal Analysis of the Dermatology Index: A Contrast of Baseline and First Follovup Examination Abnormalities (Original Comparisons Only) L-16 Figure L-l Location of Post-SEA and Pre- and Post-SEA Acne by Group (Original Comparisons Only) L-17 L-2 Location of Post-SEA Acne by Group (Original Comparisons Only) ... L-18 L-i �TABLE Unadjusted Exposure Index Analyses for Dennatological Variables by Occupation Variable Low Number Percent Exposure Index Medium High Number Percent Number Percent Contrast Est. Relative Risk ( 5 C I ) p-Value 92.. 20.3 79.7 Overall M vs. L H vs. L 0.198 09,.5 1.78 ( . 5 3 3 ) 0 0 4 .8 1.45 ( . 5 2 8 ) 0.320 07,.0 57 16 41 28.1 71.9 0.225 Overall M vs. L 0.58 ( . 7 1 2 ) O.lflO 02,.3 H vs. L 0 5 ( . 5 1 1 ) 0 1 5 .6 .4 02,.9 22.1 77.9 142 41 101 28.9 71.1 Overall M vs. L H vs. L 0.393 0 8 ( . 8 1 3 ) 0.433 .1 04,.5 1.16 ( . 9 1 9 ) 0 6 3 06,.3 .0 130 21 109 16.2 83.8 123 23 100 18.7 81.3 Overall M vs. L H vs. L 070 .2 05,.5 1.10 ( . 6 2 1 ) 0 8 4 .6 06,.4 1.31 ( . 7 2 5 ) 0 5 0 .0 14.5 85.5 65 9 56 13.8 86.2 57 8 49 1. 40 8. 60 Overall M vs. L H vs. L 094 .9 0 9 (.426) 099 . 4 03,.4 .9 0 % (.327) 099 . 03,.7 .9 20.1 79.9 163 36 127 22.1 77.9 142 33 109 23.2 76.8 Overall M vs. L H vs. L 086 .0 1.13 ( . 6 1 9 ) 0.682 06,.3 06,.9 1.20 ( . 9 2 0 ) 0.573 n Abnormal Normal 127 19 108 15.0 85.0 130 31 99 23.8 76.2 123 25 98 Enlisted Flyer n Abnormal Normal 55 23 32 4. 18 58.2 65 19 46 29.2 7. 08 Enlisted Groundcrew n Abnormal Normal 154 40 114 26.0 7. 40 163 36 127 Officer Acneiform Lesions Statistic Officer •Comedones Occupation n Abnormal Normal 127 19 108 15.0 85.0 Enlisted Flyer n Abnormal Normal 55 8 47 Enlisted Groundcrew n Abnormal Normal 154 31 123 �.___— _ _ unadjusted &Kposure Index Analvses for 1Iprmat-nlncriretl rf Variable Low Number Percent 'Obi-iaKloQ YKT (V-mmrim —. ^ -_^-_ _ _ Exposure Index Medium High Number Percent Number Percent Statistic Officer n Abnormal Normal 127 7 120 5.5 94.5 130 13 117 1. 00 9. 00 123 15 108 12.2 87.8 0.175 Overall 07,.4 M vs. L 1.91 ( . 3 4 9 ) 0.245 H vs. L 2.38 ( . 4 6 0 ) 0 0 5 09,.6 .7 Enlisted " Flyer n Abnormal Normal 55 12 43 21.8 78.2 65 12 53 18.5 81.5 57 7 50 12.3 87.7 041 .0 Overall M vs. L 0 8 ( . 3 1 9 ) 0.655 .1 03,.9 H vs. L 0.50 ( . 8 1 3 ) 0.214 01,.9 Enlisted Groundcrew n Abnormal Normal 154 22 132 14.3 85.7 163 29 134 17.8 82.2 142 33 109 23.2 76.8 Overall M vs. L H vs. L Officer • Acneiform Scars Contrast Est. Relative Risk ( K . . p-Value 9CI) Occupation n Abnormal Normal 127 14 113 11.0 8. 90 130 5 125 3.8 96.2 123 19 104 15.4 8. 46 Overall 008 .0 01,.3 M vs. L 0.32 ( . 1 0 9 ) 0.033 H vs. L 1 4 ( . 0 3 0 ) 0.352 . 8 07,.9 Enlisted Flyer Abnormal Normal 55 10 45 18.2 81.8 65 7 58 1. 08 89.2 57 8 49 1. 40 8. 60 Overall Hvs. L H vs. L 0.509 01,.4 0.54 ( . 9 1 5 ) 0.298 02,.3 0.74 ( . 7 2 0 ) 0.613 Enlisted Groundcrew n Abnormal Normal 154 15 139 9.7 90.3 163 7 156 4.3 95.7 142 17 125 12.0 8. 80 Overall M vs. L M vs. L 0.045 01,.5 0.42 ( . 7 1 0 ) 0 0 6 .7 1.26 ( . 0 2 6 ) 0.578 06,.3 0.135 07,.8 1.30 ( . 1 2 3 ) 0 4 6 .4 1.82 ( . 0 3 3 ) 0.053 10,.0 E Depigmentation �TfiBLB L-l. (continued) Unadjusted Exposure Index Analyses for Dermatological \fariables by Occupation Variable Lew Number Percent Exposure Index Medium High Number Percent Number Percent Contrast Est. Relative Risk ( 5 C I ) p-Value 9% . . 10.6 8. 94 Overall M vs. L H vs. L 0.221 1.92 ( . 8 4 1 ) 0.125 08,.9 1.25 ( . 4 2 9 ) 0.671 05,.0 57 8 49 1. 40 8. 60 Overall M vs. L H vs. L 087 .5 1.31 ( . 4 3 9 ) 0 7 3 04,.5 .8 1.33 ( . 3 4 1 ) 0 7 7 04,.3 .7 9.2 9. 08 142 16 126 11.3 8. 87 Overall 086 .3 M vs. L 0.87 ( . 2 1 8 ) 0 8 0 04,.4 .5 H vs. L 1.10 ( . 3 2 2 ) 0 8 3 05,.8 .5 130 23 107 17.7 82.3 123 20 103 16.3 83.7 Overall M vs. L H vs. L 0.922 0.97 ( . 1 1 8 ) 0 9 9 05,.4 .9 08 (.617) 0 7 9 .8 04,.0 .3 23.6 76.4 65 D 52 2. 00 8. 00 57 16 41 28.1 71.9 Overall M vs. L H vs. L 059 .7 .6 08 (.419) 0 6 2 .1 03,.3 05,.5 .6 1.26 ( . 4 2 9 ) 0 6 9 26.6 73.4 163 47 116 2. 88 71.2 142 32 110 22.5 77.5 Overall M vs. L H vs. L 0.452 1.12 ( . 8 1 8 ) 0 7 7 06,.3 .0 .2 08 (.713) 0 4 2 .0 04,.6 Statistic Officer n Abnormal Normal 127 11 116 8.7 91.3 130 20 110 15.4 8. 46 123 13 110 Enlisted ' Flyer n Abnormal Normal 55 6 49 10.9 89.1 65 9 56 13.8 86.2 Enlisted Groundcrev n Abnormal Normal 154 16 138 10.4 89.6 163 15 148 Officer • Inclusion Cysts Occupation n Abnormal Normal 127 23 104 18.1 81.9 n Abnormal Normal 55 13 42 n Abnormal Normal 154 41 113 Hyperpigmen- Enlisted tation Flyer Enlisted Groundcrev �TAHE L-l. (contiiued) Unadjusted Exposure Index Analyses for Dermatological Variables by Occupation Exposure Index Variable Low Number Percent Occupation Statistic Officer n Abnormal Normal 127 73 54 n Abnormal Normal Enlisted Groundcrew n Abnormal Normal Hsdli™ High Number Percent Number Percent Est. Relative Risk ( 5 C I ) p-Walue 92 . . 67.0 33.0 123 85 38 69.1 30.9 Overall 0.122 M vs. L 1.50 ( . 0 2 4 ) 0.125 09,.9 H vs. L 1.66 ( . 8 2 7 ) 0 0 7 09,.8 .6 65 43 22 66:2 33.8 57 33 24 57.9 42.1 Overall 001 .7 M vs. L 0.55 ( . 4 1 2 ) 0.160 02,.4 H vs. L 0.38 ( . 7 0 8 ) 0.027 01,.8 56.5 43.5 163 78 85 47.9 52.1 142 79 63 55.6 4. 44 Overall 0.237 M vs. L 0.71 ( . 5 1 1 ) 0.144 04,.0 H vs. L 0.97 ( . 1 1 5 ) 0 9 7 06,.3 .0 n 127 X) (Abnormal) 46 0 (Normal) 81 36.2 63.8 130 56 74 43.1 56.9 123 56 67 45.5 54.5 Overall 030 .0 Mvs. L 1.33 ( . 1 2 2 ) 0 3 8 08,.0 .0 H vs. L 1.47 ( . 9 2 4 ) 0.157 08,.4 Enlisted Flyer n 55 X) (Abnormal) 33 0 (Normal) 22 6. 00 4. 00 65 34 31 52.3 47.7 57 29 28 5. 09 49.1 Overall 059 .7 03,.1 Mvs. L 0.73 ( . 5 1 5 ) 0 4 2 .6 H vs. L 0 6 ( . 3 1 4 ) 0.349 .9 03,.6 Enlisted Groundcrew n 154 X) (Abnormal) 74 0 (Normal) 80 48.1 51.9 163 77 86 47.2 52.8 142 78 64 5. 49 45.1 Overall 0.348 M vs. L 0 9 ( . 2 1 5 ) 0.911 .7 06,.0 H vs. L 1.32 ( . 3 2 0 ) 0.247 08,.8 57.5 42.5 130 87 43 55 43 12 78.2 21.8 154 87 67 Officer Other Enlisted Abnormalities Flyer Dermatology Index Contrast �TABLE L-2. Index Analyses of Dermatological Variables Variable Interaction (Occupation) Stratification Acneiform Scars' Acneiform Scars Exposure Index-byPresence of pre-SEA Acne (Officer) Exposure Index-byPresence of pre-SEA Acne (Enlisted Groundcrew) p-Value n Abnormal Normal 99 0 99 00 . 100 0. 77 6 71 7.8 92.2 76 8 68 10.5 89.5 Overall M vs. L H vs. L n Abnormal Normal 25 5 20 2. 00 8. 00 49 7 42 14.3 85.7 44 6 38 13.6 8. 64 Overall H vs. L 0 7 ( . 0 2 6 ) . 3 02,.8 H vs. L 0 6 ( . 7 2 5 ) .7 01,.6 083 .3 060 .3 0.555 Black, No pre-SEA Acne n Abnormal Normal 1 1 0 100 0. 00 . 2 0 2 00 . 100 0. 1 0 1 00 . 100 0. Overall M vs. L 0 0 ( . 0 , . 9 * . 7 00154)* H vs. L 0.11 ( . 0 , 0 2 ) * 0011.7* 0.135* 0.333* 099 .9* Black, pre-SEA Acne n Abnormal Normal 1 1 0 100 0. 00 . 1 0 1 00 . 100 0. 1 1 0 100 0. 00 . Overall ' M vs. L 0.11 ( . 0 , 0 2 ) * 0011.7* H vs. L — 0.135* 099 .9* — No pre-SEA Acne n Abnormal Normal 101 16 85 15.8 84.2 105 15 90 14.3 85.7 99 16 83 16.2 83.8 M vs. L H vs. L 08 (.918) .6 03,.6 1.11 ( . 1 2 3 ) 05,.9 065 .9 079 .9 Pre-SEA Acne n Abnormal Normal 51 6 45 11.8 88.2 57 14 43 24.6 75.4 41 17 24 41.5 58.5 M vs. L H vs. L 2.24 ( . 9 6 3 ) 07,.9 5.38 ( . 5 1 . ) 14,9% 0.131 002 .1 Nonblack, No pre-SEA Acne Exposure Index-byRace Exposure Index Low Mediun Est. Relative Hitfi Statistic Number Percent Number Percent Number Percent Contrast Risk ( 5 C I ) 9* . . Nonblack, pre-Sea Acne 18.09 ( . 0 3 6 3 ) * 10,2.1* 2.9 (.044%* 46 14,3.)* 006 .0* 006 .0* 001 .0* �TABLE L-2. (continued) Index Analyses of variable Interaction (Occupation) Depigmentation Exposure Index-byPresence of pre-SEA Acne (Enlisted Groundcrev) Stratification lYmnt-n' «T Irwical \Variables — ~ Exposure Index Median Lev High Est. Relative Statistic Number Percent Number Percent Number Percent Contrast Risk (95ZC.I.) p-Value Dermatology Index • Exposure Index-byPresence of pre-SEA Acne (Officer) n Abnormal Normal 101 13 88 12.9 87.1 105 7 98 6.7 93.3 99 11 88 11.1 88.9 Overall M vs. L 0.48 ( . 9 1 2 ) 01,.7* H vs. L 0.85 ( . 6 1 9 ) 03,.9* 034 .1* 011 .6* 088 .2* Pre-SEA Acne n Abnormal Normal 51 2 49 3.9 96.1 57 0 57 00 . 100.0 41 6 35 14.6 85.4 Overall M vs. L H vs. L 006 .0* 021 .2* 0.133* Nonblack, No pre-SEA Acne Exposure Index-byRace No pre-SEA Acne n Abnormal Normal 99 32 67 32.3 67.7 77 40 37 51.9 48.1 76 35 41 46.1 53.9 Overall M vs. L 2.26 ( . 3 4 1 ) 12..8* H vs. L 1.78 ( . 6 3 3 ) 09,.1* 005 .2* 003 .1* 004 .8* Nonblack, pre-SEA Acne n Abnormal Normal 25 12 13 48.0 52.0 49 15 34 30.6 69.4 44 20 24 45.5 54.5 Overall M vs. L 0.48 ( . 8 1 2 ) 01,.9* H vs. L 0.90 ( . 4 2 4 ) 03,.1* 022 .2* 022 .0* 099 .9* Black, No pre-SEA Acne n Abnormal Normal 1 1 0 100.0 00 . 2 0 2 00 . 100.0 1 0 1 00 . 100.0 Overall M vs. L H vs. L 0.07 ( . 0 , . 9 * 00154)* 0.11 ( . 0 , 0 2 ) * 0011.7* 0.135* 033 .3* 099 .9* Black, pre-SEA Acne n Abnormal Normal 1 1 0 100.0 00 . 1 0 1 00 . 100.0 1 1 0 100.0 00 . Overall M vs. L 0.11 ( . 0 , 0 2 ) * 0011.7* — H vs. L 0.223* 099 .9* — 0.17 ( . 1 3 6 ) * 00, .7* 4.20 ( . 0 2 . 5 * 08,20) �TABLE L-2. (continued) Interaction Suamaries of Adjusted Exposure Index Analyses of Dermatological Variables FOOTNOTES, Table L-2 a Eesults presented for acneiform scars (officers) are based on stratification into the four categories shown. Unadjusted results are presented for all strata except nonblack Banch Hands with pre-SEA. acne. These results are adjusted for age. *Uhadjusted estimate of relative risk, and confidence interval, or p-value, based on stratified tables. **uhadjusted estimate of relative risk and confidence interval calculated after adding 0.5 to each cell. —-No normal participants present in contrast; estimated relative risk, confidence interval and p-value not presented. Note: Snail sample sizes may affect validity of overall p-value. ip --j Note: Results without (*) or ( * are adjusted for all other main effects in model (age, race, presence of pre-SEA acne), unless otherwise *) noted. �TABLE L-3. Unadjusted Analysis for Reported Historical Occurrence of Acne by Group (Original Comparisons Only) Acne Yes Group Number No Percent Number Percent Total Ranch Hand Original Comparison 412 40.6 604 59.4 1,016 349 36.5 606 63.5 955 Total 761 Summary Statistics Est. RR: 1.18 95% C.I. 5 (0.99,1.42) p-Value: 0.071 1,971 1,210 TABLE L-4. Unadjusted Analysis for Reported Historical Occurrence Relative to 1961 by Group* (Original Comparisons Only) Occurrence of Acne Post-1961 Group Number Percent Pre-1961 Number Percent Total Ranch Hand Original Comparison 239 58.3 171 41.7 410 192 55.2 156 44.8 348 Total 431 327 Summary Statistics Est. RR (for post1961 cases): 1.14 95* C.I.: (0.85, 1.52) p-Value: 0.418 758 *Three participants deleted due to missing data on time of occurrence. L-8 �TABLE L-5. Unadjusted Analysis for Reported Historical Occurrence of Acne Relative to SEA Tour of Duty for Post-1961 Acne by Group* (Original Comparisons Only) Post-1961 Acne Pre-SEA Group Ranch Hand Original Comparison Total Number Percent . Post-SEA Pre- and Post-SEA Total Number Percent Number Percent p-Value 0.155 58 25.4 80 35.1 90 39.5 228 48 106 26.2 81 161 44.2 54 144 29.5 183 411 *Twenty post-1961 participants with acne deleted due to missing data on time of occurrence. TABLE L-6. Adjusted Analysis for Duration of Acne (in Years) for Post-1961 Acne by Group* Group Ranch Hand Original Comparison Total Total Adjusted Mean** 219 8.18 (7.43,8.96) 175 394 7.15 Covariate Remarks (6.38,7.97) 952 C.I.** p-Value 0.070 Time Reference to SEA (p<0.001) *Seventeen participants deleted due to missing data on time of occurrence. **Converted from square root scale. L-9 �TABLE L-7. Results Discussed Narratively in Questionnaire Data Section of Chapter (Original Comparisons Only) Description p-Value • Effect due to SEA category in continuous analysis of duration of acne • Interaction between group and SEA category in continuous analysis of duration of acne <0.001 0.286 • Categorical analysis of duration of acne — p-value for group difference SEA Category pre-SEA post-SEA pre- and post-SEA 0.718 0.592 0.753 • Intersection of Venn diagram circles (temples, ears, and eyes) — p-value for group difference SEA Category post-SEA and pre- and post-SEA post-SEA L-10 0.133 0.627 �TABLE L-8. Unadjusted Analyses for Dermatological Variables by Group (Original Comparisons Only) Group Ranch Hand Number Percent Original Comparison Number Percent Est. Relative Risk ( 5 C.I.) 9% p-Value 26.5 73.5 0.90 (0.74,1.11) 0.352 954 159 795 16.7 83.3 1.14 (0.90,1.43) 0.288 14.8 85.2 954 128 826 13.4 86.6 1.12 (0.87,1.44) 0.401 1,016 102 914 10.0 90.0 954 113 841 11.8 88.2 0.83 (0.63,1.10) 0.219 n Abnormal Normal 1,016 114 902 11.2 88.8 954 118 836 12.4 87.6 0.90 (0.68,1.18) Hyperpigmentation n Abnormal Normal 1,016 228 788 22.4 77.6 954 222 732 23.3 76.7 0.95 (0.77,1.18) Variable Comedones t-1 Statistic n Abnormal Normal 1,016 • 250 766 24.6 75.4 954 253 701 Acneiform Lesions n Abnormal Normal 1,016 188 828 18.5 81.5 Acneiform Scars n Abnormal Normal 1,016 150 866 Depigmentation n Abnormal Normal Inclusion Cysts • 0.442 0.668 �TABLE L-8. (continued) Unadjusted Analyses for Dermatological Variables by Group (Original Comparisons Only) Group Ranch Hand Number Percent Original Comparison Number Percent Variable Statistic Other Abnormalities n Abnormal Normal 1,016 608 408 59.8 40.2 954 565 389 59.2 40.8 Dermatology Index n 0 1 2 3 4 1,016 533 318 121 34 10 52.5 31.3 11.9 3.3 1.0 954 497 301 116 35 5 52.1 31.6 12.2 3.7 0.5 Est . Relative Risk ( 5 C.I.) 9% 1.03 (0.86,1 .23) Overall 1 vs. 0 0 . 9 9 2 vs. 0 0 .97 3 vs. 0 0 ,91 4 vs. 0 1 .87 ( .81,1. 20) 0 ( .73,1. 29) 0 (0.56,1. 48 ^ ( .63,5. 49) 0 p-Value 0.783 0.816 0.919 0.885 0.711 0.304 �TABLE L-9. Adjusted Analyses for Dermatological Variables by Group (Original Comparisons Only) Group Variable Ranch Hand Total Original Comparison Total Adj . Relative Risk (95% C.I.) Comedones 1,007 946 0.93 (0.75,1.14) 0.469 RACE (p<0.001) AGE (p<0.001) OCC*SEA ACNE (p=0.047) Acneiform Lesions 1,007 946 1.13 (0.89,1.43) 0.315 AGE (p<0.001) RACE (p=0.034) SEA ACNE (p=0.007) Acneiform Scars 1,007 946 1.12 (0.86,1.45) 0.417 AGE (p<0.001) RACE (p<0.001) SEA ACNE (p<0.001) Depigmentation 1,016 954 0.83 (0.63,1.10) 0.202 RACE (p=0.010) Inclusion Cysts 1,016 954 0.90 (0.68,1.18)* 0.442* — Hyperpigmentation 1,007 946 0.95 (0.76,1.17) 0.608 RACE (p<0.001) SEA ACNE (p=0.008) p- Value Covariate Remarks �TABLE L-9. (continued) Adjusted Analyses for Dermatological Variables by Group (Original Comparisons Only) Group Ranch Hand Total Original Comparison Total Other Abnormalities 1,016 954 Dermatology Index 1,007 946 Variable Adj. Relative Risk (95% C.I.) 1.07 (0.89,1.30) **** p-Value 0.459 **** Covariate Remarks AGE (p<0.001) RACE (p<0.001) OCC (p=0.013) GROUP*SEA ACNE (p=0.030) I l-» *- *No significant covariates in adjusted analysis; consequently, estimated relative risk, confidence interval, and p-value are from unadjusted analysis. ****Group-by-covariate interaction; adjusted relative risk, confidence interval, and p-value are not presented. Abbreviations; OCC: Occupation SEA ACNE: Presence of pre-SEA acne �TABLE HO. Summary of Group-by-Presence of. Pre-SEA Acne Interaction for Dermatology Index (Original Comparisons Only) Group Variable Interaction Stratification Statistic Ranch Hand Number Percent n No r K Dennatology Index Group-bypresenceof-pre-SEA. acne 684 0 1 2 3 4 360 234 69 16 5 0 1 2 3 4 167 82 51 18 5 Contrast Adj. Relative Risk(95%C.I.) p-Value 653 52.6 34.2 10.1 2.3 07 . 323 Yes Original Conparison Number Percent 370 202 61 19 1 56.7 3. 09 9.3 2.9 0.2 1 vs. 0 1.19 2 vs. 0 1.17 3 vs. 0 0.87 4 vs. 0 4.16 41.6 33.4 1B.8 5.1 1.0 1 vs. 0 2 vs. 0 3 vs. 0 4 vs. 0 (0.94,1.52) (0.80,1.70) (0.44,1.71) (0.61,28.48) 0.153 0.422 0.689 0.146 293 51.7 25.4 15.8 5.6 1.5 122 98 55 15 3 0.62 (0.42,0.90) 0.68 (0.44,1.07) 0.89 (0.43,1.83) 1.20 (0.30,4.88) 0.012 0.097 0.745 0.797 �TABLE L-ll. Longitudinal Analysis of the Dermatology Index: A Contrast of Baseline and First Followup Examination Abnormalities (Original Comparisons Only) Group 1982 Baseline Exam 1985 Followup Exam Abnormal Normal Odds Ratio (OR)* Ranch Hand Abnormal Normal 241 228 136 366 1.68 Original Comparison Abnormal Normal 222 202 109 339 1.85 p-Value (ORRH vs. ORQC) 0.53 Number Normal Baseline, Abnormal Followup *OrMa Rat in; Number Abnormal Baseline, Normal Followup L-16 �Ranch Hand 53 Other Sites n=80 Original Comparison p = 0.374 (8 Categories) 36 Other Sites n=48 Figure L-1. Location of Post-SEA and Pre- and Post-SEA Acne by Group (Original Comparisons Only) L-17 p = 0.310 (7 Categories; Other Sites Deleted) �Ranch Hand 103 Other Sites n=169 Original Comparison p - 0.492 (8 Categories) p = 0.449 (7 Categories; Other Sites Deleted) 85 Other Sites n=129 Figure L-2. Location of Post-SEA Acne by Group (Original Comparisons Only) L-18 �APPENDIX M Cardiovascular Assessment �APPENDIX M: CARDIOVASCULAR ASSESSMENT Contents Table M-l M-2 M-3 M-4 M-5 M-6 M-7 Page Association Between Reported Essential Hypertension and the Covariates in the Combined Ranch Hand and Comparison Groups M-l Association Between Reported Heart Disease and the Covariates in the Combined Ranch Hand and Comparison Groups M-2 Association Between Reported Myocardial Infarction and the Covariates in the Combined Ranch Hand and Comparison Groups M-3 Cardiovascular Morbidity-Mortality Analysis (Ranch Hands and Comparisons) M-4 Summary of Group-by-Covariate Interactions for Central Cardiac Function Variables (Diabetics Excluded) M-5 Summary of Group-by-Covariate Interactions for Peripheral Vascular Function Variables (Diabetics Excluded) M-6 Unadjusted Exposure Index Analyses for Reported and Verified Heart Disease by Occupation M-7 Unadjusted Exposure Index Analyses for Central Cardiac Function Variables by Occupation M-10 Unadjusted Exposure Index Analyses for Diastolic Blood Pressure, Funduscopic Abnormalities, Carotid Bruits, and Manual Pulse Readings by Occupation M-15 M-10 Unadjusted Exposure Index Analyses for Peripheral Vascular System Doppler Pulse Readings by Occupation . . . . M-21 M-8 M-9 M-ll Association Between Central and Peripheral Abnormalities and Verified Heart Disease in the Combined Ranch Hand and Comparison Groups M-25 M-12 Unadjusted Analyses for Reported and Verified Heart Disease by Group (Original Comparisons Only) M-26 M-13 Adjusted Analyses for Reported and Verified Heart Disease (Original Comparisons Only) M-27 M-i �APPENDIX M: CARDIOVASCULAR ASSESSMENT Contents (continued) Table Page M-14 Unadjusted Analyses for Central Cardiac Function by Group (Diabetics Excluded) (Original Comparisons Only) M-28 M-15 Adjusted Analyses for Central Cardiac Function (Diabetics Excluded) (Original Comparisons Only) M-30 M-16 Summary of Group-by-Covariate Interactions for Central Cardiac Function Variables (Diabetics Excluded) (Original Comparisons Only) M-32 M-17 Unadjusted Analyses for Peripheral Vascular Function by Group (Diabetics Excluded) (Original Comparisons Only) . . . M-33 M-18 Adjusted Analyses for Peripheral Vascular Function (Diabetics Excluded) (Original Comparisons Only) M-36 M-19 Summary of Group-by-Covariate Interactions for Peripheral Vascular Function Variables (Diabetics Excluded) (Original Comparisons Only) M-38 M-20 Longitudinal Analysis of Pulse Index and Overall ECG: A Contrast of Baseline and First Followup Examination Abnormalities (Original Comparisons Only) M-39 M-ii �TABLE M-l. Association Between Reported Essential Hypertension and the Covariates in the Combined Ranch Hand and Comparison Groups Covariate Covariate Category Total Percent Abnormal p-Value Age Born >1942 Born <1942 934 1,214 20.4 29.6 <0.001 Race Black Nonblack 126 2,022 27.8 25.5 0.647 Occupation Officer Enlisted Flyer Enlisted Groundcrev 807 354 987 27.6 27.1 23.5 0.108 Current Smoking 0 >0 - 20 >20 1,262 . 463 422 28.8 22.0 20.4 <0.001 Pack-Years Smoking 0 >0 - 10 >10 512 760 869 28.7 23.6 25.7 0.120 Cholesterol <200 >200 - 230 >230 766 650 732 19.7 26.6 31.0 <0.001 HDL <40 >40 - 50 >50 719 754 675 27.5 25.3 24.0 0.309 Cholesterol-HDL Ratio <4.2 >4.2 - <5.5 >5.5 717 743 688 20.6 26.0 30.5 <0.001 Percent Body Fat <10 10 - 25 >25 10 1,758 379 0.0 21.0 47.8 <0.001 Personality Score <-5 -5-5 >5 829 731 580 26.8 26.7 22.9 0.202 Differential Cortisol <0.6 >0.6 - 4.0 >4.0 704 745 683 29.1 24.4 23.3 0.030 592 809 738 25.2 22.5 29.1 0.011 691 719 666 25.9 21.8 29.4 0.005 Current Alcohol 0 Use (Drinks/Day) >0 - 1 >1 Drink-Years Alcohol <1.25 >1.25 - 25 >25 M-l �TABLE M-2. Association Betveen Reported Heart Disease and the Covariates in the Combined Ranch Hand and Comparison Groups Covariate Covariate Category Total Percent Abnormal p-Value Age Born >1942 Born <1942 934 1,214 22.7 30.3 <0.001 Race Black Nonblack 126 2,022 31.0 26.8 0.354 Occupation Officer Enlisted Flyer Enlisted Groundcrev 807 354 987 30.6 24.9 24.8 0.014 Current Smoking 0 >0 - 20 >20 1,262 463 422 27.5 28.1 24.2 0.342 Pack- Years Smoking 0 >0 - 10 >10 512 760 869 28.1 25.9 27.3 0.667 Cholesterol <200 >200 - 230 >230 766 650 732 26.5 25.5 28.8 0.361 HDL <40 >40 - 50 >50 719 754 675 26.6 25.6 29.0 0.326 Cholesterol-HDL Ratio <4.2 >4.2 - <5.5 >5.5 717 743 688 29.6 23.0 28.6 0.009 Percent Body Fat <10 10 - 25 >25 10 1,75& 379 30.0 27.2 25.9 0.849 Personality Score <-5 -5-5 >5 829 731 580 27.5 24.9 29.1 0.214 Differential Cortisol <0.6 >0.6 - 4.0 >4.0 704 745 683 24.6 27.9 28.4 0.212 Current Alcohol 0 Use (Drinks/Day) >0 - 1 >1 592 809 738 27.4 28.6 25.1 0.297 Drink-Years Alcohol 691 719 666 26.8 25.7 28.4 0.537 <1.25 >1.25 - 25 >25 M-2 • �TABLE M-3. Association Between Reported Myocardial Infarction and the Covariates in the Combined Ranch Hand and Comparison Groups Covariate Covariate Category Total Percent Abnormal p-Value Age Born >1942 Born <1942 934 1,214 0.4 3.1 <0.001 Race Black Nonblack 126 2,022 0.8 2.0 0.523 Occupation Officer Enlisted Flyer Enlisted Groundcrew 807 354 987 1.5 2.3 2.2 0.477 Current Smoking 0 >0 - 20 >20 1,262 463 422 1.4 3.0 2.4 0.083 Pack-Years Smoking 0 >0 - 10 >10 512 760 869 0.2 1.8 3.1 0.001 Cholesterol <200 >200 - 230 >230 766 650 732 1.2 2.0 2.7 0.093 HDL <40 >40 - <50 >50 719 754 675 2.8 1.9 1.2 0.096 Cholesterol-HDL Ratio <4.2 >4.2 - <5.5 >5.5 717 743 688 1.0 1.5 3.5 0.002 Percent Body Fat <10 10 - 25 >25 10 1,758 379 0.0 1.9 2.1 0.867 Personality Score <-5 -5-5 >5 829 731 580 1.9 2.6 1.2 0.195 Differential Cortisol <0.6 >0.6 - 4.0 >4.0 704 745 683 2.0 1.9 2.0 0.973 Current Alcohol 0 Use (Drinks/Day) >0 - 1 >1 592 809 738 2.4 2.1 1.4 0.365 Drink-Years Alcohol 691 719 666 2.6 0.8 2.2 0.034 <1.25 >1.25 - 25 >25 M-3 �TABLE M-4. Cardiovascular Morbidity-Mortality Analysis (Ranch Hands and Comparisons*) Ranch Hands n-1,257 Event Endpoint Status Dead Living, Fully Compliant* Living, not Fully Compliant Event Yes Death (Any Cause) or Verified Heart Disease Comparisons* n.1,253 Estimated Percent Events Death (Any Cause) or Verified Myocardial Infarction Dead Living, Fully Compliant* Living, not Fully Compliant Estimated Percent Events Fatal or Nonfatal Verified Heart Disease Dead Living, Fully Compliant* Living, not Fully Compliant Estimated Percent Events Fatal or Nonfatal Verified Myocardial Infarction or Fatal Heart Disease *Comparison: Hands. Dead Living, Fully Compliant* Living, not Fully Compliant Estimated Percent Events Total No Total Yes 66 0 66 77 0 77 251 823 1,074 187 768 955 27.4° 89. 6C 117 27.4% 66 43. 3C No 177. 7C 221 24.5% 0 66 77 0 77 1,074 9 946 955 1.4° 115.6° 117 2.1° 218. 9C 221 6.4% 7.0% 13 1 ,061 5 61 66 3 74 77 251 823 1,074 187 768 955 27.4C 89.6° 117 22.5% 1 13 1.4C 43.3C 177.7C 221 18.6% 65 66 2 75 77 1,061 1,074 9 946 955 2.1° 218.9C 221 115.6° 117 1.2% 1.0% first Comparison of the randomly ordered set matched to the Ranch *Fully compliant at Baseline or third-year followup examination. Not fully compliant at Baseline and not fully compliant at third-year followup examination. c Estimated using event rate in living, fully compliants. M-4 �TABLE M-5. Summary of Group-by-Covariate Interactions for Central Cardiac Function Variables (Diabetics Excluded) Variable Interaction Stratification Adj. Relative Risk (95% C.I.) p-Value 4.56 (-1.91,11-03)* -16.01 (-27.52,-4.50)* 0.170 008 .0 Systolic Blood Pressure (Continuous Analysis) Group-by-Age (Blacks) Age (Baseline): 35 Age (Baseline): 53 ECG (Overall) Group-by-Pack-years Smoking Pack-years: 0 Pack-years: 30 0.70 (0.50,0.98) 1.25 (0.89,1.76) 0.038 0.197 ECG: Arrythmia Group-by-pack-years Smoking (21% Body Pat) Pack-years: 0 Pack-years: 30 0.58 (0.30,1.10) 1.59 (0.83,3.04) 0.093 0.162 Group-by-Percent Body Fat (7 Pack-years Smoking) 10% Body Fat 30% Body Fat 0.23 (0.07,0.78) 1.88 (0.66,5.34) 0.018 0.234 *Difference in group means (Ranch Hand-Comparison) and associated p-value given, rather than relative risk, for continuous analysis of dependent variables. �TABLE M-6. Sumnary of Group-by-Covariate Interactions for Peripheral Vascular Function Variables (Diabetics Excluded) Variable Interaction Stratification Adj. Relative Risk (95X C.I.) p-Value Popliteal Pulses (Manual) Group-by-Race Black Nbnblack 6.74 (0.72,63.40) 0.55 (0.28,1.12) 0.095 009 .9 Dorsalis Pedis Pulses (Manual) Group-by-Occupation Officer Enlisted Flyer Enlisted Groundcrew 0.70 ( . 5 1 0 ) 04,.9 1.70 ( . 6 3 3 ) 08,.4 1.32 ( . 4 2 0 ) 08,.6 0.116 0.124 0.222 Posterior Tibial Pulses (Manual) Group-by-Occupa t i on Officer Enlisted Flyer Enlisted Groundcrew 0.33 ( . 9 1 1 ) 00,.8 4.33 (1.14,16.50) 1.22 (0.58,2.57) 0.087 0.032 0.603 Leg Pulses (Manual) Group-by-occupation (21% Body Fat) Officer Enlisted Flyer Enlisted Groundcrew 0.62 ( . 1 0 9 ) 04,.4 1.55 (0.87,2.76) 1.24 (0.85,1.81) 0.026 0.136 0.271 Group- by-Percent Body Fat (Officer) Nonobese (<25X) Obese (>25X) 0.66 ( . 2 1 0 ) 04,.4 0.44 (0.17,1.12) 004 .8 0.072 Group- by-Percent Body Fat (Enlisted Flyer) Nonobese (<25%) Obese (>25X) 1.36 (0.73,2.55) 1.83 ( . 6 4 3 ) 07,.8 0.177 0.332 Group-by-Percent Body Fat (Enlisted Grounder ew) Nonobese (<25Z) Obese (>25Z) 1.20 (0.79,1.83) 0.91 (0.39,2.10) 0.818 0.390 Peripheral Pulses (Manual) Group-by-occupation Officer Enlisted Flyer Enlisted Groundcrev 0.64 ( . 2 0 9 ) 04,.6 1.45 (0.82,2.56) 1.16 ( . 0 1 6 ) 08,.9 000 .3 0.204 0.441 All Pulses (Manual) Group-by-Occupation Officer Enlisted Flyer Enlisted Groundcrev 0.64 (0.42,0.96) 1.45 (0.82,2.57) 1.14 (0.78,1.66) 0.030 0.204 0.490 3; �TABLE H-7. Unadjusted Exposure Index Analyses for Reported axl \ferified Heart niappgg by Occupation Variable Low Nunber Percent Exposure Index Medium Number Percent High Number Percent n Abnormal Normal 119 31 88 2. 60 74.0 122 33 89 27.0 73.0 109 32 77 29.4 70.6 Overall M vs. L H vs. L 1.05 ( . 9 1 8 ) 05,.6 1.18 ( . 6 2 1 ) 06,.1 080 .5 087 .5 0.575 Enlisted Flyer n Abnormal Normal 54 14 40 25.9 74.1 56 14 42 25.0 75.0 53 18 35 34.0 6. 60 Overall M vs. L H vs. L 0 9 (.024) . 5 04,.8 1.47 ( . 4 3 3 ) 06,.8 0.525 0.912 0.363 Enlisted Groundcrev n Abnormal Normal 148 33 115 22.3 77.7 151 42 109 27.8 72.2 130 30 100 23.1 76.9 Overall M vs. L H vs. L 1.34 ( . 9 2 2 ) 07,.7 1.05 ( . 0 1 8 ) 06,.3 040 .9 0.271 0.881 Officer Verified Essential Hypertension Statistic Officer Reported Essential Hypertension Contrast Est. Relative Risk ( 5 C I ) p-Value 9% . . Occupation n Abnormal Normal 119 23 % 19.3 80.7 122 22 100 18.0 82.0 109 27 82 2. 48 75.2 Overall M vs. L H vs. L 0.92 ( . 8 1 7 ) 04,.6 1.37 ( . 3 2 5 ) 07,.8 0.413 075 .9 0.322 Enlisted Flyer n Abnormal Normal 54 11 43 20.4 7. 96 56 8 48 14.3 85.7 53 12 41 22.6 77.4 Overall M vs. L 0.65 ( . 4 1 7 ) 02,.7 H vs. L 1.14 ( . 5 2 8 ) 04,.8 0.514 041 .0 0.772 Enlisted Groundcrev n Abnormal Normal 148 28 120 18.9 81.1 151 38 113 25.2 7. 48 130 26 104 2. 00 8. 00 Overall M vs. L H vs. L 0.375 0.194 088 .1 1.44 ( . 3 2 5 ) 08,.0 1.07 ( . 9 1 9 ) 05,.4 �TABLE H-7. (continued) Unadjusted Exposure Index Analyses for Reported and Verified Heart Disease by Occupation Exposure Index Low Number Percent Mgjjiin High. Number Percent Number Percent Contrast 109 • 28.4 31 78 71.6 Overall M vs. L H vs. L n Abnormal Normal 119 41 78 34.4 65.6 122 40 82 32.8 67.2 Enlisted Flyer n Abnormal Normal 54 15 39 27.8 72.2 56 19 37 33.9 66.1 53 11 42 n Abnormal Normal 148 42 106 28.4 71.6 151 25 126 16.6 83.4 Officer Verified Heart Disease Statistic Enlisted Groundcrew Reported Heart Disease Occupation Officer Variable n Abnormal Normal 119 37 82 31.1 68.9 122 35 87 Enlisted Flyer n Abnormal Normal 54 14 40 25.9 74.1 Enlisted Groundcrew n Abnormal Normal 148 37 'ill 25.0 75.0 Est. Relative Risk ( 5 C.I.) p-Value 9% 0.93 ( . 4 1 5 ) 05,.8 0.76 ( . 3 1 3 ) 04,.3 0.607 077 .8 0.332 2. 08 79.2 Overall M vs. L 1.34 ( . 9 3 0 ) 05,.1 H vs. L 0 6 ( . 8 1 6 ) .8 02,.6 0.306 044 .8 041 .0 130 41 89 31.5 68.5 Overall M vs. L 0 5 ( . 9 0 8 ) . 0 02,.7 H vs. L 1.16 ( . 0 1 9 ) 07,.5 008 .0 005 .1 059 .6 28.7 71.3 109 28 81 25.7 74.3 Overall M vs. L 0 8 ( . 1 1 5 ) .9 05,.5 H vs. L 0 7 ( . 3 1 3 ) .7 04,.7 065 .6 0.682 0.368 56 16 40 28.6 71.4 53 9 44 17.0 8. 30 Overall M vs. L 1.14 ( . 9 2 6 ) 04,.5 H vs. L 0.58 ( . 3 1 5 ) 02,.0 0.335 0.757 023 .6 151 16 135 10.6 89.4 130 32 98 24.6 75.4 Overall M vs. L 0.36 ( . 9 0 6 ) 01,.7 H vs. L 0 9 ( . 7 1 7 ) .8 05,.6 002 .0 001 .0 0.994 �TKBLE M-7. (continued) Unadjusted Exposure Index Analyses for Reported aid Verified Heart Disease by Occupation Occupation Statistic Officer Variable n Abnormal Normal Low Number Percent Exposure Index Median Number Percent Higi Number Percent Contrast Est. Relative Risk ( 5 C I ) p-Value 9% .. 08 . 99.2 122 4 118 3.3 96.7 109 1 106 09 . 99.1 Overall M vs. L 4 0 ( . 4 3 . ) . 0 04,63 H vs. L 1 0 ( . 7 1 . ) . 9 00,77 026 .5 0.219 092 .5 54 0 00 . 54 100.0 56 1 55 08 . 98.2 53 4 49 7.6 92.4 Overall M vs. L H vs. L 001 .6 — — 3.1 96.9 Overall M vs. L 0 6 ( . 1 3 9 ) . 5 01,.4 H vs. L 1 5 ( . 4 6 9 ) . 3 03,.8 119 1 118 n Abnormal Normal n Abnormal Normal 148 3 145 2.0 9. 80 151 2 149 1.3 98.7 130 4 126 Officer n Abnormal Normal 119 0 00 . 119 1 0 0 0. 122 4 118 3.3 96.7 109 0 00 . 109 100.0 Enlisted Flyer n Abnormal Normal 54 0 00 . 54 1 0 0 0. 56 1 55 1.8 98.2 53 1 52 1.9 98.1 Overall Mvs. L — H vs. L — Enlisted Groundcrev Verified Myocardial Infarction Enlisted Flyer Enlisted Groundcrev Reported Myocardial Infarction n Abnormal Normal 151 00 . 0 151 100.0 130 1 129 08 . 99.2 Overall M vs. L H vs. L 148 2 146 — Analysis not performed due to sparse cells. 1.4 98.6 — — Overall Mvs. L — H vs. L — — — 052 .9 068 .3 052 .8 — — - — — ___ — — �TABLE M-8. Unadjusted Exposure Index Analyses for Central Cardiac Function Variables ty Occupation Exposure Index Low Number Percent Medium Number Percent HisJi Number Percent n Abnormal Normal 119 8 111 6.7 93.3 122 8 114 6.6 93.4 109 7 102 64 . 93.6 Overall M vs. L 0.97 ( . 5 2 6 ) 03,.9 H vs. L 0.95 ( . 4 2 6 ) 03,.3 09 .% 090 .6 098 .2 Enlisted Flyer n Abnormal Normal 54 4 50 7.4 92.6 56 4 52 7.1 92.9 53 3 50 5.7 94.3 Overall M vs. L 0 % ( . 3 4 0 ) . 02,.5 H vs. L 0.75 ( . 6 3 5 ) 01,.3 097 .2 090 .6 0.719 n Abnormal Normal 148 5 143 3.4 96.6 151 13 138 86 . 91.4 130 8 122 6.2 93.8 Overall M vs. L H vs. L 2.69 ( . 4 7 7 ) 09,.6 1.88 ( . 0 5 8 ) 06,.8 016 .6 006 .6 020 .8 Officer Heart Sounds Statistic Enlisted Groundcrev Systolic Blood Pressure n Abnormal Normal 119 4 115 3.4 96.6 122 4 118 3.3 96.7 109 5 104 46 . 95.4 Overall H vs. L 0.97 ( . 4 3 9 ) 02,.9 H vs. L 1.38 ( . 6 5 2 ) 03,.9 084 .4a 098 .6 068 .3 Enlisted Flyer n Abnormal Normal 54 1 53 1.8 98.2 56 2 54 3.6 96.4 Enlisted Groundcrew n Abnormal Normal 148 6 142 40 . 96.0 150 2 148 1.3 98.7 53 0 00 . 53 1 0 0 0. 130 7 123 5.4 94.6 Contrast Est. Relative Risk ( 5 C I ) p-Value 9%.. Occupation Officer Variable — — Overall M vs. L 0.32 ( . 6 1 6 ) 00,.1 H vs. L 1.35 ( . 4 4 1 ) 04,.2 — — 0.167 018 .6 0.603 �TABLE H-8. (continued) Ubadjusted Exposure Index Analyses for Central Cardiac Ruction Variables by Occupation Exposure Index Medium Number Percent n Abnormal Normal 119 13 ; 106 10.9 89.1 122 21 101 17.2 82.8 109 14 95 12.8 87.2 Overall M vs. L H vs. L 1.70 ( . 1 3 5 ) 08,.7 1.20 ( . 4 2 6 ) 05,.9 037 .4 0.165 063 .5 Enlisted Flyer n Abnormal Normal 54 9 45 16.7 83.3 56 13 43 23.2 76.8 53 9 44 17.0 8. 30 Overall M vs. L H vs. L 1.51 ( . 9 3 9 ) 05,.0 1.02 ( . 7 2 8 ) 03,.2 064 .1 0.395 098 .6 n Abnormal Normal 148 17 131 11.5 88.5 151 10 141 6.6 93.4 130 15 115 11.5 88.5 Overall M vs. L 0.55 ( . 4 1 2 ) 02,.4 H vs. L 1.01 ( . 8 2 1 ) 04,.0 026 .6 0.147 092 .9 Officer Right Bundle Branch Block Statistic Enlisted Groundcrew EGG (Overall) Occupation Officer Variable Low Number Percent n Abnormal Normal 119 0 00 . 119 100.0 122 1 121 08 . 99.2 109 0 00 . 109 100.0 — 53 0 00 . 53 100.0 — Enlisted Flyer n Abnormal Normal 54 1 53 1.8 98.2 56 00 . 0 56 1 0 0 0. Enlisted Groundcrew n Abnormal Normal 148 1 147 0.7 99.3 151 0 00 . 151 100.0 High Number Percent 130 2 128 Contrast Est. Relative Risk ( 5 C I ) p-Value 92.. — — — 1.5 98.5 — �TABLE M-8. (continued) Uiadjusted Exposure Index Analyses for Central Cardiac Ruction Variables Variable Low Number Percent Exposure Index Median Number Percent High Number Percent Statistic Officer Left Bundle Branch Block Occupation n Abnormal Normal 119 00 . 0 0. 119 1 0 0 . 122 00 . 0 122 1 0 0 0. n Abnormal Normal 54 00 . 0 0. 54 1 0 0 56 00 . 0 0. 56 1 0 0 53 0 00 . 53 1 0 0 0. Est. Relative Risk ( 5 C I ) p-Value 9% . . 109 0 00 . 109 1 0 0 0. Enlisted Flyer Contrast — — 4 Enlisted Groundcrew 148 0 00 . 143 1 0 0 0. 151 0 00 . 151 1 0 0 0. 130 0 00 . 130 100=0 Officer Nonspecific Sr-THave Changes n Abnormal Normal n Abnormal Normal 119 8 111 6.7 93.3 122 15 107 12.3 87.7 109 11 96 10.1 8. 99 Overall M vs. L H vs. L Enlisted Flyer n Abnormal Normal 54 4 50 7.4 92.6 56 11 45 19.6 8. 04 53 7 46 13.2 8. 68 Overall M vs. L 3.06 ( . 1 1 . 8 09,02) H vs. L 1.90 ( . 2 6 9 ) 05,.2 0.171 0.072 0.327 Enlisted Groundcrew n Abnormal Normal 148 15 133 10.1 89.9 151 7 144 46 . 95.4 130 7 123 5.4 94.6 Overall M vs. L 0.43 ( . 7 1 0 ) 01,.9 H vs. L 0.50 ( . 0 1 2 ) 02,.8 016 .2 005 .7 0.150 1.94 ( . 9 4 7 ) 07,.7 1.56 ( . 0 4 0 ) 06,.3 030 .4 0.147 033 .6 �T&BLE M-8. (continued) Unadjusted Exposure Index .Analyses for Central Cardiac Function Variables by Peculation Exposure Index Variable law Number Percent Medium Number Percent High Number Percent Contrast Est. Relative Risk (952C.I.) p-Value 119 7 112 5.9 94.1 122 8 114 66 . 93.4 109 7 102 6.4 93.6 Overall M vs. L 1.12 ( . 9 3 2 ) 03,.0 H vs. L 1.10 ( . 7 3 2 ) 03,-4 095 .7 086 .2 085 .6 Enlisted Flyer n Abnormal Normal 54 1 53 1.8 98.2 56 5 51 8.9 91.1 53 2 51 3.8 96.2 Overall M vs. L H vs. L 0.205 0.139 0.555 n Abnormal Normal 148 9 139 6.1 93.9 151 2 . 1.3 149 98.7 130 4 126 3.1 96.9 Overall M vs. L 0.21 ( . 4 0 9 ) 00,.8 H vs. L 0.49 ( . 5 1 6 ) 01,.3 Officer Tachycardia n Abnormal Normal Enlisted Groundcrew G Statistic Officer Bradycardia Occupation n Abnormal Normal 119 0 00 . 119 100.0 122 0 00 . 122 100.0 109 0 00 . 109 100.0 — — — — Enlisted Flyer n Abnormal Normal 54 0 00 . 54 100.0 56 0 00 . 56 100.0 53 00 . 0 53 100.0 — — — — Enlisted Groundcrew n Abnormal Normal 148 0 00 . 148 100.0 151 00 . 0 151 100.0 130 00 . 0 130 100.0 5.20 ( . 9 4 . 4 05,60) 2.08 ( . 8 2 . 3 01,36) 008 .7 007 .4 026 .4 _ — — — — �TAHE H-8. (continiad) unadjusted Exposure Index Analyses for Central Cardiac Ruction Variables by Occupatic Low Number Percent Exposure Index Medium Number Percent High Nunber Percent Est. Relative Risk (95%C.I.) p-Value n Abnormal Normal 119 14 105 11.8 88.2 122 21 101 17.2 8. 28 109 12 97 11.0 8. 90 Overall M vs. L 1.56 ( . 5 3 2 ) 07,.3 H vs. L 0 9 ( . 1 2 1 ) .3 04,.0 0.311 0.234 087 .5 Enlisted Flyer n Abnormal Normal 54 9 45 16.7 83.3 56 5 51 89 . 91.1 53 4 49 7.6 92.4 M vs. L 0 4 ( . 3 1 5 ) .9 01,.7 01,.2 H vs. L 0.42 ( . 2 1 4 ) 0.265 0.230 0.159 n Abnormal Normal 148 13 135 8.8 91.2 151 9 142 60 . 9. 40 130 10 120 7.7 92.3 M vs. L 0 6 ( . 7 1 5 ) .6 02,.9 H vs. L 0 8 ( . 7 2 0 ) .7 03,.4 065 .4 0.352 0.741 Officer n Abnormal Normal 119 6 113 5.0 9. 50 122 3 119 2.5 97.5 109 1 108 09 . 99.1 Overall .7 01,.4 M vs. L 0 4 ( . 2 1 9 ) H vs. L 0.17 ( . 2 1 4 ) 00,.7 016 .6a 028 .9 0.110 Enlisted Flyer n Abnormal Normal 54 2 52 3.7 96.3 56 0 00 . 56 1 0 0 0. 53 1 52 1.9 98.1 Enlisted Groundcrew Arrhythmia Statistic Enlisted Groundcrew Other BOG Diagnoses Occupation Officer Variable n Abnormal Normal 143 7 141 4.7 95.3 130 7 123 5.4 94.6 a 151 4 147 Overall analysis; sparse cells, chi-square test may not be valid. — Analysis not performed due to sparse cells. 2.6 97.4 Contrast — — Overall M vs. L H vs. L 0.55 ( . 6 1 9 ) 01,.1 1.15 ( . 9 3 3 ) 03,.6 — — 042 .8 037 .4 052 .8 �TABLE H-9. 4 Unadjusted Exposure Index Analyses for Diastolic Blood Pressure, Fuoduscopic Abnormalities, Carotid Bruits, and Manual Pulse Readings by Occupation Exposure Index Lew Number Percent Medium Number Percent High Number Percent Est. Relative Risk (95%C.I.) p-Value n Abnormal Normal 119 8 111 6.7 93.3 122 7 115 5.7 94.3 109 10 99 9.2 9. 08 Overall .4 03,.1 M vs. L 0 8 ( . 0 2 4 ) H vs. L 1.40 ( . 3 3 6 ) 05,.9 0.585 079 .4 047 .9 Enlisted Flyer n Abnormal Normal 54 5 49 9.3 90.7 56 5 51 89 . 91.1 53 6 47 11.3 88.7 Overall M vs. L H vs. L 0 % (.635) . 02,.3 1.25 ( . 6 4 3 ) 03,.8 093 .0 0.952 0.726 n Abnormal Normal 148 11 137 7.4 92.6 151 17 134 11.3 88.7 130 10 120 7.7 92.3 Overall M vs. L H vs. L 1.58 ( . 1 3 5 ) 07,.0 1.04 ( . 3 2 5 ) 04,.3 044 .3 0.258 096 .3 Officer Funduscopic Examination Statistic Misted Groundcrew Diastolic Blood Pressure Occupation Officer Variable n Abnormal Normal 119 2 117 1.7 98.3 122 0 00 . 122 1 0 0 0. 1.8 98.2 — — 0.337a — — Enlisted Flyer n Abnormal Normal 54 1 53 1.8 98.2 56 0 00 . 56 1 0 0 0. 53 0 00 . 53 IflO.O — — Enlisted Groundcrew n Abnormal Normal 148 1 147 — _ — 0.7 99.3 150 Q 150 00 . 100 0. 109 2 107 130 1 129 Contrast Overall 08 . 99.2 �TABLE K-9. Continued) Unadjusted Exposure Index Analyses for Diastolic Blood Pressure, Funduscopic Abnormalities, Carotid Bruits, and Manual Pulse Readings by Occupation Variable Occupation Statistic Officer n Abnormal Normal Low Number Percent 120 3 117 1.7 96.3 Exposure Index Medium Number Percent 122 1 121 08 . 99.2 High Number Percent 109 0 00 . 109 1 0 0 0. Contrast Est. Relative Risk ( 5 C.I.) p-Value 9% Overall — — 038 .8a — — n Abnormal Normal 54 0 00 . 0. 54 1 0 0 56 0 00 . 56 1 0 0 0. 53 2 51 3.8 96.2 — — n Abnormal Normal 148 0 00 . 0. 148 1 0 0 150 0 00 . 150 100.0 130 2 128 — — _ 1.5 98.5 — — — — Officer Radial Pulses Enlisted Flyer Enlisted Groundcrev Carotid Bruits n Abnormal Normal 118 2 116 1.7 98.3 119 0 00 . 119 1 0 0 0. 108 2 106 — — 0.342a — — .. . ,,_ Overall 1.8 98.2 Enlisted Flyer n Abnormal Normal 54 00 . 0 0. 54 1 0 0 55 0 00 . 55 1 0 0 0. 53 0 00 . 53 1 0 0 0. Enlisted Groundcrev n Abnormal Normal 146 0 00 . 146 100.0 149 0 00 . 149 1 0 0 0. 127 00 . 0 127 1 0 0 0. — — — — _ — — — — �TAKE M-9. (continued) Unadjusted Exposure Index Analyses foe Dutstolic Blood Pressure, Funduscopic AbnoociLities, Carotid Bruits, and Manual Pulse Readings by Occupation Exposure Index Variable Low timber Percent n Abnormal Normal 118 1 117 Enlisted Flyer n Abnormal Normal Enlisted Groundcrev Officer . Popliteal Pulses (Manual) Statistic Officer Femoral Pulses (Manual) Occupation Medium Number Percent 0.8 99.2 119 1 118 54 1 53 1.8 98.2 55 2 53 n Abnormal Normal 146 ' 5 141 3.4 96.6 n Abnormal Normal 118 0 00 . 118 1 0 0 0. Bilisted Flyer n Abnormal Normal 54 2 52 Enlisted Groundcrev n Abnormal Normal 146 4 142 3.7 96.3 2.7 97.3 Higjh Number Percent Contrast Overall 08 . 99.2 108 3 105 2.8 97.2 3.6 96.4 53 4 49 7.6 92.4 149 0 00 . 149 1 0 0 0. 127 3 124 2.4 97.6 119 1 118 08 . 99.2 108 1 107 09 . 99.1 55 0 00 . 55 1 0 0 0. 53 3 50 149 2 147 1.3 98.7 127 3 124 Est. Relative Risk ( 5 C I )p-Value 9% .. — — 0.334 Overall M vs. L 2 0 ( . 8 2 . ) 0.575 . 0 01,27 H vs. L 4 3 ( . 7 3 . ) 0.197 . 3 04,98 Overall H vs. L — 6 ( . 6 2 9 ) 0 8 01,.1 . — — Overall 5.7 94.3 2.4 97.6 0.379a — — — — Overall M vs. L 0 4 ( . 9 2 6 ) . 8 00,.8 H vs. L 0 8 ( . 9 3 9 ) .6 01,.2 008 .8 063 .— 0 — — 0.224a — — 0.692a 047 .0 081 .4 �TABLE M-9. (continued) Unadjusted Exposure Index Analyses foe Diastolic Blood Pressure, Funduscopic Abnormalities, Carotid Bruits, and Manual Pulse Readings by Occupation Low Number Percent Exposure Index Medium Number Percent High Number Percent Est. Relative Risk (95%C.I.) p-Value Statistic n Abnormal Normal 118 15 103 12.7 87.3 119 13 106 10.9 89.1 108 8 100 7.4 92.6 Overall 03,.6 M vs. L 0.84 ( . 8 1 8 ) H vs. L 0.55 ( . 2 3 3 ) 02,.5 048 .1 067 .6 0.194 Enlisted Flyer n Abnormal Normal 54 6 48 11.1 88.9 55 9 46 16.4 83.6 53 8 45 15.1 8. 49 Overall M vs. L H vs. L 1.57 ( . 2 4 7 ) 05,.5 1.42 ( . 6 4 4 ) 04,.2 0.716 040 .3 0.542 Enlisted Groundcrew Dorsalis Pedis Pulses (Manual) Occupation Officer Variable n Abnormal Normal 146 17 129 11.6 88.4 149 12 137 80 . 9. 20 127 14 113 11.0 8. 90 Overall M vs. L 0.66 ( . 1 1 4 ) 03,.5 H vs. L 0.94 ( . 4 1 9 ) 04,.9 0.555 0.303 081 .4 n Abnormal Normal 118 1 117 08 . 99.2 119 2 117 1.7 98.3 108 0 00 . 108 1 0 0 0. Enlisted Flyer n Abnormal Normal 54 3 51 5.6 9. 44 55 3 52 Enlisted Groundcrew n Abnormal Normal 146 4 142 2.7 97.3 149 5 144 Officer Posterior Tibial Pulses (Manual) 5.4 94.6 53 4 49 3.4 96.6 127 5 122 Contrast — — — — 7.6 92.4 Overall M vs. L 0.98 ( . 9 5 0 ) 01,.9 H vs. L 1.39 ( . 0 6 5 ) 03,.2 0.879a 094 .8 064 .7 3.9 96.1 Overall M vs. L H vs. L 089 .5 0.757 0.582 1.23 ( . 2 4 6 ) 03,.8 1.45 ( . 8 5 5 ) 03,.4 �TABLE M-9. (continued) unadjusted Exposure Index Analyses for Diastolic Blood Pressure* Funduscopic Abnomalities, Carotid Bruits, and Manual Pulse BRad-jiggf" by Occupation Exposure Index Low Nunber Percent HpfHim High Nunber Percent Nunber Percent Est. Relative Risk ( 5 C I ) p-Value 9% . . n Abnormal Normal 118 15 103 12.7 87.3 119 14 105 11.8 88.2 108 12 96 11.1 8. 90 Overall M vs. L H vs. L 0 9 (.219) . 2 04,.9 08 (.819) .6 03,.3 0.932 086 .2 071 .1 Enlisted Flyer n Abnormal Normal 54 9 45 16.7 83.3 55 11 44 2. 00 8. 00 53 11 42 20.8 79.2 Overall M vs. L H vs. L 1.25 ( . 7 3 3 ) 04,.1 1.31 ( . 9 3 4 ) 04,.8 088 .4 063 .5 059 .8 n Abnormal Normal 146 23 123 15.8 84.2 149 18 131 12.1 87.9 127 18 109 14.2 85.8 Overall M vs. L H vs. L 0.73 ( . 8 1 4 ) 03,.3 08 (.517) .8 04,.2 069 .5 0.363 0.719 Officer Peripheral Pulses (Manual) Statistic Enlisted Groundcrev All Leg Pulses (Manual) Occupation Officer Variable n Abnormal Normal 118 16 102 13.6 86.4 119 14 105 11.8 88.2 108 13 95 12.0 8. 80 Overall M vs. L 0 8 ( . 9 1 8 ) .5 03,.3 H vs. L 0.87 ( . 0 1 9 ) 04,.1 094 .0 067 .9 0.734 Enlisted Flyer n Abnormal Normal 54 9 45 16.7 83.3 55 11 44 2. 00 8. 00 53 11 42 20.8 79.2 Overall M vs. L H vs. L 1.25 ( . 7 3 3 ) 04,.1 1.31 ( . 9 3 4 ) 04,.8 088 .4 063 .5 059 .8 Enlisted Groundcrev n Abnormal Normal 146 23 123 15.8 84.2 149 18 131 12.1 87.9 127 18 109 14.2 8. 58 Overall M vs. L H vs. L 0.73 ( . 8 1 4 ) 03,.3 08 (.517) .8 04,.2 069 .5 0.363 0.719 Contrast �TABLE H-9. (continued) Unadjusted Exposure Index Analyses for Diastolic Blood Pressure, Fuoduscopic Abnornalities, Carotid Bruits, aid Manual Pulse Readings by Occupation Variable Low timber Percent Exposure Index Medium Number Percent High Number Percent Statistic Officer n Abnormal Normal 118 16 102 13.6 8. 64 119 14 105 11.8 88.2 106 13 95 12.0 8. 80 M vs. L 0 8 ( . 9 1 8 ) .5 03,.3 B vs. L 0 8 ( . 0 1 9 ) .7 04,.1 094 .0 067 .9 0.734 Misted Flyer n Abnormal Normal 54 9 45 16.7 83.3 55 11 44 2. 00 8. 00 53 11 42 20.8 79.2 M vs. L H vs. L 1.25 ( . 7 3 3 ) 04,.1 1.31 ( . 9 3 4 ) 04,.8 088 .4 0.653 059 .8 Enlisted Groundcrev All Pulses (Manual) n Abnormal Normal 146 23 123 15.8 84.2 149 18 131 12.1 87.9 127 18 109 14.2 8. 58 M vs. L 0.73 ( . 8 1 4 ) 03,.3 H vs. L 0 8 ( . 5 1 7 ) .8 04,.2 069 .5 0.363 079 .1 a Overall analysis; sparse cells, chi-square test may not be valid. — Analysis not performed due to sparse cells. Contrast Est. Relative Risk ( 5 C.I.) p-Value 9% Occupation �TABLE H-10. Unadjusted Exposure Index Analyses for Peripheral Vascular System Popple** Pulse BparKr^gg by Occupation Occupation Statistic Officer Variable n Abnormal Normal Exposure Index Medium Number Percent High Number Percent 08 . 99.2 122 0 00 . 122 100.0 109 0 0.0 109 100.0 — — 54 0 00 . 54 100.0 56 0 00 . 56 100.0 53 0 0.0 53 100.0 — — — — Low Number Percent 119 I 118 Est. Relative Contrast Risk ( 5 C.I.) 9% p-Value — — n Abnormal Normal n Abnormal Normal 148 1 147 0.7 99.3 151 0 00 . 151 100.0 130 I 129 0.8 99.2 — — — — Officer Femoral Pulses Enlisted Flyer Enlisted Groundcrew Radial Pulses n Abnormal Normal 119 1 118 08 . 99.2 122 0 00 . 122 100.0 109 0 0.0 109 100.0 — — — — — — _ — — Enlisted Flyer n Abnormal Normal Enlisted Groundcrew n Abnormal Normal 54 0 00 . 54 100.0 148 1 147 0.7 99.3 56 1 55 1.8 98.2 53 1 52 151 0 00 . 151 100.0 130 2 128 .— .,_ 1.9 98.1 Overall 1.5 98.5 — — 034 .0a — — �TABLE MHO. (continued) Unadjusted Exposure Index Analyses for Peripheral Vascular Systen Doppler Pulse Readings by Occupation Occupation Statistic Officer Variable n Abnormal Normal Low Number Percent Exposure Index Medium Number Percent Higfc timber Percent 08 . 99.2 122 2 120 1.6 98.4 54 0 00 . 54 100.0 56 2 54 3.6 96.4 53 1 52 Contrast 1.9 98.1 119 1 118 109 00 . 0 109 100.0 Etet. Relative ) Risk (95% C.I. p-Value — — — — — — — — 0.123* — — n Abnormal Normal n Abnormal Normal 148 1 147 0.7 99.3 151 00 . 0 151 100.0 130 3 127 2.3 97.7 Officer Dorsalis Pedis Pulses Enlisted Flyer Enlisted Groundcrew Popliteal Pulses n Abnormal Normal 119 33 86 27.7 72.3 121 35 86 28.9 71.1 109 30 79 27.5 72.5 Overall M vs. L H vs. L 097 .6 1.06 ( . 0 1 8 ) 0 8 1 06,. 6 .4 .6 0.99 ( . 5 1 77) 0 9 8 05,. Enlisted Flyer n Abnormal Normal 54 11 43 20.4 79.6 56 13 43 23.2 76.8 52 13 39 25.0 75.0 Overall M vs. L H vs. L 088 .4 1.18 ( . 8 2 93) 0 7 9 04,. .1 1.30 ( . 2 3 25) 0 5 9 .6 05,. Enlisted Groundcrew n Abnormal Normal 147 32 115 21.8 78.2 151 32 119 21.2 78.8 129 29 100 22.5 77.5 Overall M vs. L H vs. L 097 .6 0.97 ( . 6 1 68) 0 9 4 .0 05,. 1.04 ( . 9 1 ,84) 0 8 9 .8 05,. . Overall — — �TABLE M-1D. (continued) Unadjusted Exposure Index Analyses for Vascular Systen Doppler Pulse BptytiT^gg by Qccufiaticn Posterior Tibial Pulses Occupation Statistic Officer Variable n Abnormal Normal Low Number Percent Exposure Index Medium Number Percent 1.7 98.3 121 5 116 54 0 00 . 54 100.0 56 3 53 119 2 117 High. Number Percent 4.1 95.9 109 2 107 5.4 9. 46 52 2 50 Contrast Est. Relative Risk ( 5 C I ) p-Value 9% . . Overall M vs. L H vs. L 040 .1 2.52 ( . 8 1 . ) 0 2 6 04,33 .7 1.09 ( . 5 7 9 ) 0 9 8 01,.0 .2 3.8 96.2 1.8 98.2 — — 0.248a — — 0.193* — — — — n Abnormal Normal Enlisted Groundcrew n Abnormal Normal 147 2 145 1.4 98.6 151 0 00 . 151 1 0 0 0. 130 3 127 2.3 97.7 Officer Leg Pulses Enlisted Flyer n Abnormal Normal 119 33 86 27.7 72.3 121 38 83 31.4 68.6 109 32 77 29.4 70.6 Overall M vs. L H vs. L 082 .2 1.19 ( . 8 2 0 ) 0.535 06,.8 1.08 ( . 1 1 9 ) 0 7 7 06,.3 .8 Enlisted Flyer n Abnormal Normal 54 11 43 20.4 79.6 56 15 41 26.8 73.2 52 14 38 26.9 73.1 Overall M vs. L H vs. L 066 .6 1.43 ( . 8 3 4 ) 0 4 0 05,.8 .3 1.44 ( . 8 3 5 ) 0 4 0 05,.5 .3 Enlisted Groundcrew n Abnormal Normal 147 33 114 22.4 77.8 151 32 119 21.2 78.8 129 29 100 22.5 77.5 Overall M vs. L H vs. L 095 .5 09 (.416) 075 .3 05,.1 .9 1.00 ( . 7 1 7 ) 1.000 05,.7 Overall Overall �TAHE tt-10. (cmtinued) ttiadjusted Exposure Index Analyses for Peripheral Vascular Systen Doppler Pulse Readings by Occupation Low Variable Number Percent Exposure Index Medium Number Percent High Number Percent Statistic Officer n Abnormal Normal 119 33 86 27.7 72.3 121 38 83 31.4 68.6 109 32 77 29.4 70.6 Overall M vs. L H vs. L 082 .2 1.19 ( . 8 2 0 ) 0.535 . 06,.8 1.08 ( . 1 1 9 ) 0 7 7 06,.3 .8 Enlisted Flyer n Abnormal Normal 54 11 43 20.4 79.6 56 15 41 26.8 73.2 52 14 38 26.9 73.1 Overall M vs. L H vs. L 066 .6 1.43 ( . 8 3 4 ) 0 4 0 05,.8 . 3 1.44 ( . 8 3 5 ) 0 4 0 05,.5 .3 Enlisted Groundcrew n Abnormal Normal 147 34 113 23.1 76.9 151 32 119 21.2 78.8 129 30 99 23.3 76.7 Overall M vs. L H vs. L 084 .9 08 (.215) 069 .9 05,.4 .8 1.01 ( . 8 1 7 ) 6 9 4 05,.6 .8 Officer n Abnormal Normal 119 33 86 27.7 72.3 121 38 83 31.4 68.6 109 32 77 29.4 70.6 Overall M vs. L H vs. L 082 .2 1.19 ( . 8 2 0 ) 0 5 5 06,.8 . 3 1.08 ( . 1 1 9 ) 0 7 7 06,.3 . 8 Enlisted Flyer n Abnormal Normal 54 11 43 20.4 79.6 56 15 41 26.8 73.2 52 14 38 26.9 73.1 Overall M vs. L H vs. L 066 .6 1.43 ( . 8 3 4 ) 0 4 0 05,.8 . 3 1.44 ( . 8 3 5 ) 0 4 0 05,.5 . 3 Enlisted Groundcrew Peripheral Pulses n Abnormal Normal 147 34 113 23.1 76.9 151 32 119 21.2 78.8 129 30 99 23.3 76.7 Overall M vs. L H vs. L 084 .9 08 (.215) 069 .9 05,.4 .8 1.01 ( . 3 1 7 ) 0 9 4 03,.6 .8 _ All Pulses a Overall analysis; sparse cells, chi-square test may not be valid. — Analysis not performed due to sparse cells. Contrast Est. Relative Risk (95%C.I.) p-Value Occupation �TABLE H-ll. Association Between Central and Peripheral Abnormalities and Verified Heart Disease in the Combined Ranch Hand and Comparison Groups Variable Stratification Total .Hypertension Essential, Percent Percent p-Value Abnormal Abnormal Heart Disease* Percent p-Value Abnormal Myocardial Infarction p-Value Systolic Pressure 2,002 145 17.5 61.4 <0.001 21.7 24.1 0.556 1.0 2.1 0.386 Diastolic Pressure s i NJ <140 XL40 <90 >90 1,966 180 14.8 81.7 <0.001 22.4 15.6 0.043 1.1 0.6 0.790 ECG Normal Abnormal 1,858 290 18.8 30.7 <0.001 17.4 50.3 <0.00i 0.3 5.9 <0.001 Heart Sounds Normal Abnormal 2,084 63 20.1 31.8 0.036 21.4 34.9 0.017 0.9 6.4 <0.001 Funduscopic Examination Normal Abnormal 2,134 13 20.2 53.8 0.008 21.6 53.8 0.014 1.0 0.0 0.999 Carotid Bruits Normal Abnormal 2,132 14 20.3 42.9 0.080 21.8 28.6 0.775 1.0 0.0 0.999 Peripheral Pulses (Manual) Normal Abnormal 1,806 314 20.0 21.7 0.547 21.5 22.6 0.725 0.9 1.6 0.391 Peripheral Pulses (Doppler) Normal Abnormal 1,611 529 20.6 19.8 0.754 20.4 26.1 0.007 0.9 1.5 0.306 Ul *Excluding hypertension. �TABLE M-12. Unadjusted Analyses for Reported'and Verified Heart Disease by Group (Original Comparisons Only) Group Variable Statistic Ranch Hand Number Percent Original Comparison Number Percent Est. Relative Risk (95% C.I.) p-Value Reported Essential Hypertension 942 247 695 26.2 73.8 888 231 657 26.0 74.0 1.01 (0.82,1.24) 0.920 Verified Essential Hypertension a: to n Yes No n Yes No 942 195 747 20.7 79.3 888 184 704 20.7 79.3 1.00 (0.80,1.25) 0.992 Reported Heart Disease (Excluding Hy pe r t ens ion) n Yes No 942 265 677 28.1 71.9 888 234 654 26.4 73.6 1.09 (0.89,1.34) 0.390 Verified Heart Disease (Excluding Hypertension) n Yes No 942 224 718 23.8 76.2 888 180 708 20.3 79.7 1.23 (0.98,1.53) 0.070 Reported Myocardial Infarction n Yes No 942 20 922 2.1 97.9 888 18 870 2.0 98.0 1.05 (0.55,1.99) 0.889 Verified Hyocardial Infarction n Yes No 942 9 933 1.0 99.0 888 11 877 1.2 98.8 0.77 (0.32,1.87) 0.562 �TABLE M-13. Adjusted Analyses for Reported and Verified Heart Disease (Original Comparisons Only) Adj. Relative Risk (95% C.I.) p-Value Covariate Remarks* Reported Essential Hypertension 1.12 (0.90,1.40) 0.312 AGE (p<0.001), CSMOK (p<0.001), CHOL (p<0.001), %BFAT (p<0.001), ALC (p=0.003) Verified Essential Hypertension 1.10 (0.86,1.39) 0.447 CSMOK (p=0.015), CHOL (p<0.001), %BFAT (p<0.001), PS (p=0.024), ALC (p=0.028) Reported Heart 1.14 (0.92,1.40) Disease 0.230 AGE (p<0.001), RACE (p=0.046) Verified Heart 1.28 (1.02,1.60) Disease 0.032 AGE (p<0.001) Reported Myocardial Infarction 1.16 (0.58,2.32) 0.682 AGE (p<0.001), OCC (p=0.005), CHOL (p»0.040) Verified Myocardial Infarction 0.90 (0.36,2.24) 0.810 AGE (p<0.001), CHOL (p=0.021) Variable Abbreviations; CSMOK: Current smoking CHOL: Cholesterol 2BFAT: Percent body fat ALC: Current alcohol use (drinks/day) PS: Personality score OCC: Occupation M-27 �TABLE H-14. Unadjusted Analyses for Central Cardiac Function by Group (Diabetics Excluded) (Original Comparisons Only) Group Variable Statistic Ranch Hand Number Percent Original Comparison Number Percent Est. Relative Risk (95% C.I.) p-Value Systolic Blood Pressure n Abnormal Normal 942 60 882 6.4 93.6 888 67 821 7.6 92.4 0.83 (0.58,1.20) 0.322 Heart Sounds n Abnormal Normal 941 31 910 3.3 96.7 888 27 861 3.0 97.0 1.09 (0.64,1.84) 0.757 n Abnormal Normal 942 121 821 12.8 87.2 888 125 763 14.1 85.9 0.90 (0.69,1.18) 0.441 n Abnormal Normal 942 5 937 0.5 99.5 888 6 882 0.7 99.3 0.79 (0.24,2.58) 0.689 n Abnormal Normal 942 0 942 0.0 100.0 888 0 888 0.0 100.0 — — n Abnormal Normal 942 85 857 9.0 91.0 888 78 810 8.8 91.2 1.03 (0.75,1.42) 0.857 i CO ECG (Overall) ECG: RBBB ECG: LBBB . ECG: Nonspecific T-Have Changes �TABLE M-14. (continued) Unadjusted Analyses for Central Cardiac Function by Group (Diabetics Excluded) (Original Comparisons Only) Group Ranch Hand Number Percent Original Comparison Number Percent Est. Relative Risk (95% C.I.) p- Value 1.12 (0.72,1.75) 0.610 Variable Statistic ECG: Bradycardia n Abnormal Normal 942 45 897 4.8 95.2 888 38 850 4.3 95.7 n Abnormal Normal 942 0 942 0.0 100.0 888 0 888 0.0 100.0 n Abnormal Normal 942 31 911 3.3 96.7 888 32 856 3.6 96.4 0.91 (0.55,1.50) 0.711 n Abnormal Normal 942 97 845 10.3 89.7 888 92 796 10.4 89.6 0.99 (0.74,1.34) 0.968 ECG: Tachycardia KJ SO ECG: Arrhythmia ECG: Other Diagnoses * —No relative risk, given, since no abnormals are present. — — �TABLE M-15. Adjusted Analyses for Central Cardiac Function (Diabetics Excluded)* (Original Comparisons Only) Adj. Relative Risk (95% C.I.) p-Value Systolic Blood 0.95 (0.65,1.38) Pressure (Discrete) 0.772 Variable Covariate Remarks** AGE (p<0.001) CHOL (p=0.001) %BFAT (p<0.001) PS (p=0.001) Systolic Blood -0.152 (-1.401,1.097)* 0.811" Pressure (Continuous) AGE*OCC (p=0.018) CSMOK (p<0.001) CHOL (p<0.001) %BFAT (p<0.001) PS (p=0.001) ALC (p<0.001) Heart Sounds AGE (p<0.001) RACE (p=0.002) OCC (p»0.048) CHOL/HDL (p<0.001) ECG (Overall) ECG: 1.26 (0.73,2.17) **** 0.407 **** AGE (p<0.001) RACE (p=0.023) %BFAT (p<0.001) GRP*PACKYR (p=0.034) RBBB 0.85 (0.26,2.81) 0.787 AGE (p»0.015) ECG: Nonspecific ST-T-Wave Changes 1.20 (0.85,1.69) 0.294 AGE (p<0.001) RACE (p=0.033) CHOL (p-0.009) JSBFAT (p<0.001) ECG: Bradycardia 1.16 (0.74,1.82) 0.509 OCC (p=0.019) CHOL/HDL (p<0.001) M-30 �TABLE M-15. (continued) Adjusted Analyses for Central Cardiac Function (Diabetics Excluded)* (Original Comparisons Only) Adj. Relative Risk (95% C.I.) p-Value Covariate Remarks** EGG: Arrhythmia **** **** AGE (p<0.001) OCC (p<0.001) GRP*PACKYR (p=0.034) GRP*%BFAT (p=0.016) ECG: Other Diagnoses **** **** AGE (p<0.001) GRP*OCC (p=0.036) Variable *Some adjusted analyses did not explore effects of all covariates due to sparse number of abnormalities (see text). **Additional Abbreviations; CHOL/HDL: cholesterol to HDL ratio GRP: group PACKYR: pack-years of smoking ****Group-by-covariate interaction—relative risk/difference in group means, 95% confidence interval, and p-value not presented. "Difference in group means (Ranch Hand-Original Comparison) and associated p-value given, rather than relative risk, for continuous analysis of dependent variables. M-31 �TABLE M-16. Summary of Group-by-Covariate Interactions for Central Cardiac Function Variables (Diabetics Excluded) (Original Comparisons Only) Variable Interaction Stratification Adj. Relative Risk (95% C.I.) p-Value ECG (Overall) 0.76 (0.53,1.10) 1.24 (0.87,1.78) 0.142 0.234 Group-by-Pack- Years Smoking (21% Body Fat) Pack- Years: 0 Pack- Years: 30 0.59 (0.30,1.18) 1.55 (0.72,3.33) 0.134 0.263 Group-by-Percen t Body Fat (7 Pack- Years Smoking) to Pack- Years: 0 Pack- Years: 30 ECG: Arrhythmia 3C Group-by-Pack- Years Smoking 10% Body Fat 30% Body Fat 0.17 (0.05,0.62) 2.46 (0.76,7.99) 0.007 0.136 Officer Enlisted Flyer Enlisted Groundcrew 1.73 (1.06,2.85) 0.75 (0.38,1.47) 0.73 (0.44,1.22) 0.030 0.401 0.234 ECG: Other Diagnoses Group-by-Occupation �TABLE M-17. Unadjusted Analyses for Peripheral Vascular Function by Group (Diabetics Excluded) (Original Comparisons Only) Group Variable Statistic Ranch Hand Number Percent Original Comparison Number Percent Est. Relative Risk (95% C.I.) p- Value Diastolic Blood Pressure 942 79 863 8.4 91.6 887 76 811 8.6 91.4 0.98 (0.70,1.36) 0.889 Funduscopic Examination 3. i 0 0 u> n Abnormal Normal n Abnormal Normal 941 7 934 0.7 99.3 888 6 882 0.7 99.3 1.10 (0.37,3.29) 0.865 Carotid Bruits n Abnormal Normal 941 7 934 0.7 99.3 887 6 881 0.7 99.3 1.10 (0.37,3.28) 0.865 Radial Pulses (Manual) n Abnormal Normal 929 4 925 0.4 99.6 876 7 869 0.8 99.2 0.54 ( . 6 1 8 ) 01,.4 0.322 Radial Pulses (Doppler) n Abnormal Normal 942 3 939 0.3 99.7 886 4 882 0.4 99.6 0.70 (0.16,3.16) 0.646 Femoral Pulses (Manual) n Abnormal Normal 929 20 909 2.2 97.8 876 21 855 2.4 97.6 0.90 (0.48,1.66) 0.726 �TABLE M-17. (continued) Unadjusted Analyses for Peripheral Vascular Function by Group (Diabetics Excluded) (Original Comparisons Only) Group Variable Statistic Ranch Hand Number Percent Original Comparison Number Percent Est. Relative Risk (95% C.I.) p-Value Femoral Pulses (Doppler) n Abnormal Normal 942 6 936 0.6 99.4 887 3 884 0.3 99.7 1.89 (0.47,7.57) 0.368 Popliteal Pulses (Manual) n Abnormal Normal 929 16 913 1.7 98.3 876 21 855 2.4 97.6 0.71 (0.37,1.38) 0.312 Popliteal Pulses (Doppler) n Abnormal Normal 942 10 932 1.1 98.9 887 6 881 0.7 99.3 1.58 (0.57,4.35) 0.379 n Dorsalis Pedis Pulses Abnormal (Manual) Normal 929 102 827 11.0 89.0 876 95 781 10.8 89.2 1.01 (0.71,1.45) 0.928 n Dorsalis Pedis Pulses Abnormal (Doppler) Normal 938 228 710 24.3 75.7 885 203 682 22.9 77.1 1.08 (0.87,1.34) 0.490 Posterior n Tibial Pulses Abnormal (Manual) Normal 929 27 902 2.9 97.1 876 24 852 2.7 97.3 1.06 (0.61,1.85) 0.834 Posterior n Tibial Pulses Abnormal (Doppler) Normal 939 19 920 2.0 98.0 885 18 867 2.0 98.0 0.99 (0.52,1.91) 0.984 �TABLE M-17. (continued) Unadjusted Analyses for Peripheral Vascular Function by Group (Diabetics Excluded) (Original Comparisons Only) Group Variable Statistic Ranch Hand Number Percent Original Comparison Number Percent Est. Relative Risk (95* C.I.) p-Value Leg Pulses (Manual) n Abnormal Normal 929 131 798 14.1 85.9 876 134 742 15.3 84.7 0.91 (0.70,1.18) 0.472 Leg Pulses (Doppler) n Abnormal Normal 938 237 701 25.3 74.7 885 212 673 23.9 76.1 1.07 (0.87,1.33) 0.516 Peripheral Pulses (Manual) n Abnormal Normal 929 133 796 14.3 85.7 876 138 738 15.8 84.2 0.89 (0.69,1.16) 0.395 Peripheral Pulses (Doppler) n Abnormal Normal 938 239 699 25.5 74.5 885 214 671 24.2 75.8 1.07 (0.87,1.33) 0.522 All Pulses (Manual) n Abnormal Normal 929 133 796 14.3 85.7 876 138 738 15.8 84.2 0.89 (0.69,1.16) 0.395 All Pulses (Doppler) n Abnormal Normal 938 239 699 25.5 74.5 884 214 670 24.2 75.8 1.07 (0.86,1.32) 0.529 Ul �TABLE H-1B. Adjusted Analyses for Peripheral Vascular Ruction (Diabetics Excluded)* (Original Comparisons Only) Variable Statistical/Clinical Analysis Adj. Relative Risk (95% C.I.) p-Value Covariate Remarks** Discrete 1.04 ( . 4 1.46) 07, 0.818 CHOL ( < . 0 ) p001 %BFAT (p«O.C01) Continuous 0.250 (-0.530, 1.030)a 0.537* AGE (p=0.016) CSMOK ( < . 0 ) p001 CHOL ( < . 0 ) p001 %BFAT (p<0.001) ALC ( = . 0 ) p004 Rmduscopic Examination 1.31 (0.42,4.09) 0.646 AGE ( < . 0 ) p001 RACE ( < . 0 ) p001 CHOL/HDL ( = . 1 ) p000 ALC (p=0.032) Carotid Bruits 0.95 ( . 0 2 9 ) 03,.7 098 .2 AGE (p=0.018) DRK2R ( < . 0 ) p001 Diastolic Blood Pressure Radial Pulses Manual Doppler 0.57 ( . 7 1 % 01,.) 07 (.631) .0 01,.6b 0.373 066 .4b AGE (p=0.047) Femoral Pulses Manual 1.20 ( . 0 2 3 ) 06,.6 0.710 AGE (p<0.001) OCC (p=0.050) CHOL/HDL (p=0.043) %BFAT (p<0.001) DIFOCRT (p=O.C05) Doppler 1.87 (0.45,7.74) 0.384 AGE (p=0.001) CSMCK (p=0.001) CHOL/HDL (p=0.032) Popliteal Pulses AGE (p=O.C02) RACE (p*Q.C02) PACm (p=0.006) CHOL/HDL (p=Q.021) GRP*OCC (p=0.048) rCrfrfft Doppler Dorsalis Pedis Pulses 1.66 (0.59,4.72) 0.337 AGE (p<D.C01) RACE (p=0.015) CSMOK (p<0.001) Manual 1.06 (0.78,1.44) 0.704 AGE (p=O.C04) Doppler 1.07 ( . 6 1 3 ) 08,.3 0.569 AGE ( = . 3 ) p006 RACE ( 4 . 0 ) p)07 %BFAT ( = . 1 ) p000 M-36 �TABLE M-18. (continued) Adjusted Analyses for Peripheral Vascular Function (Diabetics Excluded)* (Original Comparisons Chly) Variable Posterior Tibial Pulses Statistical/Clinical Analysis Manual Doppler Leg Pulses **** **** AGE (p=0.003) RACE (p<D.001) PACKXR ( = . 1 ) p001' GRP*OCC (p=0.026) 0.849 AGE (p<0.001) RACE (p<0.001) CSMCK (p=0.010) CHOL/HDL (p=0.002) **** AGE (p=O.C01) GRP*OCC (p=0.038) GRP**BFAT 1.07 (0.54,2.13) 1.07 ( . 6 1 3 ) 08,.3 1.06 ( . 6 1 3 ) 08,.2 0.575 **** Manual Doppler 0.516 **** Manual Doppler All Pulses p-Value Manual Doppler Peripheral Pulses Adj. Relative Risk (95% C.I.) 1.06 ( . 6 1 3 ) 08,.2 0.582 Covariate Remarks** AGE (p=0.016) RACE (p=0.024) 2BFAT (p=0.034) AGE (pO.OOl) GRP*%BFAT (p=0.036) AGE (p=£.028) RACE (p=0.042) 2BFAT AGE (p<0.001) GRP*%BFAT (p=0.036) AGE (p=0.027) RACE (p=0.042) %BFAT (p=0.038) *Some adjusted analyses did not explore effects of all covariates due to sparse number of abnormalities (see text). **Additional Abbreviations; DRKBR: Drink-years of alcohol DIFCORT: Differential cortisol "Difference in group means (Ranch Hand - Original Comparison) and associated p-value given, rather than relative risk, continuous analysis of dependent variables. Unadjusted for any covariates—same as for unadjusted results. —No covariates significant. ****Group-by-covariate interaction—relative risk, 95% confidence interval, and p-value not presented. M-37 �TABLE H-19. Summary of Group-by-Covariate Interactions for Peripheral Vascular Function Variables (Diabetics Excluded) (Original Comparisons Only) Variable Interaction Stratification Adj. Relative Risk (95% C.I.) p-Value Popliteal Pulses (Manual) Officer Enlisted Flyer Enlisted Groundcrev 0.21 (0.02,2.54) 4.68 (0.53,41.1) 0.99 (0.38,2.55) 0.219 0.165 0.976 Posterior Tibial Pulses (Manual) Group-by-Occupation Officer Enlisted Flyer Enlisted Groundcrew 0.31 (0.05,1.82) 3.68 (0.96,14.1) 1.27 (0.56,2.87) 0.197 0.057 0.575 Leg Pulses (Manual) 3: i Group-by-Occupation Group-by-occupation (21% Body Fat) Officer Enlisted Flyer Enlisted Groundcrew 0.65 (0.41,1.01) 1.52 (0.83,2.77) 1.23 (0.81,1.87) 0.054 0.174 0.327 Peripheral Pulses (Manual) Group-by-Percent Body Fat 10% Body Fat 30% Body Fat 1.87 (0.94,3.72) 0.58 (0.34,0.98) 0.073 0.042 All Pulses (Manual) Group-by-percent Body Fat 10% Body Fat 30% Body Fat 1.87 (0.94,3.72) 0.58 (0.34,0.98) 0.073 0.042 CO �TABLE M-20. Longitudinal Analysis of Pulse Index and Overall EGG: A Contrast of Baseline and First Follovup Examination Abnormalities (Original Comparisons Only) Variable Group 1982 Baseline Exam 1985 Followup Exam Odds* Ratio p-Value (OR) (ORRH vs ORQC) Abnormal Normal Ranch Hand Abnormal Normal 50 104 72 743 1.44 Original Comparison Abnormal Normal 35 115 51 668 2.25 Ranch Hand Abnormal Normal 86 43 192 650 0.22 Original Comparison All Pulses (Manual) Abnormal Normal 83 44 147 598 0.30 0.05 ECG • (Overall) 0.23 Number Normal Baseline, Abnormal Follovup *OHH<3 Rat-In? . Number Abnormal Baseline, Normal Followup M-39 �APPENDIX N Hematological Evaluation �APPENDIX N: Hematological Evaluation Contents Table N-l Summary Statistics for Hematological Covariates by Group • N-l N-2 Summary of Group-by-Covariate Interactions for Hematological Variables N-2 N-3 Interaction Summaries for Adjusted Continuous Exposure Index Analyses for Hematological Variables N-4 N-4 Unadjusted Categorical Analyses for Hematological Variables by Group (Original Comparisons Only) N-6 N-5 Unadjusted Continuous Analyses for Hematological Variables (Contrast of Group Means) (Original Comparisons Only) N-7 N-6 Adjusted Categorical Analyses for Hematological Variables (Abnormal Versus Normal, Adjusted for Age, Race, Occupation, and Smoking) (Original Comparisons Only) N-8 N-7 Adjusted Continuous Analyses for Hematological Variables (Ranch Hand-Original Comparison Group Differences) (Original Comparisons Only) N-9 N-8 Summary of Group-by-Covariate Interactions for Hematological Variables (Original Comparisons Only) N-10 N-9 Longitudinal Analyses for MCV, MCH, and PLT: A Contrast of Baseline and First Followup Examination Test Means (Original Comparisons Only) N-15 N-i �TABLE N-l. Summary Statistics for Hematological Covariates by Group Group Covariate Covariate Category Ranch Hand Comparison Percent Percent Race Black Nonblack 5.9 94 .1 6.4 93.6 0.679 Occupation Officer Enlisted Flyer Enlisted Groundcrew 37 .4 17 .5 45 .2 37.5 16.2 46.3 0.724 Mean±SE Mean±SE p-Value Age (At Baseline) — 43.94±0.24 43.86±0.22 0.804 Current Smoking* — 10.75±0.48 9.13±0.40 Smoking History (Pack-Years) — 13.24±0.52 12.98±0.47 0.708 "Equivalent cigarettes per day. —Covariate not categorized for these results. N-l 0.010 �TABLE N-2. Summary of Group-by-Covariate Interactions for Heoatological Variables RBC Occupation Smoking History Ranch Hand-Comparison Group Difference ± SB p-Value - . 0 + 0.030 009 -0.014 ± 0.035 0.775 0.686 -0.066 ± 0.053 -0.001 ± 0 0 6 .4 0.212 0.975 Enlisted Groundcrew Nonsmoker 30 Pack-Years Variable Interaction Covariates -0.010 + 0.029 -.9 ± 006 003 .3 0.736 000 .1 Nonblack Nonsmoker 35 53 -0.010 + 0.020 0 0 2 ± 0.021 .0 069 .0 0.942 0.115 + 0.031 -0.017 ± 0.019 <0.001 0.369 RH>C -0.550 + 0.184 -0.143 ± 0.186 0.003 0.446 ORH Enlisted Flyer 35 53 -0.341 ± 0.172 0.067 ± 0.134 0.050 0.618 ORH Enlisted Groundcrew 35 53 -0.121 ± 0.074 0.286 ± 0.149 0.104 0.057 Stratification Direction of Result* Officer Nonsmoker 30 Pack- Years Enlisted Flyer Nonsmoker 30 Pack- Years z: i to VBC* Race Smoking History ORH Occupation Age (At Baseline) 30 Pack- Years 35 53 Black Officer 35 53 RH>C �TABLE N-2. (continued) Summary of Group-by-Covariate Interactions for Hematological Variables Variable PLTa Interaction Covariates Race Stratification Ranch Hand-Comparison Group Difference ± SE Direction p-Value of Result* 0.003 ± 0.012 0.789 0.103 ± 0.042 0.014 Black Nonsmoker 30 Pack-Years, Currently 1 Pack/Day Smoking (Smoking History and Current Smoking) Nonblack Nonsmoker 30 Pack-Years, Currently 1 Pack/Day -0.050 ± 0.050 0.317 0.258 ± 0.096 0.007 Variable mean in Ranch Hand group greater than variable mean in Comparison group. ORE: Variable mean in Comparison group greater than variable mean in Ranch Hand group. a Units are on log scale. RH>C RH>C �TABLE N-3. Interaction Summaries for Adjusted Continuous Exposure Index Analyses for Hematological Variables . Interaction Variable (Occupation) Stratification Contrast Exposure Index Level Difference ± SE p-Value Direction of Result* HGB HGB HCT 0.014 + 0.295 -0.180 ± 0.372 0.962 0.630 M vs. L H vs. L -0.047 + 0.402 0.516 ± 0.502 0.908 0.304 M vs. L H vs. L -0.119 + 0.128 0.261 ± 0.151 0.356 0.086 H > L M vs. L H vs. L 0.033 + 0.102 -0.172 ± 0.102 0.748 0.093 L > H Nonblack M vs. L H vs. L -0.007 + 0.122 0.060 ± 0.123 0.955 0.626 Black RBC M vs. L H vs. L Black RBC Nonblack M vs. L H vs. L 0.089 + 0.874 2.227 ± 0.960 0.919 0.021 Exposure Index35 by-Age (At Baseline) (Enlisted Flyer) 53 M vs. L H vs. L -0.854 + 0.389 0.626 ± 0.457 0.029 0.172 M vs. L H vs. L 0.425 + 0.308 -0.339 ± 0.308 0.169 0.272 Exposure Indexby-Race (Officer) Nonblack M vs. L H vs. L -0.093 + 0.359 0.206 ± 0.362 0.795 0.569 Black M vs. L -0.849 + 2.569 5.611 ± 2.823 0.741 0'4 .08 Exposure Indexby-Race (Officer) Exposure Index35 by-Age (At Baseline) (Enlisted Flyer) 53 Exposure Indexby-Race (Officer) H vs. L H > L L > M H > L �TABLE N-3. (continued) Interaction Summaries for Adjusted Continuous Exposure Index Analyses for Hematological Variables Variable Stratification Contrast Exposure Index Level Difference ± SE p- Value Exposure Index 35 by-Age (At Baseline) (Enlisted Flyer) 53 PLT* PUT* M vs. L H vs. L -2.332 ± 1.141 1.921 ± 1.341 0.042 0.154 M vs. L H vs. L 1.074 + 0.904 -0.887 ± 0 9 4 .0 Exposure Indexby -Current Smoking (Packs per Day) (Enlisted Flyer) 0 M vs. L H vs. L 0.097 ± 0.054 -0.005 ± 0.053 0.075 0.932 1 M vs. L H vs. L -0.015 ± 0.042 -0.062 ± 0.045 0.710 0.169 M vs. L H vs. L -0.128 ± 0 0 8 .6 -0.118 ± 0.078 0.064 0.132 Nonblack M vs. L H vs. L -0.015 ± 0.025 -0.004 ± 0.026 0.557 0.885 Black M vs. L H vs. L 0.156 ± 0.079 -0.036 ± 0.081 0.049 0.659 Exposure Indexby-Race (Enlisted Groundcrew) Variable Variable Variable Variable Units are in log *H L L M > > > > L: H: M: L: Direction of Result* 0.236 0.328 2 HCT z i Interaction (Occupation) mean in mean in mean in mean in scale. L >M M > L L >H M > L high exposure index category greater than variable mean in low exposure index category. low exposure index category greater than variable mean in high exposure index category. low exposure index category greater than variable mean in medium exposure index category medium exposure index category greater than variable mean in low exposure index category �TABLE N-4. Unadjusted Categorical Analyses for Hanatological Variables by Group (Original Comparisons Only) Abnormally Low Variable Group RBC WBC HGB HOT MCV MCH MCHC PIT Normal Number Percent Number Percent Abnormally High Number Percent Total p-Valuea Ranch Hand 30 Original Comparison 36 3.0 976 96.2 8 0.8 1,014 3.8 910 95.4 8 0.8 954 Ranch Hand 45 Original Comparison 47 4.4 906 89.4 62 6.1 1,013 4.9 857 89.8 50 5.2 954 3.8 752 74.2 . 223 22.0 1,014 3.6 717 75.2 203 21.3 954 Ranch Hand 11 Original Comparison 13 1.1 1,001 98.7 2 0.2 1,014 1.4 938 98.3 3 0.3 954 Ranch Hand 10 Original Comparison 11 1.0 857 84.5 147 14.5 1,014 1.2 809 84.8 134 14.0 954 Ranch Hand Original Comparison 7 0.7 943 93.0 64 6.3 1,014 7 0.7 890 93.3 57 6.0 954 1 0.1 1,013 99.9 0 0.0 1,014 0 0.0 954 100 0. 0 0.0 954 5 0.5 987 97.4 21 2.1 1,013 2 0.2 935 98.0 17 1.8 954 Ranch Hand 39 Original Comparison 34 Ranch Hand Original Comparison Ranch Hand Original Comparison 0.598 0.631 0.867 0.722 0.904 0.947 0.760 a Chi-square test, 2 d.f., except for HCT and PLT, which were obtained from continuity adjusted chi-square tests on 1 d.f. (Abnormally high category pooled with normal, and abnormally low category pooled with normal for HCT and PLT, respectively.) —Only one abnormal MCHC value; p-value not given. N-6 �TABLE N-5. Unadjusted Continuous Analyses for Hematological Variables (Contrast of Group Means) (Original Comparisons Only) Group Mean±SE Variable Ranch Hand Original Comparison Difference ±SE RBC 4.964±0.012 4.974±0.012 WBC1 7.003 6.844 t-Statistic p-Value -0.010±0.017 -0.61 0.540 1.78 0.076 0.50 0.619 HGB 15.624±0.033 15.600±0.034 — 0.024±0.047 HCT 45.904±0.097 45.874±0.097 0.030±0.137 0.22 0.827 MCV 92.596±0.150 92.355+0.157 0.241±0.217 1.11 0.268 MCH 31.544±0.055 31.435±0.058 0.108±0.080 1.35 0.177 MCHC 34.040±0.021 34.009±0.613 0.030±0.029 1.06 0.291 1.39 0.165 PLT1 265.2 261.7 — 1 Means transformed from log scale. —Difference and standard errors (SE) not presented, since variables were analyzed on logarithmic scale. N-7 �TABLE N-6. Adjusted Categorical Analyses for Hematological Variables (Abnormal Versus Normal, Adjusted for Age, Race, Occupation, and Smoking) (Original Comparisons Only) Abnormally Low vs. Normal Abnormally High vs. Normal Variable Adj. Relative Risk (95% C.I.) p-Value Adj. Relative Risk (95% C.I.) p-Value RBC 0.81 (0.51,1.30) 0.380 0.95 (0.41,2.17) 0.898 WBC 0.95 (0.63,1.43) 0.730 1.16 (0.80,1.70) 0.431 HGB 1.10 (0.70,1.72) 0.692 1.05 (0.85,1.30) 0.662 HCT 0.84 (0.41,1.71)* 0.628a MCV Nonblack Black 1.92 (0.64,5.76) 0.32 (0.09,1.11) 0.242 0.072 1.09 (0.84,1.42) 0.41 (0.12,1.34) 0.503 0.139 MCH 0.96 (0.40,2.30) 0.920 1.06 (0.74,1.52) 0.734 PLT Sparse Data 1.13 (0.63,2.05)b 0.682b "Abnormally low versus normal/abnormally high. 'Abnormally high versus normal/abnormally low. N-8 Sparse Data �TABLE N-7. Adjusted Continuous Analyses for Hematological Variables (Ranch Hand-Original Comparison Group Differences) (Original Comparisons Only) Variable Ranch Hand-Original Comparison Group Difference ±SE p-Value Covariate Remarks* AGE (p<0.001) CSMOK (p=0.001) OCC*PACKYR (p=0.043) RBC -0.018 ± 0.017 0.278 WBC **** **** GRP*RACE*OCC (p<0.001) GRP*AGE*RACE (p=0.013) GRP*AGE*PACKYR (p=0.013) GRP*RACE*PACKYR (p=0.022) PACKYR*CSMOK (p<0.001) HGB **** **** GRP*RACE*AGE (p=0.030) GRP*RACE*OCC (p=0.020) OCC*PACKYR (p=0.023) CSMOK (p<0.001) HCT **** **** GRP*RACE*AGE (p=0.026) GRP*RACE*OCC (p=0.011) AGE*OCC (p=0.009) CSMOK (p<0.001) MCV **** **** GRP*AGE*PACKYR (p=0.041) GRP*AGE*CSMOK (p=0.012) RACE*OCC (p=0.021) AGE*RACE (p<0.001) MCH **** **** GRP*AGE*CSMOK (p=0.026) AGE*RACE (p=0.001) OCC (p=0.001) MCHC 0.030 ± 0.028 0.296 PLT **** **** RACE (p<0.001) GRP*RACE*PACKYR (p=0.011) GRP*AGE (p=0.040) OCC (p=0.005) *CSMOK: current level of smoking (cigarettes per day) PACKYR: smoking history (pack years) OCC: occupation GRP: group ****Group-by-covariate interaction—group difference, standard error (SE), and p-value not presented. N-9 �TABLE N-8. Summary of Group-by-Covariate Interactions for Hematological Variables (Original Comparisons Only) Ranch Hand-Original Comparison Group Difference ±SE p-Value -0.014 + 0.021 0.014 ± 0.022 0.511 0.508 0.107 + 0.033 -0.011 ± 0.019 0.001 0.551 Black -0.053 ± 0.058 (No significant interactions) 0.364 Variable Interaction Covariates Stratification WBCa Nonblack Nonsmoker 35 53 Race Smoking History Age (At Baseline) 30 Pack-Years 35 53 I l-» o Direction of Result* RH>OC �TABLE N-8. (continued) Summary of Group-by-Covariate . Interactions for Hematological Variables (Original Comparisons Only) HGB Interaction Covariates Direction of Result* Race Occupation Ranch Hand-Original Comparison Group Difference +SE p-Value 0.151 ± 0.113 -0.035 ± 0.082 0.181 0.667 Enlisted Flyer 35 53 Variable 0.087 ± 0.132 -0.100 ± 0.121 0.509 0.410 Enlisted Groundcrew 35 53 0.026 + 0 0 0 .8 -0.161 ± 0.107 0.744 0.134 -1.034 + 0.627 0.209 ± 0.629 0.099 0.740 OORH Enlisted Flyer 35 53 -1.398 + 0.613 -0.155 ± 0.481 0.023 0.747 OORH Enlisted Groundcrew 35 53 -0.014 + 0.283 1.230 ± 0.526 0.962 0.020 RH>OC Stratification Nonblack Officer 35 53 Age (At Baseline) z Black Officer 35 . 53 �TABLE N-8. (continued) Summary of Group-by-Covariate Interactions for Hematological Variables (Original Comparisons Only) HCT Interaction Covariates Race Occupation Ranch Hand-Original Comparison Group Difference ±SE p-Value 0.506 + 0.325 -0.174 ± 0.237 0.120 0.462 Enlisted Flyer 35 53 Variable 0.390 + 0.377 -0.291 ± 0.347 0.301 0.403 Enlisted Groundcrew 35 53 0.063 + 0.230 -0.617 ± 0.307 0.783 0.045 -2.662 + 1.801 0.825 ± 1.805 0.140 0.648 Enlisted Flyer 35 53 -4.365 + 1.753 -0.878 ± 1.378 0.013 0.524 OORH Enlisted Groundcrew 35 53 -0.050 + 0.806 3.436 ± 1.497 0.950 0.022 RH>OC Stratification Nonblack Officer 35 53 Direction of Result* Age (At Baseline) as »-» to Black Officer 35 53 OORH �TABLE N-8. (continued) Summary of Group-by-Covariate Interactions for Hematological Variables (Original Comparisons Only) Variable Interaction Covariates MCV MCH p-Value 0.692 + 0.418 -0.136 ± 0.417 0.098 0.743 0.057 ± 0.635 -0.582 ± 0.412 0.928 0.158 Nonsmoker 35 53 0.126 + 0.141 0.158 ± 0.132 0.373 0.232 1 Pack/Day 35 53 0.254 + 0.144 -0.275 ± 0.149 0.079 0.066 RH>OC OORH 2 Pack/ Day 35 53 0.381 + 0.273 -0.707 ± 0.281 0.162 0.012 OORH Smoking Nonsmoker (Smoking History 35 and Current Smoking) 53 Age (At Baseline) Current Smoking Age (At Baseline) Direction of Result* Ranch Hand-Original Comparison Group Difference ±SE Stratification 30 Pack- Years, Currently 1 Pack/Day 35 53 RH>OC �TABLE N-8. (continued) Summary of Group-by-Covariate Interactions for Hematological Variables (Original Comparisons Only) Variable PLTa Interaction Covariates Race Smoking History Stratification Nonblack Nonsmoker 35 53 Ranch Hand-Original Comparison Group Difference ±SE p-Value Direction of Result* 0.030 + 0.016 -0.017 ± 0.018 0.061 0.354 RH>OC 0.050 + 0.020 0.003 ± 0.016 0.012 0.858 RH>OC -0.082 + 0.051 -0.129 ± 0.054 0.111 0.016 OORH 0.261 + 0.103 0.214 ± 0.102 0.011 0.036 RH>OC RH>OC Age (At Baseline) 3Q Pack- Years 35 53 I (-* *- Black Nonsmoker 35 53 30 Pack-Years 35 53 *RH>OC: Variable mean in Ranch Hand group greater than variable mean in Original Comparison group. OORH: Variable mean in Original Comparison group greater than variable mean in Ranch Hand group. a Units are on log scale. �TABLE N-9. Longitudinal Analyses for MCV, MCH, and PLT: A Contrast of Baseline and First Followup Examination Test Means (Original Comparisons Only) Means Variable Group Total Baseline Difference Followup (Followup-Baseline) Error* 88.89 92.60 +3.71 872 88.56 92.35 +3.79 Ranch Hand 971 30.81 31.55 +0.74 872 30.63 31.44 +0.81 Ranch Hand 971 276.90 271.50 -5.4 Original Comparison PLT 971 Original Comparison MCH Ranch Hand Original Comparison MCV 872 272.10 267.30 -4.8 p-Value (Equality of Difference) *Error - HSubj * Time/Group mean squares. N-15 2.808 0.65 0.873 0.21 35.50 0.80 �APPENDIX 0 Renal Assessment �TABLE 0-1. Unadjusted Exposure Index Analyses for Renal Variables by Occupation Occupation Statistic Officer Variable n Number /% Abnormal Normal Low 127 10 7.9% - 117 92.1% Exposure Index Medium High Contrast Est. Relative Risk (95% C.I.) p- Value 130 123 Overall 0.226 11 8.5% 119 91.5% 17 13.8% 106 86.2% M vs. L H vs. L 1.08 (0.44,2.64) 0.999 1.88 (0.82,4.28) 0.155 65 57 Overall 0.165 7 12.7% 48 87.3% 8 12.3% 57 87.7% 2 3.5% 55 96.5% M vs. L H vs. -L 0.96 (0.33 ,2.85) 0.999 0.25 ( . 5 ,1.26) 0.091 00 n Number /% Abnormal Normal 153 163 141 Overall 0.214 18 11.8% 135 88.2% 11 6.7% 152 93.3% 10 7.1% 131 92.9% M vs. L H vs. L 0.54 (0.25 ,1.19) 0.172 0.57 (0.26 ,1.29) 0.233 n Number/% Abnormal Normal 127 130 123 Overall 0.696 Officer Urinary Protein n Number/% Abnormal Normal 55 Enlisted Flyer Enlisted Groundcrev Kidney Disease 4 3.1% 123 96.9% 2 1.5% 3 2.4% 128 98.5% 120 97.6% M vs. L H vs. L 0.48 ( . 9 ,2.67) 0.443 00 0.77 (0.17 ,3.51) 0.999 n Number/% Abnormal Normal 55 65 57 Overall 0.708 Enlisted Flyer 2 3.6% 53 96.4% I 1.5% 64 98.5% 1 1.8% 56 98.2% M vs. L H vs. L 0.41 ( . 4 ,4.69) 0.593 00 0.47 ( . 4 ,5.37) 0.615 00 n Number/% Abnormal Normal 154 163 Overall 0.726 Enlisted Groundcrew 8 5.2% 146 94.8% 9 6.3% 7 4.3% 156 95.7% 133 93.7% M vs. L H vs. L 0.82 (0.29 ,2.32) 0.794 1.24 (0.46 ,3.29) 0.804 142 �TABLE 0-1. (continued) Unadjusted Exposure Index Analyses for Renal Variables by Occupation Variable Occupation Statistic Low Exposure Index Medium High Est. Relative Contrast Risk (95% C.I.) p-Value n Number/% Abnormal Normal 127 130 123 Overall 0.334 17 13.4% 110 86.6% 15 11.5% 115 88.5% 22 17.9% 101 82.1% H vs. L H vs. L 0.84 ( .40,1.77) 0.708 0 1.41 (0.71,2.81) 0.385 n Number/% Abnormal Normal 55 65 57 Overall 0.325 Enlisted Flyer 12 21.8% 43 78.2% 14 21.5% 51 78.5% 7 12.3% 50 87.7% M vs. L H vs. L 0.98 ( .41,2.35) 0.999 0 0.50 ( .18,1.39) 0.214 0 n Number /% Abnormal Normal 154 163 141 Overall 0.144 Enlisted Groundcrev 25 16.2% 129 83.8% 34 20,9% 129 79.1% 36 25.5% 105 74.5% H vs. L H vs. L 1.36 (0.77,2.41) 0.315 1.77 (1.00,3.13) 0.061 n Number/% Abnormal Normal 127 130 123 Overall 0.558 11 8.7% 116 91.3% 7 5.4% 123 94.6% 10 8.1% 113 91.9% M vs. L H vs. L 0.60 ( .23,1.60) 0.337 0 0.93 (0.38,2.28) 0.999 n Number /% Abnormal Normal 55 65 57 Overall 0.763 Enlisted Flyer 7 12.7% 48 87.3% 8 12.3% 57 87.7% 5 8.8% 52 91.2% M vs. L H vs. L 0.96 (0.33,2.85) 0.999 0.66 ( .20,2.22) 0.554 0 n Number /% Abnormal Normal 154 163 142 Overall 0.447 Enlisted Groundcrev 20 13.0% 134 87.0% 15 148 19 13.4% 123 86.6% M vs. L H vs. L 0.68 (0. 33,1.38) 0.290 1.04 (0. 53,2.03) 0.999 Officer Urinary Occult Blood o I Officer Urinary White Blood Cell Count 9.2% 90.8% �TABLE 0-1. (continued) Unadjusted Exposure Index Analyses for Renal Variables by Occupation Exposure Index Statistic Low Medium High Contrast n Mean* 95* C.I.* 127 14.62 (14.05, 15.19) 130 14.82 (14.20, 15.46) 123 14.74 (14.14, 15.35) Overall M vs. L H vs. L 0.891 0.633 0.775 Enlisted Flyer n Mean* 95* C.I.* 55 13.84 (12.79, 14.93) 65 13.80 (12.87, 14.77) Overall 57 13.93 M vs. L (13.15, H vs. L 14.72) 0.983 0.963 0.895 Enlisted Groundcrew n Mean* 95* C.I.* 154 14.08 (13.50, 14.67) 163 13.72 (13.15, 14.30) 142 14.00 (13.48, 14.52) Overall M vs. L H vs. L 0.639 0.389 0.841 Officer n Mean 95% C.I. 127 1.0144 (1.0132, 1.0156) 130 1.0149 (1.0137, 1.0161) Overall 123 1.0150 M vs. L (1.0137, H vs. L 1.0162) 0.786 0.554 0.541 Enlisted Flyer n Mean 95* C.I. 55 1.0159 (1.0139, 1.0178) 65 1.0155 (1.0138, 1.0173) 57 Overall 1.0141 M vs. L (1.0123, H vs. L 1.0158) 0.358 0.803 0.178 Enlisted Groundcrew Blood Urea Nitrogen Occupation Officer Variable n Mean • 95* C.I. 154 1.0167 (1.0156, 1.0178) 163 1.0168 (1.0158, 1.0179) 142 Overall 1.0166 M vs. L (1.0154, H vs. L 1.0179) 0.967 0.885 0.907 o Urine Specific Gravity ^Transformed from square root scale. —No relative risk given for variables analyzed continuously. Risk (95* C.I.) p-Value �TABLE 0-2. Interaction Summaries of Adjusted Exposure Index Analyses for Renal Variables Variable Interaction (Occupation) Stratification Statistic Diabetic Urinary Protein Exposure Index-byDiabetic Impaired flagg (Officer) Normal Nonblack Urinary Protein Exposure Index-byRace (Enlisted Groundcreu) Black Bom XL942 Urinary Occult Blood Exposure Index-byAge (Enlisted Groundcreu) Bom <1942 Lou Exposure Index Medium High Adj. Relative 9% Contrast Risk ( 5 C.I.) p-Value n Nunber/% Abnormal Normal 8 8 14 Overall 000 .1* 3 37.5% 5 62.5% 0 00 .% 8 100 0.% 0 0.0% M vs. L 0.% 14 1 0 0 H vs. L 0.09 ( . 0 , . 6 * 0 2 0 00421)* . 0 * 0.05 ( . 0 , . 2 * 0 0 6 00212)* .3* n Number/% Abnormal Normal 11 n Nunfcer/% Abnormal Normal 106 108 0 0.0% 0.% 108 1 0 0 2 106 n Number/% Abnormal Normal 138 149 4 2.9% 134 97.1% 6 143 n Number/% Abnormal Normal 16 14 4 25.0% 12 75,0% 1 13 n Number/* Abnormal Normal 87 n Number /% Abnormal Normal 67 1 10 11 76 14 53 14 9.1% 90.9% • 12.6% 87.4% 16 Overall 0.247* 0 00 .% 14 1 0 0 0.% 0 0.0% M vs. L 16 1 0 0 H vs. L 0.% 0.24 ( . 1 6 5 ) * 0 4 0 00,.3* .4* 0.21 ( . 0 , . 1 * 0 4 7 00857)* . 0 * Overall 0.190* 3 3.2% M vs. L 90 96.8% H vs. L 5.09 ( . 4 1 7 3 ) * 0 4 8 02,0.6* .9* 8.39 ( . 3 1 4 6 ) * 0 0 7 04,6.1* .9* Overall 0.259* 93 1.9% 98.1% 129 4.0% 96.0% 13 7.1% 92.9% 17.8% 82.2% 11 23 Overall 0.751* 0.158* 002 .9* 0 0.0% M vs. L 0.23 ( . 2 2 3 ) 0 0 , . 7 * 0.336* 0.% 13 1 0 0 H vs. L 0.10 ( . 0 , . 1 * 0.107* 00521)* 20 24.7% H vs. L 61 75.3% H vs. L 1.56 ( . 1 3 4 ) 07,.1 2.32 ( . 3 7 3 ) 07,.2 0.252 046 .6 1.73 ( . 8 4 4 ) 06,.2 1.36 ( . 9 3 1 ) 05,.2 0.265 0.152 60 34 20.9% 79.1% 1.41 ( . 9 5 0 ) 03,.9* 2.51 ( . 5 8 3 ) 07,.7* 81 129 23 106 9 7.0% M vs. L 120 93.0% H vs. L 32.4% 67.6% 16 26.7% M vs. L 44 73.3% H vs. L �TABLE 0-2. (continued) Interaction Summaries of Adjusted Exposure Index Analyses for Renal Variables Variable Interaction (Occupation) Stratification Statistic Low Exposure Index Medium High Adj. Relative Contrast Risk ( 5 C.I.) 9% p-Value Age £ 1 0 Overall 0 0 — 1 100 0.% 0 0.0% 0 0 — M vs. L H vs. L n Hunter/* Abnormal Nonnal 1 1 0 Overall 0.157* Bom >1942 Impaired Exposure Index-by- n Number/% Abnormal Normal 0 Born >1942 Diabetic 0 0.0% 1 100 0.% 1 100 0.% 0 0.0% 0 — 0 — M vs. L H vs. L 9.00 ( . 0 8 1 7 ) * 0 9 9 01,3.8* . 9 * n Number/% Abnormal Normal 10 16 9 Overall 0.065* Bom >1942 Normal 0 0.0% 16 1 0 0 0.% 1 11.1% M vs. L 8 88.9% B vs. L 0.07 ( . 0 , . 2 * 0 0 6 00314)* . 4 * 0.29 ( . 2 3 4 ) * 0.582* 00,.8* n Number/% Abnormal Normal 1 8 Bom <1942 Diabetic 0 0.0% 1 100 0.% 1 7 n Number/% Abnormal Normal 8 6 Bom <1942 Impaired 0 0.0% 8 100 0.% 2 4 n Number/% Abnormal Nonnal 35 33 Bom <1942 Nonnal 4 11.4% 31 88.6% 4 29 Urinary Unite Blood Cell Count ( ) a Exposure Index-byDiabetic Class (Enlisted Flyer) 3 7 30.0% 70.0% 5 12.5% 87.5% Overall 068 .6* 0 0.0% H vs. L 5 1 0 0 H vs. L 0.% 0.60 ( . 2 2 . 7 * 0 9 9 00,30)* . 9 * — — 6 33.3% 66.7% Overall 0.222* 1 16.7% 5 83.3% H vs. L H vs. L 9.44 ( . 7 2 2 1 ) * 0.165* 03,4.8* 4.64 ( . 6 1 5 5 ) * 0 4 9 01,3.7* .2* 37 12.1% 87.9% Overall 3 8.1% M vs. L 34 91.9% H vs. L 085 .0 0.79 ( . 6 3 9 ) 01,.0 0.58 ( . 1 3 0 ) 01,.3 0.776 055 .1 �TABLE 0 2 (continued) -. Interaction Summaries of Adjusted Exposure Index Analyses for Benal Variables Variable Interaction (Occupation) Low Stratification Statistic Exposure Index Medium High Adj. Relative Contrast Risk ( 5 C.I.> ] 9Z p-Value Nonblack Blood Urea Nitrogen Exposure Index-byRace (Officer) n . 125 127 121 M vs. L Adj. Mean(c) 1 . 5 46 14.71 14.55 H vs. L 95XC.I.(c) ( 3 9 , 5 4 ) ( 3 9 , 5 4 ) ( 3 8 , . 7 1.11.1 1.61.9 1.412) (b) (b) 087 .7 0.814 Black n Adj. Mean 95XC.I. (b) (b) 0.035 0.494 2 2 3 M vs. L 16.93 19.51 H vs. L 10.65 (22,23) (.21.9 (44,53) 1.42.8 76,41) 1.52.3 ^unadjusted estimate of relative risk and confidence interval, or p-value, based on stratified tables. **Uhadjusted estimate of relative risk and confidence interval calculated after adding 0.5 to each cell. —Zero counts in cells do not allow for calculation of percent, relative risk, confidence interval, or p-value. (a)Results presented for urinary white blood cell count are based on stratification into the six categories shown. Unadjusted results are presented for all strata except Ranch Bands in normal diabetic class, born before 1 4 . These results have been adjusted for race. 92 (b)No relative risk given for variables analyzed continuously. (c)Transforned from square root scale. Note: Small sample sizes may affect validity of overall p-value. Note: Results without (*) or ( * are adjusted for all other main effects in model (age, race, and diabetic class), unless otherwise noted. *) �TABLE 0-3. Unadjusted Analyses for Renal Variables by Group (Original Comparisons Only) Group Original Comparison Number Percent Variable Kidney Disease n Yes No 1,014 94 920 9.3 90.7 954 94 859 Urinary Protein o •-J Statistic Ranch Hand Number Percent n Abnormal Normal 1,016 37 979 3.6 96.4 Urinary Occult Blood n Abnormal Normal 1,015 182 833 Urinary White Blood Cell Count n Abnormal Normal 1,016 102 914 Blood Urea Nitrogen n Mean* 95% C.I.* 1,016 14 .21 (13.99,14.43) 955 14.37 (14.15,14.59) 0.318 Urine Specific Gravity n Mean 95% C.I. 1,016 1.0157 (1.0153,1.0162) 954 1.0154 (1.0150,1.0159) 0.365 : Est. Relative Risk ( 5 C.I.) 9% p-Value 9.9 90.1 0.93 (0.69,1.26) 0.701 954 28 926 2.9 97.1 1.25 (0.76,2.06) 0.449 17.9 82.1 954 155 799 16.2 83.8 1.13 (0.89,1.43) 0.338 10.0 90.0 954 83 871 8.7 91.3 1.17 (0.86,1.59) 0.316 *Converted from square root scale. —No relative risk given for variables analyzed continuously. �TABLE 0-4. Adjusted Analyses for Renal Variables by Group (Original Comparisons Only) Variable Stratification Statistic Group Ranch Original Hand Comparison Adj. Relative Risk (95% C.I.) Covariate p-Value Remarks* Kidney Disease 1,014 950 0.95 (0.70,1.28) 0.726 AGE (p=0.001) Urinary Protein 0 00 n n 1,016 942 **** **** OCC (p=0.007) AGE*DIAB (p=0.037) GRP*DIAB (p=0.003) Urinary Occult Blood GRP*OCC*RACE (p=0.020) Nonblack n 955 887 1.14 (0.89,1.46) 0.299 AGE (p=0.003) OCC (p<0.001) Black Enlisted n 53 51 1.43 (0.54,3.76) 0.465 GRP*OCC (p=0.051) AGE*DIAB (p=0.027) �TABLE 0-4. (continued) Adjusted Analyses for Renal Variables by Group (Original Comparisons Only) Group Variable Stratification Ranch Hand Statistic Original Comparison Adj. Relative Risk (95* C.I.) p-Value GRP*AGE*RACE (p=0.048) GRP*OCC*RACE (p=0.022) Urinary White Blood Cell Count Nonblack VO n 956 n Black Enlisted o i Covariate Remarks* * 893 **** **** OCC (p=0.010) GRP*DIAB (p=0.054) GRP*AGE-(p=0.053) 53 51 **** **** GRP*OCC (p=0.015) DIAB (p-0.017) GRP*AGE (p=0.010) **** AGE (p<0.001) OCC (p=0.025) RACE*DIAB (p=0.049) GRP*RACE (p=0.077) Blood Urea Nitrogen n Adj . Mean 95* C.I. 1,016 **** **** 952 **** **** • Urine Specific Gravity n Adj. Mean 952 C.I. 1,016 **** **** 951 **** **** **** DIAB (p=0,011) OCC (p<0.001) GRP*RACE (p=0.037) ^Abbreviations: OCC: occupation GRP: group DIAB: diabetic class ****Group-by~covariate interaction—adjusted relative risk/group means, p-value, and confidence interval are not presented. —No relative risk given for variables analyzed continuously. �TABLE 0-5. Summary of Group-by-Covariate Interactions for Renal Variables (Original Comparisons Only) Group Variable Interaction Stratification Statistic Diabetic Urinary Protein o Group-byDiabetic Class Impaired Normal Officer Urinary Occult Blood Group-byOccupation (Black) Enlisted Flyer Enlisted Groundcrev n Number/% Abnormal Normal Ranch Hand n Number/% Abnormal Normal n Number/% Abnormal Normal 7 25 807 0 7 n Number/% Abnormal Normal n Number/% Abnormal Normal 4.7% 95.3% 3.0% 97.0% 14.4% 85.5% 0.57 (0.20,1.60) 0.286 11 125 8.1% 91.9% 0.58 (0.19,1.75) 0.332 6 733 0.8% 99.2% 3.82 (1.56,9.37) 0.004 7 0.0% 100.0% 43 13 30 11 65 739 10 3 7 p-Valn*e 136 832 n Number/% Abnormal Normal 9.0% 91.0% 106 5 101 Adj. Relative Risk (95% C.I.) 76 78 7 •71 Original Comparison 3 4 42.9% 57.1% — 8.3% 91.7% — 12 30.0% 70.0% 1 11 39 30.2% 69.8% 9 30 23 U% 76.9% — — �TABLE 0-5. (continued) Summary of Group-by-Covariate Interactions for Renal Variables (Original Comparisons Only) Group Variable Interaction Adj. Relative Risk (95% C.I.) p-Value 16.7% 83.3% 0.90 (0.38,2.10) 0.799 33 4 29 12.1% 87.9% 4.12 (2.06,8.24) <0.001 11.0% 89.0% 298 13 285 4.4% 95.6% 2.08 (1.16,3.71) 0.014 57 6 51 10.5% 89.5% 57 12 45 21.1% 78.9% 0.38 (0.14,1.04) 0.059 n Abnormal Normal 80 14 66 17.5% 82.5% 89 6 83 6.7% 93.3% 1.77 (0.75,4.17) 0.191 n Abnormal Normal 439 31 408 7.1% 92.9% 404 37 367 9.2% 90.8% 0.89 (0.55,1.44) 0.639 Original Comparison Ranch Hand Stratification Statistic Born >1942 Diabetic n Abnormal Normal 14 I 13 7.1% 92.9% 12 2 10 Born <1942 Impaired n Abnormal Normal 22 2 20 9.1% 90.9% Born >1942; Normal n Abnormal Normal 344 38 306 Born <1942 Diabetic n Abnormal Normal Born <1942 Impaired Born <1942 Normal . Nonblack: Group-byAge o i Urinary White Blood Cell Count Group-byDiabetic Class 4 �TABLE 0-5. (continued) Summary of Group-by-Covariate Interactions for Renal Variables (Original Comparisons Only) Group Variable Group-byOccupation (Black) o i Stratification Statistic Ranch Band Officer Urinary Uhite Blood Cell Count Interaction n Number/% Abnormal Normal 7 Enlisted Flyer Enlisted Groundcrew Born >1942 Urinary Uhite Blood Cell Count Group-byAge (Black Enlisted) Born <1942 0 7 n Number/% Abnormal Normal n Number/% Abnormal Normal n Number/% Abnormal Normal 29 n Number/% Abnormal Normal 0.0% 100.0% 43 24 7 36 4 25 6 18 Adj. Relative Risk (95% C.I.) p-Value 7 10 3 7 Original Comparison 2 5 28.6% 71.4% 12 30.0% 70.0% 0 12 0.0% 100 0.% 39 16.3% 83.7% 7 32 17.9% 82.1% 26 13.8% 86.2% 5 21 19.2% 80.8% 0.65 (0.15,2.78) 0.565 8.0% 92.0% 4.99 (0.79,31.56) 0.088 25 25.0% 75.0% 2 23 �TABLE 0-5. (continued) Summary of Group-by-Covariate Interactions for Renal Variables (Original Comparisons Only) Group Variable Stratification Statistic Nonblack Blood Urea Nitrogen Interaction n Adj. Mean* 95* C.I.* Group-byRace Black o i Nonblack Urine Specific Gravity Group-byRace Black •Ranch Hand 956 14.16 (13.86,14.47) Original Comparison 894 14.22 (13.92,14.53) Adj. Relative Risk (95% C.I.) p-Value 0.726 n Adj. Mean* 95Z C.I.* 60 12.43 (11.41,13.51) 58 13.63 (12.66,14.64) 0.057 n Adj. Mean 95X C.I. 956 1.0163 (1.0157,1.0169) 893 1.0158 (1.0151,1.0164) 0.104 n Adj. Mean 95Z C.I. 60 1.0154 (1.0135,1.0173) 58 1.0177 (1.0158,1.0196) 0.081 —No relative risk given for variables analyzed continuously. Also, no relative risk or p-value given for interactions involving occupation for urinary occult blood and urinary white blood cell count; tables presented only for illustrative purposes. ^Transformed from square root scale. �APPENDIX P Endocrine Assessment �APPENDIX P: Endocrine Assessment Contents Table Page P-l Summary of Group-by-Covariate Interactions for Endocrinological Variables P-l P-2 Unadjusted Exposure Index Analyses for Endocrinological Variables by Occupation P-2 P-3 Interaction Summaries of Adjusted Exposure Index Analyses for Endocrinological Variables P-6 P-4 Unadjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group (Original Comparisons Only)... P-8 P-5 Adjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group (Original Comparisons Only)... P-10 P-6 Summary of Group-by-Covariate Interactions for Endocrinological Variables (Original Comparisons Only) P-12 �TABLE P-l. Summary of Group-by-Covariate Interactions for Endocrinological Variables Group Comparison Differential Group-byCortisol Race-byAge n Mean Ad j . Mean 95% C.I. 4 6 522.6 1044.8 654.4 1042.8 0.012 (484.7,849.5) (803.5,1313.2) 10-25% n Mean Ad j . Mean 95% C.I. 815 627.8 603.3 (583.4,623.5) 1026 605.7 582.4 0.023 (563.9,601.3) n Mean Ad j . Mean 95% C.I. 179 470.5 463.0 (435.0,491.9) 257 470.3 456.7 0.706 (432.9,481.2) Nonblack Born >1942 Testosterone Group-by% Body Fat Ranch Hand >25% Interaction Statistic <10% Variable n Mean Ad j . Mean 95% C.I. 374 2.07 1.58 (0.64,2.51) 500 2.44 1.97 (1.06,2.88) 0.156 Nonblack Born <1942 n Mean Ad j . Mean 95% C.I. 572 2.53 2.08 (1.18,2.97) 697 2.46 1.99 (.028) 11,.8 0.708 Black Born >1942 n Mean Adj. Mean 95% C.I. 32 0.17 -0.46 (-2.11,1.18) 47 2.78 2.33 (0.89,3.76) Black Born <1942 n Mean Adj . Mean 95% C.I. 26 3.24 2.94 (1.22,4.67) 36 2.30 1.89 (0.32,3.46) Stratification p-Value 0.003 0.312 �TABLE P-2. Unadjusted Exposure Index Analyses for Endocrinological Variables by Occupation Low Exposure Index High Medium Contrast n Mean 95% C.I. 126 124 122 28.01 28.35 27.96 (27.66,28.36) (28.00,28.71) (27.61 ,28.32) Overall M vs. L H vs. L 0.250 0.177 0.857 Enlisted Flyer n Mean 95% C.I. 55 65 55 27.59 27.31 27.83 (26.82,27.81) (27.14,28.06) (27.32 ,28.34) Overall M vs. L H vs. L 0.362 0.411 0.155 n Mean 95% C.I. 153 142 161 27.68 27.65 27.53 (27.37,28.00) (27.34,27.95) (27.21 ,27.85) Overall M vs. L H vs. L 0.775 0.865 0.493 Officer •ja Statistic Enlisted Groundcrev T3 % Uptake Occupation Officer Variable n Mean 95% C.I. 126 124 122 1.167 1.278 1.181 (1.089,1.286) (1.076,1.271) (1.173 ,1.397) Overall M vs. L H vs. L 0.283 0.843 0.203 Enlisted Flyer n Mean 95% C.I. 55 65 55 1.153 1.033 1.204 (0.924,1.162) (1.035,1.292) (1.068 ,1.367) Overall M vs. L H vs. L 0.181 0.178 0.075 Enlisted Groundcrev n Mean 95% C.I. 153 161 1.117 1.124 (1.043,1.199) (1.051,1.204) Overall M vs. L B vs. L 0.738 0.899 0.465 NJ TSH 142 1.160 (1.079 ,1.250) p-Value �TABLE P-2. (continued) Unadjusted Exposure Index Analyses for Endocrinological Variables by Occupation Variable Low Exposure Index Medium High n Mean 95% C.I. 125 118 128 599.4 547.4 558.0 (563.8,636.1) (513.8,582.1) (522.7,594.4) Overall M vs. L H vs. L 0.099 0.041 0.112 Enlisted Flyer n Mean 95% C.I. 55 63 588.7 637.4 (531.6,648.6) (581.8,695.6) 57 584.9 (529.1,643.6) Overall M vs. L H vs. L 0.363 0.244 0.929 Enlisted Groundcrew n Mean 95% C.I. 162 139 153 641.4 603.4 609.3 (574.7,644.9) (606.9,676.9) (567.3,640.6) Overall M vs. L H vs. L 0.276 0.205 0.820 Officer *j3 Statistic Officer Testosterone Occupation n • Mean 95% C.I. 124 130 11.97 11.43 (11.33,12.65) (10.83,12.07) 121 12.28 (11.61,12.99) Overall M vs. L H vs. L 0.186 0.239 0.531 Enlisted Flyer n Mean 95% C.I. 57 55 65 11.97 11.08 11.13 (11.09,12.91) (10.33,11.88) (10.33,11.99) Overall M vs. L H vs. L 0.276 0.001 0.092 Enlisted Groundcrew n Mean 95% C.I. 142 154 161 11.50 11.69 11.33 (11.09,12.32) (10.76, 11. 92> (10.89,12.15) Overall M vs. L H vs. L 0.699 0.397 0.678 UJ Initial Cortisol Contrast p-Value �TABLE P-2. (continued) Unadjusted Exposure Index Analyses for Endocrinological Variables by Occupation Low Exposure Index Medium High Contrast Differential Cortisol n Mean 95% C.I. 124 9.27 (8.72,9.86) 130 9.16 (8.63,9.73) 121 9.69 (9.10,10.32) Overall M vs. L H vs. L 0.416 0.789 0.324 Enlisted Flyer n Mean 95% C.I. 55 8.43 (7.67,9.26) 65 9.48 (8.69,10.34) 57 9.06 (8.25,9.94) Overall M vs. L H vs. L 0.200 0.074 0.286 n' Mean 95% C.I. 154 161 9.64 9.29 (9.11,10.19) (8.79,9.81) 142 9.17 (8.65,9.72) Overall M vs. L H vs. L 0.453 0.358 0.229 Officer *a i Statistic Enlisted Groundcrev 2-Hour Cortisol Occupation Officer Variable n Mean 95% C.I. . 124 2.69 (2.02,3.36) 130 2.31 (1.66,2.96) 121 2.59 (1-91,3.26) Overall M vs. L H vs. L 0.709 0.424 0.829 Enlisted Flyer n Mean 95% C.I. 55 3.43 (2.50,4.36) 65 1.20 (0.34,2.06) 57 2.30 (1.39,3.22) Overall M vs. L H vs. L 0.003 <0.001 0.092 Enlisted Groundcrev n Mean 95% C.I. 154 1.96 (1.26,2.66) 161 2.12 (1.44,2.81) 142 2.37 (1.64,3.09) Overall M vs. L H vs. L 0.726 0.740 0.425 p-Value �TABLE P-2. (continued) Unadjusted Exposure Index Analyses for Endocrinological Variables by Occupation Variable Low Exposure Index Medium High •jo Ul Statistic Officer 2-Hour Postprandial Glucose Occupation n Mean 95% C.I. 121 113 124 109.0 107.5 107.6 (104.3,113.8) (103.0,112.3) (102.9,112.6) Overall M vs. L H vs. L 0.892 0.666 0.694 Enlisted Flyer n Mean 95% C.I. 54 62 56 100.9 118.0 110.9 (92.1,110.6) (108.3,128.5) (101.4,121.4) Overall M vs. L H vs. L 0.051 0.015 0.149 Enlisted Groundcrew n Mean 95% C.I. 150 156 140 105.5 106.7 109.1 (100.1,111.1) (101.4,112.4) (103.3,115.2) Overall M vs. L H vs. L 0.674 0.750 0.380 Contrast p-Value �TAHUEP-3. Variable Interaction Stratifi(Occupation) cation <LO% Testosterone Exposure Index-by. % Body Fat (Enlisted Groundcrew) io_25% >25% 10-25% 2-8ax Cbrtisol Exposure Index-by%Body Fat (Officer) >25% • Nonblack 2-Hour Cortisol Exposure Index-byRace (Enlisted Groundcrev) Black Low Statistic Exposure Index Medium High Contrast p-Value n Adj. Mean 95% C.I. 1 2 3 72.7 645 8. M vs. L 888 2. 57818.) 454 958 ( . , 7 . ) ( 3 . , 1 2 5 ( 6 . , 4 . ) H vs. L 03230 O01 .0 O01 .0 n Adj. Mean 95% C.I. 125 128 105 600 3. 646.2 647 4. M vs. L ( 8 . , 7 . ) ( 0 . , 9 . ) ( 9 . , 9 . ) H vs. L 561655 616625 582 629 0.527 055 .8 n Adj. Mean 95% C.I. 27 31 29 501.8 467.5 M vs. L 530 0. ( 3 . , 7 . ) ( 3 . , 7 . ) ( 0 . , 3 . ) H vs. L 439547 476530 407 5 9 6 090 .8 0.471 n Adj. Mean 95% C.I. 102 110 8.31 85 .0 (.095) (.197) 72,.8 74,.6 99 9.05 M vs. L ( . 6 1 . 3 H vs. L 78, 0 4 ) 069 .2 009 .7 n Adj. Mean 95% C.I. 22 20 10.03 7.46 (.71.8 (.291) 82,21) 61,.0 20 M vs. L 8.26 ( . 7 1 . 7 H vs. L 67, 00) 006 .0 008 .6 n 129 138 146 Adj. Mean 9.30 9.32 M vs. L 9.33 95% C.I. ( . 0 1 . 2 ( . 8 1 . 1 ( . 3 1 . 3 H vs. L 83,04) 83,04) 83, 0 4 ) 091 .2 0.961 11 14 8.41 M vs. L 9.36 ( 0 5 , 5 6 ) ( . 1 1 . 2 ( . 8 1 . 8 H vs. L 1.91.7 76,15) 74, 04) 003 .1 002 .0 n 16 Adj. Mean 12.89 95% C.I. �• — —- x t Interaction granaries of Adjusted Exposure Index Analyses for Endocrinological Variables Variable Interaction Stratifi(Occupation) cation Initial Cortisol Exposure Index-by%Body Fat (Officer) Statistic Low Exposure Index Medium Higfr Contrast p-Value n Adj. Mean 95% C.I. 102 110 99 11.51 11.19 12.25 M vs. L ( 0 1 , 3 0 ) ( . 9 1 . 6 , ( 0 8 , 3 9 ) H vs. L 1.31.8 98,26) 1.01.1 0.514 0.154 n Adj. Mean 95% C.I. 22 20 20 12.28 9.59 M vs. L 10.26 ( 0 3 , 4 6 ) ( . 3 1 . 5 ( . 9 1 . 6 M vs. L 1.31.0 80,14) 85,22) 000 .1 000 .6 Nonblack TypeA n Adj. Mean 952: C.I. 49 63 1.94 0% . (07,.6 (.834) -.426) 03,.9 53 2.00 M vs. L ( . 2 3 6 ) H vs. L 03,.8 0.237 0.223 Nonblack TypeB 76 n 89 83 1.36 M vs. L Adj.. Mean 2.29 1.64 95% C.I. ( . 0 3 7 ) ( 0 1 , . 4 ( . 4 3 1 ) H vs. L 08,.8 -.228) 01,.4 0.158 0.336 Black TypeA 5 n 8 4 Adj.. Mean -5.49 3.27 3.10 M vs. L 95% C.I. ( 8 7 , 2 2 ) ( 0 7 , . 5 ( 1 3 , . 9 H vs. L -.0-.8 -.172) -.074) <D.001 001 .0 Black TypeB n 8 9 7 M vs. L Adj,. Mean 1.16 2.99 0.48 95% C.I. ( 2 0 , . 7 ( 0 0 , . 5 ( 2 8 , . 6 H vs. L -.443) -.760) -.037) 0.384 0.762 10-25% Differential Cortisol Exposure Ihdex-bySace-byPersonality Type (Enlisted Groundcrew) >25K �TABLE P-4. Unadjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group (Original Comparisons Only) Group Variable Statistic Ranch Hand Original Comparison T3 % Uptake n Mean • 95% C.I. Number/% Low Normal High 1,003 27.79 (27.67,27.91) 936 27.68 (27.55,27.80) 7 969 27 13 898 25 n Mean 95% C.I. Number/% Normal High 1,003 1.16 (1.13,1.19) n Mean 95% C.I. Number/% Low Normal High 952 1,000 597.31 570.90 (583.97,610.80) (557.69,584.27) TSH 00 Testosterone Initial Cortisol n • Mean 95% C.I. Number/% Low Normal High 996 7 38 949 13 0.7% 96.6% 2.7% 99.3% 0.7% 3.8% 94.9% 1.3% 1,009 11.62 (11.39,11.85) 52 950 7 5.2% 94.2% 0.7% 1.4% 95.9% 2.7% 936 1.11 (1.08,1 -14) 931 5 37 904 11 99.5% 0.5% 3.9% 95.0% 1.2% 949 11.67 (11.44,11.91) 44 898 7 4.6% 94.6% 0.7% Contrast Est. Relative Risk (95% C.I.) p-Value 0.203 Overall Low vs. Normal 0.50 (0.20,1.26) High vs. Normal 1.00 (0.58,1.74) — 1.31 (0.41,4.14) — Overall Low vs. Normal 0.98 (0.62,1.55) High vs. Normal 1.13 ( . 0 2.53) 05, — Overall Low vs. Normal 1.12 (0.74,1.69) High vs. Normal 0.95 (0.33,2.71) 0.322 0.177 0.999 0.040 0.775 006 .0 0.955 0.999 0.839 0.729 0.864 0.603 0.999 �TABLE P-4. (continued) Unadjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group (Original Comparisons Only) Group Variable Ranch Hand Original Comparison 2-Hour Cortisol n Mean 95% C.I. Number/% Low Normal High 1,009 9.30 (9.10,9.51) 949 9.32 (9.11,9.52) 0 0.0% 1,005 99.6% 4 0.4% 0 946 3 Differential Cortisol n Mean 95% C.I. 1,009 2.30 (2.05,2.55) 949 2.38 (2.12,2.64) 2-Hpur Postprandial Glucose T3 ** Statistic n Mean 95% C.I. Number/% Normal Impaired Diabetic 976 909 107.93 110.03 (105.90,110.01) (108.03,112.07) n Number/% Yes No 1,016 • Diabetes (Composite Indicator) 836 106 34 85.7% 10.9% 3.5% 74 7.3% 942 92.7% 748 137 24 Contrast p-Value 0.924 0.0% 99.7% 0.3% 82.3% 15.1% 2.6% Est. Relative Risk (95% C.I.) 1.26 (0.28,5.62) 0.999 0.692 Overall Impaired vs. Normal Diabetic vs. Normal 0.154 0.69 (0.53,0.91) 1.27 (0.75,2.16) 0.017 0.009 0.423 1.04 (0.74,1.47) 0.861 955 67 888 7.0% 93.0% — No relative risk or confidence interval given for continuous analysis. �TABLE P-5. T fOriariral flnBrarianrv: (Wlv^ •* —' u — ___ s __ _.—-j ^ Group Variable Statistic Ranch Hand Original Comparison T3 % Uptake n 998 933 TSH n Adj. Mean 95%C.I. 1,003 1.16 (1.13,1.19) 936 1.11 (1.08,1.14) Contrast Adj. Relative Risk (95% C.I.) Overall — Low vs. Normal 0.52 (0.45,2.36) High vs. Normal 1.03 (0.59,1.78) — High vs. Normal 1.31 (0.41,4.14) n Adj. Mean 95%C.I. 1,000 n Testosterone 1,000 OOC (p=0.025) PERSTYPE (p=0.035) 0.020 AGE (pO.OOl) 0.775 RACE (p=0.015) AGE*BFAT (p=0.039) GRP*BFAT (p=0.022) AAA'A — 951 Initial Cortisol n Adj. Mean 95%C.I. 1,004 11.35 (10.49,12.27) 945 11.38 (10.52,12.32) Differential Cortisol 1,004 945 AAA'A AAAA AAAA Covariate Remarks 0.332 0.153 0.916 951 Overall Low vs. Normal 1.01 90.63,1.61) High vs. Normal 1.02 (0.42,2.52) n Adj. Mean 95% C.I. p-Value — 0.998 0.980 0.958 AGE (pO.OOl) %BFAT (pO.OOl) 0.820 AGE (pO.OOl) %BFAT (pO.OOl) PERSTYPE (p=O.028) OOC*RACE (p=0.026) **** GRP*AGE*RACE (p=0.01< PERSTYPE (p=0.002) XBFM- fn=0.006^ �TAKE P-5. (coatimad) (Original Cnmarisons Only) Group Variable Statistic Ranch Band Original Comparison 2-flour Postprandial Glucose n Adj. Mean 95% C I .. 976 114.8 (0.,2.) 172129 906 160 1. (0.,2.) 163142 Adj. Relative Risk ( 5 C I ) 9% . . p-Value Covariate Remarks 1,016 954 040 .3 %BFAT (pO.OOl) OCC ( = . 0 ) p002 AGE*RACE ( = . 0 ) p002 Overall Impaired vs. Normal 0 7 ( . 4 0 9 ) . 1 05,.3 Diabetic vs. Normal 1 2 ( . 1 2 3 ) . 9 07,.8 - Diabetes (Composite Indicator) Contrast 003 .2 0.014 045 .0 AGE (p<D.C01) RACE ( = . 2 ) p007 %BFAT ( < . 0 ) p001 1.13 ( . 0 1 6 ) 08,.1 044 .8 AGE ( < . 0 ) p001 %BFAT (pO.OOl) RACE ( = . 1 ) p005 interaction—adjusted mean/relative risk, confidence interval, and p-value are not presented. —No relative risk or confidence interval given for continuous analyses. GRP: Group OCC: Occupation FERSHFE: Personality type (A or B) %BFAT: Percent body fat �TABLE P-6. Variable Summary of Group-by-Covariate Interactions for Endocrinological Variables (Original Comparisons Only) Group Original Comparison Stratification Statistic Ranch Hand p-Value Interaction <10% 3 1173.8 1074.7 0.014 (794.4,1397.2) 10-25% n Mean Ad j . Mean 95% C.I. 815 627.8 602.3 (581.2,623.7) 757 595.9 577.7 0.013 (556.7,599.1) n Mean Ad j . Mean 95% C.I. 179 470.5 461.5 (432.9,491.1) 191 470.7 459.1 0.891 (431.8,487.2) Nonblack Born >1942 Differential Group-by Cortisol Race-by Age 6 522.6 640.1 (472.2,833.5) >25% Testosterone Group-by % Body Fat n Mean Ad j . Mean 95% C.I. n Mean Ad j . Mean 95% C.I. 374 2.07 1.36 (.823) 03,.4 342 2.34 1.65 (.726) 06,.4 0.331 n Mean Ad j . Mean 95% C.I. 572 2.53 1.85 (.227) 09,.9 545 2.36 1.66 (.126) 07,.0 0.419 n Mean Adj. Mean 95% C.I. 32 0.17 -0.68 (-2.35,0.99) 27 3.05 2.41 (0.64,4.17) 004 .0 n Mean Ad j . Mean 95% C.I. 26 3.24 2.72 (.844) 09,.7 31 2.22 0.274 1.54 (-0.14,3.23) Nonblack Born <1942 Black Born >1942 Black Born <1942 �APPENDIX Q Immunological Evaluation �APPENDIX Q: Immunological Evaluation Contents Table Page Q-l Summary of Group-by-Covariate Interactions for Immunological Variables " Q-l Q-2 Unadusted Exposure Index Analyses for Cell Surface Markers by Occupation Q-4 Q-3 Unadjusted Exposure Index Analyses for Functional Stimulation Tests by Occupation Q-8 Q-4 Interaction Summaries of Adjusted Exposure Index Analyses for Immunological Variables Q-10 Q-5 Unadjusted Analyses for Cell Surface Markers by Group (Original Comparisons Only) Q-ll Q-6 Adjusted Analyses for Cell Surface Markers by Group (Original Comparisons Only) Q-12 Q-7 Summary of Group-by-Covariate Interactions for Cell Surface Markers (Original Comparisons Only) Q-14 Q-8 Unadjusted Analyses for Functional Stimulation Tests by Group (Original Comparisons Only) Q-15 Q-9 Adjusted Analyses for Functional Stimulation Tests by Group (Original Comparisons Only) Q-16 Q-10 Summary of Group-by-Covariate Interactions for Functional Stimulation Tests (Original Comparisons Only) Q-17 �•DfflLEQ-1. Sunmary of Gcoup-by-Covariate Interactions for Innuiological Variables Group Comparison p-Value Variable Interaction Stratification Statistic Ranch Hand Total TCells Group-by-Race Nonblack n Adj. Mean 95% C.I. 424 532 0.619 1,616 1,599 (1,564, 1,669) (1,554, 1,646) Black n Adj. Mean 95% C.I. 18 35 0.039 1,566 1,888 (1,340, 1,810) (1,705, 2,080) 0 n Adj. Mean 95% C.I. 128 149 153.8 189.9 0.004 (135.6, 173.3) ( 7 . , 210.0) 109 X)-20 n Adj. Mean 95% C.I. 265 200 199.5 0.998 199.6 ( 8 . , 216.9) (185.3, 214.3) 129 >20-40 n Adj. Mean 95% C.I. 74 94 200.4 172.8 0.109 ( 7 . , 228.7) (151.5, 195.4) 140 >40 n Adj. Mean 95% C.I. 32 29 219.9 214.4 0.857 ( 7 . , 266.3) (172.5, 260.9) 180 0 n Adj. Mean 95% C.I. 35 30 44.19 0.060 32.33 (34.72, 56.23) ( 4 8 , 42.03) 2.8 XX-2 n Adj. Mean 95% C I .. 100 40.43 ( 4 3 , 47.59) 3.5 BCells Group-by(Nonblacks lifetime Smoking Only) (Pack-years) Monocytes Group-byOfficer Occupation Group-byCurrent Alcohol Use Officer (Drinks/day) Officer >2-4 n Adj. Mean 95% C.I. Officer >4 Enlisted 0 Flyer 131 38.38 0.569 (30, 4.4 3.0 46) 31 37 41.53 44.35 0.689 ( 2 0 , 53.91) (35.02, 56.17) 3.0 n Adj. Mean 95% C.I. 13 37.65 (25.58, 55.43) n Adj. Mean 95% C.I. 21 27 35.65 (26.32, 48.28) 45.10 0.232 (34.36, 59.20) Q-l 17 3.2 31 0.608 (25.18, 43.56) �TABLE Q-l. (continued) Group Comparison Ranch Hand Misted X)-2 Flyer n Adj. Mean 95% C.I. 35 50 34.95 0.159 43.22 3.0 (27.53, 44.36) ( 5 0 , 53.37) 8 9 0.097 32.71 56.23 (20.35, 52.56) (36.38, 86.89) n Adj. Mean 95% C.I. 11 5 36.17 0.729 41.10 2.9 ( 2 2 , 75.80) ( 3 9 , 54.54) 2.8 Enlisted 0 Groundcrew n Adj. Mean 95% C.I. % 76 0.421 45.04 48.98 ( 7 8 , 53.54) ( 1 7 , 57.50) 3.9 4.3 Enlisted >0-2 Groundcrew n Adj. Mean 95% C.I. 120 81 089 .3 42.42 41.59 ( 5 8 , 50.24) ( 6 1 , 47.91) 3.1 3.0 Enlisted >2-4 Groundcrew n Adj. Mean 95% C.I. 17 19 37.37 0.397 45.15 2.8 (33.05, 61.67) ( 7 0 , 51.57) Enlisted >4 Groundcrew HLA-CR Cells Statistic Enlisted >4 Flyer Monocytes (Cbnt.) Interaction Stratification Enlisted >2-4 n Flyer Adj. Mean 95% C.I. Variable n Adj. Mean 95% C.I. 16 23 35.32 0.003 68.86 2.2 ( 9 0 , 96.68) ( 6 4 , 47.23) 4.4 n Adj. Mean 95% C.I. 142 158 0.650 593.1 581.0 (555.1, 632.2) ( 4 . , 618.9) 544 n Adj. Mean 95% C.I. 223 306 0.232 559.5 582.8 (529.5, 590.3) ( 5 . , 609.2) 570 >2-4 n Adj. Mean 95% C.I. 65 58 0.750 553.4 565.9 ( 9 . , 611.9) (512.7, 621.7) 478 >4 n Adj. Mean 95% C.I. 51 36 0.052 563.7 472.7 ( 9 . , 640.6) (417.7, 531.0) 417 Group-by0 Current Alcohol Use (Drinks/day) >0-2 0-2 p-Value �TfiBUSQ-l. (continued) Smmary of Gcoup-hy-Covariate Interactions for Inmnological Variables Variable Interaction MLC Net Response Statistic Group Ranch H a n d C o m p a r i s o n n Adj. Mean 95% C.I. 129 156 68,921 77,232 0.053 (56,625, 82,424) (64,572, 91,025) X)-20 n Adj. Mean 95% C.I. 201 277 67,976 74,333 0.057 ( 6 9 0 79,930) (62,947, 86,665) 5,9, >2CMO n Adj. Mean 95% C.I. 71 88 76,511 67,758 018 .2 ( 3 0 7 91,216) (55,676, 81,025) 6,9, >40 n Adj. Mean 95% C.I. 29 29 71,116 63,991 0.444 ( 4 7 4 89,576) ( 8 8 7 81,189) 5,8, 4,3, Stratification Group-byLifetime Smoking (Pack-years) Q-3 p-Value �TABLE Q-2. Unadjusted Exposure Index Analyses for Cell Surface Markers by Occupation Exposure Index Medium High Contrast p-Value 63 1,550 (1429,1676) 56 1,578 (1449,1713) Overall M vs. L H vs. L 0.885 0.626 0.865 24 1,677 (1477,1889) 25 1,659 (1464,1866) 28 1,622 (1439,1815) Overall M vs. L H vs. L 0.923 0.901 0.697 n Mean 95% C.I. 69 1,759 (1623,1901) 78 1,555 (1434,1680) 59 1,586 (1446,1732) Overall M vs. L H vs. L 0.076 0.032 0.091 Officer n Mean 95% C.I. 62 55 63 852.7 838.3 839.6 (755.8,925. 1) (757.7,925.8) (764.5,945.8) Overall M vs. L H vs. L 0.970 0.983 0.821 Enlisted Flyer n Mean 95% C.I. 27 24 26 861.7 917.5 888.3 (789.3,1055 ,3)(767. 1,1018. 5)(744.5,987.5) Overall M vs. L H vs. L 0.832 0.755 0.545 Enlisted Groundcrev n Mean 95% C.I. 59 69 76 874.6 865.4 948.9 (867.6,1033 .9X791.3,942.7) (790.4,963.1) Overall M vs. L H vs. L 0.296 0.146 0.227 Statistic Lov n Mean 95% C.I. 62 1,594 (1471,1723) Enlisted Flyer n Mean 95% C.I. Enlisted Groundcrev Total T Cells Occupation Officer Variable o Helper T Cells �TABLE Q-2. (continued) Unadjusted Exposure Index Analyses for Cell Surface Markers by Occupation Occupation Statistic Low Officer n Mean 95% C.I. Enlisted Flyer Exposure Index Medium High p-Value 62 64 56 522.6 461.4 497.3 (467.9,583.6) (413.8,514.4) (442.7,558.7) Overall M vs. L H vs. L 0.287 0.117 0.546 n Mean 95% C.I. 28 24 26 557.7 533.9 528.1 (443.1,629.3) (451.1,631.9) (474.1,656.0) Overall M vs. L H vs. L 0.891 0.930 0.656 n Mean 95% C.I. 69 77 59 575.5 502.3 505.5 (518.9,638.3) (455.4,554.0) (451.9,565.3) Overall M vs. L H vs. L 0.123 0.063 0.097 Officer Suppressor T Cells Contrast Enlisted Groundcrew Variable n Mean 95% C.I. 55 62 64 167.4 154.1 176.8 (143.5,193.2) (131.5,178.4) (150.7,205.0) Overall M vs. L H vs. L 0.453 0.445 0.618 Enlisted Flyer n Mean 95% C.I. 24 25 26 233.2 228.6 183.3 (185.4,286.6) (182.1,280.3) (142.7,228.9) Overall M vs. L H vs. L 0.262 0.897 0.144 Enlisted Groundcrew n Mean 95% C.I. 69 73 59 202.4 196.7 199.2 (173.3,233.8) (168.8,226.8) (168.1,233.0) Overall M vs. L H vs. L 0.966 0.792 0.889 o Ul B Cells �TABLE Q-2. (continued) Unadjusted Exposure Index Analyses for Cell Surface Markers by Occupation Occupation Statistic Low Officer n Mean 95% C.I. Enlisted Flyer Exposure Index Medium p-Value 62 64 55 47.39 43.32 46.85 (35.48 ,52.89) (38.49,57.02) (38.34,58.58) Overall M vs. L H vs. L 0.799 0.584 0.546 n Mean 95% C.I. 24 26 27 37.52 41.76 36.91 (27.27 ,49.97) (28.05,50.20) (31.38,55.56) Overall M vs. L H vs. L 0.816 0.940 0.563 n Mean 95% C.I. 69 76 59 48.18 46.97 47.03 (39.65 ,55.64) (40.02,55.27) (40.12,57.87) Overall M vs. L H vs. L 0.975 0.992 0.842 Officer Monocytes Contrast Enlisted Groundcrew Variable n Mean 95% C.I. 62 63 56 564.9 527.5 544.1 (463.9 ,595.1) (480.0,612.2) (495.8,638.7) Overall M vs. L H vs. L 0.750 0.727 0.449 Enlisted Flyer n Mean 95% C.I. 24 26 28 568.3 570.3 547.1 (475.4 ,673.9) (457.7,644.5) (480.3,663.7) Overall M vs. L H vs. L 0.932 0.739 0.977 Enlisted Groundcrew n Mean 95% C.I. 69 58 76 580.7 578.0 587.3 (521.7 ,637.1) (533.2,644.0) (519.4,645.6) Overall M vs. L H vs. L 0.973 0.820 0.949 o I HLA-DR Cells High �TABLE Q-2. (continued) Unadjusted Exposure Index Analyses for Cell Surface Markers by Occupation o I Statistic Low n Mean 95% C.I. Enlisted Flyer Enlisted Groundcrew T4 /T ?8 Ratio Occupation Officer Variable Exposure Index Medium High Contrast p-Value 62 55 63 1.732 2.002 1.758 (1.500,1.964) (1.773,2.232) (1.512,2.004) Overall M vs. L H vs. L 0.207 0.106 0.881 n Mean 95% C.I. 27 24 26 1.625 1.793 1.781 (1.478,2.108) (1.478,2.083) (1.328,1.922) Overall M vs. L H vs. L 0.692 0.955 0.449 n Mean 95% C.I. 59 69 76 1.780 1.751 1.780 (1.580,1.922) (1.617,1.943) (1.595,1.965) Overall M vs. L H vs. L 0.965 0.812 0.822 �TABLE Q-3. Unadjusted Exposure Index Analyses for Functional Stimulation Tests by Occupation Low Exposure Index Medium High Contrast p-Value 64 1,428 (1249,1632) 56 1,809 (1568,2087) Overall M vs. L H vs. L 0.047 0.071 0.557 24 1,668 (1315,2114) 26 1,973 (1571,2478) 28 1,642 (1318,2045) Overall M vs. L H vs. L 0.466 0.320 0.926 n Mean 95% C.I. 69 1,814 (1583,2079) 78 1,923 (1691,2186) 58 1,843 (1588,2139) Overall M vs. L H vs. L 0.818 0.541 0.877 Officer Variable n Mean 95% C.I. 61 213,383 (192,608, 234,158) 64 205,057 (184,775, 225,339) 55 209,661 (187,782 231,539) Overall M vs. L H vs. L 0.853 0.575 0.809 Enlisted Flyer n Mean 95% C.I. 24 245,179 (212,269, 278,089) 26 209,656 (178,038, 241,275) 28 219,358 (188,890, 249,827) Overall M vs. L H vs. L 0.297 0.131 0.263 Enlisted Groundcrew n Mean 95% C.I. 69 227,062 (208,219, 245,905) 78 221,793 (204,070, 239,515) 58 208,537 (187,984, 229,089) Overall M vs. L H vs. L 0.414 0.690 0.194 Statistic Officer n Mean 95% C.I. 61 1,705 (1487,1955) Enlisted Flyer n Mean 95% C.I. Enlisted Groundcrev Unstimulated Response (PHA) Occupation 0 00 PHA Net Response �TABLE Q-3. (continued) Unadjusted Exposure Index Analyses for Functional Stimulation Tests by Occupation Exposure Index Medium High Contrast p-Value 64 86,250 (71,074, 102,894) 56 96,107 (78,992, 114,899) Overall M vs. L H vs. L 0.633 0.947 0.388 24 92,847 (71,008, 117,610) 26 101,256 (79,227, 125,985) 28 55,480 (40,058, 73,410) Overall M vs. L H vs. L 0.004 0.619 0.011 n Mean 95% C.I. 69 77,859 (64,872, 92,030) 78 92,188 (78,815, 106,610) 58 87,540 (72,540, 103,950) Overall M vs. L H vs. L 0.345 0.151 0.360 Officer Variable n Mean 95% C.I. 58 74,175 (59,443, 90,536) 63 78,069 (63,516, 94,123) 54 76,486 (61,000, 93,722) Overall M vs. L H vs. L 0.940 0.727 0.841 Enlisted Flyer n Mean 95% C.I. 23 71,959 (50,336, 97,436) 26 74,217 (53,426, 98,416) 26 62,391 (43,471, 84,721) Overall M vs. L H vs. L 0.724 0.892 0.550 Enlisted Groundcrew n Mean 95% C.I. 68 69,460 (56,591, 83,647) 77 85,901 (72,345, 100,620) 57 81,091 (65,912, 97,842) Overall M vs. L H vs. L 0.248 0.102 0.275 Low MLC Net Response n Mean 95% C.I. 62 85,479 (70,144, 102,328) Enlisted Flyer n Mean 95% C.I. Enlisted Groundcrew o Statistic Officer "Pokeweed Net Response Occupation �TABLE Q-4. Interaction Summaries of Adjusted Exposure Index Analyses for Immunological Variables Variable Interaction (Occupation) Stratificaton Statistic Low Exposure Index Medium High Contrast p-Value __ 4 3,273 0 — 5 1,557 M vs. L H vs. L 0.031 n Adj. Mean 22 118 2, 22 1,975 16 2,770 M vs. L H vs. L 0.655 0.125 n Adj. Mean 24 2,752 22 2,069 15 2,788 M vs. L H vs. L 0.071 0.941 n Adj. Mean 10 2,357 17 2,120 16 1,892 M vs. L H vs. L 0.628 0.319 >100 n Adj. Mean 1 983 1, 2 1,555 3 6,700 M vs. L H vs. L 0.725 0.049 Born >1942 n 18 Adj. Mean 261, 397 9 153,534 7 299 ,002 M vs. L H vs. L 0.002 0.302 Born n 40 1923-1941 Adj. Mean 234, 921 43 236,262 46 223 ,373 M vs. L H vs. L 0.943 0.534 Born <1922 11 229,757 1 191 ,111 M vs. L H vs. L 0.095 0.632 0 Exposure Index-byLi fe time Alcohol Use >0-5 (Drink/years) (Officer) >5-30 o i PHA Net Response n Adj. Mean >30-100 Unstimulated Response (PHA) Exposure Index-byAge (Officer) n Adj. Mean 3 139,307 �TABLE Q-5. Unadjusted Analyses for Cell Surface Markers by Group (Original Comparisons Only) Group Variable Statistic Ranch Hand Original Comparison p-Value Total T Cells n Mean 95% C.I. 464 1,606 (1551,1662) 424 1,592 (1537,1648) 0.711 Helper T Cells n Mean 95% C.I. 461 868.1 (833.3,903.6) 423 861.0 (826.4,896.3) 0.765 Suppressor T Cells n Mean 95% C.I. 465 521.9 (498.5,546.5) 425 529.7 (505.7,554.8) 0.630 B Cells n Mean 95% C.I. 457 185.2 (173.8,197.0) 418 188.1 (176.5,200.1) 0.708 n Mean 95% C.I. 462 46.19 (43.05,49.56) 424 45.47 (42.39,48.79) 0.744 n Mean 95% C.I. 462 571.6 (547.6,596.2) 424 569.9 (546.1,594.3) 0.917 n Mean 95% C.I. 461 1.597 (1.528,1.669) 421 1.558 (1.491,1.629) 0.410 Monocytes HLA-DR Cells T4/T8 Ratio Q-ll �TABLE Q-6. Adjusted Analyses for Cell Surface Markers by Group (Original Goqparisons Only) Variable Statistic Total TCells n Adj. Mean 95% C.I. Helper Cells Group Original Ranch Hand Comparison 442 p-Value 414 **** n 439 413 Adj. Mean 829 6. 867.5 95% C.I. (2.,8 6 6 (3., 9 1 1 898 9 . ) 846 0 . ) Suppressor TCells n 465 Adj. Mean 95% C.I. BCells n Adj. Mean 95% C.I. Monocytes n 440 Adj. Mean 95% C.I. HLA-CR Cells n Adj. Mean 95% C.I. 085 .3 425 455 416 155 8. 111 9. 041 .5 (174.2, 197.1) (179.6, 202.9) Covariate Remarks* BATCH ( < . 0 ) p001 GRP*RACE (p=0.028) CRKXR*PACKYR ( = . 0 ) p004 CSMCK (p<0.001) ALC (p=0.005) AGE (p=0.049) BATCH (p=0.014) DAY (BATCH) (p=0.007) AGE (p<0.001) CSMCK (p<0.001) ALC*OCC (p=0.025) (=.0) p001 BATCH (p<0.001) GRP*RACE (p=0.010) OCC (p=0.021) A3E (p=O.C06) CSMCK (p<0.001) BATCH (p<0.001) OOC (p=0.041) AGE (p=0.005) ALC (p=0.002) CSMCK (p<0.001) 414 BATCH (p<0.001) MY (BATCH) (p<0.001) GRP*AGE (p=0.040) OCC (p=0.010) EKCZR (p=0.008) CSMCK (p<0.001) PACKYR (p=0.020) 461 423 570.5 568.2 0.887 ( 4 . , 594.5) ( 4 . , 592.0) 569 550 BATCH (p<0.001) DAY (BATCH) (p4).014) OCC (p=0.017) CSMCK (p<0.001) AGE*PACKXR (p=0.007) CH12 �TAHE Q-6. (continued) Adjusted Analyses for Cell Surface Markets by Group (Original Gon|Brisans Only) Group Variable Statistic Ranch Hand T /T Ratio n 461 Adj. Mean **** 95% C.I. **** Original Comparison 421 **** **** p-Value **** Covariate Remarks* MICH (p<0.001) OOC (p=0.002) GRP*CSMCK (p=0.016) ^Abbreviations; BA3XH: batch-to-batch variation among examination groups DAY (MTCH): blood-draw day variation ALC: current alcohol use CSMGK: current smoking OOC: occupation GRP: group EHOR: lifetime alcohol use (drink-years) PACKER: lifetime smoking (pack-years) ****Group-by-covariate interaction - adjusted mean, confidence interval, and p-value not presented. Q-13 �TAHEQ-7. Sunnary of Group-4y-0ovarlate Interactions for Cell Surface Markers (Original Comparisons Only) Group Original Comparison p-Value Variable Interaction Stratification Statistic Banch Hand Total TCells Group-by-Race Nonblack n Adj. Mean 95% C.I. 424 391 0.324 1,601.7 1,567.3 (1,554.6, 1,649.6) (1,518.6, 1,161.6) Black n Adj. Mean 95% C.I. 18 23 1,558.3 1,888.6 0.043 (1,338.6, 1,794.8) (1,673.1, 2,117.1) Nonblack n Adj. Mean 95% C I .. 444 530.0 ( 0 . , 555.6) 555 402 530.8 ( 0 . , 556.6) 562 0.960 n Adj. Mean 95% C.I. 21 497.5 ( 1 . , 602.2) 410 23 708.8 ( 8 . , 852.2) 596 0.008 n Adj. Mean 95% C.I. 174 41.39 (68, 4.4 3.8 64) 162 48.02 ( 2 8 , 53.76) 4.9 0.048 Bom 1923-1941 n Adj. Mean 95% C.I. 247 48.20 ( 4 0 , 52.76) 4.4 232 42.82 ( 9 0 , 46.99) 3.2 0.058 Born <L922 n Adj. Mean .95% C.I. 19 47.88 ( 5 0 , 65.40) 3.6 20 46.90 (34.85, 63.13) 0.924 n Adj. Mean 95% C.I. 277 1.44 (1.36, 1 5 ) .2 273 1.48 (.0 1 5 ) 14, . 6 0.454 n Adj. Mean 95% C.I. 78 1.70 (.4 1 8 ) 15, . 8 69 1.73 (1.56, 1 9 ) .2 0.811 n Adj. Mean • 95% C.I. 90 1.84 (.8 2 0 ) 16, . 1 71 1.51 (1.36, 1 6 ) .7 0.004 n Adj. Mean 95% C.I. 16 1.70 (1.38, 2 1 ) .1 8 1.75 (1.30, 2 3 ) .5 0.895 Suppressor Group-by-Pace TCells Black Monocytes Group-by-Age T4/T 8 Ratio Group-byCurrent Smoking (Cigarettes/ day) Born >1942 0 X)-20 >2CWO >40 0-14 �TABLE Q-8. Unadjusted Analyses for Functional Stimulation Tests by Group (Original Comparisons Only) Group Ranch Hands Original Comparison p-Value 464 1,669 (1,578, 1,755) 426 1,640 (1,559, 1,724) 0.608 n Mean 95% C.I. 463 212,481 425 206,796 0.168 Pokeweed Net Response n Mean 95% C.I. 465 85,559 (81,373, 89,849) 426 83,724 (79,579, 87,974) 0.522 MLC n Net Response Mean 95% C.I. 452 79,451 (75,485, 83,519) 415 82,387 (78,363, 86,512) 0.285 Variable Statistic Unstimulated n Response (PHA) Mean 95% C.I. PHA Net Response (206,424, 218,539) (200,733, 212,858) Q-15 �TABLE Q-9. Adjusted Analyses for Functional Stimulation Tests by Group (Original Comparisons Only) Group Ranch Hand Original Comparison p-Value Unstimulated n Response Adj. Mean (PHA) 95% C.I. 464 1,813 (1,645, 1,997) 426 1,783 (1,620, 1,962) 0.613 PHA Net n 461 423 Response Adj. Mean **** Variable Statistic **** 95% C.I. Pokeweed n 465 426 Net Response Adj. Mean 92,684 91,738 95% C I . . (86,712, 98,855) (85,660, 98,023) MLCNet Response n 431 405 Adj. Mean 79,412 82,916 95% C I . . (73,680, 85,359) ( 6 8 8 89,214) 7,4, Covariate Remarks BATCH (p<0.001) DAY (BATCH) (p<0. 0 ) 01 RACE (p<0.001) AGE (p<0.001) BATCH (p<O.C01) DAY (BATCH) (p<0.001) GRP*OCC (p=0.017) RACE (p=0.005) AGE (p4).001) AUXSMOK (p*0.037) BATCH (p<0.001) 0.746 DAY (BATCH) (p<0.001) OCC (p=0.046) CSMOK (p<0.001) BATCH (p<0.001) 0 1 7 DAY (BATCH) (p<0.001) .9 AUC (p<0.001) EBK2R (p=0.001) CSMCK (p<0.001) ****Group-by-covariate interaction - adjusted mean, confidence interval, and p-value not presented. Q-16 �TABLE Q-10. Sunnary of Group-by-Covarlate Interactions for Functional Stimulation Tests (Original Comparisons Only) Group Original Comparison Variable Interaction Stratification Statistic Ranch Hand PHA Net Pesponse Group-byOccupation Officer n Adj. Mean 95% C.I. 180 163 217,397 210,363 0.280 (201,412, 233,382) (193,843, 226,883) Enlisted Flyer n Adj. Mean 95% C.I. 78 77 230,597 202,077 0.003 (212,032, 249,162) (183,749, 220,405) Enlisted Groundcrev n Adj. Mean 95% C.I. 203 183 208,347 212,473 0.504 (193,386, 223,308) (197,550, 227,397) 0 n Adj. Mean 95% C.I. 129 109 78,900 87,110 0.107 (69,737, 88,627) (77,309, 9 , % 74) >0-20 n Adj. Mean 95% C.I. 201 206 76,689 83,468 009 .7 (69,735, 83,973) (76,169, 91,101) n Adj. Mean 95% C.I. 71 67 86,904 75,456 0.086 (77,241, 97,135) (65,854, 85,711) n Adj. Mean 95% C.I. 29 23 80,525 75,984 0.679 (66,260, 96,180) ( 0 8 6 92,850) 6,0, MLC Net Response Group-byLifetime Smoking (Pack-years) >40 Q-17 p-Value �APPENDIX R Pulmonary Disease �APPENDIX R: Pulmonary Disease Contents Table R-l R-2 R-3 R-4 R-5 R-6 R-7 R-8 R-9 R-10 Page Unadjusted Analyses of Respiratory Variables by Group (Original Comparisons Only). . . R-l Adjusted Analyses of Respiratory Variables by Group (Original Comparisons Only) R-3 Summary of Group-by-Covariate Interactions for Respiratory Variables (Original Comparisons Only) R-5 Cross Tabulation of Bronchitis- (Abnormal, Total) byExposure Index Category-by-Age Category-by-Pack-Years for Ranch Hand Officers R-9 Cross Tabulation of Pneumonia- (Abnormal, Total) byExposure Index Category-by-Age Category-by-Pack-Year Category for Ranch Hand Officers R-10 Cross Tabulation by Hyperresonance- (Abnormal, Total) byExposure Index Category-by-Pack-Year Category for Ranch Hand Officers R-ll Cross Tabulation of Thorax- and Lung- (Abnormal, Total) by-Exposure Index Category for Ranch Hand Officers R-12 Cross Tabulation of X Ray- (Abnormal, Total) by-Exposure Index Category-by-Age Category-by-Pack-Year Category for Ranch Hand Officers R-13 Cross Tabulation of Bronchitis- (Abnormal, Total) by-Exposure Index Category-by-Age Category-by-Pack-Year Category for Ranch Hand Enlisted Flyers R-14 Cross Tabulation of Pleurisy- (Abnormal, Total) byExposure Index Category-by-Age Category-by-Pack-Year Category for Ranch Hand Enlisted Flyers R-15 R-ll Cross Tabulation of Pneumonia- (Abnormal, Total) byExposure Index Category-by-Age Category for Ranch Hand Enlisted Flyers R-16 R-12 Cross Tabulation of Hyperresonance- (Abnormal, Total) byExposure Index Category-by-Pack-Year Category for Ranch Hand Enlisted Flyers R-17 �APPENDIX R: Pulmonary Disease Contents (continued) Table R-13 Cross Tabulation of Thorax and Lung Abnormal!ties(Abnormal, Total) by-Exposure Index Category for Ranch Hand Enlisted Flyers R-18 R-14 Cross Tabulation of Asthma- (Abnormal, Total) byExposure Index Category-by-Age Category-by-Pack-Year Category for Ranch Hand Enlisted Groundcrev R-19 R-15 Cross Tabulation of Pleurisy- (Abnormal, Total) byExposure Index Category-by-Age Category for Ranch Hand Enlisted Groundcrew. R-20 R-16 Cross Tabulation of Wheezes- (Abnormal, Total) byExposure Index Category-by-Age Category for Ranch Hand Enlisted Groundcrew R-21 R-17 Cross Tabulation of Wheezes- (Abnormal, Total) byExposure Index Category-by-Pack-Year Category for Ranch Hand Enlisted Groundcrew R-22 R-18 Cross Tabulation of Bronchitis- (Abnormal, Total) byExposure Index Category for Ranch Hand Enlisted Groundcrew R-23 �TABLE R-l. Unadjusted Analyses of Respiratory Variables by Group (Original Comparisons Only) Group Ranch Hand Original Comparison Number Percent Est. Relative Risk (95% C.I.) p-Value Variable Statistic Number Percent Asthma n Abnormal Normal 1,016 44 972 4.3 95.7 954 42 912 4.4 95.6 0.98 (0.64,1.51) 0.92 Bronchitis n Abnormal Normal 1,015 129 886 12.7 87.3 954 131 823 13.7 86.3 0.91 (0.71,1.19) 0.50 Pleurisy n Abnormal Normal 1,016 47 969 4.6 95.4 953 45 908 4.7 95.3 0.98 (0.65,1.49) 0.92 Pneumonia n Abnormal Normal 1,016 195 821 19.2 80.8 953 191 762 20.0 80.0 0.95 (0.76,1.18) 0.64 Tuberculosis n Abnormal Normal 1,015 7 1,008 0.7 99.3 954 5 949 0.5 99.5 1.32 (0.43,3.97) 0.64 Thorax & Lungs n Abnormal Normal 1,015 61 954 6.0 94.0 955 45 910 4.7 95.3 1.29 (0.87,1.92) 0.20 �TABLE R-l. (contined) Unadjusted Analyses of Respiratory Variables by Group (Original Comparisons Only) Group Ranch Hand Original Comparison Number Percent n Asymmetrical Exp. Abnormal Normal 1,015 2 1,013 0.2 99.8 955 2 953 0.2 99.8 0.94 (0.16,5.49) 0.95 Hyperresonance n Abnormal Normal 1,015 30 985 3.0 97.0 955 28 927 2.9 97.1 1.01 (0.60,1.69) 0.98 Dullness n Abnormal Normal . 1,015 2 1,013 0.2 99.8 955 0 955 0.0 100.0 Wheezes n Abnormal Normal 1,015 24 991 2.4 97.6 955 16 939 1.7 98.3 1.42 (0.76,2.67) 0.28 Rales n Abnormal Normal 1,015 6 1,009 0.6 99.4 955 4 951 0.4 99.6 1.41 (0.42,4.70) 0.59 X ray n Abnormal Normal 1,012 102 910 10.1 89.9 951 113 838 11.9 88.1 0.83 (0.63,1.10) 0.20 Statistic Percent Est. Relative Risk (95% C.I.) p-Value Number Variable 0.17 �TABLE R-2. Adjusted Analyses of Respiratory Variables by Group (Original Comparisons Only) Group Variable Ranch Hand Original Comparison Adj. Relative Risk ( 5 C.I.) 9% p-Value Covariate Remarks* **** PACKYR (p=0.02) GRP*PACKYR (p=0.03) 0.44 GRP*PACKYR (Borderline: p=0.09) **** GRP*PACKYR (p=0.0006) 0.97 (0.77,1.22) 0.74 AGE (p=0.002) 953 1.32 (0.43,3.98) 0.60 GRP*PACKYR (Borderline: p=0.06) 1,011 954 1.33 (0.88,2.01) 0.14 AGE (p<0.0001) PACKYR (p<0.0001) Asymmetrical Exp. 1,011 954 0.94 (0.16,5.50) 0.97 AGE*PACKYR (Borderline: p=0.08) Hyperresonance 1,011 954 1.00 (0.58,1.74) 0.87 AGE (p<0.0001) PACKYR (p<0.0001) GRP*PACKYR (Borderline: p=0.09) As thma 1,012 953 Bronchitis 1,011 953 Pleurisy 1,012 952 Pneumonia 1,012 952 Tuberculosis 1,011 Thorax and Lungs **** 0.91 (0.70,1.18) **** �TABLE R-2. (continued) Adjusted Analyses of Respiratory Variables by Group (Original Comparisons Only) Group Variable Ranch Hand Original Comparison Adj. Relative Risk (95% C.I.) Dullness 1,011 954 ¥heezes 1,011 954 — 1.43 (0.75,2.76) 0.25 PACKYR (p<0.0006) Rales 1,011 954 1.42 (0.39,5.18) 0.59 AGE*PACKYR (p=0.04) GRP*AGE (p=0.09) X Ray 1,008 950 0.86 (0.64,1.16) 0.31 AGE (p<0.0001) PACKYR (p=0.02) GRP*PACKYR (Borderline: p=0.07) 50 **Abbreviations PACKYR: lifetime smoking history (pack-years) GRP: group p-Value Covariate Remarks* None �TABIER-3. Suomary of Group-by-Covariate Interactions for Respiratory Variables x~—o— —~ ~ i isons *—--j f ~ Group Variable Interaction Asthma Group-byPack-Year Stratification 0 Statistic Ranch Hand Number Percent Original Comparison Number Percent Adj. Relative Risk ( 5 C I ) p-Value 9%.. >10 £ Bronchitis Group-byPack-Year 0 >0-10 >10 291 11 280 3.78 96.22 268 3 265 1.12 98.88 4.78 ( . 9 1 . 6 0 0 13,55) .4 n Abnormal Normal 284 16 268 5.63 94.37 280 12 268 4.29 95.71 1.33 ( . 3 2 8 ) 0 4 06,.3 .6 Abnormal Normal XKLO n Abnormal Normal 437 17 420 3.89 96.11 405 27 378 6.67 93.33 0.57 ( . 1 1 0 ) 0 0 03,.5 . 7 n Abnormal Normal 291 38 253 1.6 30 8.4 69 268 25 243 9.33 90.67 1 4 (.624) 0 1 . 6 08,.8 .6 n Abnormal Normal 284 38 246 13.38 86.62 280 40 240 1.9 42 85.71 0.93 ( . 8 1 4 ) 0 7 05,.9 . 5 n Abnormal Normal 436 52 384 11.93 8.7 80 405 66 339 16.30 8.0 37 0.70 ( . 7 1 0 ) 0 0 04,.3 . 7 n �TABIE R-3. (continued) Sunnary of Group-by-Covariate Interactions for Respiratory Variables Comparisons Only) Group Variable Interaction Pleurisy Group-byPack-Year Stratification 0 XKLO >10 Tuberculosis Group-byPack-Year 0 >0-10 >10 Statistic Ranch Hand Number Percent Original Comparison Number Percent Adj. Relative Risk ( 5 C I ) p-Value 9% .. n Abnormal Normal 291 8 283 2.75 97.25 267 5 262 1.87 98.13 1.48 ( . 0 4 3 ) 0 4 05,.7 . 9 n Abnormal Normal 284 18 266 6.34 9.6 36 280 4 276 1.43 98.57 4.67 ( . 1 1 . 2 O 0 1 16,30) . 0 n Abnormal Normal 437 21 416 48 .1 95.19 405 36 369 88 .9 91.11 0.52 ( . 0 0 9 ) 0 0 03,.0 . 2 n Abnormal Normal 290 I 289 0.34 99.66 268 3 265 1.12 9.8 88 0.31 ( . 5 2 3 ) 0 2 00,.4 . 8 n Abnormal Normal 284 4 280 1.41 98.59 280 0 00 .0 280 1 0 0 0.0 n Abnormal Normal 437 2 435 04 .6 99.54 405 2 403 04 .9 99.51 — 00 .5 0.93 ( . 6 5 4 ) 0 9 01,.2 . 2 �TABLE R-3. (continued) (Original Comparisons Oily) Group Variable Interaction Byperresonnance Group-byPack-Year Stratification Adj. Relative Risk ( 5 C I ) p-Value 9%.. 291 2 289 0.69 99.31 268 00 .0 0 268 1 0 0 0.0 n Abnormal Normal 283 8 275 2.83 97.17 281 3 278 1.07 98.93 2.70 ( . 5 9 2 ) 0 1 07,.7 . 3 n Abnormal Normal 437 20 417 4.58 95.42 405 25 380 6.17 93.83 0.73 ( . 0 1 3 ) 0 3 04,-3 .0 n Abnormal Normal 384 1 383 0.26 9.4 97 341 0 00 .0 341 1 0 0 0.0 Born 1922-1942 n Abnormal Normal 600 5 595 0.83 99.17 577 4 573 >10 Group-byAge Original Comparison Number Percent n Abnormal Normal 0 XKLO Bales Statistic Ranch Hand Number Percent Bom >1942 Born <L922 n Abnormal Normal 27 0 00 .0 27 1 0 0 0.0 0.69 99.31 36 0 00 .0 36 1 0 0 0.0 — — 01 .7 03 .5 1.20 ( . 4 4 2 ) 0 7 03, . 1 .8 — — �TfiBLE R-3. (continued) Sunnary of Gxxip-by-Gowariate Interactions for Respiratory \feriables (Original Gonjjariscns Only) Group Variable Interaction XRay Group-byPack-Year Stratification 0 >0-10 >10 Statistic Ranch Band Hunter Percent Original Comparison Number Percent Adj. Relative Risk ( 5 C I ) p-Valxe 9% .. n Abnormal Normal 290 15 275 51 .7 9.3 48 266 29 237 1.0 09 8.0 91 n Abnormal NornBl 282 28 254 9.93 90.07 281 26 255 9.25 90.75 1.08 (0.62, 1.89) 0.79 n Abnornal Normal 436 59 377 13.53 86.47 403 58 345 14.39 85.61 0.93 (0.63, 1.37) 0.72 0 4 (.408) 0 0 . 5 02,.5 . 1 �OFFICERS The statistically significant bronchitis-by-exposure-by-age-by-pack-year shown in Table 20-5 is described here in Table R-4. interaction (p=0.009) TABLE R-4. Cross Tabulation of Bronchitis- (Abnormal, Total) by-Exposure Index Category-by-Age Category-by-Pack-Year Category for Ranch Hand Officers Exposure Index Pack-Years Age Low Abnormal Total Percent Medium Abnormal Total Percent High Abnormal Total Percent 0 50 VO Born >1942 Born 1922-1942 Born <1922 4 3 0 21 32 3 19 9 0 0 7 0 10 36 3 0 19 0 1 8 0 3 44 0 33 10 >0-10 Born >1942 Born 1922-1942 Born <1922 1 2 0 14 15 0 7 13 2 5 0 7* 19 3 29 26 0 0 1 0 0 26 1 4 0 Born >1942 Born 1922-1942 Born <1922 0 8 0 3 39 0 0 21 2 3 1 4 43 5 50 7 20 0 4 0 4 44 1 0 9 0 >10 �The statistically significant pneumonia-by-exposure-by-age-by-pack-year Table 20-5 is described here in Table R-5. interaction (p=0.040) shown in TABLE R-5. Cross Tabulation of Pneumonia- (Abnormal, Total) by-Exposure Index Category-byAge Category-by-Pack-Year Category for Ranch Hand Officers Exposure Index Low Pack-Years Age Abnormal Total Medium Percent Abnormal Total High Percent Abnormal Total Percent 0 Born >1942 Born 1922-1942 Born <1922 6 6 0 21 22 3 29 19 0 0 6 1 10 ' 36 3 0 17 33 1 10 0 3 44 0 33 23 >0-10 Born. >1942 Born 1922-1942 Born <1922 1 4 0 14 15 0 7 27 1 4 0 7 19 3 14 21 0 0 9 1 0 26 1 35 100 Born >1942 Born 1922-1942 Born <1922 0 9 0 3 39 0 0 23 1 6 1 4 43 5 25 14 20 0 8 0 4 44 1 0 18 0 >10 �The statistically significant hyperresonance-by-exposure-by-pack-year (p=0.07) seen in Table 20-5 is tabulated in Table R-6. interaction TABLE R-6. Cross Tabulation of Hyperresonance- (Abnormal, Total) by-Exposure Index Categoryby-Pack-Year Category for Ranch Hand Officers Exposure Index Low Pack-Years Abnormal Total Percent Medium Abnormal Total Percent High Abnormal Total Percent 0 0 56 0 0 49 0 1 47 2 >0-10 0 29 0 1 29 3 1 27 4 >10 5 42 12 1 52 2 0 49 0 �The statistically significant thorax- and lung-by-exposure category interaction (p=0.05) shown in Table 20-5 is tabulated in Table R-7. TABLE R-7. Cross Tabulation of Thorax and Lung- (Abnormal, Total) by- Exposure Index Category for Ranch Hand Officers Exposure Index Low Abnormal Total 9 l-» NS 127 Percent 7 Medium Abnormal Total Percent 3 130 2 High Abnormal Total Percent 5 123 4 �The statistically borderline significant x ray-by-exposure-category-byage-by-pack-year interaction (p=0.08) seen in Table 20-5 is tabulated in Table R-8. TABLE R-8. Cross Tabulation of X Ray- (Abnormal, Total) by-Exposure Index Categoryby-Age Category-by-Paek-Year Category for Ranch Hand Officers Exposure Index Pack-Years Age Low Abnormal Total Percent Medium Abnormal Total Percent Abnormal High Total Percent 0 50 h-» 00 Born >1942 Born 1922-1942 Born <1922 1 3 0 21 32 3 5 9 0 0 1 0 10 36 3 0 3 0 0 2 0 3 44 0 0 5 >0-10 Born >1942 Born 1922-1942 Born <1922 0 3 0 14 15 0 0 20 0 3 1 6 19 3 0 16 33 0 2 0 0 26 1 8 0 Born >1942 Born 1922-1942 Born <1922 0 7 0 3 39 0 0 18 0 7 1 4 43 5 0 16 20 0 2 1 4 44 1 0 5 100 >10 �ENLISTED FLYERS The statistically significant bronchitis-by-exposure-by-age-by-pack-year seen in Table 20-6 is tabulated in Table R-9. interaction (p=0.005) TABLE R-9. Cross Tabulation of Bronchitis- (Abnormal, Total) by-Exposure Index Category-by-Age Category-by-Pack-Year Category for Ranch Hand Enlisted Flyers Exposure Index Pack-Years Age Low Abnormal Total Percent Medium Abnormal Total Percent High Abnormal Total Percent 0 50 Born >1942 Born <1942 2 0 3 6 67 0 0 0 3 6 0 0 0 1 3 11 0 9 >0-10 Born >1942 Born <1942 1 1 4 13 25 8 1 1 6 9 17 11 2 0 2 7 100 0 >10 Born >1942 Born <1942 0 3 2 26 0 12 0 6 7 32 0 19 0 3 2 31 0 10 �The statistically borderline significant pleurisy-by-exposure-by-age-by-pack-year interaction (p=0.08) shown in Table 20-6 is tabulated in Table R-10. TABLE R-10. Cross Tabulation of Pleurisy- (Abnormal, Total) by-Exposure Index Category-byAge Category-by-Pack-Year Category for Ranch Hand Enlisted Flyers Exposure Index Pack-Years Age Low Abnormal Total Percent Medium Abnormal Total Percent High Abnormal Total Percent 0 Born >1942 Born <1942 0 0 3 6 0 0 0 0 3 6 0 0 0 0 3 11 0 0 >0-10 Born >1942 Born <1942 1 1 4 13 25 8 0 1 6 9 0 11 1 0 2 7 50 0 >10 Born >1942 Born <1942 1 1 2 26 50 4 0 1 7 32 0 3 0 2 2 32 0 6 �The statistically borderline significant pneumonia-by-exposure-by-age interaction (p=0.08) shown in Table 20-6 is tabulated in Table R-ll. TABLE R-ll. Cross Tabulation of Pneumonia- (Abnormal, Total) -by-Exposure Index Category-by-Age Category for Ranch Hand Enlisted Flyers Exposure Index Age Born >1942 Born <1942 JO . Low Abnormal Total Percent 4 10 9 45 44 22 Medium Abnormal Total Percent 2 10 16 47 13 21 High Abnormal Total 0 13 7 50 Percent 0 26 �The statistically borderline significant hyperresonance-by-exposure-by-pack-year category interaction (p=0.08) shown in Table 20-6 is tabulated in Table R-12. TABLE R-12. Cross Tabulation of Hyperresonance- (Abnormal, Total) by-Exposure Index Category-by-Pack-Year Category for Ranch Hand Enlisted Flyers Exposure Index Pack-Years Abnormal 0 >0-10 >10 I l-» -J 0 0 3 Low Total 9 17 28 Percent 0 0 11 Medium Abnormal Total Percent 0 2 4 9 15 39 0 13 10 High Abnormal Total 0 0 0 14 9 34 Percent 0 0 0 �The statistically significant (p=0.04) thorax and lung-by-exposure interaction shown in Table 20-6 is tabulated in Table R-13. TABLE R-13. Cross Tabulation of Thorax and Lung Abnormalities- (Abnormal, Total) by-Exposure Index Category for Ranch Hand Enlisted Flyers Exposure Index Low Abnormal Total 7 00 54 Percent 13 Medium Abnormal Total Percent 10 63 16 High Abnormal Total 2 57 Percent 4 �ENLISTED GROUND The statistically significant asthma-by-exposure-by-age-by-pack-year interaction (p=0.02) shown in Table 20-7 is tabulated in Table R-14. TABLE R-14. Cross Tabulation of Asthma- (Abnormal, Total) by-Exposure Index Category-by-Age Category-by-Pack-Year Category for Ranch Hand Enlisted Groundcrev Exposure Index Pack-Years 50 Age Low Abnormal Total Percent Medium Abnormal Total Percent High Abnormal Total Percent 0 Born >1942 Born 1922-1942 Born <1922 1 0 0 24 18 0 4 0 1 0 0 35 6 0 3 0 1 0 0 15 9 0 7 0 >0-10 Born >1942 Born 1922-1942 Born <1922 0 1 0 37 16 0 0 16 2 1 0 47 7 0 4 14 4 0 0 36 13 1 11 0 0 >10 Born >1942 Born 1922-1942 Born <1922 0 4 0 22 33 3 0 12 0 4 0 0 46 21 1 9 0 0 0 2 1 25 37 5 0 5 20 �The statistically significant pleurisy-by-exposure-by-age interaction (p=0.03) shown in Table 20-7 is tabulated in Table R-15. TABLE R-15. Cross Tabulation of Pleurisy-(Abnormal, Total) by-Exposure Index Category-byAge Category for Ranch Hand Enlisted Groundcrev Exposure Index Age Low Abnormal Total Born >1942 Born 1922-1942 Born <1922 to o 5 7 0 83 67 3 Percent 6 10 0 Medium Abnormal Total Percent 5 0 0 128 34 1 4 0 0 High Abnormal Total 0 5 0 76 59 6 Percent 0 8 0 �The statistically significant wheeze-by-exposure-by-age interaction (p=0.009) shown in Table 20-7 is tabulated in Table R-16. TABLE R-16. Cross Tabulation of Wheezes- (Abnormal, Total) by-Exposure Index Category-byAge Category for Ranch Hand Enlisted Groundcrev Exposure Index Age Low Abnormal Total Born >1942 Born 1922-1942 Born <1922 2 1 0 83 67 3 Percent 2 1 0 Abnormal 1 3 0 Medium Total Percent Abnormal High Total Percent 1 9 0 3 2 0 76 59 6 4 3 0 127 34 1 �The statistically significant wheeze-by-exposure-by-pack-year interaction (p=0.02)' shown in Table 20-7 is tabulated in Table R-17. TABLE R-17. Cross Tabulation of Wheezes- (Abnormal, Total) by-Exposure Index Category-byPack-Year Category for Ranch Hand Enlisted Groundcrew Exposure Index Low Pack-Years Abnormal Total Percent Medium Abnormal Total Percent High Abnormal Total Percent 0 42 0 3 41 0 0 24 0 >0-10 to to 0 2 53 4 3 53 6 1 50 2 >10 1 58 2 1 68 1 4 67 6 �The statistically borderline significant bronchitis-by-exposure category interaction (p=0.08) shown in Table 20-7 is tabulated in Table R-18. TABLE R-18. Cross Tabulation of Bronchitis- (Abnormal, Total) by-Exposure Index Category for Ranch Hand Enlisted Groundcrev Exposure Index Lov AbnormalTotalPercent 25 153 16 Medium Abnormal Total Percent 19 163 12 ' High Abnormal Total 11 141 Percent 8 �APPENDIX S Glossary of Abbreviations �APPENDIX S: Glossary of Abbreviations Contents Table Page S-l Glossary of Technical and Medical Abbreviations S-l S-2 Glossary of Covariate and Statistical Abbreviations S-5 S-i �TABLE S-l GLOSSARY OF TECHNICAL AND MEDICAL ABBREVIATIONS CHECO - replacement Comparisons interviewed by Air Force staff contemporary historical evaluation and combat operations CHOL cholesterol CMI -A- Cornell Medical Index ACA® - Automated Chemical Analyser CNS central nervous system ACTH - adrenocorticotropic hormone COPRO coproporphyrins AFHS - Air Force Health Study CPK creatine phosphokinase ALA - delta-aminolevulinic acid cpm counts per minute ALKPHOS - alkaline phosphatase CUSUM cumulative sum CV coefficient of variation -BBA - Bachelor of Arts BMDP-4F® - log-linear program DBF diastolic blood pressure BMDP-LR® - logistic regression program DNA deoxyribonucl'eic acid BS - Bachelor of Science BSL-2 - National Institutes of Health Biosafety Level 2 ECG electrocardiograph - blood urea nitrogen EEC elec t roencephalogram -D- -E- erythema BUN -C- -F- CAD - coronary artery disease FC fully compliant CAP - College of American Pathology FEF 25-75 forced midexpiratory flow rate CATI - computer-aided telephone interview FEV forced expiratory volume CBC - complete blood count CPU - colony forming unit CHD - coronary heart disease forced expiratory volume in one second FIR CUSUM S-l - fast initial response cumulative sum �TABLE S-l (continued) GLOSSARY OF TECHNICAL AND MEDICAL ABBREVIATIONS FSH follicle stimulating hormone FTA fluorescent treponemal antibody FTI free thyroxine index FVC forced vital capacity ICD-9-CMD - International Classification of Disease 9th Revision, Clinical Modification IQ -K- K-S -GGED General Equivalency Diploma GGTP - Kolmogorov-Smirnov -L- gamma-glutamyl transpeptidase GLUC - intelligence quotient glucose LBBB - left bundle branch block LDH - lactic dehydrogenase LH - leuteinizing hormone -HHB8Ag hepatitis B surface antigen HCT hematocrit HD Hodgkin's Disease HDL high density lipoprotein HGB hemoglobin HLA histocompatibility antigens HPF HRB -M- MCH - mean corpuscular hemoglobin MCHC - mean corpuscular hemoglobin concentration MCV - mean corpuscular volume MI - myocardial infarction ML - malignant lymphoma high-power field MLC - mixed lymphocyte culture Halstead-Reitan battery MMPI - Minnesota Multiphasic Personality Inventory MMPID - MMPI depression scale -II induration IARC International Agency for Research on Cancer -N- ICD International Classification of Disease S-2 NIOSH - National Institute for Occupational Safety and Health �TABLE S-l (continued) GLOSSARY OP TECHNICAL AND MEDICAL ABBREVIATIONS NC - noncorapliant PULM - FEVj/FVC ratio NCI - National Cancer Institute PULSE - pulse index NCVA - nerve conduction velocity above the ankle PWM - pokeweed mitogen NCVE - nerve conduction velocity above the elbow NHL - non-Hodgkin's lymphoma NKC - natural killer cell NORC - National Opinion Research Center NOS - not otherwise specified -QQA - quality assurance QC - quality control QRC - Quality Review Committee -RR RBBB -0- - replacement Comparisons - right bundle branch block 0 - original Comparisons RBC - red blood cell OHR - Optical Mark Recognition RIA - radioimmunoassay RPR - rapid plasma reagin RVN - Republic of Vietnam -PPBM - peripheral blood mononuclear PC - partially compliant PCB - polychlorinated biphenyl PCT - porphyria cutanea tarda PHA - phytohemagglutinin PLAT - platelet count PLT - platelet count PMR - proportionate mortality ratio POMS -' profile of mood states PTSD - post-traumatic stress disorder -SS SAIC - Science Applications International Corporation SAS® - Statistical Analysis System SAS®-GLM - Statistical Analysis System general linear model SCRF - Scripps Clinic and Research Foundation SEA S-3 - shifted original Comparisons - Southeast Asia �TABLE S-l (continued) GLOSSARY OF TECHNICAL AND MEDICAL ABBREVIATIONS SED sedimentation rate SEMEN semen count WAIS - Wechsler Adult Intelligence Scale SCOT serum glutamic-oxaloacetic transaminase WBC - white blood cell SGPT serum glutamic-pyruvic transaminase SIRL Scripps Immunology Reference Laboratory SKIN skin index STS -W- soft tissue sarcoma -T- triiodothyronine 4 total thyroxine TBILI total bilirubin TCDD 2,3,7,8-tetrachlorodibenzo-p-dioxin TEST testosterone TLC total lymphocyte count TSH thyroid stimulating hormone -U- USG urine specific gravity -V- VA Veterans Administration S-4 �TABLE S-2. Glossary of Covariate and Statistical Abbreviations d.f. Adj. RR - differential cortisol DRKYR - current alcohol use (drinks/day) - diabetic class DIFCORT ALC - degrees of freedom DIAB -A- - lifetime alcohol history (drink-years) - adjusted relative risk -B- BATCH - batch-to-batch variation among examination groups BUN - blood urea nitrogen -E- EDUC - education Est. RR - estimated relative risk -C-G- C.I. - confidence interval CC - continuous dependent variable, adjusted by continuous covariates GRP CD - group -H- - continuous dependent variable, adjusted by discrete covariates H - high exposure index level HDL - high density lipoprotein cholesterol CHOL - cholesterol CHOL/HDL - cholesterol to HDL ratio CI - combat index 1C - exposure to industrial chemicals CSMOK - current smoking (packs/day) INS - exposure to insecticides -I- -D- -L- DAY(BATCH) - blood-draw day variation vtihin batch LOG - discrete dependent variable, adjusted by discrete covariates - average lifetime residential latitude - logarithmic - exposure to degreasing chemicals DD - low exposure index level LAT DC L S-5 �TABLE S-2. (continued) Glossary of Covariate and Statistical -M- M medium exposure index level -N- NS not significant -0- OCC occupation OR odds ratio -P- PACKYR lifetime smoking history (pack-years) 2BFAT percent body fat PERSTYPE personality type PS personality score PTSD post-traumatic stress disorder -R- RR relative risk -S- SE standard error SEA ACNE presence of pre-SEA acne SKIN skin color SORT square root SUNREAC sun-reaction index -tfWINE wine consumption S-6 Abbreviations � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 059 Folder The folder containing the original item. 1586 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. William H. Wolfe Joel E. Michalek Judson C. Miner Michael R. Peterson Stella G. Machado Theodore G. Karrison William D. Grubbs Wanda F. Thomas Date A point or period of time associated with an event in the lifecycle of the resource 1987-10-01 Title A name given to the resource Air Force Health Study: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Volume II, First Followup Examination Results, January 1985-September 1987 Subject The topic of the resource Air Force Health Study dioxin morbidity trends ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/281d8a128a8712e513d8bba495ef4193.pdf 2917c78f2f135bda34d4006615f46b59 PDF Text Text Item ID Number 01535 Author Lathrop, George D. Corporate Author Report/Article TltlB Air Force Health Study: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Volume I, First Follow/up Examination Results, January 1985-September 1987 Journal/Book Title Year 1987 Month/Day October Color n Number of Images 632 nnsnrlntnn NntHg * Contract no. F41689-85-D-0010 and SAIC Project no. 2816-XX-195/254-XX . Wednesday, May 23,2001 Page 1586 of 1608 �Air Force Health Study An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides SAIC Team Air Force Team George D. Lathrop, M.D., M.P.H., Ph.D. Stella G. Machado, Ph.D. Theodore G. Karrison, Ph.D. William D. Grubbs, Ph.D. Wanda F. Thomas, M.S. COL William H. Wolfe, M.D., M.P.H. Joel E. Michalek, Ph.D. LTC Judson C. Miner, D.V.M., M.P.H. LTC Michael R. Peterson, D.V.M., M.P.H., Dr.P.H. Project Manager: Program Manager: R.W. Ogershok W.F. Thomas SCIENCE APPLICATIONS INTERNATIONAL CORPORATION 8400 Westoark Drive McLean, Virginia 22102 EPIDEMIOLOGY DIVISION USAF School of Aerospace Medicine Human Systems Division (AFSC) Brooks Air Force Base, Texas 78235 October 1987 VOLUME I First Followup Examination Results January 1985 to September 1987 Contract Number F41689-85-D-0010 SAIC Project Number: 2-816-XX-195/254-XX (Distribution Unlimited) �SECURITY llAttif ICATI6N OE TMiTPAGT Form Approved OMB No. 0704-0188 REPORT DOCUMENTATION PAGE 1a. REPORT SECURITY CLASSIFICATION 1b. RESTRICTIVE MARKINGS Unclassified 3 . DISTRIBUTION/AVAILABILITY OF-REPORT . SECURITY CLASSIFICATION AUTHORITY 2b. DECLASSIFICATION/DOWNGRADING SCHEDULE Approved for public release; distribution unlimited 4. PERFORMING ORGANIZATION REPORT NUMBER{S) 5. MONITORING ORGANIZATION REPORT NUMBER(S) 6*. NAME OF PERFORMING ORGANIZATION 7a. NAME OF MONITORING ORGANIZATION Science Applications International Corp. Life Sciences & Systems Dept 6b, OFFICE SYMBOL (If applicable) Human Systems Division 6c ADDRESS (C/ty, Start, and ZIP Cock) (HSD) 7b. ADDRESS (C/ty, State, and ZIP Code) McLean. Virginia 22102 Brocks Air Force Base. Texas 78235-5000 8a. NAME OF FUNDING/SPONSORING ORGANIZATION 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER 8b. OFFICE SYMBOL (If applicable) F41689-S5-D-0010 Be. ADDRESS (C/ty, Staf«, and ZIP Code) 10. SOURCE OF FUNDING NUMBERS PROGRAM ELEMENT NO. PROJECT NO. TASK NO WORK UNIT ACCESSION NO. 6530F 0003 2767 1. TITLE (Include Security Classification) An EpidearLologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides. First Follcwup Examination Results 2. PERSONAL AUTHOR(S) G.D. Lathrop. SAIC; W.H. Wolfe, USAF; S.G. Machado, SAIC; J.E. MLchalek, USAF; T.G. Karri con. TT. of C: J.C. MLt»r. USAF: W.D. Grtibbs. SAICi M.R. Peterson. USAFt W.F. Thntiaa. SATH. a. TYPE OF REPORT 15. PAGE COUNT !4. DATE OF REPORT (Year, Month. Day) 13b. TIME COVERED J Interim 1982-1985 FROM 1/85 TO 9/87 1987 Oct 1 16 11 r 16. SUPPLEMENTARY NOTATION 17. COSATI CODES FIELD 06 GROUP 1 18. SUBJECT TERMS (Continue on reverse if necessary and identify by block number) EpidearLologic Investigation Fnenoxy Herbicides Herbicide Orange SUB-GROUP 05 Dicodn Ranch Hand Air Force Health Study Morbidity 19. ABSTRACT (Continue on reverse if necessary and identify by block number) Bos report presents the results of the health assessment of the 1,016 Ranch Hands and the 1,293 Comparisons who participated in the 1985 f ollowup examination of the Air Force Health Study. The purpose of the study is to determine whether Iccg-tenn health effects exist and can be attributed to occupational exposure to herbicides. The result showed a subtle but consistent narrowing of medical differences between the two groups since the Baseline study in 1982; however, the Ranch Hands continue to manifest slightly more minor adverse health conditions than the Comparisons. Continued surveillance of these two groups is indicated. The report concludes that there is not sufficient evidence to implicate a causal relationship between herbicide exposure and adverse health in the Ranch Hand group. 20. DISTRIBUTION / A VAll ABILITY OF ABSTRACT 0 UNCLASSIFIED-UN'tiMlTEO 21 ABSTRACT SECURITY CLASSIFICATION D SAME AS RPT. Q 01-. NAME OF RESPONSIBLE INDIVIDUAL 'OL WILLIAM H. WOLFE, USAF, MC 00 Form 1473, JUN 86 ;..»$ 22b TELEPHONCr/nc/ude Area Code) 512-530-2604 Previous edir i>f obsolete 22c. OFFICE SYMBOL USAFSAM/EK SECURITY CLASSIFICATION Of THIS PAGE �EXECUTIVE SUMMARY FIRST FOLLOWUP MORBIDITY STUDY The Air Force Health Study is an epidemiological study conducted to determine whether adverse health effects exist and can be attributed to occupational exposure to Herbicide Orange. The study consists of mortality and morbidity components, based on a matched cohort design in a nonconcurrent prospective setting with followup studies. The Baseline study was conducted in 1982, and the first followup morbidity study was performed in 1985. The purpose of this report is to present the results of the first followup study. In the Baseline morbidity study, each living Ranch Hand was matched to the first living and compliant member of a randomly selected Comparison mortality set based on age, race, and military occupation, producing an approximate 1:1 contrast. The Comparisons had served in numerous flying organizations that transported cargo to, from, and within Vietnam but were not involved in the aerial spray operations of Herbicide Orange. Recruitment for the first followup was in accordance with the Study Protocol: All previous participants and refusals, newly located study members, and replacements (matched to noncompliant Comparisons on self-perception of health) were invited. Of the living Baseline study participants, 99.2 percent were contacted to enroll in the followup on a strictly voluntary basis. Participation was very high, with 93 percent of both the Ranch Hands and the Comparisons fully compliant at Baseline also participating in the followup. Overall, the 2,309 followup participants (1,016 Ranch Hands and 1,293 Comparisons) represented a loss to the study of 159 individuals but a gain of 199 new participants since Baseline. Statistical analyses of selection and participation bias supported the use of the total Comparison group for the main analyses presented in this report. The followup study was conducted under contract to the Air Force by Science Applications International Corporation, in conjunction with the Scripps Clinic and Research Foundation and the National Opinion Research Center. Most of the data were collected through face-to-face interviews and physical examinations conducted at the Scripps Clinic in La Jolla, California. Other data sources included medical and military records and the 1982 Baseline data base. As a contract requirement, all data collection personnel were blind to exposure status, and all phases of the study were monitored by stringent quality control. The statistical analyses were based on analysis of variance and covariance, chi-square tests, Fisher's exact tests, general linear models, Kolmogorov-Smirnov tests, logistic regression, proportional odds models, t-tests, and log-linear models. The questionnaire and physical examination data were analyzed by major organ system. The primary focus was on the assessment of differences between the Ranch Hand and Comparison groups based on data from the first followup. Additionally, dose-response relationships within the Ranch Hand group were examined, and longitudinal assessments of differences in the changes of the two groups between the examinations were conducted for selected variables. iii �In terms of general health, Ranch Hand enlisted groundcrew rated their health as fair or poor more frequently than their enlisted Comparisons; differences were not observed for the enlisted flyers or the officers. Physician examiners detected no differences for appearance of illness or distress or for the appearance of relative age. The Ranch Hands had significantly lower percent body fat. They also had a higher proportion of sedimentation rate abnormalities than the Comparisons, but mean sedimentation rates were not statistically different between the two groups. No significant differences between the Ranch Hand and Comparison groups were seen in the 1982-1985 interval for skin or systemic cancers. However, when overall lifetime basal cell carcinoma rates were adjusted for risk factors involved in the cause of such cancers (e.g., sun exposure, skin color, skin reaction to sun), Ranch Hands had a significantly higher proportion of basal cell carcinoma than Comparisons. No group differences were observed for systemic cancer, although two cases of possible dioxin-related cancer were noted in Ranch Hands, bringing the lifetime total to two of these cancers in each group. Overall, the cancer findings were not viewed as disturbing but as reason for continued medical surveillance. The neurological assessment of cranial nerve function, peripheral nerve function, and central nervous system coordination did not reveal any consistently significant group differences, although abnormalities tended to aggregate in the Ranch Hands. The Babinski reflex (found adverse in the Ranch Hands at the 1982 Baseline examination) was equal in both groups at the 1985 followup. Age, alcohol, and diabetes showed classical effects with many neurological measures. In the psychological evaluation based on the Minnesota Multiphasic Personality Inventory, the Comparisons had significantly more abnormalities for the denial and masculinity/femininity scales, whereas the Ranch Hands manifested marginally more abnormalities in the hysteria and social introversion scales. The Ranch Hands showed more abnormalities on the Cornell Medical Index scales than did-the Comparisons, but no differences were detected between the two groups on the functionally oriented Halstead Reitan Battery. There were no group differences for current or past neuroses or psychoses. Age, educational level, and alcohol history showed strong and expected effects on the psychological measures. Both the interval and the lifetime history of liver disease were equal in both groups, as was a lifetime history of peptic ulcer disease. Of nine liver function and two porphyrin laboratory tests, the Comparisons had significantly higher serum glutamic pyruvic transaminase and uroporphyrin means, whereas the Ranch Hands had a significantly higher mean alkaline phosphate level and a borderline elevated coproporphyrin value. There was no evidence to suggest an increased likelihood of porphyria cutanea tarda in the Ranch Hand group. In the dermatological assessment, not one case of chloracne was diagnosed on examination, nor was historical acne anatomically distributed in a pattern that suggested past chloracne in the Ranch Hand group. Exposure and longitudinal analyses were also essentially negative. The cardiovascular evaluation showed no significant group differences for reported or verified hypertension, reported heart disease, or reported or iv �verified heart attacks. However, the frequency of verified heart disease was significantly greater in the Ranch Hands than the Comparisons. The assessment of the central cardiac function by systolic blood pressure and electrocardiogram did not reveal any meaningful group differences. Evaluation of peripheral pulses by the Doppler technique revealed group equivalence in marked contrast to the Baseline examination, which found significant pulse deficits in the Ranch Hands. This change was likely due to required tobacco abstinence before the'pulse measurements. Overall, the groups were remarkably similar in cardiovascular health. The assessment of eight hematological measures showed no significant group differences. In fact, the groups were more similar at the followup examination than at the Baseline examination. Age, race, and smoking were significant risk factors for most hematological measures. The groups did not differ significantly in reported past kidney disease, although the Baseline questionnaire noted such in the Ranch Hands. Five laboratory measures of renal function were similar between groups in the unadjusted analyses. No pattern of results suggested a detriment to either group in the adjusted analyses. For the endocrine function, TSH and testosterone means were significantly higher in the Ranch Hands, but these results were not supported by the categorical tests. The impaired category of the glucose tolerance test revealed an excess in the Comparison group. Examination results for past thyroid disease, thyroid and testicular abnormalities, and additional tests for cortisol level and T3 % Uptake were similar in both groups. Age, race, occupation, percent body fat, and personality type were often significant adjusting variables. Overall, the endocrine health status was comparable in both groups. Comprehensive immunological tests composed of six cell surface marker studies and three functional stimulation studies showed no significant group differences in the unadjusted analyses. Age, smoking, and alcohol usage were generally strong covariates. The assessment of delayed hypersensitivity by skin testing was declared invalid because of excessive reader variation and shifting diagnostic criteria. The pulmonary assessment, consisting of past history, physical examination, and x-ray results did not indicate any consistently different disease patterns in the two groups. Age and lifetime smoking history were important risk factors for most pulmonary measures. The exposure index analyses, which were stratified by occupation, revealed sporadic differences between exposure levels; however, there were no consistent dose-response relationships that supported an herbicide effect for any clinical area. Longitudinal analyses were conducted for 19 variables, and 5 showed significant differences in the changes of the groups between the Baseline and followup examinations. Of these 5 variables, 1 (sedimentation rate) was believed to be related to a change in laboratory methods, and the other 4 (Babinski reflex, depression, platelet count, and manual all pulse index) were attributed to true changes over time for the groups. In comparing all results between the examinations as well as the formal longitudinal analyses, �a subtle, but consistent, decrease in group differences over the 3-year period has been observed. The process of inferring causality is complex and must be based on careful consideration of many factors. Any interpretations of the data must consider the biological plausibility, clinical significance, specificity and consistency of the findings, and a host of statistical factors, such as strength of the association, lack of independence of the measurements, and multiple testing. By direct and indirect evidence, it is concluded that this study is free of overt bias and that the measurement systems used to obtain the data were accurate and valid. By an overall pattern assessment, it is further concluded that the Ranch Hand and Comparison populations are similar. Finally, this first followup examination report concludes that there is insufficient evidence to support a cause and effect relationship between herbicide exposure and adverse health in the Ranch Hand group at this time. The study has revealed a number of minor medical findings that require continued surveillance. In full context, the results of this study must be viewed as additional reassuring evidence that, at this time, the current state of health of the Ranch Hand participants is unrelated to herbicide exposure in Vietnam. vi �PREFACE The release of this 1987 followup Morbidity Report marks more than 8-1/2 years of intensive Air Force research into the herbicide question. Since the commitment to Congress in October 1978 to conduct an epidemiologic investigation of Air Force personnel who aerially disseminated herbicides in the Vietnam War (code-named Operation Ranch Hand), the United States Air Force Surgeon General has issued the following publications: a Study Protocol, four annual mortality reports, the Baseline Morbidity Report, and this first followup morbidity report. Within the next 2 years, the second followup morbidity report, other annual mortality reports, and an expanded birth defects study are expected for publication. This level of commitment has used approximately $40 million of contract research funds, excluding significant Air Force in-house expenditures. Nearly 100 Government, academic, and industry scientists have guided and contributed to the Air Force Health Study (AFHS) since its inception. The Air Force's current advisory committee, chaired by Dr. Robert W. Miller of the National Cancer Institute, is responsible for providing assistance on all scientific and medical matters pertaining to the AFHS. The distinguished panelists are listed in Appendix A. There are numerous scientific strengths in the AFHS, beginning with the unequivocal exposure status of the Ranch Hand population, estimated to have been, on the average, 1,000 times that experienced by an unclothed man directly beneath a spraying aircraft. In the other direction, the Ranch Hand population was probably less exposed to dioxin than many studied industrial populations (based upon a lack of chloracne), and may not develop adverse health consequences because of a possible threshold mechanism. However, the participants of the AFHS have a more defined exposure than the ground troops and constitute a larger population under study than industrial cohorts. The chief strength of the AFHS is its design. The interwoven study elements of multiple mortality assessments, a Baseline morbidity study, and five followup morbidity studies over 20 years provide a comprehensive approach to the detection of attributable adverse health effects. The weakest feature of the design is the mortality assessment which, in the absence of significant case clustering, cannot detect group differences for very rare conditions (e.g., soft tissue sarcoma) because of the inherent constraints of the limited size of the Ranch Hand population. To some extent, this problem may be offset for the more prevalent cancers by combining both living and fatal cancers for future analyses. The strength of the mortality studies should increase with the aging of the study population and the concomitant increase in death with the passage of time. All four mortality assessments have shown that the Ranch Hand population is faring about the same as the Comparison group, with no unusual causes of death, increased frequency of death, or evidence suggesting death at younger ages. Because of the healthy veteran effect, both groups are surviving significantly longer than similarly aged civilians. The morbidity assessment, released in 1984, disclosed only minor differences between the Ranch vii �Hands and the Comparisons, and these differences were not traditional indicators of dioxin-related disease. Both the content and the progress of the AFHS has been presented on many occasions to Congress, to the media, and to scientific meetings around the world. On the whole, the AFHS has been very well received in these circles, giving additional strength and credence to this work. This report of the first followup study is important as it marks the sustained commitment of Congress and the Air Force to pursue the Agent Orange question to its logical scientific conclusion. From the medical and scientific perspectives, this followup examination gives the first opportunity to confirm or refute some of the Baseline findings, and to explore subtle longitudinal changes,while controlling for confounding factors. The fifth-year followup examination, which will have been initiated when this report is released, will be conducted at an average time of 20 years postexposure for the Ranch Hands, a critical period for the emergence of attributable cancer. Followup studies such as these provide the most powerful scientific means of detecting emerging herbicide effects. This report differs slightly from the Baseline Morbidity Report in several ways. The populations under study have changed slightly (see Chapter 2), since some Ranch Hands and Comparisons have voluntarily dropped out of the study, and additional study participants have joined (via the Comparison replacement strategy, or the addition of formerly noncompliant participants). Further, a greater variety of statistical techniques are used to explore bias considerations, subgroup categorical differences (see Chapter 7), and "best" model fitting via the use of two- and three-way interactions. In addition, specific medical tests were included in this examination to clarify whether less specific Baseline findings were relevant (e.g., Doppler measurement of arterial pulses). Early in both the examination and analysis phases of this followup examination, it became clear that a joint Air Force-contractor approach to the analysis of the data was required. The Air Force elected to perform much of the analytical work of this report (e.g., bias, compliance, longitudinal, and pulmonary analyses). Thus, this study has transitioned from "independent" contract work to a genuine team effort between the Science Applications International Corporation (SAIC) and the Air Force scientific staffs. In the spirit of this enriching teamwork, SAIC has listed the Air Force scientific staff co-equally on the cover page of this report. Because of the highly professional scientific interchanges on many challenging aspects of the analytical work, it is believed that this report represents a scientific product unattainable by either team independent of the other. A brief explanation of this report to the reader is in order. This report is written primarily for clinical epidemiologists, clinicians, and biostatisticians so that they may fully evaluate the data and analytic techniques herein. There are segments of this report that will be difficult for even the most experienced of these specialists to understand. Complete familiarity with the Study Protocol and prior mprtality and morbidity reports is essential in the full understanding of this report. Thus, this report is not intended for rapid distillation by the layman or by media representatives. It should be noted that the intent of the introductions of the clinical chapters is to provide only a broad overview of the literature with respect to dioxin endpoints. In addition, the statistical analyses in this report were generally prescribed by the Air.Force (based primarily upon viii �analyses performed for the Baseline Morbidity Report) and are not ad hoc analyses. The report format has been established to be complete, rigorous, and straightforward on all issues so that maximum scientific credibility will be maintained. As with the Baseline Report, the contractor, with Air Force authority, or the Air Force itself, will respond to telephone or written inquiries about the content of this report. This report, prepared by Science Applications International Corporation, is submitted as partial fulfillment of Contract No. F41689-85-D-0010. ix �ACKNOWLEDGMENTS The authors of the report gratefully acknowledge the outstanding support of all the contributors to this project. To all the individuals, named and unnamed, whose dedication and hard work over the past 2-1/2 years have made this report possible, the authors wish to express their sincere appreciation. U.S. Air Force Coinvestigators: Lt. Col. F. Page Armstrong, USAF (Ret.), Nurse Epidemiologist Vincent V. Elequin, Medical Record Librarian Alton Rahe, Mathematical Statistician Lt. Col. John Silva, Consultant, Immunology Support in conducting the statistical analysis: Michael B. Lustik, SAIC Paul Meier, Ph.D., University of Chicago Dung B. Phan, SAIC Wai-Kouk W. Yu, SAIC Data processing and management support: Cristina E. Buchholz, SAIC Melody Darby, USAF Christie L. Dyer, SAIC Steven C. Fullerton, SAIC Task Manager Dawnelle Gonzenbach, USAF George Sacerich, USAF Conduct of the medical records coding: Calvin E. Hollman, USAF Maricella Luna, USAF Earl A. Metts, USAF Janie E. Ridgill, SAIC Consultant Marion B. Yonce, SAIC Consultant Edward E. Zimmerman, USAF xi �Conduct of the physical examinations: Maung H. Aung, M.D., SCRF Dianna M. Cooper, SCRF Roger C. Cornell, M.D., SCRF Karen Curd, M.D., SCRF Donald J. Oalessio, M.D., SCRF Roberta M. Davidson, R.N., SCRF William R. Dito, M.D., SCRF Betty Greene, SCRF Gene T. Izuno, M.D., SCRF L. Dee Jacobsen, Ph.D., SCRF Sharon Lav, SIRL Tony P. Lopez, M.D., SCRF David A. Mathison, M.D., SCRF Anthony P. Moore, M.D., SCRF Robert M. Nakamura, M.D., SCRF Shirley M. Otis, M.D., SCRF Roy F. Perkins, M.D., SCRF John S. Romine, M.D., SCRF Kathleen Rooney, SCRF Stephen K. Sargeant, M.D., SCRF Stanley G. Seat, M.D., SCRF Abbas Sedaghat, M.D., SCRF Marjorie E. Seybold, M.D., SCRF Robert B. Sigafoes, M.D., SCRF Jack C. Sipe, M.D., SCRF Ernest S. Tucker, M.D., SIRL Tonia Vyenielo, M.D., SCRF David E. Williams, M.D., SCRF, Medical Project Director Questionnaire administration and scheduling: Mary Catherine Burich, NORC Terrence D. Callier, NORC Ellwood Carter, NORC Charlene Harris, NORC Celia E. Romans, NORC Suzanne Turner, NORC Logistical arrangements: Joyce A. Douglass, SAIC, Task Manager Jacqueline P. Kirk, SAIC Martha Jean Perkins, SAIC Editorial support and report production: Thelma M. Bailey, SAIC Bernadette A. Bannister, SAIC L. Jean Massie, SAIC Consultant Grace Verchek, SAIC Lenore C. Wagner, SAIC Anna B. Wittig, SAIC Consultant xii �Management and quality review: Patrick A. Bannister, SAIC Leon B. Ellwein, Ph.D., SAIC Consultant Charles Pricker, SAIC Consultant Michael J. Higgins, SAIC Laurence C. Novotney, SAIC Carole J. O'Toole, SAIC James F. Striegel, Ph.D., SAIC Contractual and administrative support: Marie H. Manber, SAIC Lloyd E. Payne, Jr., USAF Joyce C. Standish, SAIC Support and Encouragement: Ranch Hand Association Members And, for making this study possible: All Study Participants xiii �TABLE OF CONTENTS EXECUTIVE ill PREFACE vii ACKNOWLEDGEMENTS xi 1. BACKGROUND STUDY DESIGN PURPOSE REFERENCES 1-1 1-2 1-3 1-4 2. 2-1 2-1 2-2 POPULATION BASELINE CANDIDATE IDENTIFICATION FOLLOWUP CANDIDATE IDENTIFICATION PARTICIPANT SELECTION 2-2 ENROLLMENT ; PERSONAL CHARACTERISTICS AND HABITS OF FOLLOWUP POPULATION LONGITUDINAL LOSSES AND GAINS SUMMARY REFERENCES ' 2-2 2-4 2-10 2-13 2-14 3. QUESTIONNAIRE METHODOLOGY QUESTIONNAIRE DEVELOPMENT INTERVIEWER TRAINING TELEPHONE SURVEY SCHEDULING OF PARTICIPANTS DATA COLLECTION DATA PROCESSING REFERENCES 3-1 3-1 3-2 3-2 3-3 3-3 3-5 3-6 4. PHYSICAL EXAMINATION METHODOLOGY EXAMINATION CONTENT CONDUCT OF EXAMINATIONS 4-1 4-2 4-2 5. 5-1 5-1 STUDY SELECTION AND PARTICIPATION BIAS INTRODUCTION AND BASELINE SUMMARY The Protocol 5-1 The Baseline Replacement Operation 5-1 The Baseline Selection Bias Analyses 5-2 The Baseline Compliance Bias Analyses THE FIRST FOLLOWUP SCHEDULING AND REPLACEMENT OPERATION FIRST FOLLOWUP COMPLIANCE^ FACTORS KNOWN OR SUSPECTED TO INFLUENCE STUDY PARTICIPATION THE TELEPHONE SURVEY. xv 5-3 5-3 5-4 5-7 5-7 �TABLE OF CONTENTS (continued) REPLACEMENT COMPARISONS VERSUS THE NONCOMPLIANT COMPARISONS THEY REPLACED Baseline Replacement First Followup Replacement .0 SCHEDULING AT FIRST FOLLOWUP NEW REPLACEMENTS VERSUS OLD REPLACEMENTS ORIGINAL COMPARISONS VERSUS SHIFTED ORIGINAL COMPARISONS PARTIALLY COMPLIANT VERSUS FULLY COMPLIANT PARTICIPANTS CONCLUSIONS 6. QUALITY CONTROL ADMINISTRATIVE QUALITY ASSURANCE QUESTIONNAIRE QUALITY CONTROL PHYSICAL EXAMINATION QUALITY CONTROL LABORATORY QUALITY CONTROL Quality Control Procedures for the Immunology Laboratory.... DATA MANAGEMENT QUALITY CONTROL Overview of Quality Control Procedures Data Processing System Design Design and Administration of Physical and Psychological Examination Forms. Data Completeness Checks Data Validation Techniques Medical Records Coding Quality Control STATISTICAL ANALYSIS QUALITY CONTROL REFERENCES 7. STATISTICAL METHODS STATISTICAL STUDY DESIGN STATISTICAL ISSUES Intervening Variables Power Multiple Endpoints and Comparisons Paired Versus Unpaired Analyses Mortality and Morbidity Data Cutpoints Exclusions OVERVIEW OF STATISTICAL PROCEDURES Preliminary Analysis « Core Analysis . Continuous Dependent Variables. Categorical Dependent Variables Modeling Strategy. EXPOSURE INDEX ANALYSES LONGITUDINAL ANALYSES General Continuous Data. Categorical Data REFERENCES xv i 5-12 5-12 5-15 5-16 5-16 5-23 5-30 5-33 6-1 6-1 6-1 6-3 6-4 6-6 6-7 6-7 6-8 6-8 6-10 6-11 6-11 6-12 6-13 7-1 7-1 7-2 7-3 7-4 7-4 7-7 7-8 7-8 7-8 7-8 7-14 7-14 7-14 7-15 7-15 7-16 7-17 7-17 7-17 7-18 7-19 �TABLE OF CONTENTS (continued) 8. EXPOSURE INDEX REFERENCES 8-1 8-5 9. GENERAL HEALTH INTRODUCTION Baseline Summary Results Parameters of the 1985 General Health Assessment RESULTS AND DISCUSSION Subjective Assessments Self-Perception of Health Appearance of Illness or Distress Appearance of Relative Age Objective Assessments Erythrocyte Sedimentation Rate Percent Body Fat EXPOSURE INDEX ANALYSES Self-Perception of Health Appearance of Relative Age., Erythrocyte Sedimentation Rate Percent Body Fat LONGITUDINAL ANALYSES SUMMARY AND CONCLUSIONS REFERENCES 9-1 9-1 9-1 9-1 9-2 9-2 9-2 9-7 9-7 9-10 9-10 9-12 9-14 9-14 9-16 9-17 9-19 9-20 9-21 9-24 10. MALIGNANCY INTRODUCTION Baseline Summary Results Parameters of the 1985 Cancer Assessment RESULTS AND DISCUSSION General Questionnaire Data Physical Examination Data Statistical Analysis Baseline-Followup Interval Interval Skin Neoplasms Interal Systemic Neoplasms Lifetime (Baseline and Interval) Lifetime Skin Neoplasms Lifetime Systemic Neoplasms Comparison of Baseline, Interval, and Lifetime Results... Malignant Skin Neoplasms Malignant Systemic Neoplasms Baseline Participants EXPOSURE INDEX ANALYSES DISCUSSION Skin Cancer Systemic Cancer All Cancer SUMMARY AND CONCLUSIONS REFERENCES 10-1 10-1 10-5 10-6 10-6 10-6 10-6 10-7 10-7 10-8 10-8 10-24 10-30 10-33 10-43 10-51 10-51 10-51 10-53 10-55 10-63 10-64 10-66 10-68 10-68 10-73 xvii �TABLE OF CONTENTS (continued) 11. NEUROLOGICAL ASSESSMENT INTRODUCTION. Baseline Summary Results Parameters of the 1985 Neurological Assessment. RESULTS AND DISCUSSION General Questionnaire Data Physical Examination Data Cranial Nerve Function Peripheral Nerve Status Central Nervous System Coordination EXPOSURE INDEX ANALYSES LONGITUDINAL ANALYSES SUMMARY AND CONCLUSIONS REFERENCES 11-1 11-1 ' 11-2 11-3 11-3 11-3 11-5 11-6 11-9 11-14 11-14 11-18 11-25 11-25 11-30 12. PSYCHOLOGICAL ASSESSMENT INTRODUCTION Baseline Summary Results Parameters of the 1985 Psychological Assessment RESULTS AND DISCUSSION Questionnaire Data Psychological Examination Data Statistical Analysis Minnesota Multiphasic Personality Inventory (MMPI) Cornell Medical Index (CMI) Halstead-Reitan Battery (HRB) EXPOSURE INDEX ANALYSES LONGITUDINAL ANALYSES DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES 12-1 12-1 12-3 12-4 12-4 12-4 12-5 12-7 12-7 12-19 12-24 12-25 12-38 12-40 12-42 12-45 13. GASTROINTESTINAL ASSESSMENT INTRODUCTION Baseline Summary Results Parameters of the 1985 Gastrointestinal Assessment... RESULTS AND DISCUSSION Questionnaire Data, Liver Disorders Peptic Ulcer Diseases. Mortality Count Data Physical Examination Data. General Laboratory Examination Data Statistical Analyses. Serum Glutamic-Oxaloacetic Transaminase (SCOT) Serum Glutamic-Pyruvic Transaminase (SGPT) Gamma-Glutamyl Transpeptidase (GGTP) Alkaline Phosphatase 13-1 13-1 13-2 13-3 13-3 13-4 13-8 13-9 13-11 13-11 13-14 13-14 13-22 13-22 13-23 xviii �TABLE OF CONTENTS (continued) Total Bilirubin Direct Bilirubin Lactic Dehydrogenase (LDH) Cholesterol Triglycerides Uroporphyr in Coproporphyrin Discussion Questionnaire-Laboratory Correlations: Porphyria Cutanea Tarda EXPOSURE INDEX ANALYSES SCOT SGPT GGTP Alkaline Phosphatase Total Bilirubin Direct Bilirubin. LDH Cholesterol Triglycerides Uroporphyrins and Coproporphyrin EXPOSURE INDEX ANALYSES LONGITUDINAL ANALYSES SUMMARY AND CONCLUSIONS REFERENCES 14. DERMATOLOGICAL EVALUATION INTRODUCTION Baseline Summary Resul ts Parameters of the 1985 Dermatological Evaluation RESULTS AND DISCUSSION General Questionnaire Data Occurrence of Acne Duration of Acne Location of Acne Physical Examination Data Preliminary Dependent Variables and Covariate Relationships Analyses of Individual Dependent Variables Biopsy Results EXPOSURE INDEX ANALYSES LONGITUDINAL ANALYSES DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES xix 13-24 13-24 .. 13-25 13-25 13-26 13-26 13-27 13-27 13-28 13-30 13-31 13-31 13-31 13-42 13-42 13-42 13-42 13-42 13-43 13-43 13-43 13-43 13-44 13-48 14-1 14-1 14-2 14-2 14-3 14-3 14-3 14-5 14-6 14-7 14-8 14-8 14-11 14-25 14-27 14-33 14-34 14-34 14-37 �TABLE OF CONTENTS (continued) 15. CARDIOVASCULAR EVALUATION INTRODUCTION 15-1 15-1 Baseline Summary Results .. Parameters of the 1985 Cardiovascular Examination RESULTS AND DISCUSSION. 15-2 15-3 15-4 Questionnaire Data: Reported and Verified Heart Disease.... 15-4 Morbidity-Mortality Analysis 15-5 Physical Examination Data 15-11 Central Cardiac Function 15-11 Peripheral Vascular Function 15-20 Diastolic Blood Pressure 15-20 EXPOSURE INDEX ANALYSES 15-36 Reported and Verified Heart Disease Central Cardiac Function Peripheral Vascular System Association of Cardiovascular Examination Findings with Verified Heart Disease LONGITUDINAL ANALYSES DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES 15-36 15-36 15-42 15-42 15-49 15-50 15-51 15-56 16. HEMATOLOGICAL EVALUATION INTRODUCTION....'.. 16-1 16-1 Baseline Summary Results Parameters of the 1985 Hematological Evaluation RESULTS AND DISCUSSION 16-2 16-3 16-3 General Unadjusted Categorical Analyses Unadjusted Analyses of Continuous Data Dependent Variable and Covariate Relationships Adjusted Categorical Analyses. Adjusted Analyses of Continuous Data. Discussion Red Blood Cell Count (RBC) White Blood Cell Count (WBC) Hemoglobin Concentration (HGB) 16-3 16-4 16-6 16-7 16-8 16-9 16-9 16-9 ,.. 16-11 16-11 Hematocrit (HCT). 16-12 Mean Corpuscular. Volume (MCV) Mean Corpuscular Hemoglobin (MCH) 16-12 16-13 Mean Corpuscular Hemoglobin Concentration (MCHC) Platelet Count (PLT) EXPOSURE INDEX ANALYSES LONGITUDINAL ANALYSES ... .. SUMMARY AND CONCLUSIONS.. REFERENCES ... ....',....... xx 16-13 16-14 16-14 16-20 16-21 16-24 �TABLE OF CONTENTS (continued) 17. RENAL ASSESSMENT INTRODUCTION Baseline Summary Results Parameters of the 1985 Renal Assessment RESULTS AND DISCUSSION Questionnaire Data Physical Examination Data Laboratory Data Urinary Protein Urinary Occult Blood Urinary White Blood Cell Count Blood Urea Nitrogen (BUN) Urinary Specific Gravity EXPOSURE INDEX ANALYSES LONGITUDINAL ANALYSES SUMMARY AND CONCLUSIONS REFERENCES 17-1 17-1 17-1 17-2 17-3 17-3 17-5 17-5 17-6 17-9 17-12 17-15 17-16 17-17 17-23 17-23 17-27 18. ENDOCRINE ASSESSMENT INTRODUCTION Baseline Summary Results Parameters of the 1985 Endocrine Assessment RESULTS AND DISCUSSION Questionnaire Data Physical Examination Data Laboratory Test Data General Thyroid Function: T3 % Uptake and Thyroid Stimulating Hormone (TSH) Testosterone Cortisol: Initial, 2-Hour, and Differential Glucose Metabolism: 2-Hour Postprandial Glucose and Composite Diabetes Indicator EXPOSURE INDEX ANALYSES LONGITUDINAL ANALYSES SUMMARY AND CONCLUSIONS REFERENCES 18-1 18-1 18-2 18-2 18-3 18-3 18-3 18-6 18-6 19. IMMUNOLOGICAL EVALUATION INTRODUCTION Baseline Summary Results Parameters of the 1985 Immunologic Profile Rationale of the Immunologic Measurements Immunology Methodologies Cell Surface Marker. Analysis Phytohemagglutinin (PHA) and Pokeveed Mitogen Stimulation Assays Mixed Lymphocyte Reaction Natural Killer Cell Assays Interpretive Considerations 19-1 19-1 19-2 19-3 19-3 19-6 19-6 xx i 18-10 18-16 18-17 18-18 18-24 18-25 18-30 18-34 19-6 19-6 19-7 19-7 �TABLE OF CONTENTS (continued) RESULTS AND DISCUSSION 19-8 Cell Surface Marker (Phenotypic) Studies 19-8 Total T Cells (T ) 19-14 Helper T Cells (T4) 19-14 Suppressor T Cells ( . T) 19-15 B Cells Monocy tes 19-15 19-16 HLA-DR Cells T4/T8 Ratio 19-17 19-18 Functional Stimulation Studies Unstimulated Response (PHA) 19-18 19-20 PHA Net Response 19-23 Pokeweed Net Response Net Response to MLC Stimulation 19-24 19-24 Discussion EXPOSURE INDEX ANALYSES 19-25 19-25 Cell Surface Markers Functional Stimulation Tests. 19-32 19-32 SKIN TESTING RESULTS. 19-33 General Statistical Analyses and Interpretations 19-33 19-34 SUMMARY AND CONCLUSIONS REFERENCES 19-42 19-45 20. PULMONARY DISEASE; INTRODUCTION 20-1 20-1 Baseline Summary Results 20-1 Parameters of the 1985 Pulmonary Examination RESULTS AND DISCUSSION 20-2 20-2 Mortality Experience Unadjusted Morbidity Analyses. 20-2 20-2 Adjusted Morbidity Analyses EXPOSURE ANALYSES SUMMARY AND CONCLUSIONS REFERENCES 20-3 20-3 20-11 20-13 21. INTERPRETIVE CONSIDERATIONS DIOXIN ENDPOINTS EXPOSURE TYPES OF MEASUREMENTS BASELINE-FOLLOWUP EXAMINATION DIFFERENCES STUDY BIASES. GROUP INTERACTIONS: PATTERN RECOGNITION CLASSICAL COVARIATES .„ MULTIPLE COMPARISONS CAUSALITY. . 21-1 21-1 21-2 21-3 21-3 21-4 21-5 21-9 . xxii 21-9 21-10 �TABLE OF CONTENTS (continued) 22. CONCLUSIONS INTRODUCTION STUDY PERFORMANCE ASPECTS POPULATION CHARACTERISTICS Patterns of Results CLINICAL ASPECTS General Health Malignancy ' Neurological Assessment Psychological Assessment Gastrointestinal Assessment Dermatological Evaluation Cardiovascular Evaluation. Hematological Evaluation Renal Assessment Endocrine Assessment Immunological Evaluation Pulmonary Disease CONCLUSION 23. FUTURE DIRECTIONS FIFTH-YEAR FOLLOWUP EXAMINATION EXPOSURE INDEX REFINEMENTS ADDITIONAL ANALYSES AND STUDIES xxiii .. . .. 22-1 22-1 22-1 22-1 22-2 22-2 22-2 22-3 22-3 22-4 22-4 22-5 22-5 22-5 22-6 22-6 22-7 22-7 22-7 23-1 23-1 23-1 23-2 �LIST OF APPENDICES Appendix A Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants B Questionnaire Methodology. C Physical Examination Methodology D Study Selection and Participation Bias. E Statistical Methods F Exposure Index G General Health H Neoplasia I Neurological Assessment J Psychological Assessment K Gastrointestinal Assessment L Dermatological Evaluation M Cardiovascular Assessment N Hematological Evaluation 0 Renal Assessment P Endocrine Assessment Q Immunological Evaluation. R Pulmonary Disease. .. .... .... S Glossary of Abbreviations xxiv , A-l B-l C-l D-l E-l F-l ,. G-l H-l . 1-1 J-l K-l L-l M-l . N-l . 0-1 P-l Q-l R-l S-l �LIST OF TABLES Table 2-1 2-2 2-3 2-4 2-5 2-6 2-7 2-8 2-9 2-10 4-1 4-2 5-1 5-2 5-3 5-4 5-5 5-6 5-7 5-8 5-9 5-10 5-11 Candidate Followup Participants by Group and Baseline Compliance Status Participants Enrolled in the Followup Study by Group and Baseline Compliance Status Age (in 1985) of Participants of the Followup Examination by Group History of Tobacco and Alcohol Use of Participants of the Followup Examination by Group Average Use of Tobacco Products and Alcohol for Those Reporting Use of These Substances: Participants of the Followup Examination by Group Educational Background of Participants of the Followup Examination by Group Religious Preference of Participants of the Followup Examination by Group Military Status of Participants of the Followup Examination by Group Risk-Taking Behavior of Participants of the Followup Examination by Group Losses/Gains of Participants Between the Baseline and Followup Examinations 2-3 2-5 2-6 2-7 2-8 2-9 2-9 2-10 2-11 2-12 Elements of the Followup Physical Examination Laboratory Test Procedures of the Followup Physical • Examination 4-3 Baseline Versus First Followup Sample Sizes Reasons for Nonparticipation in the First Followup of 56 Ranch Hands and 50 Comparisons Who Were Fully Compliant at Baseline Reported Health Status of 35 Ranch Hands and 42 Comparisons Fully Compliant at Baseline and Noncompliant at First Followup Baseline Status of Newly Examined Participants Summary of Reasons for Noncompleted Telephone Interviews... Summary of Results to the Telephone Questionnaire Contrast of Interviewer's Remark from Telephone Interviews and Reported Health Status Self-Reported Health of Previously Uncontacted Comparisons, in 1986, Versus Self-Reported Health Status of Original Comparisons at Baseline Noncompliant Original Comparisons and Replacement Comparisons Versus Their Baseline Replacements: Reported Health Status at Baseline Noncompliant Original Comparisons and Replacement Comparisons Versus Their Baseline Replacements: Medication Use at Baseline Noncompliant Original Comparisons and Replacement Comparisons Versus Their Baseline Replacements: Income at Baseline 5-4 xxv 4-4 5-5 5-5 5-6 5-8 5-9 5-11 5-12 5-13 5-13 5-14 �LIST OF TABLES (Continued) Table 19-5 19-6 19-7 19-8 19-9 19-10 19-11 19-12 19-13 19-14 20-1 20-2 20-3 20-4 20-5 20-6 20-7 20-8 21-1 21-2 Page Adjusted Analyses for Cell Surface Markers by Group Unadjusted Analyses for Functional Stimulation Tests by Group Association Between Functional Stimulation Test Variables and the Covariates in the Combined Ranch Hand and Comparison Groups „ ...... Adjusted (Directionality Shown) for Functional Stimulation Tests by Group Adjusted Exposure Index Analyses for Cell Surface Markers by Occupation Adjusted Exposure Index Analysis for Functional Stimulation Tests by Occupation Summary of Exposure Index by Covariate Interactions for Functional Stimulation Tests Clinical Interpretation Categories of Skin Test Results by Specific Measurement Criteria at SCRF Induration Erythema Relationships in Average Percentage Over Four Skin Tests, by Reader.. Overall Sumary Results of Unadjusted and Adjusted Analyses of Immunologic Variables Unadjusted Analyses of Reported History of Respiratory Illness by Group Unadjsted Analyses of Radiological and Clinical Respiratory System Findings by Group Adjustd Analyses of Respiratory Variables by Group.. Summary of Group-by-Covariate Interactions for Respiratory Variables Exposure Index Analysis Results for Officers p-Values of Dependent Variable-by-Covariate Association Exposure Index Analysis Results for Enlisted Flyers p-Values of Dependent Variable-by-Covariate Association.. Exposure Index Analysis Results for Enlisted Groundcrew: p-Values of Dependent Variable by Covariate Association.. Overall Summary Results of Unadjusted and Adjusted Analyses of Pulmonary Disease Summary Associations of Adverse Health Effects to TCDD Exposure Reported in the Literature Summary of Significant Covariate Strata (or Covariate Level Difference) Found Within Significant Two- and Three-Factor Group-by-Covariate Interactions by Clinical Chapter and Dependent Variable (Group Direction and p-Value) xxxvi 19-12 19-20 19-21 19-22 19-26 19-30 19-33 19-34 19-41 19-43 20-4 20-5 20-6 20-7 20-10 20-10 20-11 20-12 21-1 21-6 �LIST OF TABLES (Continued) Table 9-1 9-2 9-3 9-4 9-5 9-6 9-7 9-8 9-9 9-10 9-11 9-12 9-13 9-14 9-15 9-16 9-17 9-18 9-19 9-20 9-21 10-1 10-2 Page Unadjusted Analysis for Self-Perception of Health by Group Association Between Self-Perception of Health and Age, Race, Occupation, and Personality Score in the Combined Ranch Hand and Comparison Groups Adjusted Relative Risks of Self-Perception of Health by Occupation Frequency of Self-Perception of Health by Occupation and Group Unadjusted Analysis for Appearance of Acute Illness or Distress by Group Unadjusted Analysis for Appearance of Relative Age by Group Association Between Appearance of Relative Age and Age, Race, and Occupation in the Combined Ranch Hand and Comparison Groups Adjusted Relative Risks of Appearance of Relative Age by Occupation Association Between Sedimentation Rate and Age, Race, Occupation, and Personality Score in the Combined Ranch Hand and Comparison Groups Unadjusted Analysis for Sedimentation Rate by Group Association Between Percent Body Fat and Age, Race, and Occupation in the Combined Ranch Hand and Comparison Groups Unadjusted Exposure Index Analysis of Self-Perception of Health by Occupation Adjusted Relative Risk of Self-Perception of Health by Occupation and Exposure Contrast Unadjusted Exposure Index Analysis of Appearance of Relative Age by Occupation Adjusted Mean Sedimentation Rates by Occupation Unadjusted Exposure Index Analysis of Sedimentation Rate by Occupation Adjusted Relative Risk of Sedimentation Rate by Occupation and Exposure Contrast Unadjusted Means of Percent Body Fat by Occupation Unadjusted Exposure Index Analysis of Percent Body Fat by Occupation Longitudinal Analysis of Self-Perception of Health and Sedimentation Rate: A Contrast of Baseline and First Followup Examination Abnormalities Overall Summary Results of Unadjusted and Adjusted Analyses of General Health Variables Unadjusted Analyses of Followup Participants with Verified and Suspected Neoplasms in the Baseline-Followup Interval by Group (Nonblacks and Blacks) Unadjusted Analyses of Nonblack Followup Participants with Verified and Suspected Malignant Skin Neoplasms in the Baseline-Followup Interval by Cell Type and Group xxvii 9-3 9-4 9-5 9-6 9-7 9-8 9-9 9-9 9-11 9-12 9-13 9-15 9-15 9-16 9-17 9-18 9-18 9-19 9-20 9-21 9-22 10-9 10-10 �LIST OF TABLES (Continued) Table 10-3 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 10-12 10-13 10-14 10-15 10-16 10-17 10-18 10-19 Unadjusted Analyses of Nonblack Followup Participants with Verified and Suspected Malignant Skin Neoplasms in the Baseline-Followup Interval by Anotomic Site and Group Unadjusted Analyses of Nonblack Followup Participants with Nonmelanoma Skin Neoplasms and Sun Exposure-Related Skin Malignancies in the Baseline-Followup Interval Occurring on the Face, Head, or Neck, by Occupation .. Covariates for Analysis of Malignant Skin Neoplasms Summary of Associations Between Incidence Rates of Basal Cell Carcinoma and Sun Exposure-Related Skin Malignancies and the Covariates, in Baseline-Followup Interval for Combined Followup Ranch Hand and Comparison Nonblack Participants Adjusted Analyses of Nonblack Followup Participants for Malignant Skin Neoplasm Incidence During BaselineFollowup Interval Summary of Followup Participants with Verified Malignant Systemic Neoplasms in Baseline-Followup Interval by Group Summary of Followup Participants with Suspected Malignant Systemic Neoplasms at Physical Examination by Group Unadjusted Analyses of Followup Participants with Verified and Suspected Malignant Systemic Neoplasms in the Baseline-Followup Interval by Group Summary of Associations Between Incidence Rates of All Malignant Systemic Neoplasms and the Covariates in the Baseline-Followup Interval for Combined Followup Ranch Hand and Comparison Groups Adjusted Analyses of Followup Participants for the Incidence of All Malignant Systemic Neoplasms During the Baseline-Followup Interval Unadjusted Analyses of Followup Participants with Lifetime Occurrence of Verified or Suspected Neoplasms by Group (Blacks Nonblacks) .Unadjusted Analyses of Nonblack Followup Participants with Lifetime Occurrence of Verified and Suspected Malignant Skin Neoplasms by Cell Type and Group Association Between Lifetime Incidence of Suspected Basal Cell Carcinoma and the Covariates for the Combined Followup Ranch Hand and Comparison Nonblack Participants. Adjusted Analyses of Nonblack Followup Participants for Lifetime Occurrence of Malignant Skin Neoplasms Incidence Summary of Followup Participants with Lifetime Incidence of Verified Malignant Systemic Neoplasms by Group Summary of Followup Participants with Lifetime Soft Tissue Sarcoma, Leukemia or Lymphoma by Group Unadjusted Analyses of Lifetime Incidence Rates of All Malignant Systemic Neoplasms Combined, by Group xxviii 10-12 10-13 10-15 10-19 10-23 10-25 10-26 10-28 10-29 10-31 10-32 10-34 10-36 10-42 10-44 10-45 10-46 �LIST OF TABLES (Continued) Table 10-20 Association Between Lifetime Incidence of All Malignant Systemic Neoplasms and the Covariates for Combined Followup Ranch Hand and Comparison Participants 10-21 Adjusted Analyses for Lifetime Incidence of All Malignant Systemic Neoplasms Combined 10-22 Unadjusted and Adjusted Analyses of the Incidence of All Verified Malignant Skin and Systemic Neoplasms and Basal Cell Carcinoma: Baseline-Followup Interval, and Lifetime Occurrence 10-23 Fully Compliant Baseline Participants by Status at Followup Examination and Group 10-24 Fully Compliant Baseline Participants Who Did Not Participate in Followup Examination by Status and Group.. 10-25 Fully Compliant Baseline Participants Also in Followup Examination by Malignant Neoplasm Status 10-26 Adjusted Exposure Index Analysis for Followup Participants for Occurrence of Malignant Neoplasms in the BaselineFollowup Interval 10-27 Adjusted Exposure Index Analysis for Followup Participants for Occurrence of Malignant Neoplasms 10-28 Computed Risks of Basal Cell Carcinoma by Group at Varying Levels of Four Risk Factors, Relative to Comparisons at Low Risk 10-29 Overall Summary Table: Unadjusted and Adjusted Analysis of Interval and Lifetime Skin and Systemic Cancer Incidence 11-1 11-2 11-3 11-4 11-5 11-6 11-7 11-8 11-9 11-10 11-11 Exclusions and Missing Data for Neurological Assessment by Group Unadjusted Analysis for Verified Neurological Disease by Group—1982-1985 Unadjusted Analysis for Verified Neurological Disease by Group—Baseline and First Followup Studies Combined Association Between Seven Neurological Variables and Three Summary Indices and the Covariates in the Combined Ranch Hand and Comparison Groups Unadjusted Analyses for Cranial Nerve Function by Group.... Adjusted Analyses for Selected Variables of Cranial Nerve Function by Group Summary Table of Group-by-Covariate Interactions for Neurological Variables '. Unadjusted Analyses for Peripheral Nerve Function by Group Adjusted Analyses for Selected Variables of Peripheral Nerve Function by Group Unadjusted Analyses for CNS Coordination Variables by Group '. Adjusted Analyses for Selected Variables of CNS Coordination by Group xxix 10-47 10-51 10-52 10-53 10-54 10-56 10-57 10-60 10-67 10-69 11-4 11-5 11-6 11-8 11-10 11-12 11-13 11-15 11-16 11-17 11-18 �LIST OF TABLES (Continued) Table Page 11-12 Adjusted Exposure Index Analyses for Neurological Variables by Occupation 11-13 Summary of Exposure Index-by-Covariate Interactions for Neurological Variables 11-14 Longitudinal Analysis of Romberg Sign and Babinski Reflex; A Contrast of Baseline and First Followup Examination Abnormalities 11-15 Overall Summary Results of Unadjusted and Adjusted Analyses of Neurological Variables 12-1 12-2 12-3 12-4 12-5 12-6 12-7 12-8 12-9 12-10 12-11 12-12 13-1 13-2 13-3 13-4 13-5 13-6 13-7 13-8 Unadjusted Analyses for Reported Psychological Illnesses by Group: Baseline and First Followup Studies Combined.. Unadjusted Analyses for MMPI by Group Association Between MMPI Variables and the Covariates in the Combined Ranch Hand and Comparison Groups Adjusted Analyses for MMPI by Group Unadjusted Analyses for the Cornell Medical Index (CMI) by Group. Association Between CMI Variables and the Covariates in the Combined Ranch Hand and Comparison Groups Adjusted Analyses for CMI Variables by Group Summary Results for the Halstead-Reitan Battery Impairment Index Analyses Adjusted Exposure Index Analyses for Psychological Variables by Occupation. Summary of Exposure Index-by-Covariate Interactions in Adjusted Analyses of Psychological Variables Longitudinal Analysis of Depression and Denial: A Contrast of Baseline and First Followup Examination Abnormalities Overall Summary Results of Adjusted and Unadjusted Analyses of Psychological Variables Number of Other Liver Conditions Reported by Study Participants at Followup by Group (Verified by Medical Record Review) Unadjusted Analyses for Baseline and Interval History of Liver Disease by Group (Verified by Medical Record Review) Medical Record Verification of Reported Liver Symptoms and Diseases by Group (Baseline and Interval Questionnaires Combined) Unadjusted Analysis of Blood Type by Group » Frequency of Diagnosed and Reported Ulcer Disease by Group Unadjusted Analyses of Peptic Ulcer Disease by Blood Type by Group :.............. Frequency of Digestive System Mortality by Group Unadjusted Analysis of Enlarged Livers Diagnosed at Physical Examination by Group xxx 11-20 11-25 11-26 11-27 12-5 12-9 12-11 12-12 12-20 12-21 12-22 12-26 12-27 12-37 12-39 12-43 13-5 13-6 13-7 13-8 13-9 13-10 13-11 13-12 �LIST OF TABLES (Continued) Table 13-9 13-10 13-11 13-12 13-13 13-14 13-15 13-16 13-17 13-18 13-19 14-1 14-2 14-3 14-4 14-5 14-6 14-7 14-8 14-9 14-10 14-11 Page Laboratory Norms for Nine Hepatic Function Variables and Two Porphyrin Determinations Unadjusted Continuous 'and Categorical Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Group. Assocation Between Nine Hepatic Function Variables and Two Porphyrin Determinations and Six Covariates in the Combined Ranch Hand and Comparison Groups Adjusted Continuous and Categorical Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Group Unadjusted Analysis for Interval History of Skin Bruises, Skin Patches, and Skin Sensitivity by Group Unadjusted Analysis for Interval Porphyrin Abnormalities by Group and Skin Patch, Bruise, or Sensitivity Reported at Followup Questionnaire Unadjusted Analysis for Interval Uroporphyrin Abnormalities by Group and Skin Patch, Bruise, or Sensitivity Reported at Followup Questionnaire Adjusted Categorical Exposure Index Analyses (Main Effects Model) Results for Hepatic Function Variables by Occupation Adjusted Continuous Exposure Index Analyses (Main Effects Model) for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Longitudinal Analyses for SCOT, SGPT, and GGTP: A Contract of Baseline and First Followup Examination Test Means Overall Summary Results of Unadjusted and Adjusted Analyses of Nine Hepatic Function Variables and Two Porphyrin Metabolite Tests Unadjusted Analysis for Reported Historical Occurrence of Acne by Group Unadjusted Analysis for Reported Historical Occurrence of Acne Relative to 1961 by Group Unadjusted Analysis for Reported Historical Occurrence of Acne Relative to SEA Tour of Duty for Post-1961 Acne by Group Adjusted Analysis for Duration of Acne (in Years) for Post-1961 Acne by Group Association Between Dermatological Variables and Age, Race, Occupation, and Pre-SEA Acne in the Combined Ranch Hand and Comparison Groups Unadjusted Analysis for Comedones by Group Adjusted Analysis for Comedones by Group Unadjusted Analysis for Acneiform Lesions by Group Adjusted Analysis for Acneiform Lesions by Group Unadjusted Analysis for Acneiform Scars by Group Adjusted Analysis for Acneiform Scars by Group xxx i 13-13 13-15 13-17 13-18 13-28 13-29 13-30 13-32 13-36 13-44 13-46 14-5 14-6 14-7 14-7 14-11 14-12 14-12 14-13 14-14 14-14 14-15 �LIST OP TABLES (Continued) Table Page 14-12 14-13 14-14 14-15 14-16 14-17 14-18 14-19 14-20 14-21 14-16 14-16 14-17 14-17 14-18 14-19 14-20 14-21 14-22 Unadjusted Analysis for Depigmentation by Group Adjusted Analysis for Depigmentation by Group Unadjusted Analysis for Inclusion Cysts by Group.... Adjusted Analysis for Inclusion Cysts by Group..... Unadjusted Analysis for Hyperpigraentation by Group Adjusted Analysis for Hyperpigmentation by Group..... Unadjusted Analysis for Other Abnormalities by Group Adjusted Analysis for Other Abnormalities by Group Unadjusted Analysis for the Dermatology Index by Group Association Between the Dermatology Index and Age, Race, Occupation, and Presence of Pre-SEA Acne in the Combined Ranch Hand and Comparison Groups 14-22 Adjusted Analysis for the Dermatology Index by SEA Acne Class and Group 14-23 Adjusted Relative Risks for Contrasts of Dermatology Index by Pre-SEA Class 14-24 Summary of Histologic Descriptions of Skin Biopsy by Group 14-25 Adjusted Exposure Index Analysis for Dermatological Variables by Occupation 14-26 Summary of Exposure Index by Covariate Interactions Encountered in Adjusted Analysis of Dermatological Variables 14-27 Longitudinal Analysis of the Dermatology Index: A Contrast of Baseline and First Followup Examination Abnormalities 14-28 Overall Summary Results of Unadjusted and Adjusted Analyses of Questionnaire and Physical Examination Dermatological Variables 15-1 15-2 15-3 15-4 15-5 15-6 15-7 15-8 Unadjusted Analyses for Reported and Verified Heart Disease by Group Association Between Verified Essential Hypertension and the Covariates in the Combined Ranch Hand and Comparison Groups Association Between Verified Heart Disease and the Covariates in the Combined Ranch Hand and Comparison Groups Association Between Verified Myocardial Infarctions and the Covariates in the Combined Ranch Hand and Comparison Groups „. Adjusted Analyses for Reported and Verified Heart Disease.. Unadjusted Analyses for Central Cardiac Function By Group (Diabetics Excluded) Association Between Central Cardiac Function Variables and the Covariates in the Combined Ranch Hand and Comparison Groups (Diabetics Excluded) Adjusted Analyses for Central Cardiac Function (Diabetics Excluded) xxx ii 14-23 14-24 14-25 14-26 14-28 14-32 14-33 14-35 15-6 15-7 15-8 15-9 15-10 15-13 15-15 15-16 �LIST OP TABLES (Continued) Table 15-9 15-10 15-11 15-12 15-13 15-14 15-15 15-16 15-17 15-18 15-19 15-20 16-1 16-2 16-3 16-4 16-5 16-6 16-7 16-8 16-9 Page Unadjusted Analyses for Peripheral Vascular Function by Group (Diabetics Excluded) Assocation Between Peripheral Vascular Function Variables and the Covariates in the Combined Ranch Hand and Comparison Groups (Diabetics Excluded) Adjusted Analysis for Peripheral Vascular Function by Group (Diabetics Excluded) Agreement Between Manual and Doppler Pulse Assessments (McNemar's r Test) Summary Statistics for Cardiovascular Covariates by Group.. Adjusted Exposure Index Analyses for Reported and Verified Heart Disease by Occupation Adjusted Exposure Index Analyses for Central Cardiac Function Variables by Occupation Adjusted Exposure Index Analyses for Diastolic Blood Pressure Funduscopic Abnormalities and Carotid Bruits by Occupation Adjusted Exposure Index Analyses for Peripheral Vascular System Manual Pulse Readings by Occupation Adjusted Exposure Index Analyses for Peripheral Vascular System Doppler Pulse Reading by Occupation Longitudinal Analyses of All Pulses Index and Overall ECG's: A Contrast of Baseline and First Followup Examination Abnormalities Overall Summary Results of Unadjusted and Adjusted Analyses Cardiovascular Variables Laboratory Parameters for Hematological Test Variables Unadjusted Categorical Analyses for Hematological Variables by Group Unadjusted Continuous Analyses for Hematological Variables (Contrast of Group Means) Association Between Hematological Variables and Age, Race, Occupation, and Smoking History in the Combined Ranch Hand and Comparison Groups Adjusted Categorical Analyses for Hematologic Variables (Abnormal versus Normal), Adjusted for Age, Race, Occupation, and Smoking) Adjusted Continuous Analyses for Hematological Variables, (Ranch Hand-Comparison Group Differences) Unadjusted Categorical Exposure Index Analyses for Hematological Variables by Occupation Adjusted Categorical Exposure Index Analyses (Log-Linear Models) for Hematological Variables by Occupation (p-Values) Unadjusted Continuous Exposure Index Analyses for Hematological Variables by Occupation (Analysis of Variance) xxxiii 15-21 15-24 15-28 15-31 15-34 15-37 15-39 15-43 15-44 15-47 15-49 15-52 16-4 16-5 16-6 16-7 16-8 16-10 16-15 16-17 16-18 �LIST OF TABLES (Continued) Table 16-10 16-11 16-12 17-1 17-2 17-3 17-4 17-5 17-6 17-7 17-8 17-9 17-10 Summary of Exposure Index-by-Covariate Interactions Encountered in Adjusted Continuous Analyses of Hematological Variables (General Linear Models) Longitudinal Analyses for MCV, MCH, and PLT: A Contrast of Baseline and First Followup Examination Test Means.... Overall Summary Results of Unadjusted and Adjusted Analyses of Hematological Variables Unadjusted Analysis of History of Kidney Disease/Kidney Stones by Group Association Between Kidney Disease/Kidney Stones and Age, Race, Occupation, and Diabetic Class in the Combined Ranch Hand and Comparison Groups Laboratory Norms for Five Renal Variables Summary of Renal Laboratory Variables by Group Association Between Urinary Protein and Age, Race, Occupation, and Diabetic Class in the Combined Ranch Hand and Comparison Groups Frequency of Urinary Protein by Diabetic Class and Group... Adjusted Relative Risks for Urinary Protein by Diabetic Class Association Between Urinary Occult Blood and Age, Race, Occupation, and Diabetic Class in the Combined Ranch Hand and Comparison Groups Adjusted Analysis for Urinary Occult Blood for Nonblacks by Group Frequency of Urinary Occult Blood for Blacks by Group Frequency of Urinary Occult Blood for Blacks by Occupation and Group 17-12 Frequency of Urinary WBC/HPF for Nonblacks by Group........ 17-13 Adjusted Analyses for Urinary WBC/HPF for Nonblacks by Age Category and Group. 17-14 Frequency of Urinary WBC for Blacks by Occupational Category and Group 17-15 Adjusted Analyses for Urinary WBC/HPF for Black Enlisted Flyers and Groundcrew by Age and Group 17-16 Adjusted Analysis of BUN by Race and Group 17-17 Adjusted Analysis of Urine Specific Gravity by Race, Occupation, and Group 17-18 Adjusted Categorical Exposure Index Analyses for Renal Variables by Occupation 17-19 Adjusted Continuous Exposure Index Analyses for Renal Variables. ,. 17-20 Summary of Exposure Index^by-Covariate Interactions for Renal Variables 17-21 Longitudinal Analysis of BUNs A Contrast of Baseline and First Followup Examination Laboratory Means 16-19 16-20 16-22 17-3 17-4 17-5 17-6 17-7 17-8 17-8 17-9 17-10 17-11 17-11 xxx iv 17-11 17-12 17-13 17-14 17-14 17-16 17-17 17-19 17-21 17-23 17-24 �LIST OF TABLES (Continued) Table 17-22 Page Overall Summary Results of Unadjusted and Adjusted Analyses for Renal Variables Unadjusted Analysis for Reporting of Thyroid Symptoms/Disease by Questionnaire Method by Group 18-2 Medical Record Verification Results of Reported Thyroid Disease by Group 18-3 Unadjusted Analysis for Thyroid and Testicular Conditions by Group 18-4 Laboratory Endocrinological Variables: SCRF Normal and Abnormal Ranges 18-5 Unadjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group 18-6 Adjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group 18-7 Association Between T3 % Uptake and Age, Race, Occupation, and Personality Type in the Combined Ranch Hand and Comparison Groups . 18-8 Adjusted Categorical Analysis for T, % Uptake 18-9 Adjusted Categorical Analysis for TSH 18-10 Adjusted Continuous Analysis for TSH by Group 18-11 Adjusted Categorical Analysis for Testosterone... 18-12 Adjusted Continuous Analysis for Initial Cortisol by Group. 18-13 Association Between Differential Cortisol and Age, Race, Occupation, Percent Body Pat, and Personality Score in the Combined Ranch Hand and Comparison Groups 18-14 Adjusted Categorical Analysis for 2-Hour Postprandial Glucose 18-15 Adjusted Continuous Analysis for 2-Hour Postprandial Glucose by Group 18-16 Adjusted Analysis for Diabetes (Composite Indicator) 18-17 Adjusted Exposure Index Analyses for Endocrinological Variables by Occupation 18-18 Summary of Exposure Index-by-Covariate Interactions Encountered in Analyses of Endocrinological Variables.... 18-19 Longitudinal Analysis for Testosterone, T. % Uptake, and TSH: A Contrast of Baseline and First Followup Examination Test Means 18-20 Overall Summary Results of Unadjusted and Adjusted Continuous and Categorical Analyses of Endocrinological Variables 17-24 18-1 19-1 19-2 19-3 19-4 Medical Significance of the Immunologic Data Frequencies of Participants Who Took the Immunological Tests and the Skin Tests, by Group Unadjusted Analyses for Cell Surface Markers by Group Association Between Cell Surface Markers Variables and the Covariates in the Combined Ranch Hand and Comparison Groups (Directionality Shown) xxxv 18-4 18-5 18-5 18-6 18-7 18-11 18-13 18-14 18-15 18-15 18-16 18-19 18-19 18-23 18-23 18-24 18-26 18-30 18-31 18-32 19-4 19-8 19-10 19-11 �LIST OF TABLES (Continued) Table 19-5 19-6 19-7 19-8 19-9 19-10 19-11 19-12 19-13 19-14 20-1 20-2 20-3 20-4 20-5 20-6 20-7 20-8 21-1 21-2 Page Adjusted Analyses for Cell Surface Markers by Group Unadjusted Analyses for Functional Stimulation Tests by Group Association Between Functional Stimulation Test Variables and the Covariates in the Combined Ranch Hand and Comparison Groups Adjusted (Directionality Shown) for Functional Stimulation Tests by Group Adjusted Exposure Index Analyses for Cell Surface Markers by Occupation Adjusted Exposure Index Analysis for Functional Stimulation Tests by Occupation Summary of Exposure Index by Covariate Interactions for Functional Stimulation Tests Clinical Interpretation Categories of Skin Test Results by Specific Measurement Criteria at SCRF Induration Erythema Relationships in Average Percentage Over Four Skin Tests, by Reader..... Overall Sumary Results of Unadjusted and Adjusted Analyses of Immunologic Variables Unadjusted Analyses of Reported History of Respiratory Illness by Group Unadjsted Analyses of Radiological and Clinical Respiratory System Findings by Group Adjustd Analyses of Respiratory Variables by Group......... Summary of Group-by-Covariate Interactions for Respiratory Variables Exposure Index Analysis Results for Officers p-Values of Dependent Variable-by-Covariate Association Exposure Index Analysis Results for Enlisted Flyers p-Values of Dependent Variable-by-Covariate Association.. Exposure Index Analysis Results for Enlisted Groundcrew: p-Values of Dependent Variable by Covariate Association.. Overall Summary Results of Unadjusted and Adjusted Analyses of Pulmonary Disease Summary Associations of Adverse Health Effects to TCDD Exposure Reported in the Literature Summary of Significant Covariate Strata Found Within Significant Two- and Three-Factor Group-by-Covariate Interactions by Clinical Chapter and Dependent Variable (Group Direction and p-Value) xxx vi 19-12 19-20 19-21 19-22 19-26 19-30 19-33 19-34 19-41 19-43 20-4 20-5 20-6 20-7 20-10 20-10 20-11 20-12 21-1 21-6 �LIST OF FIGURES Figure 3-1 Page Selection Procedure for the Questionnaire, Physical Examination, and Followup Study 3-4 Flow Diagram of Day One Followup Interview and Physical Examination 4-6 Flow Diagram of Day Two Followup Interviews and Physical Examinat ion 4-7 4-3 Mark-Sense Form for Followup Neurological Examination 4-8 4-4 Mark-Sense Form for Followup Dermatological Examination and Biopsy 4-9 Percent Completed Physical Examination by Calendar Date for All Comparisons 5-17 5-2 Percent Completed Physical by Calendar Date 5-17 5-3 Percent Completed Physical Examination By Calendar Date for Unrestricted New and Old Replacement Comparisons 5-18 Two Levels of Quality Control Applied to All Collected Data Prior to Statistical Analysis 6-9 14-1 Occurrence of Acne by Time for First Followup Participants 14-4 14-2 Location of Post-SEA and Pre- and Post-SEA Acne by Group 14-9 14-3 Location of Post-SEA Acne by Group 14-10 18-1 Mean Cortisol Levels by Personality Type, Adjusted for Age and Percent Body Fat, by Time of. Specimen Collection 18-20 Mean Cortisol Levels by Percent Body Fat, Adjusted for Age and Personality Type, by Time of Specimen Collection 18-21 Relationship of Induration Measurements to Erythema Measurements for the Mumps Skin Test Reader 1 Results 19-35 4-1 4-2 5-1 6-1 18-2 19-1 19-2 Relationship of Induration Measurements to Erythema Measurements for the Trichophyton Skin Test Reader 1 Results... 19-36 19-3 Relationship of Induration Measurements to Erythema Measurements for the Mumps Skin Test Reader 2 Results 19-4 Relationship of Induration Measurements to Erythema Measurements for the Trichophyton Skin Test Reader 2 Results... xxxvii 19-37 19-38 �LIST OF FIGURES (continued) Figure Page 19-5 Relationship of Induration Measurements to Erythema Measurements for the Mumps Skin Test Reader 3 Results 19-39 19-6 Relationship of Induration Measurements to Erythema Measurements of the Trichophyton Skin Test Reader 3 Results.... 19-40 xxxviii �UNCLASSIFIED TH>t MCVAlTV REPORT DOCUMENTATION PAOf U A«*OAf iCCUAIt* CIAUI'ICATIOM Unclassified ~f«> MCUAltt ClAMl'«CATlON AvTMOAltV 9. O<«1Ai«wTiOM'AVA(tA,«iuTV 0» AttfOAT Distribution ~~~" authorized to U.S. Government agencies only By reason of administrative use (review), 15 July 1987. Other requests for this document must be referred to HQ HSD/YAC, Brooks Air Force Base, Texas. ft* O«UAMI'*CATICN/OOWNOAAO4NO tCxCOvtl 4. M A'OAMIKO OAOAHHATIOK At *>AT MUMMAtftl •* MAMi 0» f»A*OAM«l«O OAOANUATIO* Science Applications International Corporation, Life Sciences and Systems Department •4. AOOMM «Cfly. <«*» *« «/ ft*J McLean, Virginia "1 1 MOMlTOAIMO OAOAHKA1IOM A«»OAt *WM»t Alfl - ww*»w Human Systems Division (BSD) * AOOAttt (Crfy. «MW «*« Itf t»4*l . • 22102 Brooks Air Force Base, Texas •>.0f )CttVM*Ok to. NAMC or »g*oi*»0/VO*»OA«M« OAOAMlXATlOM 78235-5000 •. f AOCUACMINT IN»TAuM«Mf iOCNTlVlCATlOH HVMtCA <r/v/««**it> F41689-85-D-0010 1*. tOUACC O* »VNOtNO MOt. MOOAAM »A«J«<T •*. AOOACtt (C«* <MM *»4 *W C«*f •UMCMTMO. 65306F »*4. 2767 TA«« WOA< v*nt — 0003 no. MO M. tiTvt ifMHi* *M«ni» «IM««A««I»«»* An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides. First Pollowup Examination Results ta.MA*OMAl AvTHOAWI G.D- Lathrop, SAIC; S.G. Machado, SAIC; T,G. Karrison, D.of C; H.D. Grubbs, SAIC; W.F. Thomas, SAIC; W.H. Wolfe,' OSAF; J.E. Michalek, OSAF; J.C. Miner, USAF; M.R. Peterson, DSAF. 14. OAK Of At*OAT >V» . »*.. 9«v( 1 «ft >*C« COuiT Vi* IVM o* M^OAT I «}* TIMC eovo»to 1987 July 15 J^ 1,000 Interim 1982-1985 | »MOM 1/85 TO 9/87 M, (WMvtMlNtAAT NOTATION »». JlffkO 06 * • It >W*^tCt TCAMl NT««MiM •• «TM«W if ******* •** **»ltft >| MM* ««««l«r« Epidemiologic Investigation Dioxin Pnenory Herbicides Ranch Hand Herbicidti Orange Air Force Health Study Morbidity CO4ATI COO«* 0»OW* ' «Ut. OA, 05 A A t T A A ^ T I^A«f«AM« Amft Mm^rmtf if •JJJiajri AA^ ***ltf» ft* itaife ««j»U»J This report presents the results of the health assessment of the 1,016 Ranch Hands and the 1,293 Comparisons who participated in the 1985 follovup examination of the Air Force Health Study. The purpose of the study is to determine whether long-term health effects exist and can be attributed to occupational exposure to herbicides. The result showed a subtle but consistent narrowing of medical differences between the two groups since the Baseline study in 1982; however, the Ranch Hands continue to manifest slightly more minor adverse health conditions than the Comparisons. Continued surveillance of these two groups is indicated. The report concludes that there is not sufficient evidence to implicate a causal relationship between herbicide exposure and adverse health in the Ranch Hand group. 1. AMTAACT SCCUAtTV UMClASllfllO/VNllMtTf O 0 lift. N A M C Of ACS»OHSl«l( R.W. Ogershok W.H. Wolfe 00 FOAM 1473. 83 APR *AM( AS A»T. O OTKvUAS O unclassified NVM«A m (512) 536-2274 (5121 536-2604 «0«TiON O' 1 MM n H Ol$0llt(. UNCLASSIFIED IICVAIT* 0» . �CHAPTER 1 BACKGROUND This chapter briefly describes the background of the Air Force Health Study (AFHS) and provides an overview of the study design and purpose of this report. Portions of this chapter have been paraphrased from the Baseline Morbidity Report, 24 February 1984. In January 1962, President John F. Kennedy approved a program of aerial herbicide dissemination, for the purpose of defoliation and crop destruction, in support of tactical military operations in the Republic of Vietnam (RVN). Under this program, code-named Operation Ranch Hand and in operation from 1962 to 1971, approximately 19 million gallons of herbicides were dispersed on an estimated 10 to.20 percent of South Vietnam.1'2 Approximately 11 million gallons of Herbicide Orange, the primary defoliant of the six herbicides utilized in the program, were disseminated. Operation Ranch Hand was the subject of intense scrutiny from the start due to the controversial nature of the program and political sensitivity to chemical warfare charges contained in enemy propaganda. The concerns, which were initially based on military, political, and ecological issues, shifted during 1977 to health issues. Numerous claims of exposure to herbicides, particularly Herbicide Orange and its dioxin contaminant, and subsequent adverse health effects among U.S. military service personnel have resulted in class action litigation and substantial controversy. Social concern for the Herbicide Orange issue continues to be manifest by continuing scientific research, media presentations, congressional hearings, and legal action. The U.S. Air Force Medical Service's concern for the health of Air Force personnel exposed to herbicides was demonstrated in October 1978 when the Air Force Deputy Surgeon General made a commitment to Congress and to the White House to conduct a health study on the Ranch Hand population, the aviators who disseminated the majority of the defoliants in the RVN. The prevailing reasons for the study commitment included the availability of a definitive occupational exposure to herbicides, a sufficient sample size for survey and clinical research, the ability to ascertain the population at risk, and an opportunity for the Air Force Medical Corps to fulfill its adage "we care" to the Air Force community. the Air Force School of Aerospace Medicine, Brooks Air Force Base, Texas, was tasked by the Surgeon General to develop the Study Protocol. In 1982, after extensive peer review, the epidemiologic study began, and the Protocol was published. Since 1978, numerous animal and human studies of dioxin effects have been planned or initiated by governmental agencies, universities, and industrial firms. The key scientific issue in these studies was the extent of exposure, e.g., who was exposed and how much each individual was exposed. Unfortunately, population identification and exposure estimation, which are critical for a valid study of ground troops, have been scientifically elusive. 1-1 �It is believed that of all the military personnel who served in the RVN, the Ranch Hand population was the most highly exposed to herbicides. Exposure estimates indicate that the average Ranch Hand received 1,000 times more exposure to Herbicide Orange and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) during his tour in the RVN than an average male would receive standing unclothed under a spraying aircraft in an open field. Based on the principle of doseresponse, the Ranch Hands should manifest more and/or earlier evidence of adverse health. Thus, the results of the AFHS should serve as an indicator of herbicide effects in ground personnel. STUDY DESIGN The purpose of the study is to determine whether adverse health effects exist and can be attributed to occupational exposure to Herbicide Orange. The study, consisting of mortality and morbidity components, is based on a matched cohort design in a nonconcurrent prospective setting with followup studies. Complete details on the design are provided in the Study Protocol. The nonconcurrent aspect of the design results from the fact that the Ranch Hands were exposed over time between 1962 and 1971. This staggered exposure is accounted for in the design of the studies to address latency considerations. For the Baseline study, the population ascertainment process identified 1,264 Ranch Hand personnel who served in the RVN between 1962 and 1971. By the time the first followup began in 1985, an additional 11 Ranch Hands had been identified, bringing the total Ranch Hand population to 1,275. A Comparison group was formed, consisting of individuals assigned to selected Air Force units with missions of flying cargo to, from, and within the RVN during the same period. Using a computerized nearest neighbor selection procedure, a maximum of 10 Comparisons was selected for each Ranch Hand, matching on age, race, and military occupation. After personnel record reviews, each Ranch Hand who was determined to be eligible and fully suitable for study had an average of 8.2 Comparison subjects. The mortality component addresses mortality from the time of the RVN assignment. A Baseline mortality study was conducted in 1982, and the mortality followup consists of annual mortality updates for 20 years. For the Baseline study and the first four updates, five individuals were randomly selected from the matched Comparison set for a 1:5 design. Subsequent to 1986, the design will be expanded to include all of the individuals in the Comparison set. The Baseline morbidity component, begun in 1982, reconstructed the medical history of each participant by reviewing and coding past medical records. A cross-sectional element, designed to assess the participant's current state of mental and physical health, was based on comprehensive questionnaires and physical examinations given to the participants. For this component of the study, each living Ranch Hand and the first living member of his Comparison set were selected to participate in the examination. Sequential questionnaires, medical record reviews, and physical examinations in 1985, 1987, 1992, 1997, and 2002 comprise the morbidity study followup. 1-2 �PURPOSE The 1985 morbidity followup is the subject of this report. The objective of the morbidity followup is to continue the investigation of the possible long-term health effects following exposure to TCDD-containing herbicides. This report describes the procedures and results of the first morbidity followup of the AFHS. Analysis of reproductive and fertility data will be conducted by the U.S. Air Force and is not part of this report. 1-3 �CHAPTER 1 \ REFERENCES 1. Young, A.L., J.A. Calcagni, C.E. Thalken, and J.W. Tremblay. 1978. The toxicology, environmental fate, and human risk of herbicide orange and its associated dioxin. Technical report OEHL-TR-78-92, USAF Occupational and Environmental Health Laboratory, Brooks AFB, Texas. 247 pp. 2. Buckingham, W.A., Jr. 1982. Operation Ranch Hand: The Air Force and herbicides in Southeast Asia, 1961-1971. Office of Air Force History, United States Air Force, Washington, D,C. pp. 9-69, 199-201. 3. Lathrop, G.D., W.H. Wolfe, R.A. Albanese, and P.M. Moynahan. 1982. Epidemiologic investigation of health effects in Air Force personnel following exposure to herbicides: Study protocol. Technical report 82-44, USAF School of Aerospace Medicine. 172 pp. Available from NTIS, Springfield, Virginia. 1-4 �CHAPTER 2 POPULATION This chapter provides a description of participant selection, the enrollment process, and the demographic characteristics of the population that participated in the clinical and questionnaire portions of the first followup morbidity study in 1985. BASELINE CANDIDATE IDENTIFICATION The study population for the first followup was defined by the Air Force investigators as part of the Baseline study design. Using detailed searches through Air Force and other Government record systems, a total of 1,264 personnel who had participated in Operation Ranch Hand was identified. Using the same historical data sources, a Comparison population of 24,971 individuals that had been assigned to a variety of military cargo missions in Southeast Asia during the same time period was identified. The Ranch Hand and the Comparison populations were matched after all individuals who had been killed in the Vietnam conflict were removed. The matching process was conducted using a computer program employing iterative nearest-neighbor statistical techniques in order to associate each Ranch Hand with 10 Comparisons by race (Black/nonblack), closest date of birth, and occupational category during Vietnam service (officer-pilot, officernavigator, officer-nonflying, enlisted flyer, and enlisted groundcrew). For each Ranch Hand, 1 of the 10 matched Comparisons was selected at random and designated the Original Comparison. The resulting exposed and multiple matched Comparison study design was used for the Baseline effort. During the questionnaire administration of the Baseline study, it was discovered that 18 percent of the Comparison population had been misselected with respect to their Southeast Asia military experience. After eliminating these ineligible Comparisons, the remaining Comparison set was collapsed to a 1:8 study design, which was used for all subsequent eligibility determinations. During the course of the Baseline morbidity study, five new Ranch Hands were verified as eligible for the study and were added to the exposed group. In addition, two Ranch Hands who had been misclassified as-Comparisons were identified during the questionnaire administration. These individuals were reclassified as exposed and new Comparisons were assigned appropriately. Following the completion of the Baseline morbidity study, 10 additional Operation Ranch Hand participants were located and added to the study population for the followup phases. 2-1 �FOLLOVUP CANDIDATE IDENTIFICATION One of the preliminary tasks associated with the followup study was to conduct a telephone survey of uncontacted replacement candidates. The purpose of the survey was to obtain new information on the candidate's general health, economic situation, and willingness to participate in the study. The Air Force address file, assembled and maintained since 1981, provided the basis for the telephone survey contact list, A location algorithm described in Chapter 3 was developed in order to find those individuals no longer at the address and telephone number indicated in the Baseline file. A total of 7,411 candidate replacements out of the candidate file of 7,963 was located, interviewed using computer-aided telephone interview (CATI) techniques, and confirmed as eligible candidate study participants. Of the 552 candidates who could not be interviewed, 26 were deceased, 335 refused, 190 were unlocatable, and 1 respondent had not served in Southeast Asia and was therefore ineligible for the study. Table 2-1 provides the number of candidate participants by Baseline compliance category for the Ranch Hand and Comparison groups. PARTICIPANT SELECTION The participant selection protocol used for the followup was similar to that used at Baseline with one important exception. If the Original Comparison declined to participate, the next randomly ordered candidate for the corresponding Ranch Hand with the same self-perception of health was contacted and recruited for the study. This matching process was not feasible at Baseline because the addresses of the Comparison pool were not fully ascertained. Perception of health was subjectively determined by the candidate during the telephone interview. The rationale for matching replacement Comparisons on self-perceived health status was an attempt to minimize any bias that might result from differential compliance. All candidates who had been contacted and invited to participate during the Baseline, including those who were refusals and partial compilers, were contacted and invited to the followup along with newly verified or located Ranch Hands and their Comparisons. ENROLLMENT The enrollment of candidates was based on the Baseline lists and health status information from the telephone survey. Recruitment was conducted for questionnaire interviews and clinical examinations that began in May 1985 and ended in March 1986. Approximately 70 individuals were examined each week in two groups of 35. A total of 2,309 Ranch Hands and Comparisons participated in both the questionnaire and clinical examination portions of the AFHS followup. Since the followup questionnaire was administered at the physical examination site, there were no "partially compliant" participants at followup. 2-2 �TABLE 2-1. Candidate Followup Participants by Group and Baseline Compliance Status Number Category Candidate Ranch Hands (by Baseline Status) 1,045 Ranch Hands Who Completed Both Baseline Questionnaire and Physical Examination (Fully Compliant) 129 Ranch Hands Who Completed Only Baseline Questionnaire (Partially Compliant) 32 Ranch Hands Who Declined to Take Part in Baseline (Noncompliant) 10 Newly Verified or Located Ranch Hands 1,216 Total Candidate Comparisons (by Baseline Status) 936 Original Comparisons Who Completed Both Baseline Questionnaire and Physical Examination (Fully Compliant) 220 Original Comparisons Who Completed Only Baseline Questionnaire (Partially Compliant) 79 288 Original Comparisons Who Declined to Take Part in Baseline (Noncompliant) Replacement Comparisons Who Completed Both Baseline Questionnaire and Physical Examination (Fully Compliant) 88 Replacement Comparisons Who Completed Only Baseline Questionnaire (Partially Compliant) 49 Replacement Comparisons Who Declined to Take Part in the Study (Noncompliant) 7,411 Replacement Comparisons Who Had Not Been Contacted Previously 9,071 Total 2-3 �Enrollment was managed using an automated scheduling and tracking system to maintain and record all candidate recruitment contacts, actions, and status; clinical examination group scheduling; schedule modifications, cancellations, and completions; and a comprehensive set of logistic management reports. An effort was made to successfully recruit every individual eligible for the study. The number of participants who participated in the physical examination and questionnaire of the first followup is provided in Table 2-2. Of the 1,016 Ranch Hands, all but 53 had matched Comparisons who also participated in the study. Due to the selection strategy used and the recruitment of previous noncompliants, several of the Ranch Hands had multiple Comparisons. -The selection strategy resulted in 79 Ranch Hands having 2 Comparisons, 9 having 3 Comparisons, and 1 Ranch Hand having a total of 5 Comparisons completing the followup. In accordance with the Study Protocol, eligible Comparisons were enrolled without regard to the compliance status of the corresponding Ranch Hand. There were 229 Comparisons in the followup study whose matched Ranch Hand did not participate. PERSONAL CHARACTERISTICS AND HABITS OF FOLLOWUP POPULATION The data on personal characteristics of the Ranch Hand and Comparison individuals were obtained from the followup questionnaire. The areas of tobacco, alcohol, and marijuana use; personal and family income; education; religious preference; active duty/retired/separated status; and risk-taking behavior received particular attention. These variables were examined to assess the similarity of the two groups in social and behavioral characteristics, which were not included in the statistical matching process. The participants in the study were matched on age. The age characteristics of the study population are shown in Table 2-3. The mean and median ages of the Ranch Hand and Comparison groups were nearly identical. The smoking and alcohol-use habits of the study subjects are displayed in Table 2-4. More participating Ranch Hands smoked cigarettes at the time of the followup physical examination than did the Comparisons (40.1% versus 35.OX). This difference in current smoking behavior was statistically significant (p=0.01). In the intervening years since the Baseline examination, 5.6 percent of the Ranch Hands and 4.6 percent of the Comparisons had stopped smoking. The proportions of participants who ever smoked cigarettes, pipes, or cigars were not significantly different in the two groups. Similarly, the number of participants who drank alcohol in the years since 1982 was not statistically different between groups. Data concerning the use of marijuana were gathered by different methods in the two interviews. In the Baseline questionnaire in 1982, confidentiality of response was given to all participants, but answers were identifiable for each participant. At the 1985 followup, random response techniques1 were used on the marijuana questions to overcome the problem of participants either refusing to respond or giving misleading replies to these highly sensitive and personal questions. With this technique, a coin was flipped by the respondent, who then answered either the marijuana question or a neutral unrelated question, which had an answer of known probability. The outcome of 2-4 �TABLE 2-2. Participants Enrolled in the Followup Study by Group and Baseline Compliance Status Number Category Enrolled Ranch Hands (by Baseline Status) 971 Ranch Hands Who Completed Both Baseline Questionnaire and Physical Examination (Fully Compliant) 39 Ranch Hands Who Completed Only Baseline Questionnaire (Partially Compliant) 0 Ranch Hands Who Declined to Take Part in Baseline (Noncompliant) 6 Newly Verified or Located Ranch Hands 1,016 Total Enrolled Comparisons (by Baseline Status) 872 Original Comparisons Who Completed Both Baseline Questionnaire and Physical Examination (Fully Compliant) 61 Original Comparisons Who Completed Only Baseline Questionnaire (Partially Compliant) 10 Original Comparisons Who Declined to Take Part in Baseline (Noncompliant) 12 New Original Comparisons 267 Replacement Comparisons Who Completed Both Baseline Questionnaire and Physical Examination (Fully Compliant) 32 Replacement Comparisons Who Completed Only Baseline Questionnaire (Partially Compliant) 11 Replacement Comparisons Who Declined to Take Part in Baseline (Noncompliant) 28 New Replacement Comparisons 1,293 Total 2-5 �TABLE 2-3. Age (in 1985) of Participants of the Follovup Examination by Group Group Ranch Hand Age Category Number Comparison Percent Number Percent 43 or Less 412 40.6 549 42.5 44 to 62 568 55.9 693 53.6 63 or More Total 1,016 3.9 51 3.5 36 100.0 1,293 100.0 Group Ranch Hand Comparison Range 35-72 Years 35-77 Years Mean 46.9 Years 46.8 Years Median 47 Years 46 Years the coin flip was unknown to the interviewer. Thus, no given reply could be traced, although the proportion of the population that had the sensitive characteristic (marijuana use) could be estimated. There were no statistically significant differences between the Ranch Hand and Comparison groups in the reported use of marijuana in the 30 days preceding the examination (7.8% and 9.2%, respectively). A much higher percentage, 26.3 percent of the Ranch Hands and 31.0 percent of the Comparisons, reported smoking marijuana at some time in the past. At Baseline, only 5.1 percent of each group reported ever using marijuana. These differences over time were most likely due to a greater sense of confidentiality generated by the random response techniques used in the 1985 questionnaire. The mean usage levels of tobacco and alcohol among those participants who did indulge in these habits are shown in Table 2-5 as pack-years, cigaryears, pipe-years, or drink-years. Mean alcohol use per day was 6.26 drinks per day for the Ranch Hands and 6.42 for the Comparisons. In most of the cumulative measurements, the median level of use was lower than the mean level, indicating that the heavy users of these substances skewed the distributions. Eighty-nine percent of both groups reported having consumed alcohol since the last physical examination. Differences in these calculated variables might have been due to either actual changes in behavior or to differences in the questionnaires used to collect the basic data. 2-6 �TABLE 2-4. History of Tobacco and Alcohol Use of Participants of the Followup Examination by Group Group Comparison Ranch Hand Habit Yes Percent No Percent Yes Percent No Percent p-Value Current Use of Cigarettes 407 40.1 609 59.9 453 - 35.0 840 65.0 0.01 Past History of Cigarettes 752 74.0 264 26.0 944 73.0 349 27.0 0.58 Past History of Cigar Use 249 24.5 767 75.5 345 26.7 948 73.3 0.24 Past History of Pipe Use 265 26.1 751 73.9 340 26.3 953 73.7 0.92 Past History of Marijuana Use* 26.3 73.7 31.0 69.0 0.15 Marijuana Use* within Past 30 Days 7.8 92.2 9.2 90.8 0.52 88.7 146 11.3 0.98 Use of Alcohol since Last Interview 901 88.7 115 11.3 1,147 ^Estimates based on random response technique. 2-7 �TABLE 2-5. Average Use of Tobacco Products and Alcohol for Those Reporting Use of These Substances: Participants of the Followup Examination by Group Group Ranch Hand Comparison Mean Median Mean Median 26.54 25.00 25.77 25.00 Cigarettes, Pack- Years (Cumulative) 17.69 13.00 17.61 13.00 Cigar-Years (Cumulative) 11.25 1.30 10.96 1.00 Pipe-Years (Cumulative) 20.03 6.10 16.90 4.00 Alcohol Drinks per Day (Current Use) 6.26 6.00 6.42 5.00 Drink-Years (Since Last Interview) 1.81 0.80 1.89 0.74 26 .59 12 .80 25 .04 13.00 Substance Cigarettes per Day (Current Use) Drink-Years (Cumulative) Educational background and religious preference for the two groups are presented in Tables 2-6 and 2-7. The current military status of each individual was classified as active duty, retired, separated, reserve duty, or deceased. There were no significant differences between the two groups. These data are presented in Table 2-8 and showed equivalence of the two groups in these social variables. Data on income were collected in a categorical form, and the median income levels of the Ranch Hand and Comparison groups were comparable. The median personal income in both groups was in the $25,000 to $30,000 range, and the median total family income ranged from $40,000 to $45,000 in each group. Risk-taking behavior patterns of the study population were assessed by a series of questions that emphasized participation in potentially dangerous recreational activities. These data are summarized in Table 2-9. In motorvehicle racing (automobiles, boats, and motorcycles) and scuba diving, there were group differences of borderline significance (p=0.07 and p=0.09, respectively). Slightly more Comparisons were scuba divers (12.4% versus 10.1%), and more Ranch Hands raced motor vehicles (12.9% versus 10.4%). There was a significant difference in scuba diving at Baseline (p=0.04), when more Comparisons were scuba divers (12.7% versus 9.9%). 2-8 �TABLE 2-6. Educational Background of Participants of the Follovup Examination by Group Group Ranch Hand Comparison Number Percent Number High School/GED 522 51.4 655 Associate Degree 84 8.3 114 8.8 BA/BS Degree 194 19.1 271 21.0 Graduate Degree 203 20.0 239 18.5 13 1.3 14 1.1 Educational Level Unknown Percent 50.7 p=0.64 TABLE 2-7. Religious Preference of Participants of the Follovup Examination by Group Group Comparison Ranch Hand Religious Preference Number Percent Number Protestant 671 66.0 856 66.2 Catholic 215 21.2 281 21.7 9 0.9 15 1.2 Other 37 3.6 54 4.2 None 84 8.3 87 6.7 Jewish p=0.60 2-9 Percent �TABLE 2-8. Military Status of Participants of the Followup Examination by Group Group Ranch Hand Military Status Comparison Number Percent Number Percent 89 8.8 118 9.1 Retired 553 54.4 683 52.8 Separated 313 30.8 420 32.5 55 5.4 65 5.0 6 0.6 7 0.5 Active Duty Reserve Forces Deceased* p=0.90 "Died after the followup examination. These data reflected the overall equivalence of the two groups in social and behavioral characteristics. The differences observed when these data were contrasted to similar data at Baseline might have reflected differences in data collection methods or slight changes in the cohorts rather than changes in behavior among group members. LONGITUDINAL LOSSES AND GAINS A total of 2,269 Ranch Hands and Comparisons was fully compliant with the Baseline study. The study population of 2,309 for the followup included a loss of 159 participants and the addition of 199 individuals. Loss to the followup occurred either because the participant was deceased, refused to participate, or was unlocatable. The loss to followup was 7 percent in both the Ranch Hand and Comparison groups. Of the 69 Comparisons lost to the followup study due to refusal or inability to locate, 17 were replaced. For the remaining 52, no replacement who satisfied the health status matching criterion and was willing to participate was identified from the candidate replacements. The categories of these individuals are provided in Table 2-10. A total of 199 new participants were recruited into the study based on the selection methodology used. Information on the new participants is provided in Table 2-10. 2-10 �TABLE 2-9. Risk-Taking Behavior of Participants of the Follovup Examination by Group Group Ranch Hand Activity Yes Percent Comparison No Percent Yes Percent No Percent p-Value Scuba Diving 103 10.1 913 89 .9 160 12.4 1,133 87.6 0.09 Auto, Boat, or Motorcycle Racing 131 12.9 885 87 .1 135 10.4 1,158 89.6 0.07 Acrobatic Flying 43 4.2 973 95 .8 43 3.3 1,250 96.7 0.25 Sky Diving 22 2.2 994 97 .8 32 2.5 1,261 97.5 0.62 Hang Gliding 11 1.1 1,005 98 .9 14 1.1 1,279 98.9 1.00 Mountain Climbing 82 8.1 934 91 .9 102 1,191 92.1 0.86 Surfboard Riding 81 8.0 935 92 .0 91 7.0 1,202 93.0 0.40 Long-Di stance Sailing 54 5.3 962 94 .7 55 4.3 1,238 95.7 0.23 Fast Downhill Skiing* 170 16.7 846 83.3 184 14.2 1,108 85.8 0.10 p=0.10 *0ne Comparison was unwilling to respond. 2-11 7.9 �TABLE 2-10. Losses/Gains of Participants Between the Baseline and Followup Examinations Losses Category Number 5 Ranch Hands Deceased Ranch Hand Refusals Ranch Hands Unlocatable 74 Total Ranch Hands Lost 16 Comparisons Deceased 55 14 Comparison Refusals Comparisons Unlocatable 85 Total Comparisons Lost 10 59 Gains Category Number 39 6 Ranch Hands Partially Compliant at Baseline Newly Verified or Located Ranch Hands 45 Total Ranch Hands Added to Study 61 Partially Compliant Original Comparisons at Baseline Partially Compliant Replacement Comparisons at Baseline Newly Selected Original Comparisons (For Newly Verified Ranch Hands) Replacements for Compliant Comparisons Who Refused Followup Noncompliant Original Comparisons Who Agreed to Attend Followup Noncompliant Replacement Comparisons Who Agreed to Attend Followup Original Comparison Not Locatable at Baseline but Found at Followup Replacement Comparisons Not Locatable at Baseline but Found at Followup Replacement Comparisons Not Contacted at Baseline 32 11 16 10 11 1 3 9 154 Total Comparisons Added to Study 2-12 �SUMMARY Participants were recruited for the first followup in accordance with the Study Protocol. All participants (Ranch Hands and Comparisons) who were contacted for enrollment at Baseline were recruited for this phase of the study. Newly verified and located Ranch Hands, since Baseline, and their respective Comparisons were invited to join the study. Due to refusals among the Comparisons,'replacements from the previously uncontacted Comparisons were selected for enrollment. The replacements were matched to the refusing Comparisons on self-perception of health; health status data were obtained in the telephone survey. Personal characteristics of the two groups were compared, based on data obtained from the followup questionnaire. Contrasts of age, educational background, religious preference, current military status, and income revealed no significant differences between the Ranch Hand and Comparison groups. Significantly more Ranch Hands smoked cigarettes at the time of the followup examination than did Comparisons, although there were no significant differences found for past history of cigarettes, cigars, or pipe use or for recent or past use of marijuana. A much higher percentage of both groups reported smoking marijuana at some time in the past at the followup than at Baseline. This difference was most likely due to a greater sense of confidentiality generated by the random response techniques used in 1985. The use of alcohol since the Baseline examination was not significantly different between the two groups. The difference in the risk-taking behavior patterns of the Ranch Hands and the Comparisons was marginally significant. Slightly more Ranch Hands than Comparisons raced motor vehicles, and more Comparisons were scuba divers. The followup study population included the loss of 159 participants (74 Ranch Hands and 85 Comparisons) who were fully compliant at Baseline and the addition of 199 participants (45 Ranch Hands and 154 Comparisons). The 199 newly examined study subjects consisted of 132 participants (39 Ranch Hands, 61 Original Comparisons, and 32 replacement Comparisons) who were partially compliant at Baseline, 21 participants (10 Originals and 11 replacements) who refused at Baseline, and 46 participants (6 Ranch Hands, 12 Originals, and 28 replacements) who were new to the study. Thus, the study population for the first followup of the AFHS consisted of 2,309 individuals: 1,016 who had been associated with Operation Ranch Hand and 1,293 Comparisons. 2-13 �CHAPTER 2 REFERENCES 1. Greenberg, B.C., A-L.A. Abdul-Ela, W.R. Simmons, and D.G. Horvitz. 1969. The unrelated question randomized response model: Theoretical framework. J. Am. Stat. Assoc. 64(326)-.520-539. 2-14 �CHAPTER 3 QUESTIONNAIRE METHODOLOGY This chapter discusses the development and the implementation of the questionnaires used in the study: the participant interval questionnaire, the spouse interval questionnaire, the Baseline participant and spouse questionnaires, and the telephone survey of previously uncontacted Comparisons. The participant interval questionnaire was designed to capture the study participant's health history in the 3 years since his participation in the Baseline study. Data collection was comparable to the Baseline effort: The questionnaire was very similar, and it was administered using the same faceto-face methodology to virtually the same population. In the Baseline study, interviews were conducted in the participants' homes and the followup interview was conducted at the physical examination site. The revised methodology was more efficient and better subject to quality control. The spouse interval questionnaire collected reproductive data similar to those collected at Baseline from spouses for the interval since Baseline. The spouse interval questionnaires were mailed to the spouses to be selfadministered, or were completed in La Jolla, California, if the spouse accompanied the participant to the physical examination site. Analysis of the spouse data is not included in this report. Since some study subjects refused to participate in 1982 and other participants were new to the study, Baseline questionnaires were administered to these new participants and their spouses. The same procedures used at « Baseline were used to administer the Baseline questionnaires in the homes of these individuals. The elements of each questionnaire are identified in Table B-l of Appendix B. Questionnaire development and administration and scheduling of participants were conducted by the National Opinion Research Center (NORC), a social science research center at the University of Chicago. QUESTIONNAIRE DEVELOPMENT The goal of questionnaire development was to maintain to the maximum extent possible the question wordings, context, and procedures that were used in the 1982 Baseline study. The largest task of questionnaire development was asking for interval histories on crucial questionnaire items to update the information provided by the 1982 Baseline questionnaires. For the participant interval questionnaire, new questions were also developed on risk factors for skin cancer, since the Baseline Morbidity Report found Ranch Hands to have an excess of nonmelanoma skin cancer.1'3 Because the chemical constituents of Herbicide Orange had not previously been associated with skin cancer in the literature, no questions had been included in the Baseline participant questionnaire to collect information on risk factors for this condition. 3-1 �New questions were added to determine personality type, since Type A behavior is associated with coronary heart disease. The Jenkins Activity Scale was administered to collect these data. Enhancements were also made to improve data collection for birth defects, smoking habits, and drinking habits. A copy of the participant interval questionnaire is provided in Appendix B. An information sheet containing a computer-generated summary of key respondent answers to the Baseline survey was used to provide bounded recall for participants. Even when given a precise "starting date," respondents frequently repeat information given earlier, neglect to report new information because they thought they had previously reported it, and otherwise misplace events in time or forget them completely. The best means of preventing such errors is through the use of bounded recall, in which the respondent is reminded of information he has already reported and new information is sought with reference to an updated information sheet. Among the data elements included were date of birth, highest educational degree, military status at last interview, marital status at last interview, and name of spouse. The questionnaire was pretested on 8 ineligible individuals who had been interviewed during Baseline, and on 10 men who participated in the pretest examination. INTERVIEWER TRAINING Twelve interviewers were recruited and trained by NORC's field management and Chicago office staffs in May 1985 to administer the interval questionnaires. The onsite NORC interview staff was not informed of the exposure status of any study participant either before or after contract completion. The site supervisor reported to the Project Director in Chicago on a weekly basis, and quarterly visits were made to the site by the Director. The site supervisor observed a sample of interviews, at least one per interviewer per week, and reviewed and edited interview questionnaires before shipping them to Chicago for further processing. In early 1985, personal interviewers were recruited to conduct Baseline interviews for new participants in their homes. The interviewers were trained in the Chicago NORC office, using questionnaires and procedures established for the Baseline survey. They were supervised by an assistant survey director in the NORC office, who edited each completed questionnaire and talked with each interviewer weekly. TELEPHONE SURVEY The telephone survey of uncontacted Comparisons was intended to gather data on the general health status of the 7,963 replacement candidates for the active Comparison group. The sample consisted of men who served in C-130 units in Southeast Asia between 1962 and 1971, but who did not participate actively in the Baseline phase of the study. A total of 7,411 cases (93%) was completed by NORC computer-assisted telephone interviewers. The telephone survey was conducted prior to the scheduling of the physical examinations. 3-2 �The key question asked was, "Compared to other people your age, would you say that your health is...excellent, good, fair, poor?" Other questions asked about current medications, severity of illness or injury during the last 6 months, and income. Locating and refusal conversion algorithms similar to the Baseline data collection efforts were used. The data from the telephone survey of uncontacted Comparisons were used to select a replacement whose self-reported health status matched that of the noncompliant Comparison. If a willing replacement was not found by this method, the perception of health status variable was dichotomized into excellent/good versus fair/poor, and a new replacement was selected from the Comparison set. If this second attempt at identifying a suitable replacement failed, no replacement was made. The selection procedure is provided in Figure 3-1. In this example, the first randomly ordered Comparison was contacted but refused to participate. In the second attempt, the Comparison was deceased. The third Comparison volunteered to participate in the morbidity study. SCHEDULING OF PARTICIPANTS NORC recruited and trained four schedulers to perform the initial contacts with study subjects. Their training included background information on the details and purpose of the study, simulation of the actual scheduling of calls, documentation of results, and conversion of refusals. An initial letter was sent by the Air Force to each study subject, informing him of the upcoming interval physical examination. The NORC scheduler then followed this letter with a call to attempt to schedule the participant. Refusals occurred at a number of steps in the scheduling process. A team of conversion specialists was assigned to contact refusing study subjects and attempt conversions. Help in conversion was also received from individuals in the U.S. Air Force School of Aerospace Medicine and the Ranch Hand Association. Many more participants were scheduled, but due to "no-shows" at the examination site, and passive refusals who rescheduled frequently, the final figure stood at 2,309. The Baseline interviewer contacted the potential study participant by telephone for scheduling the in-home interview. Toward the end of the physical examination phase, the Baseline questionnaire was administered at the examination site by one of the interviewers who had been trained in administering that questionnaire. Of the 106 participant Baseline questionnaires administered during the first followup, 21 had to be conducted at the examination site. The supervisor of the Baseline interviewers conducted the locating efforts for new and interval participants. Procedures similar to those used in 1982 were followed: a postal search, followed by a local telephone directory search, a motor vehicle registration search, and personal locating efforts in the area of last known residence when appropriate. The Air Force also provided locating support through its records. DATA COLLECTION Upon arrival at the Scripps Clinic and Research Foundation (SCRF), the participant received a schedule including the time and place for the interval interview, and a race-matched interviewer was appointed to conduct the 3-3 �Comparison Individuals (Randomly Ordered) Randomly Selected Mortality Controls Matched Ranch Hand t — + * ** 111 1:1 Unwilling Deceased Volunteered Replacement Candidates Figure 3-1. Selection Procedure for the Questionnaire, Physical Examination, and Followup Study 3-4 �interview. Because of scheduling problems and the unavailability of a Black interviewer, 65 of the 143 Black study participants were interviewed by whites. As in all of the personal interviews for the AFHS, interviewers were required to ask questions exactly as written, were not allowed to interpret questions or inject personal commentary, and were not allowed to skip between sections of the questionnaire. They were also instructed to probe "don't know" answers at least once. During the interview, medical record release forms were signed. The respondent was also asked to give the current name and address for each former spouse listed in the questionnaire, so that spouse questionnaires could be mailed to these individuals. The spouse interval survey was mailed to current spouses at the time the study subject was at the SCRF. Two NORC Chicago telephone interviewers were trained to prompt refusing spouses to return the questionnaire, or to administer the spouse interview by telephone as part of the refusal conversion effort. If the spouse also traveled to La Jolla, the questionnaire was completed under the supervision of a site interviewer. Of the 1,898 completed spouse interval questionnaires, 1,066 were returned by mail, 348 were completed by telephone, and 484 were completed in La Jolla. DATA PROCESSING All completed interviews were sent to the NORC Chicago office following editing by the site supervisor, who retrieved missing data from study subjects while they were still onsite; any further retrieval of critical items was conducted from the Chicago office through telephone contacts. Critical items were those for which missing data were unacceptable. The questionnaires were coded for data entry by a staff of five coders who received a week of training on the various AFHS instruments. Data entry was programmed to provide value and range checks as the data were being entered, to perform logic checks and arithmetic checks, to flag important missing items, and to verify the key entry of 10 percent of each questionnaire. Then the data were run through an automated cleaning program to detect a wide range of data errors that were corrected by pulling the hard copy questionnaires and reviewing each situation on a case-by-case basis. No changes were ever made in the hard copy data; corrections were entered into the data tape, and the tape was run against the cleaning program until no errors were detected. 3-5 �CHAPTER 3 REFERENCES 1. Vitaliano, P., and F. Urbach. 1980. The relative importance of risk factors in nonmelanoma carcinoma. Arch. Dermatol. 116:454-456. 2. Stern, R.S., and K. Momtaz. 1984. Skin typing for assessment of skin cancer risk and acute response to UV-B and oral methoxalen photochemotherapy. Arch. Dermatpi. 120:869-873. 3. Scotto, J., and T.R. Fears. 1978. Skin cancer epidemiology: Research needs. Natl. Cancer Inst. Monogr. 50:169-177. 4. Jenkins, C.D., R.H. Rosenman, and M. Friedman. 1967. Development of an objective psychological test for the determination of the coronary-prone behavior pattern in employed men. J. Chronic Pis. 20:371-379. 3-6 �CHAPTER 4 PHYSICAL EXAMINATION METHODOLOGY The first followup examination was provided to four categories of individuals: those who had taken the Baseline questionnaire and Baseline physical examination; those who had been invited to the Baseline events but chose not to participate, only took the questionnaire, or were unlocatable; those Comparisons who had not been invited previously, but who were selected as replacements for Baseline Comparisons noncompliant to this followup examination; and the six newly identified Ranch Hands. . As noted in the Baseline Report, all potential study participants were verified as eligible for the AFHS following a detailed review of military personnel records. Replacement individuals were carefully selected, by matching data on the self-perception of health from the noncompliant Comparison (obtained from the telephone survey) with those of the replacement candidate (see Chapter 3 for details). The followup examination differed logistically from the Baseline examination in one significant way: All structured interval questionnaires were administered at the examination site as contrasted to the in-home interviews conducted at Baseline. The followup examination consisted of the following major elements: • Interval Questionnaire • Combat Experience Questionnaire • Review-of-Systerns Questionnaire • Psychological Testing • Physical Examination • Specialized Testing, e.g., Doppler Arterial Studies • Laboratory Testing • Psychological and Medical Outbriefings. Details of the above examination elements were carefully prescribed by the Air Force and set forth as contractual requirements. Clinical innovations or variations were neither desired nor authorized; all proposed examination procedural changes were reviewed in detail by Air Force technical and contractual personnel. An important objective of the technical review was to ensure that bias was not created, by any procedural change. The requirement to maintain blind examinations was particularly stringent: The clinical staff was prohibited from knowing or seeking information as to the group identity (Ranch Hand, Comparison) of any participant. At the end of the examination, each participant was asked to note on the critique form whether such information was sought by any member of the clinical or paramedical staff. 4-1 �EXAMINATION CONTENT Examination content was designed by the Air Force to emphasize detection of medical endpoints suspected of being associated with exposure to phenoxy herbicides, chlorophenols, or dioxin. In addition, findings in the Baseline examination were used by the Air Force to direct changes in the followup examination (e.g., abnormal pulses at Baseline suggested the need for Doppler measurements at the followup). The general'content of the physical examination and psychological test battery is shown in Table 4-1, and the complete laboratory test series is displayed in Table 4-2. Quality control requirements for both laboratory testing and clinical procedures were extensive. Although details are provided in Chapter 6, the following categories provide an overview of the extent of the quality emphasis. For laboratory testing, single reagent lots and control standards were used when practical, duplicate specimens were routinely and blindly retested, testing overlaps were mandatory when test reagents required change, and fast initial response cumulative statistical techniques (FIR CUSUM) were used to detect rapidly any subtle test drift over time. In addition, 50 specimens from the Baseline serum bank were retested to assess the comparability of laboratory methods. The SCRF clinical team was carefully instructed to assure clinical quality. The quality control elements included: a pretest of the examination process; detailed clinical inspection techniques by SCRF, Science Applications International Corporation (SAIC), and Air Force physicians and personnel; preprinted mark-sense examination forms; clinical quality assurance meetings to detect and correct problems; and blindness of exposure status at the examination. In addition, participant rapportbuilding techniques were added to boost participation in future followup studies, such as participant critique forms and recreational opportunities afforded to the accompanying family members. CONDUCT OF EXAMINATIONS All examinations were conducted at SCRF, La Jolla, California, from May 1985 to March 1986. Except for weeks with national holidays, two groups of participants, averaging about 32 per group, were examined weekly. Midway through the study, NORC recruiters noted that a number of participants refused the examination because of weekday business commitments or because of single-parent responsibilities. Consequently, two special weekend examinations were arranged late in the examination cycle, and many of the former refusals were then able to attend. The examination was identical to the regular 2 1/2-day process, except that it was compressed into 2 days by reducing the number of participants in a group. The logistics effort required in contacting, transporting, and examining 2,309 study members was formidable. Preexamination contacts consisted of the telephone health survey, telephone recruitment to the examination if necessary, and calls by either the NORC scheduling specialists or by the travel agent to arrange transportation and determine whether special requirements existed (e.g., wheelchair assistance, weekend examination schedule). Once scheduling was reasonably firm, the SAIC logistics coordinator sent each participant a detailed information package outlining dietary requirements, inbriefing schedules, important telephone numbers, a request for medical records, and local maps designating examination-site eating and recreational facilities. 4-2 �TABLE 4-1. Elements of the Followup Physical Examination Elements Remarks General Physical Examination Internist Neurological Examination Neurologist Dermatological Examination Dermatologist Electrocardiogram Resting, 4-Hour Fasting and Nicotine Abstinence Doppler Peripheral Arterial Blood Flow Studies 4-Hour Nicotine Abstinence Chest X Ray Immunological Studies 50% Random Sample Skin Test Studies 75% Sample Psychological Evaluation: Minnesota Multiphasic Personality Inventory (MMPI) Cornell Medical Index Halstead-Reitan Battery Patient Outbriefing and Discussion of Individual Results 4-3 Medical Diagnostician, Internist, and Ph.D. Psychologist �TABLE 4-2. Laboratory Test Procedures of the Follovup Physical Examination Clinical Laboratory Fasting Glucose Blood Urea Nitrogen (BUN) Cholesterol HDL Cholesterol Triglyceride Serum Glutamic-Oxaloacetic Transarainase (SCOT) Serum Glutamic-Pyruvic Transaminase (SGPT) Gamma-Glutamyl Transpeptidase (GGTP) Alkaline Phosphatase Lactic Dehydrogenase (LDH) Thyroid Stimulating Hormone (TSH) Initial Cortisol 2-Hour Cortisol Prothrombin Time Quantitative Immunoglobulins Complete Blood Count (CBC) Leuteinizing Hormone (LH) 2-Hour Postprandial Glucose Creative Phosphokinase (CPK) Total Bilirubin Direct Bilirubin Total Protein Protein Electrophoresis Routine Urinalysis T3 % Uptake T 4 Testosterone Hepatitis B Surface Antigen Hepatitis B Surface Antibody Follicle Stimulating Hormone (FSH) Rapid Plasma Reagin (RPR) Porphyrins (Mayo Clinic) Sedimentation Rate Immunological Laboratory Cell Surface (Phenotype) Analyses Lymphocyte Mitogen Stimulation Assays Mixed Lymphocyte Culture (MLC) Natural Killer Cell Assay by Specific Cellular Cytotoxicity Using K-562 Target Cells Natural Killer Cell Assay (Using Interferon) by Specific Cellular Cytotoxicity Using K-562 Target Cells 4-4 �The logistical flow of the entire examination process was complex. Figures 4-1 and 4-2 outline participant flow for the first 2 examination days. As depicted in these figures, each group of participants (generally containing equal numbers of Ranch Hands and Comparisons) was transported early in the morning to SCRF on the first 2 days in a fasting state; tobacco, alcohol, and coffee abstinence were also required. Following initial inbriefing and blood draw on the first day, each participant was randomly assigned to the examination group or to the psychological testing group. On the second day, these groups were reversed. After randomization, each member was given an individualized 3-day schedule outlining his medical, interviewing, and laboratory appointments. The schedule carefully noted the specific required periods of fasting and tobacco abstinence (see Figures 4-1 and 4-2 for generalized periods in relation to ECG and Doppler testing). Each individual was reminded of the fact that all aspects of the examination were strictly voluntary, and that refusals would be honored without question. Both general and specific consent forms (e.g., skin biopsy), approved by the Air Force, were explained in detail. In contrast to the Baseline examination, great reliance was placed upon each individual to find the appropriate clinic area at his scheduled time. This approach had great appeal to this self-reliant population as evidenced by critique feedback. Throughout the examination day, generous time was provided for waiting-room activities, i.e., renewal of past friendships, discussions of the Vietnam War, consumption of refreshments when permitted, and completion of paperwork. Day 3 of the examination was largely spent in finishing up the specialty examinations and receiving the outbriefings from a psychologist and medical diagnostician. Only upon completion of these important debriefings were the participants paid their stipend, reimbursed for travel expenses, and transported to the airport. As noted previously, the SCRF clinical team was hand-picked for participation in this project. In total, 15 board-certified physicians in internal medicine, neurology, and dermatology participated in the general, specialty, and diagnostic examination. To reduce observer variability, turnover in the clinical or paramedical staffs was minimized during the 9 months of examinations. One SCRF physician served as the Project Medical Director, responsible for the scheduling, conduct, and quality control of the examinations. All examining physicians were introduced to the mark-sense examination forms during the pretest examination. The layout of the form was designed to parallel the flow of the clinical examination so as to minimize recording errors. Because data transcription was not permitted, each physician was responsible for filling in the bubbled form. To a large extent, these marksense forms and subsequent quality control were the primary reason for a remarkably clean data set. Two examples of the mark-sense forms are presented as Figures 4-3 and 4-4; a complete set of forms is provided in Appendix C. For the first followup, the special testing included Doppler tests, delayed hypersensitivity skin tests, and immunological tests. Doppler measurements were obtained on all participants by highly experienced technicians; results were recorded and Polaroid photographs were taken of representative oscilloscope displays. As previously noted, considerable emphasis was placed upon tobacco abstinence prior to Doppler evaluations. Skin tests for four antigens were administered in a standardized manner: Candida (1:1,000 weight/volume, 0.1 ml intradermal), mumps (2 complementfixing units), Trichophyton (1:1,000 weight/volume, 0.1 ml intradermal), and 4-5 �Day One PM AM Group A: Physical Examination tKIKttdl I !*=!• 1 1 o.» 1 £«», 1 k r I NOHC lillefvww. xHdv 1® I1 N< Wailing Room (Group Rappon) 1 1 HdhhMt) r-"*" 1 MMPl 1 NOHC XHdv Waiting Room \/ J^ Kay: Slaggeiad G*uup Movement Schectuled Movanwni by Indnndual 4 Hours Post PiattchaJ and Tobacco Abstinence Figure 4-1. Flow Diagram of Day One Followup Interview and Physical Examination 5 • Medical Videos a Hi o I 1 1 111< Group B: Psychological Testing Group Movement •„ • Refreshments Ail L*> 1 1 1 CtXIMH /\ it co..- ir ,^, , , II , >*s. �CHAPTER 5 STUDY SELECTION AND PARTICIPATION BIAS INTRODUCTION AND BASELINE SUMMARY The Protocol During the design phase, the authors of the Protocol anticipated that loss to followup would pose the greatest threat to the validity of the study. In particular, they expected differential compliance with relatively more Ranch Hands self-selecting themselves into the study than Comparisons and with health differences of unknown character between noncompliant Ranch Hands and noncompliant Comparisons. As a partial correction, the study design specified that noncompliant Comparisons would be replaced by Comparisons having the same values of the matching variables and the same health perception. In this way, the replacement Comparisons would serve as surrogates for those Comparisons who refused to participate. This, in turn, would tend to reduce the bias due to noncompliance in the Comparison group and would have the added advantage of maintaining this group's sample size. The Comparison in each randomized matched set who happened to be first asked to participate in the Baseline questionnaire and physical examination was identified as the Original Comparison for his respective Ranch Hand (in accordance with the Protocol). If the Original Comparison was noncompliant, that is, he refused to take the Baseline questionnaire or physical examination, his replacement was called a replacement Comparison. Replacement Comparisons were so distinguished to satisfy the Protocol requirement.that they be contrasted with the noncompliant Comparisons, also called refusals, they replaced. No corresponding replacement strategy for the Ranch Hands was possible since all Ranch Hands had been identified and invited to participate. The Protocol further specified that the replacements would be statistically compared with the noncompliant Original Comparisons to determine the extent to which the replacement strategy was being realized. The statistical contrast of replacements and refusals was to be based on responses to a noncompliance telephone questionnaire administered to refusals and to their potential replacements. This questionnaire assessed self-perception of health, days lost from work due to illness, and medication use, and was to serve as the basis for the health matching called for in the Protocol. Although the Protocol was not explicit on this point, it implied that the decision to include or exclude the replacements from the study would be based only on this contrast. The Baseline Replacement Operation The health-matching questions (identical to the noncompliance questionnaire) were, in fact, hot administered to any potential replacement 5-1 �Comparison before selection at Baseline, although questions regarding self-perception of health, medication use, and work loss were asked as part of the Baseline questionnaire after entry into the study. The noncompliance telephone questionnaire was offered to nohcompliant study participants, but only 79 completed the telephone questionnaire, and of these only 57 were actually replaced. Replacements were, therefore, not health matched to refusals at Baseline. Rather, they were matched only on the basic matching variables: date of birth, race, and occupation. The statistical contrast of refusals and their replacements was not performed at Baseline. During the scheduling operation at Baseline, two untoward events occurred that led to the identification of two additional categories of Comparisons, shifted Comparisons and Air Force-interviewed replacements. First, 212 of the Original Comparisons were discovered to be ineligible for participation in the study due to errors in the data base regarding their unit of assignment in Southeast Asia. These men had not served in Southeast Asia but, due to a duplication of codes, were mistakenly included in the Comparison population. They were deleted from the study. This resulted in another Comparison in each previously randomized match set being first asked to participate in the study. These new Original Comparisons were figuratively called "shifted" Comparisons, labeled S in the Baseline Report, to describe the effective movement of these Comparisons- in each matched set to fill the space left by the removed ineligible Original Comparison. The eligible Original Comparisons were labeled 0 in the Baseline report. Shifted Comparisons are more accurately referred to here as shifted Original Comparisons to emphasize that they are not replacement Comparisons and that they are the legitimate Original Comparisons for their respective Ranch Hands. Shifted Original Comparisons are not replacement Comparisons because their invitation to participate in the study was not the result of a previous refusal of another Comparison in their respective matched sets. Shifted Original Comparisons were identified to reflect concern that the process by which Comparisons were determined ineligible may not have distributed ineligible Comparisons uniformly. Second, 30 replacement Comparisons were interviewed by Air Force staff rather than by the contractor. These replacements were labeled A. All other replacement Comparisons, labeled R, were simply called "replacements." The removal of ineligible Comparisons from the study caused a pause in the scheduling operation that delayed the scheduling of the shifted Original and replacement Comparisons relative to that of the Original Comparisons. This scheduling delay is apparent in Figures V-3 and V-4 in the Baseline Report. Some study investigators speculated that this scheduling slip might cause shifted Original Comparisons and replacement Comparisons to self-select differently from Original Comparisons. Statistical analyses in Chapter V of the Baseline Report and further unpublished analyses following the release of the Baseline Report investigated the effect of this scheduling problem. The Baseline Selection Bias Analyses Since replacements were not health matched at Baseline to their corresponding noncompliant Comparisons and since differential scheduling opportunity may have created self-selection biases, statistical contrasts of the various Comparison groups were done at Baseline. In particular, the Comparisons labeled 0, S, R, and A were contrasted on the basis of self-perception of health, medication use, work loss, and five clinical variables. 5-2 �The results of these analyses suggested to some investigators that shifted Original Comparisons were not statistically distinguishable from Original Comparisons and that shifted Original Comparisons were not statistically different from replacements, but that replacement Comparisons appeared to be statistically different from Original Comparisons. The 30 Air Force-interviewed replacement Comparisons were not statistically distinguishable from other replacement Comparisons and were not investigated further as a group. Since opinions differed among Air Force principal investigators and statisticians, a management decision was reached to use only the Original Comparisons in the primary analyses and to contrast Ranch Hands with all Comparisons in the appendix of the Baseline Report. The reader is referred to Chapter V of the Baseline Report for additional detail. In retrospect, the concern with statistical distinguishability between replacement Comparisons and Original Comparisons is difficult to justify, since the only valid question regarding the replacements is their similarity to the refusals whom they replaced. The Baseline Compliance Bias Analyses Telephone questionnaire data obtained from the 57 noncompliant Comparisons, who were replaced, were not analyzed in the Baseline Report. Instead, compliance bias was analyzed by contrasting partially compliant with fully compliant participants, with adjustment for group (Ranch Hands, 0, S, R, A). These analyses were based on data from the Baseline questionnaire regarding self-perception of health, medication use, work loss, anger, anxiety, erosion, depression, liver ailments, miscarriages, and acne. Results suggested that partially compliant participants were statistically different from fully compliant participants for some of these variables. Based on these results, calculations were presented to suggest that the noncompliance bias could produce an error in relative risk of 25 percent, either overestimating or underestimating the risk, and a spurious mean shift of up to 8 percent in either direction. THE FIRST FOLLOVUP SCHEDULING AND REPLACEMENT OPERATION Matching of replacements to noncompliant Comparisons on the basis of health status was initiated with the first followup scheduling operation. This was accomplished by administering a short telephone questionnaire to all previously uncontacted Comparisons and then using their health status responses to select from among the Comparisons in a matched set the first one who was similar to the refusal regarding self-perception of health. In addition, NORC was required to schedule replacements within 5 working days of a confirmed refusal. These features were intended to correct the described Baseline scheduling deficiencies and to bring the study into Protocol compliance regarding health matching of replacements. To further minimize the possibility of scheduling bias, the entire study population was partitioned into 79 groups; these groups were then randomly scheduled for an examination time. In this way, no single group would be favored a priori for a certain scheduling period. The groupings, consisting of approximately 32 participants, corresponded to the examination groups established at Baseline. Group integrity was maintained to enhance study compliance and comradery. Study participants were given the option to remain with their group or to reschedule their examination at a time more convenient to them. 5-3 �FIRST FOLLOWUP COMPLIANCE Eighty-five percent (1,016/1,191) of the Ranch Hands and 81 percent (955/1,176) of the Original Comparisons participated in the first followup examination and questionnaire process. Of 288 replacements, 267, or 93 percent, chose to attend the first followup examination; additionally, 71 new replacements participated in the followup, yielding a total sample size of 338 replacements at followup. These counts and others are summarized in Table 5-1. In Table 5-1 and subsequently in this report, the shifted Original Comparisons were combined with the Original Comparisons, and the Air Force replacements were combined with the replacement Comparisons. TABLE 5-1. Baseline Versus First Followup Sample Sizes Group Comparison Participation Baseline Only Baseline and Followup Followup Only Ranch Hand Original Replacement 74 971 45 64 872 83 21 267 71 Although fully compliant at Baseline, 74 Ranch Hands, 64 Original Comparisons, and 21 replacement Comparisons chose not to participate in the first followup examination. In the interim, 10 of the 74 Ranch Hands and 16 of the 85 Comparisons died. An additional 5 of the 74 Ranch Hands and 14 of the 85 Comparisons were unlocatable during the scheduling operation. There were 56 of 59 remaining Ranch Hands and 50 of 55 remaining Comparisons who refused to participate in the first followup, although they were alive and locatable during scheduling, and responded to the noncompliance telephone questionnaire, giving their reported health status and reason for nonparticipation. The 3 remaining Ranch Hands and 5 Comparisons refused to participate in the telephone survey. Reasons for nonparticipation given in the telephone survey are summarized in Table 5-2. The totals in Table 5-2 do not correspond to Table 5-1 because some participants gave more than one reason for nonparticipation. Of the 56 living locatable Ranch Hands and the 50 Comparisons who took the noncompliance telephone questionnaire, only 35 Ranch Hands and 42 Comparisons responded to the question regarding health status. The reported health status of these 77 nonparticipants is summarized in Table 5-3. 5-4 �TABLE 5-2. Reasons for Nonparticipation in the First Follovup of 56 Ranch Hands and 50 Comparisons Who Were Fully Compliant at Baseline* Group Comparison Ranch Hand Reason Number Fear of Physical Job Commitment Dissatisfaction with USAF No Time or Interest Travel Distance, Family Confidentiality Health Reasons Passive Refusal Dissatisfaction with Baseline Financial Hardship Other Total Number Percent Percent 0 13 10 7 13 0 8 11 5 0 17 13 9 17 0 11 15 7 2 9 9 6 12 1 3 6 2 4 16 16 11 21 2 5 11 4 3 5 4 7 0 7 0 12 75 57 *Some participants gave more than one reason for nonparticipation. TABLE 5-3. Reported Health Status of 35 Ranch Hands and 42 Comparisons Fully Compliant at Baseline and Noncompliant at First Follovup Group Reported Health Ranch Hand Status Number Percent Excellent Good Fair Poor Total 5 22 6 2 14 63 17 6 Number Percent 10 22 8 2 24 52 19 5 42 35 p=0.72 5-5 Comparison �Among the individuals responding to the health status question, there was no statistically significant difference between noncompllant Ranch Hands and Comparisons regarding reported health (p=0.72). Further detail regarding the 45 Ranch Hands, 83 Originals, and 71 replacements newly examined at followup is shown in Table 5-4, which gives the Baseline status of these participants. Taking the questionnaire but not the physical examination at Baseline were 39 of the 45 Ranch Hands newly examined at followup. Five of the 45 Ranch Hands who were identified too late to be invited at Baseline were simply described as having had "no action" taken. TABLE 5-4. Baseline Status of Newly Examined Participants Group Comparisons Baseline Status Interview Only, Refused Physical Examination No Interview, No Physical Examination Unlocatable No Action Proxy New to Study Total Ranch Hand Original Replacement 39 61 32 0 10 11 0 5 1 0 1 11 0 0 3 16 0 9 45 83 71 Of the 71 newly examined replacements, 43 (32+11) were either partially compliant at Baseline or were at least contacted at Baseline and, therefore, identified as replacements, although not health matched to a noncompliant Comparison. The remaining 28 newly examined replacements were not previously contacted. Of these, 14 were health-matched replacements and 2 were replacements added to the study in August 1985 after completion of the Baseline physical examination. Thus, of the 71 replacements who took the physical examination for the first time at followup, only 14 were new health-matched replacements. All 71 replacements may be regarded as new to the study, even though 43 had been previously contacted at Baseline and knew that they were potential study participants. The 28 replacements who had not been previously contacted may be regarded as new in a more restrictive sense since they did not know of their potential involvement in this study before they were recruited for the first followup examination. This set of 71 replacement Comparisons and the subset of 28 are distinguished from each other using 5-6 �the unrestricted and restricted definitions of "new" to provide data regarding changes in replacement self-selection, an issue explored later in this chapter. FACTORS KNOWN OR SUSPECTED TO INFLUENCE STUDY PARTICIPATION A multitude of factors may be considered to influence self-selection. These may be broadly classified as health, logistic, operational, publicity, or demographic factors. The Baseline Report contains a list of specific factors within each of these categories. For example, health factors are thought to include self-perception of health as well as demonstrable health indicators, such as medication use and work days lost due to illness or injury. Logistic factors are thought to include distance to the examination site, reluctance to spend time away from family or job, income, and occupation. Demographic factors might include flying status, age, race, or military duty status (active, retired, separated). Operational factors include any aspect of study operation that may cause differential compliance, such as differential treatment of participants during scheduling, physical examination, interview, or debriefing. Publicity factors have to do with national attitudes and media presentations regarding the Agent Orange issue, the Vietnam war, veteran health care, or health care in general. Additionally, these considerations may affect people differently and, in particular, may influence Ranch Hands differently than Comparisons. The decision to volunteer for this study is admittedly complex, making statistical assessment of compliance bias difficult and necessarily crude in that many of the factors contributing to self-selection cannot be measured directly. Instead, compliance bias was investigated at first followup, as in the Baseline Report. Specifically, it was investigated with respect to selfperception of health, medication use, daily aspirin use, work days lost due to illness or injury, and income in comparing partially compliant with fully compliant participants. In other selection bias assessments, such as statistical contrasts of Original and shifted Original Comparisons, these same factors and 26 variables taken from the physical examination and psychometric testing were analyzed. THE TELEPHONE SURVEY In April 1985, all previously uncontacted living Comparisons were identified for telephone contact -to assess their current health. This health status information was necessary for the matching of replacements to noncompliant Comparisons. From a total of 9,982 available Comparisons, 7,963 were included in the telephone survey. The 2,019 nonselected Comparisons included 488 deceased, as of 1 August 1985, and 1,531 who had been previously contacted. The group of 1,531 previously contacted Comparisons comprised all Comparisons who were fully compliant, partially compliant, or noncompliant at Baseline. The survey questionnaire is, shown in Appendix D. In brief, it queried the respondent regarding self-perception of health (excellent, good, fair, poor), current prescribed medication use (yes, no), work days lost due to illness or injury, special health care needs (wheelchair, nurse, or other special equipment), and income (less than $20,000, $20,000 to $40,000, or more than $40,000). If the respondent indicated that he was taking 5-7 �prescribed medication, he was asked to identify the illness for which the medication was prescribed. If work days were lost due to illness or injury, the respondent was asked to identify the causing illness or injury. If special health care or equipment was needed, he was asked to specify the illness or condition requiring the special care. He was further asked to distinguish conditions requiring special care from those that were previously identified in response to the medication and days lost from work questions. The telephone interview was accomplished via CATI. Of the 7,963 cases fielded, 7,411 telephone surveys were actually completed. The nature of the 552 noncompletions is summarized in Table 5-5. TABLE 5-5. Summary of Reasons for Noncompleted Telephone Interviews Reason Number Percent of 7,963 Deceased Active Refusal Passive Refusal Unlocatable Ineligible 26 93 242 190 1 0.3 1.2 3.0 2.4 0.0 Total 552 6.9 Several questionnaires that could not be administered by telephone were accomplished by mail; these numbered 540 out of the 7,411 completed. Summaries of the responses to each of the five questions are shown in Table 5-6. Of the 1,271 respondents who reported that they had lost work days due to illness or injury, 550 (43%) lost 1 to 5 days, 197 (15%) lost between 6 and 10 days, and 524 (41%) lost more than 10 days. The maximum number of days reported lost was 965. The 56 respondents who reported more than 180 days lost misinterpreted the question; it referred only to the past 6 months. The telephone interviewer reported whether the respondent was friendly, cooperative but not interested, impatient, or hostile. The association between the interviewer's remark and the self-reported health of the respondent was investigated. The results are displayed in Table 5-7. The association between the interviewer's remark and reported health status is statistically significant (p=0.02), with hostile repondents reporting poorer health than friendly, cooperative, or impatient respondents. Other analyses of these data, not shown here, demonstrated significant associations between health perception and income (p=0.001), rank (p=0.001), age (p=0.001), medication use (p=0.001), and need for special health care (p=0.001). Positive health perception increased with income and rank and 5-8 �TABLE 5-6. Summary of Results to the Telephone Questionnaire Self-Assessment of Health Compared to Others Same Age Response Excellent Good Fair Poor Do Not Know Missing Total Number Percent 2,882 3,306 972 245 3 3 38.89 44.61 13.11 3.31 0.04 0.04 7,411 100.00 Taking Medication for Current Illness Response Yes No Refused Missing Total Number Percent 2,129 5,277 1 4 28.73 71.20 0.01 0.05 7,411 100.00 Illness or Injury Absence From Job During Last 6 Months Response Yes No Refused Missing Total Number Percent 1,271 6,135 ' 3 2 17.15 82.78 0.04 0.03 7,411 100.00 5-9 �TABLE 5-6. (continued) Summary of Results to the Telephone Questionnaire Need Assistance in Daily Activities Response Number Percent Yes No Refused Missing 114 7,291 4 2 1.54 98.38 0.05 0.03 7,411 100.00 Total Earned Income From Any Job During 1984 Response Number Percent Yes No Refused Missing 6,636 755 17 3 89.54 10.19 0.23 0.04 7,411 100.00 Total Income Level Response Less than $20,000 $20,000-$40,000 More than $40,000 Not Applicable Refused Do Not Know Missing Total Number Percent 2,015 3,034 1,411 774 161 9 7 27.19 40.94 19.04 10.44 2.17 0.12 0.10 7,411 100.00 5-10 �TABLE 5-7. Contrast of Interviewer's Remark from Telephone Interviews and Reported Health Status Reported Health Status Remark Excellent Good Fair PerPerNumber cent Number cent Friendly 2,209 39 622 38 Cooperative Impatient 42 40 9 21 Hostile Total 2,882 39 Number Total Poor PerPerPercent Number cent Number cent 44 46 45 63 730 229 10 3 13 14 9 7 191 44 6 4 3 3 6 9 5,606 1,650 106 43 3,306 45 972 13 245 3 7,405 2,476 755 48 27 76 22 1 0 p=0.02 decreased with age, medication use, and special health care. Further, there was no significant association between health perception and the duration of the telephone interview (p=0.17) or the time of day of the interview (p=0.98). There was no significant health-by-duration-by-time of day interaction (p=0.77). These data were also used to assess the self-reported health of 773 Original Comparisons (excluding shifted Original Comparisons) fully compliant at Baseline relative to the reported health of the 7,411 previously uncontacted Comparisons who completed the telephone survey. The selfreported health status of the Original Comparisons from the Baseline questionnaire was contrasted with that of the previously uncontacted Comparisons on a three-category scale (excellent, good, fair/poor) with an adjustment for date of birth (born during or before 1942, born after 1942). The results are displayed in Table 5-8. Previously uncontacted Comparisons who completed the survey are indicated by T (telephone); Original Comparisons are labeled 0. Data are missing for 12 Original Comparisons and 16 telephone-surveyed Comparisons. There was no statistically significant difference between these groups regarding health perception after adjustment for age (p=0.14), and this equivalence did not change with age (p=0.80). Additionally, there was a statistically significant age effect (p=0.001), as expected. These results suggested that the Original Comparisons were representative of the entire Comparison cohort with respect to health perception. 5-11 �TABLE 5-8. Self-Reported Health of Previously Uncontacted Comparisons, in 1986, Versus Self-Reported Health Status of Original Comparisons at Baseline Health Perception Excellent Age Fair/Poor Good Group* Number Percent Number Percent Number Percent Total Born >1942 T 0 1,847 203 39 39 2,003 239 43 46 837 83 18 16 4,687 525 Born <1942 T 0 1,034 91 38 39 1,298 120 48 51 376 25 14 11 2,708 236 *T = previously uncontacted Comparisons 0 = Original Comparisons. REPLACEMENT COMPARISONS VERSUS THE NONCOMPLIANT COMPARISONS THEY REPLACED Baseline Replacement These analyses are refinements of the analyses in Chapter V of the Baseline Report. Of 288 Comparisons replaced at Baseline, only 57 responded to the short noncompliance telephone questionnaire shown in the appendix. These 57 comprised 38 Original Comparisons and 19 replacements. As in the followup telephone survey, the short noncompliance telephone questionnaire queried respondents on health status, work days lost due to illness, medication use, and income level. In accordance with the Protocol, replacements were statistically contrasted with the noncompliant Comparisons they replaced based on their reported health status (excellent, good, fair, poor), medication use (yes, no), and income level (less than $20,000, $20,000 to $40,000, more than $40,000). This contrast, with adjustment for group membership (Original, replacement) of the noncompliant Comparison, is shown in Table 5-9. There was no significant difference between the reported health patterns in the upper and lower panels of Table 5-9. When these two tables were merged, no statistically significant difference was found between the health status of noncompliant Comparisons and their non-health-matched replacements (p=0.99). It is noteworthy that 53 percent of Original and replacement noncompliant Comparisons were matched, by chance, perfectly to their replacements on the basis of reported health status. Only 7 percent (4/57) were mismatched by two categories and one replacement was mismatched by three categories. These same groups were contrasted on medication use; the results are shown in Table 5-10. 5-12 �TABLE 5-9. Noncompliant Original Comparisons and Replacement Comparisons Versus Their Baseline Replacements: Reported Health Status at Baseline Health Status of Replacements Group Noncompliant Original Comparison Health Status Excellent Good Fair Poor Total Noncompliant Replacement Excellent Good 1 3 9 1 1 24 4 7 1 0 12 Fair Poor Total 2 0 0 0 0 0 0 0 19 16 2 1 2 0 38 7 3 1 0 5 3 0 0 0 0 0 0 0 0 0 0 12 6 1 0 1 1 8 0 0 19 Excellent Good Fair Poor Total TABLE 5-10. Noncompliant Original Comparisons and Replacement Comparisons Versus Their Baseline Replacements: Medication Use at Baseline Medication Use of Replacements Group Noncompliant Original Comparison Medication Use Yes .. No Total Yes No 0 3 4 31 4 34 3 35 38 0 1 1 17 1 18 18 19 Total Noncompliant Replacement Yes No Total 5-13 �Due to sparseness these data were not analyzed. It is interesting to note, however, that there was 82 percent agreement in the upper panel of Table 5-9 (31/38) and 89 percent in the lower panel (17/19), with 84 percent agreement in the combined table (48/57), close to expected within group percentages of 83 and 90 percent, respectively, due purely to chance. Work loss was not analyzed due to slight differences between the way the work loss question was worded in the noncompliance telephone and telephone survey questionnaires. The contrast regarding income level is shown in Table 5-11. TABLE 5-11. Noncompliant Original Comparisons and Replacement Comparisons Versus Their Baseline Replacements: Income at Baseline Income Level of Replacements (in thousands) Group Income Level Noncompliant Original Comparison <$20 $20-$40 >$40 <$20 1 6 0 Total Noncompliant Replacement $20-$40 3 6 7 >$40 Total 0 3 6 4 15 13 32* 16 <$20 $20-$40 >$40 0 1 1 0 7 3 2 0 5 10 Total 2 8 9 19 *Six noncompliant Original Comparisons were unwilling to respond. The patterns of income matching in the first and second panels of Table 5-11 were not significantly different (p>0.10). In the combined table, replacements reported significantly lower income than the Comparisons they replaced (p<0.05) although 49 percent (25/51) were perfectly categorically matched. These analyses suggested that the Baseline replacements were very similar to the noncompliant Comparisons they replaced regarding reported health status, medication use, and income. These analyses were also pertinent to the question of whether there was selection bias due to noncompliance in the Comparison group. The predominantly negative findings suggested that there was little or no Comparison selection bias. These 5-14 �results suggested that the upper-bound bias calculations reported in Chapter V of the Baseline Report are overestimates of reality. However, lack of clinical data for the noncompliant Comparisons precluded refining those Baseline bias calculations at this time. Accordingly, the Baseline selection bias calculations may be viewed as crude bounds to an unknown bias that must await future data for proper recalculation. First Follovup Replacement Replacements were matched to noncompliant Comparisons at first followup on the basis of the matching variables—date of birth, race, and occupation— and self-reported health status (excellent, good, fair, poor), as recorded in the telephone survey. This was accomplished by recording the self-reported health status of the noncompliant Comparison during the attempt to schedule and matching that status against those of the other Comparisons in the same matched set. A Comparison in a matched set was considered to replace a noncompliant Comparison if he had the same health status as that recorded for the noncompliant Comparison during the attempt to schedule him. If no willing Comparison reporting the same health status could be found in the matched set, health status was dichotomized to excellent or good versus fair or poor. A willing Comparison with the same health status as the refusal on the dichotomized scale was then accepted as a replacement, if no willing Comparison could be found using the dichotomized scale, attempts to find a replacement were terminated. During this process, 14 Comparisons were health matched to noncompliant Comparisons. The results are summarized in Table 5-12. TABLE 5-12. Health Status of Refusals and Their Matched Replacements Refusal's Health Replacement's Health Excellent Good Fair Poor Total Excellent Good Fair Poor Total 1 5 0 0 2 6 0 0 0 0 0 0 0 0 0 0 3 11 0 0 6 8 0 0 14 All refusals reported good or excellent health. This implied that bias due to noncompliance in the Comparison group could possibly bias the study away from finding an herbicide effect. The inclusion of health-matched replacements tended to correct for this by replacing healthy noncompliant Comparisons with healthy replacement Comparisons. The relatively small number of new health-matched replacements minimized the actual effect of this bias "correction," however. 5-15 �SCHEDULING AT FIRST FOLLOWUP The schedulers were required to find and schedule a willing healthmatched replacement within 5 working days of a confirmed refusal to correct scheduling differences experienced at Baseline. This constraint proved impractical to implement since Comparisons would vacillate, forcing a series of repeated telephone calls. Rather than terminate the process at 5 days, as required by the contract, the schedulers continued their recruiting attempts, sometimes for several months. Hence, new health-matched replacements were brought into the study much later than other participants. The percent completing the physical examination by calendar date is plotted in Figure 5-1 for all Ranch Hands, Original Comparisons, and all Comparisons. The corresponding plot for Ranch Hands, Original Comparisons, old replacements, and the 28 restricted new replacement Comparisons is shown in Figure 5-2. Additionally, schedulers experienced reticence and vacillation with other Comparisons being scheduled for the first time. In particular, as a group, the 71 unrestricted new replacement Comparisons were also scheduled later than other participants. Figure 5-3 shows the percent of Ranch Hands, Original Comparisons, "old" Comparisons, and the 71 unrestricted newly examined replacement Comparisons completing the physical examination by calendar date. During the scheduling for the 1987 followup examination, schedulers will attempt to schedule health-matched replacements within 15 working days of a refusal. NEW REPLACEMENTS VERSUS OLD REPLACEMENTS Another statistical issue of concern is the homogeneity of the replacement Comparisons. The validity of the study might be compromised if, for example, newly admitted replacements had self-selected themselves into the study differently than previously admitted replacements. This kind of difference may occur due to changes in public opinion regarding the Agent Orange issue, the national political climate, changes in national opinion regarding health care, changes in the location of the examination site, or a combination of these and other factors. This issue was addressed by comparing new with old replacements on a variety of endpoints with adjustment for the matching variables. Blacks were deleted from the analyses. Two separate series of analyses were performed, one for each of the two kinds of new replacements (unrestricted and restricted) defined earlier. First, unrestricted new replacements were identified as the 71 replacements who were examined for the first time at first followup, regardless of their compliance at Baseline. Second, analyses were restricted to the 28 replacements who were examined for the first time and who had never been contacted before the first followup; these were called restricted new Comparisons. In each of the two series of new replacement analyses, all replacements not satisfying the definition of "new" are included by referring to them as "old" replacements. All "old" replacements were at least contacted at Baseline and were fully compliant at first followup. 5-16 �10090807060- Cumulative Percent 50 • 40 - 30- 20100M r J J A S 1985 IDD O 0 N D J Date at Clinic 0 0 o All Comparisons F M 1986 o o o Original Comparisons * * * Ranch Hands Figure 5-1. Percent Completed Physical Examination by Calendar Date for All Comparisons 100 90 80 70 60 Cumulative Percent 50 • 40 30 20 10- 0M r J i J r A 1985 i S i 0 i N Date at Clinic. Restricted New Replacement Comparisons Original Comparisons \ D F 1986 o o o Old Replacements & a-a Ranch Hands Figure 5-2. Percent Completed Physical by Calendar Date 5-17 M �100 90 8070- Cumulative 60Percent 50. 40302010- 0 M T J ~r J A 1985 IDD 1 0 —r S 1 N ~r J Date at Clinic 0 Q"$ Unrestricted New Replacement Comparisons * * * Original Comparisons M —r A 1986 €> o o Old Replacements Ranch Hands Figure 5-3. Percent Completed Physical Examination by Calendar Date for Unrestricted New and Old Replacement Comparisons In each of the tvo series of analyses, new and old replacement Comparisons were contrasted on health perception (excellent, good, fair, or poor), medication use (yes, no), work loss (yes, no), and daily use of aspirin (yes, no). Blacks were deleted from all analyses. New and old replacements were then contrasted on 20 clinical determinations from the first follovup examination. Table 5-13 shows tvo cross-classifications of 313 nonblack replacements, from a total of 338 replacements fully compliant at first follovup, by group (old, nev) and reported health status. In the unrestricted sense, the reported health status of nev and old replacements differed significantly (p-0.04), vith nev replacements reporting more fair or poor health than old replacements. In the restricted sense, the difference betveen nev and old replacements vas statistically significant (p=0.001), vith nev replacements tending to declare themselves of fair or poor health more often than old replacements. 5-18 �The same groups were contrasted on medication use; the results are shown in Table 5-14. The difference between old and new Comparisons under the unrestricted definition was not statistically significant (p=0.16) as regards medication use. The difference between old and new Comparisons under the restricted definition was, however, statistically significant (p»0.003). This difference was due to the higher reported medication use of the 26 nonblack new replacements not previously contacted. New and old replacements were contrasted on work loss due to illness; the results are shown in Table 5-15. TABLE 5-13. Reported Health Status of Nonblack New and Old Replacements, According to Two Definitions of "New" Unrestricted Old Restricted New New Old Number Percent Number Percent Number Percent Number Percent Health Excellent Good Fair/Poor 142 91 19 Total 56 36 8 252 30 20 11 49 33 18 161 103 23 56 36 8 26 287 61 p=0.04 42 31 27 11 8 7 p=0.001 TABLE 5-14. Reported Medication Use of Nonblack New and Old Replacements, According to Two Definitions of "New" Restricted Unrestricted Old Medication Yes No Total New New Old Number Percent Number Percent Number Percent Number Percent 30 222 12 88 . 252 12 49 61 p=0.16 20 80 33 254 11 89 287 26 p=0.003 5-19 9 17 35 65 �TABLE 5-15. Reported Work Loss of Nonblack New and Old Replacements, According to Two Definitions of "New" Unrestricted Old Work Loss Yes No Total Restricted New Old New Number Percent Number Percent Number Percent Number Percent 47 205 19 81 252 12 49 61 p=0.99 20 80 54 233 19 81 287 5 21 19 81 26 p=0.99 The difference between new and old replacements regarding work loss under the unrestricted or restricted definition was not statistically significant (p=0.99 and p=0.99, respectively). Results of a similar contrast on daily aspirin usage are shown in Table 5-16. The difference between new and old replacements regarding daily use of aspirin under the unrestricted or the restricted definition was not statistically significant (p=0.99 and p=0.75, respectively). It is noteworthy that the differences for general health and medication use did not occur for work loss and daily aspirin usage, suggesting that some participants may have over-reported when asked less specific questions about their health. New and old replacement Comparisons were also compared on 20 clinical and psychometric variables measured during the physical examination and psychological testing. These 20 variables are a subset from 26 selected from among an entire collection of nearly 200 endpoints in this study by requiring near statistical independence within and between organ systems. Variables selection was accomplished by screening the correlation matrices of variables as an entire set and separately within each organ system, including examining partial correlations between single variables and linear combinations of other variables within organ systems. Identified first were 10 variables with pairwise correlations less than 0.10 in absolute value. This was followed by identification of 16 additional variables with pairwise correlations between 0.10 and 0.20 in absolute value, making a total of 26 variables. These variable selection screens were accomplished on Baseline data for 1,154 nonblack fully compliant Comparisons subsequent to publication of the Baseline Report. The complete set of 26 dependent variables selected as 5-20 �TABLE 5-16. Reported Daily Aspirin Usage of Nonblack New and Old Replacements, According to Two Definitions of "New" Unrestricted Old Restricted New Old New Aspirin Usage Number Percent Number Percent Number Percent Number Percent Yes No 182 69 Total 73 27 251 44 17 61 p=0.99 72 28 206 80 72 28 286 20 6 77 23 26 p-0.75 nearly statistically independent is shown in Table 5-17. correlation matrix of these 26 variables as determined on Comparison data set is shown in Table 0-1 of Appendix D. that relative statistical independence of these variables biological independence of these variables. The Baseline the entire It is recognized does not imply These 26 variables were intended to serve as the basis for statistical contrasts of Original Comparisons, shifted Original Comparisons, and replacement Comparisons in the decision regarding the inclusion of shifted Original Comparisons and replacement Comparisons in the primary analyses. Generically, the analyses first compared two groups on each of the 26 variables with adjustment for rank (officer, enlisted), age at Baseline (40 or under, over 40), occupation (officer flyer, officer nonflying, enlisted flyer, enlisted groundcrew), and race (Black, nonblack). Blacks were deleted from the analysis. The total number of significant differences on the first set of 10 dependent variables was used as the basis for a decision regarding group difference. These 10 analyses were assumed to be 10 independent repetitions of a Bernoulli trial with probability of 0.05 of success under the null hypothesis that there were no group differences for any of the 10 variables. The probability of observing three or more successes in 10 independent repetitions of a Bernoulli trial, with probability of 0.05 of success, is 0.012. The entire set of 26 analyses was then assessed to test the hypothesis of group equality. The probability of 4 or more successes in 26 independent repetitions of a Bernoulli trial, with probability of 0.05 of success, is 0.039. These 2 critical values, both probabilities below 0.05, were used to assess the analyses on the 10 and on the 26 selected variables. 5-21 �TABLE 5-17. Twenty-Six Dependent Variables Selected as Nearly Statistically Independent With the Use of Baseline Data Variables Having Pairwise Absolute Correlations Less Than 0.10 Total Bilirubin (TBILI) Diastolic Blood Pressure (DBP) White Blood Cell Count (WBC) Skin Index (SKIN) MMPI Depression Scale (MMPID) Blood Urea Nitrogen (BUN) Urine Specific Gravity (USG) Pulse Index (PULSE) Nerve Conduction Velocity Above the Elbow (NCVE) Semen Count (SEMEN) Variables Having Pairwise Absolute Correlations Greater Than 0.10 and Less Than 0.20 Red Blood Cell Count (RBC) FEV1/FVC (PULM) Glucose (GLUC) Electrocardiogram (ECG) Platelet Count (PLAT) Full IQ (IQ) Central Nervous System Index (CNS) Nerve Conduction Velocity Above the Ankle (NCVA) Cholesterol (CHOL) Alkaline Phosphatase (ALKPHOS) Coproporphyrins (COPRO) Delta-Aminolevulinic Acid (ALA) Thyroid T4 (T4) Testosterone (TEST) Sedimentation Rate (SED) Gamma-Glutamyl Transpeptidase (GGTP) 5-22 �The statistical issue of how to account for the many interactions in the 26 separate analyses was not resolved during or since the first application of this method. Only the group main effect was regarded as the basis for determining whether a particular analysis was a success. At first followup, only 20 of the 26 variables were measured. The six variables not measured were the two-nerve conduction velocities (NCVE, NCVA), semen count (SEMEN), FEV1/FVC (PULM), full IQ (IQ), and delta-aminolevulinic acid (ALA). New and old replacements were contrasted on each of the remaining 20 variables via the general linear model and log-linear model. The variables—skin index (SKIN), pulse index (PULSE), electrocardiogram (ECG), and central nervous system index (CNS)—were analyzed as dichotomous variables, with each being scored abnormal if any of its components were abnormal. All others were analyzed as continuous variables. The correlation matrix of the 20 variables, based on 1,210 nonblack Comparisons fully compliant at first followup, on first followup data is shown in Table D-2 of Appendix D. The results of these analyses contrasting new versus old replacements with "new" following the unrestrictive definition and Blacks removed from the analyses are shown in Table 5-18. There were 61 nonblack new replacements and 251 nonblack old replacements. In some analyses, the dependent variable was transformed to better approximate normality. Unadjusted means are presented when there is a significant interaction involving group. The probability of observing 2 or more successes in 8 independent repetitions of a Bernoulli trial, with probability of 0.05 of success, is 0.057. In view of the results for the first 8 dependent variables in Table 5-18, new and old replacements appeared to be statistically indistinguishable. The probability of observing 3 or more successes in 20 independent repetitions of a Bernoulli trial, with probability 0.05 of success, is 0.075; the probability of 4 or more is 0.016. Recognizing the slight correlations between the dependent variables in the lower panel of Table 5-18, and the results of the analyses, new and old replacements again appeared to be statistically indistinguishable. The same analyses were conducted to contrast new and old replacement Comparisons, with "new" defined in the restrictive sense. The results are shown in Table 5-19, with the same notations as Table 5-18. The same binominal critical values, 2 for the first panel and 4 for the entire set of 20 analyses, and the results shown in Table 5-18 indicated that there was no statistical difference between the 26 nonblack new replacements and the 287 nonblack old replacements. The negative findings shown in Tables 5-18 and 5-19 suggested very strongly that there has been no change in the way replacements self-select for entry into this study. ORIGINAL COMPARISONS VERSUS SHIFTED ORIGINAL COMPARISONS The removal of ineligible Comparisons early in the Baseline scheduling operation resulted in the exclusion of approximately 18 percent of all Comparisons from the study. Since some of these ineligibles had been randomized as Original Comparisons, some previously randomized Comparisons were allocated to the positions vacated by the removed original Comparisons and, thus, were referred to as shifted Original Comparisons. 5-23 �TABLE 5-18. Summary Results of Unrestricted New Versus Old Nonblack Replacements Contrasted on 20 Variables Replacement Group Means* (Percent Abnormal) Variable (Transformation) Old New p-Value Significant Interactions Variables With Absolute Pairwise Correlations Less Than 0.10 TBILI (LOG) DBF (SORT) WBC (LOG) SKIN MMPID (LOG) BUN (SORT) USG PULSE 0.76 79.17 7.06 (54.0) 56.21 14.15 1.014 (16.7) 0.76 79.51 7.13 (49.2) 57.19 13.79 1.014 (11.5) NS NS NS NS NS NS NS . GRP*OCC, GRP*AGE Variables With Absolute Pairwise Correlation Between 0.10 and 0.20 RBC GLUC (LOG) ECG PLAT (SQRT) CNS CHOL (SQRT) ALKPHOS (LOG) COPRO (SQRT) T4 TEST (SQRT) SED (LOG) GGTP (LOG) 5.00 109.31 (15.5) 269.5 (2.8) 212.7 87.9 116.9 7.51 601.4 4.17 31.06 GRP*OCC*AGE 5.00 101.33 (13.1) 275.0 (5.0) 208.8 87.10 122.6 7,94 605.3 4.93 29.77 NS NS NS NS NS GRP*OCC 0.03 NS NS GRP*OCC*AGE GRP*AGE *A11 means are expressed in original units. NS: LOG: SQRT: Not significant (p>0.05) Analysis performed on logarithmic scale. Analysis performed on square root scale. GRP: Group OCC: Occupation AGE: Birth year (Age) 5-24 �TABLE 5-19. Summary Results of Restricted New Versus Old Nonblack Replacements Contrasted on 20 Variables Replacement.Group Means* (Percent Abnormal) Variable (Transformation) Old New p-Value Significant Interactions Variables With Absolute Pairwise Correlations Less Than 0.10 TBILI (LOG) DBF (SORT) WBC (LOG) SKIN MMPID (LOG) BUN (SORT) USG PULSE 0.76 79.44 7.01 (52.3) 56.11 14.02 1.014 (15.3) 0.75 76.98 7.91 (61.5) 59.73 14.75 1.013 (19.2) NS NS NS NS NS NS NS NS Dependent Variables With Absolute Pairvise Correlation Between 0.10 and 0.20 RBC GLUC (LOG) ECG PLAT (SQRT) CNS CHOL (SQRT) ALKPHOS (LOG) COPRO (SQRT) T4 TEST (SQRT) SED (LOG) GGTP (LOG) 5.01 108.8 (14.3) 270.5 (2.8) 212.5 87.75 117.8 7.56 601.2 4.15 31.23 4.90 95.86 (23.1) 271.56 (7.7) 205.6 87.72 120.5 8.00 612.6 6.37 26.41 NS 0.007 GRP*AGE NS NS NS NS NS NS NS 0.03 NS *A11 means are expressed in original units. NS: Not significant (p>0.05). LOG: Analysis performed on logarithmic scale. SQRT: Analysis performed on square root scale. 5-25 �Fully compliant Original and shifted Original Comparisons were compared in the Baseline Report with respect to reported health status, medication use, and work loss. Group differences for health status were significant (p=0.001) but were not so for medication use or for work loss; the shifted Original Comparisons tended to report themselves in poorer health than the Original Comparisons but were statistically equivalent to the Originals regarding medication use and work loss. Fully compliant Original and shifted Original Comparisons were contrasted at first followup on reported health status, work loss, medication use, and daily use of aspirin. As in the Baseline Report, these analyses were done for only nonblack Comparisons. The results of the contrast of Original and shifted Original Comparisons on reported health status are shown in Table 5-20. Here, health status is evaluated on a three-category scale (excellent, good, fair/poor). The group difference between Original and shifted Original nonblack Comparisons regarding reported health status was not significant (p=0.30). The results of the contrast of Original versus shifted Original Comparisons on medication use are shown in Table 5-21. The group difference between Original and shifted Original nonblack Comparisons regarding medication use was not significant (p=0.68). The results of the contrast on work loss are shown in Table 5-22. The group difference between nonblack Original and shifted Original Comparisons regarding work loss was not significant (p=0.82). The results of the contrast on daily aspirin usage are shown in Table 5-23. The group difference between Original and shifted Original nonblack Comparisons regarding daily aspirin usage was not significant (p=0.98). Fully compliant Original and shifted Original nonblack Comparisons were also contrasted on each of the full set of 26 nearly uncorrelated variables shown in Table 5-17 on Baseline data. The results are shown in Table 5-24. Sedimentation rate (SED) was analyzed as a categorical variable with values low (0-1), medium (2-3), and high (3-4). The percents of Original Comparisons within these categories were 35.8, 33.1, and 31.1 percent, respectively; the shifted Original Comparison percents were 30.8, 36.3, and 32.9, respectively. The probability of observing 3 or more successes in 10 independent repetitions of a Bernoulli trial, with a probability of 0.05 of success, is 0.0115. The probability of observing 2 or more is 0.0861. Based on these critical values and the results shown in the upper panel of Table 5-24, there appeared to be no statistical difference between Original Comparisons and shifted Original Comparisons. The probability of observing 4 or more successes in 26 independent repetitions of a Bernoulli trial is 0.039. The probability of observing at most 2 successes in 26 independent repetitions of a Bernoulli trial, with probability 0.05 of success, is 0.86. Based on these critical values and the known slight correlation of the 16 dependent variables in the second panel of Table 5-19, these results suggested that Original and shifted Original Comparisons are not statistically distinguishable. 5-26 �TABLE 5-20. Reported Health Status of Fully Compliant Original and Shifted Original Nonblack Comparisons: First Followup Original Comparison Group Original Reported Health Shifted Original Number Percent Number Percent Excellent Good Fair/Poor 387 307 53 Total 747 52 41 7 51 45 4 76 68 6 Total 463 375 59 p-Value 0.30 897 150 TABLE 5-21. Medication Use of Fully Compliant Original and Shifted Original Nonblack Comparisons: First Followup Original Comparison Group Shifted Original Original Medication Use Yes No Number Percent Number Percent 102 645 Total 14 86 747 23 127 150 5-27 15 85 Total 125 772 897 p-Value 0.68 �TABLE 5-22. Work Loss of Fully Compliant Original and Shifted Original Nohblack Comparisons: First Followup Original Comparison Group Shifted Original Original ¥ork Loss Number Percent Number Percent 631 125 No Yes Total 116 25 83 17 756 82 18 Total 747 150 p-Value 0.82 897 141 TABLE 5-23. Daily Aspirin Use of Fully Compliant Original and Shifted Original Nonblack Comparisons: First Followup Original Comparison Group Shifted Original Original Daily Aspirin Use Yes No Number Percent 529 218 Total 71 29 747 Number Percent 107 43 150 5-28 71 29 Total 636 261 897 p-Value 0.98 �TABLE 5-24. Summary Results of Original Versus Shifted Original Nonblack Comparisons on 26 Variables at Baseline Original Comparison Group Means* (Percent Abnormal) Variable (Transformation) Original Shifted Original p-Value Significant Interactions Variables With Absolute Pairvise Correlations Less Than 0.10 TBILI DBF WBC SKIN MMPID BUN USG PULSE NCVE SEMEN (LOG) 0.61 80.46 7.52 (37.5) 56.25 14.26 1.0209 (10.7) 56.26 77.4 0.61 78.95 7.18 (43.8) 58.40 13.76 1.0205 (.) 89 55.88 72.8 GRP*OCC*AGE NS NS NS NS NS NS NS NS NS Variables With Absolute Pairvise Correlation Between 0.10 and 0.20 RBC PULM GLUC (LOG) ECG PLAT IQ CNS NCVA CHOL ALKPHOS COPRO (LOG) ALA T4 TEST SED GGTP (LOG) 5.18 5.20 0.80 0.81 97.4 94.5 (26.7) (27.6) 269.9 270.6 108.4 108.6 (31.5) (23.7) 47.59 48.17 213.1 220.7 7.84 7.60 30.4 31.1 2,497.0 2,505.3 8.35 8.42 634.6 634.3 given in text 35.53 38.43 NS NS NS NS NS NS 0.02 0.01 NS NS NS NS NS NS NS . NS *A11 means are expressed in original units. 5-29 �Taken together, the results displayed in Table 5-24 very strongly suggested that Original and shifted Original Comparisons did not differ statistically at Baseline. These analyses were repeated on the 20 available variables at the first followup. The results are shown in Table 5-25. The results in the first and second panels of Table 5-25 and the binomial critical values given above suggested that no statistical difference was present between the Original and shifted Original Comparisons. A single multivariate linear regression analysis was done on the 20 dependent variables shown in Table 5-25; no significant interactions involving group (Original, shifted Original) were noted and the group effect was not significant (p=0.28). Taken together, these analyses strongly suggested that there was also no statistical difference between Original and shifted Original Comparisons at first followup. PARTIALLY COMPLIANT VERSUS FULLY COMPLIANT PARTICIPANTS Ideally, compliance bias should be assessed by comparing the health of noncompliant and fully compliant participants with adjustment for group (Ranch Hand, Comparison) and the matching variables. The only information available on the noncompliant participants, however, is their responses to the health status questions, if they were willing to answer them, during the telephone conversation in which they refused to participate in the study. Noncompliant Comparisons were contrasted with their Baseline replacements (see noncompliance telephone questionnaire data, Tables 5-9 to 5-12). In addition, as in the Baseline Report, selection bias was studied by contrasting partially compliant with fully compliant participants with adjustment for group (Ranch Hand, Comparison). Taking the Baseline questionnaire at followup but refusing to take the physical examination or followup questionnaire were 9 Ranch Hands and 30 Comparisons who were either nonlocatable or noncompliant at Baseline. These 39 men were the only partially compliant participants at first followup. Their Baseline compliance is summarized in Table 5-26. One of these individuals, a Ranch Hand with no interview, no physical, and no telephone interview, was Black. The label "no action" indicates that these individuals were not contacted because the Baseline contract expired. Individuals labeled "new Comparisons" were added to the study after the Baseline examination but before start of the first followup. Data from these 39 partially compliant participants were statistically compared with similar data from fully compliant participants with adjustment for group (Ranch Hand, Comparison). This is shown in Table 5-27. Endpoints evaluated were reported health, medication use, and work loss. These analyses are similar to those reported in Table V-15 of the Baseline Report. Reported health status was collapsed to two categories (excellent, good/fair/poor) due to sparse data. One Black participant, a Ranch Hand, was deleted from these analyses. The health versus compliance association in these data was of borderline statistical significance (p=0.08), with partially compliant participants tending to report themselves in better health than fully compliant 5-30 �TABLE 5-25. Summary Results of Original Versus Shifted Original Nonblack Comparisons on 20 Variables: First Followup Original Comparison Group Means* (Percent Abnormal) Variable (Transformation) Original Shifted Original p-Value Significant Interactions Variables With Absolute Pairwise Correlations Less Than 0.10 TBILI (LOG) DBF (SORT) WBC (LOG) SKIN MMPID (LOG) BUN (SORT) USG PULSE 0.75 80.0 6.88 (49.7) 56.2 14.8 1.015 (16.7) GRP*OCC*AGE 0.73 79.60 6.92 (42.1) 55.1 14.04 1.015 (16.4) NS GRP*AGE NS NS NS NS NS Variables With Absolute Pairwise Correlation Between 0.10 and 0.20 RBC GLUC (LOG) ECG PLAT (SORT) CNS CHOL (SORT) ALKPHOS (LOG) COPRO (SORT) T4 TEST (SORT) SED (LOG) GGTP (LOG) 4.97 111.8 (15.3) 263.2 (2.6) 219.5 89.76 115.4 7.58 576.6 5.11 32.39 4.95 111.6 (11.9) 271.9 (2.3) 214.1 85.53 114.9 7.58 559.0 4.91 29.77 NS NS NS NS NS NS NS NS NS GRP*OCC, GRP*AGE NS NS *A11 means are expressed in original units. 5-31 �TABLE 5-26. Baseline Compliance Status of 39 Partially Compliant Participants: First Follovup Group Ranch Hand Baseline Compliance Comparison 23 No Interview, No Physical, No Telephone Interview No Interview, No Physical, Telephone Interview New Comparison 0 3 No Action 4 3 Total 9 30 TABLE 5-27. Reported Health of Partially Compliant Versus Fully Compliant Nonblack Participants Group Ranch Hands Comparisons Compliance Status Reported Health Number Percent Number Percent Total Full Excellent Good/Fair/Poor 473 482 Total Partial 43 46 955 5 3 Excellent Good/Fair/Poor Total 635 575 20 23 20 10 1,108 1,057 2,165 1,210 30 5-32 57 54 80 77 25 13 38 �participants; 66 percent of partially compliant participants reported excellent health while only 51 percent of fully compliant participants reported excellent health. This association did not change with group (p-0.91). The data on medication use and compliance status demonstrated no association (p=0.57), and this equivalence did not change with group (p=0.79). These data are shown in Table 5-28. As shown in Table 5-29, the work loss-by-compliance association in these data was significant (p=0.03), with 84 percent of fully compliant participants reporting work loss and 95 percent of partially compliant participants reporting work loss. These data are sparse and are not considered supportive or nonsupportive of the compliance bias calculations presented in the Baseline Report. The conclusions of the Baseline Report regarding the potential effects of compliance bias should be regarded as conservative overestimates, but worthy of consideration in inference formulations until more data become available. CONCLUSIONS These predominantly negative findings suggest that there has been no change in the way replacements self-select for entry into this study and, due to the obvious scheduling differences between new and old replacements, that no additional bias has been introduced at followup by scheduling differences. These data also strongly suggest that shifted Original Comparisons are not statistically distinguishable from Original Comparisons, either at Baseline or at first followup. This interpretation is also equivalent to the conclusion that no additional bias was introduced by scheduling differences between Original Comparisons and shifted Original Comparisons at Baseline. Available data on noncompliant Comparisons and their replacements suggest that, although replacements were not health-matched to refusals at Baseline, they are r.emarkably similar to refusals with respect to reported health, medication use, and income level. This result also supports a conclusion that there has been little, if any, selection bias due to nonparticipation in the Comparison group. This conclusion supports the use of the total Comparison group for all of the main analyses in the body of this report. Data regarding the few partially compliant participants at first followup are not sufficient to confirm or deny compliance bias calculations published in the Baseline Report. 5-33 �TABLE 5-28. Medication Use of Partially Compliant Versus Fully Compliant Nonblack Participants Group Comparison Ranch Hand Compliance Status Medication Use Full Number Percent Number Percent Total Yes No 123 832 Total Partial 955 1 7 Yes No Total 167 1,043 58 290 56 1,875 1,210 42 44 2,165 25 21 3 27 8 TABLE 75 79 30 4 34 38 5-29. Work Loss of Partially Compliant Versus Fully Compliant Nonblack Participants Group Ranch Hand Compliance Status Work Loss Full Yes No Number Percent Number Percent Total 796 155 Total Partial Comparison 44 44 951 Yes No 8 0 Total 1,010 200 1,210 22 0 28 2 30 5-34 56 56 1,806 355 2,161 78 100 36 2 38 �CHAPTER 6 QUALITY CONTROL During the first AFHS followup, stringent adherence to quality assurance (QA) was planned for and upheld throughout the study, from project initiation to final product delivery and acceptance by the Air Force. A quality program plan was developed for this study cycle, outlining all contract activities requiring periodic and/or systematic QA and quality control (QC) monitoring. The purpose of this chapter is to provide an overview of the specific QA measures developed and used by the project team, specifically in the areas of administrative QC; questionnaire, physical, and psychological examination QC; laboratory QC measures; data base management QA; and statistical QC. ADMINISTRATIVE QUALITY ASSURANCE In recognition of the magnitude, complexity, and importance of the AFHS, a Quality Review Committee (QRC) was established at the initiation of the third-year followup for the purpose of providing general oversight to the AFHS QA Program and advice on the appropriateness of program management and QC actions. The QRC was composed of senior corporate personnel from the prime contractor. These independent reviewers remained separate from the project management staff. The QRC met formally each quarter to review recent study progress and any issues that either had an impact on study quality or were perceived as a potential problem. Assisting the QRC in day-to-day oversight responsibilities was a QA officer responsible for reviewing procedures, performance, and work products from all task managers and key project staff. As part of the monitoring function, the QA officer received exception reports from project task managers whenever an incident occurred that appeared to affect study quality. Monthly reports were also prepared for the Air Force, documenting project compliance with project QA criteria and noting any instances of noncompliance. An additional measure of corporate QC was implemented through independent QA audits of individual project tasks. Members of the QRC determined first-hand whether QA procedures for a particular task were being conducted, whether procedures were appropriate for the task, and whether QA was complete for all aspects of each task. The remainder of this chapter comprises specific QA procedures followed for the individual tasks. QUESTIONNAIRE QUALITY CONTROL NORC used both onsite and home-office QA procedures to produce a comprehensive data set. All AFHS questionnaires were pretested to evaluate 6-1 �their completion time and participant acceptability before they were used at the SCRP. Onsite QC procedures included weekly observation and rating of each interviewer, editing of every questionnaire at the completion of the interview, and monitoring of participant evaluations. The Air Force also continuously conducted QA observations of all onsite activities. QC of data processing included manually editing each questionnaire, including a 100-percent verification of critical items for each questionnaire, computerized cleaning (with both single item and inter!tern review for range and consistency), identifying outliers, and reviewing the actual questionnaire copy to reconcile or correct detected errors. All telephone surveys were monitored for quality and accuracy of interviewer performance by NORC supervisors. The telephone survey supervisor monitored 3 percent of each interviewer's calls to assure an appropriate presentation and an accurate transcription of responses. An additional 5 percent of the participants were recontacted after the interview to evaluate interviewer performance and validate that the correct respondent had been contacted. NORC recruited and trained interviewers according to the detailed procedures described in Chapter 3. A minimum number of interviewers was selected to reduce interviewer variability. Additionally, these individuals were blinded to the participants' exposure status to avoid any bias. Interviewers were required to ask questions exactly as recorded, and in the order in which they appeared. No personal interpretation was allowed. An onsite field manager closely supervised each interviewer's work regularly, observing individual interviews weekly during the examination schedule. The field manager reported directly to the NORC Project Director weekly, and was reviewed by the Project Director during quarterly site visits, to ensure direct accountability by the home office and the field manager for promptly resolving any issues. Specifically, interviewers were checked for accuracy in questionnaire skip patterns, probing, circling of the correct code, control of the interview, voice quality, reading, and use of associated documents. When called for, the onsite manager gave immediate retraining after each observation and documented the content of this training. At weekly meetings, held with all interviewers, the field manager used generalizations from individual interviewer performance observations to train an entire group of interviewers. The NORC field manager also monitored participant evaluations of the study closely and used the information gathered to plan and implement retraining. The manager and staff edited each completed questionnaire before it was shipped to Chicago, attempting to retrieve missing data while the study participant was at the physical examination site. Missing or ambiguous data were also retrieved by telephone when necessary. Spouse fertility data were obtained independently of the participant interview by sending the mail questionnaire while the study participant was at the examination site, and by having a group meeting for wives who accompanied their spouses to the clinic site, where they could complete their questionnaires in private. The Assistant Survey Director in Chicago supervised and edited all interviews conducted at home with participants and spouses. 6-2 �Once the participant and spouse questionnaires were received in Chicago, they vere edited for completeness by a coding supervisor and staff dedicated to the AFHS for the entire project. Resolution of inconsistencies was accomplished by staff members, who standardized all responses prior to keypunching. Questionnaires were then coded, and a 10-percent recede was done on open-ended items. When a batch failed the 10-percent recede, the entire batch was receded and the coding staff was retrained. One hundred percent quality control was accomplished by the Air Force. During data entry, range validity checks were performed and 10 percent of the most important items in each questionnaire was verified. Data were then passed through a computer program that checked for inter- and intracolumn errors. When errors were detected, the-questionnaires were reviewed and the errors corrected. The process continued until no errors were detected by the cleaning program. Then, frequencies were reviewed and any anomalies or errors previously undetected were corrected by reviewing the questionnaires on a case-by-case basis. All corrections were entered into the data tape, but no changes were made to the data recorded in the questionnaires. QA reports were generated monthly, detailing the summary statistics on the number of questionnaires reviewed, the number and types of transcriptions failing QC checks, and the average number of coding errors per batch processed. PHYSICAL EXAMINATION QUALITY CONTROL QC was emphasized in the physical examination, as this data source provided most of the medical information for clinical and epidemiological analyses. Initial concern for a high-quality physical examination was addressed by a stringent SCRF selection process for all personnel who were to directly interact with the participants. Each staff member was hand-selected for the AFHS on the basis of expertise, experience, and a commitment to remain with the study throughout the examination cycle. Further, the Air Force Technical Team reviewed the credentials of all key staff members and approved their participation in the study. A complete pretest physical examination, interview, psychological test, and laboratory workup was done for 10 volunteers several weeks before the scheduled start of the study. Refresher training was given to the dermatologists to enhance their skill in diagnosing chloracne, techniques for detecting specific heart sounds were reviewed with the internists, and diagnosticians were reminded of the need to review Baseline examination data as they formulated all diagnoses. Further, all aspects of patient contact were reviewed: the initial inbriefing of the participants, the logistics of transportation and patient flow within the clinic, and the final outbriefing by the diagnostician. During the examinations, refinements continued whenever operational problems were detected by the SCRF staff and the Air Force onsite monitor, or when participants identified areas requiring improvement. Both of these types of information were addressed during the weekly clinical QA meeting of key SCRF staff, chaired by the SCRF Medical Project Director and attended by an Air Force representative. In addition, written critique forms submitted by all participants were reviewed in detail at the SCRF weekly meetings, 6-3 �providing additional insight to both temporary shortcomings of the entire logistic process as well as the numerous strong points of the programs. Following examination of each participant group, all physical examination forms were reviewed by the SCRF staff for omissions, incomplete examinations, and inconsistencies. The examiners or technicians were quickly contacted to correct the data. Special effort was made to complete this review while the participants were at the examination site. In all cases of data correction, a complete audit trail was maintained. Finally, all mark-sense physical examination forms were read by an optical scanner to ensure total continuity and sensibility of the final examination contents. (This subject is discussed in more detail in the Data Management Quality Control section of this chapter.) Compliance with all aspects of the physical examination was monitored daily by the Air Force onsite monitor and the SCRF Medical Project Director. Additional periodic inspections were conducted by the SCRF Chief of Medicine and the SAIC Principal Investigator. All such clinical reviews were done unobtrusively, and with the full consent of the participant; suggestions or corrections to the examination procedure were always discussed privately with the attending physician. These inspections emphasized aspects of clinical techniques, sequencing and completeness of the clinical data with respect to the examination forms, and the total blindness of the examinations. Of particular note were the detailed daily log entries of the five Air Force monitors. These entries ensured continuity of knowledge (the monitors rotated approximately every 2 weeks) by documenting examination procedural changes and recording events requiring followup by either the Air Force or the prime contractor. Establishment of rapport with each study participant was a primary goal of all organizations involved in this study. Although "rapport building" may not be a traditional QA parameter in most research studies, it is paramount in the AFHS because maintaining the satisfaction of participants encourages them to continue in the study, and thus a significant reduction in future statistical power or bias, or both, is avoided. Every staff member, therefore, from the initial telephone recruiter to the nurse coordinator and the Project Manager, emphasized courtesy, empathy, assistance, and personalized treatment of each participant. LABORATORY QUALITY CONTROL Before the study was begun, specific QC laboratory procedures were designed, developed, and implemented to rapidly detect problems related to test/assay performance, validity of reagents, analysis of data, and reporting of results. All laboratory assays for the study were done with state-of-theart laboratory equipment and techniques. Laboratory facilities all had the equivalent of National Institutes of Health Biosafety Level 2 (BSL-2) approval ratings and were certified by the College of American Pathology (CAP). Hematology assays were performed on Coulter S Plus® equipment; sedimentation rate determinations were performed using the large-tube Westergren method. The Dupont Automated Chemical Analyzer® (ACA) was used to perform the biochemical assays; radioimmunoassays (RIA) were done with standard test kits; and porphyrin was assayed by high-performance liquid 6-4 �chromatography at the Mayo Clinic in Rochester, Minnesota. Hepatitis B tests were performed using Abbott kits, and manually performed electrophoresis and monospecific antibodies were used for immunologlobulin assays. Blood-cell counts were performed with standard microscopy, and Clinitek, a reflectance spectometry urinalysis, was used for all urinalyses. All other assays were done using industry-approved equipment and techniques. All laboratory operations were controlled with the use of an integrated medical laboratory management information system that incorporated direct device to data base interfaces for automated testing equipment, and data entry for manual tests was performed by the laboratory technologists. An automated audit trail and a set of comments for technologist entries were kept for each test so that any QC results could be retraced. Procedural QC included using instrumentation and reagents from one lot number throughout the study. Strict standards of calibration for all automated laboratory equipment were maintained at all times. Trilevel or bilevel controls were used as the primary means for monitoring the quality of all tests. On every group of participant samples, one control (low, medium, or high) was run at the start, after every ninth sample, and at the end of each test run. Each trilevel control was used before repeating it in the run, when more than 18 experimental samples were analyzed. In addition, split aliquots were made from every tenth patient sample and were analyzed separately to measure test reproducibility. All QC data were analyzed and summarized in formal QC reports generated weekly. QC data were subjected to independent statistical analysis to produce and analyze time-dependent trends. For all equipment malfunctions or other exceptions, a formal QC exception report was prepared by the responsible individual and forwarded to the QA officer and the project management team. An additional measure of quality control introduced during the study was the CUSUM tests run with trilevel controls. In particular, the fast initial response cumulative sum (FIR CUSUM) QC technique was used. It has an advantage in detecting long-term 2 subtle drift that could have substantial adverse analytical consequences. FIR is a special case of the CUSUM QC scheme that increases the overall effectiveness of the QC procedure. Unlike QC procedures using standard control charts, which compare each observation to designated limits, these tests utilize the cumulative sum of deviations from a target value. CUSUM statistics were accumulated for each of the trilevels to quickly detect instrument calibration problems as identified by excessive drift. If an out-of-control situation was indicated, the graph showed when the change first occurred. Coefficient of variation (CV) standards were established before the study for each test. All adjacent patient samples were reanalyzed after the equipment was thoroughly checked and fresh controls were run. FIR CUSUM generally has been applied to QC in industry, particularly in high-volume, high-precision applications. To our knowledge, FIR CUSUM has not generally been applied in a biomedical setting. According to SCRF laboratory personnel, this procedure proved so successful in the AFHS that most of the SCRF clinical laboratory will begin using it in the near future. 6-5 �As the examination portion of this study ended, all laboratory outliers were analyzed for logical validity by an independent clinician. All out-ofrange test results were examined and scored as clinically explainable, clinically possible, or clinically unexplained. Quality Control Procedures for the Immunology Laboratory The QC procedures for the Cellular Immunology section of the AFHS were structured to rapidly detect any problems in four major test parameters: (1) assay performance, (2) reagent validity, (3) data analysis, and (4) results reporting. The QC measures were detailed in the Quality Procedures Plan and documented before testing started. Compliance was monitored daily by the Cellular Immunology laboratory supervisor. Key aspects of the program included instrument and equipment calibration and maintenance, assay controls, accuracy and precision determination, and system failure checks. QC measures followed in all Cellular Immunology assays included: • Blood sample from a normal, healthy control individual with each group of AFHS patient samples • Duplicate testing of one random patient sample in each assay • Quadruplicate testing of each patient sample for each variable in each of the functional assays (e.g., PHA stimulation, natural killer cell effector/target ratios) • Parallel testing and monitoring reactivity of various lots of reagents when appropriate • Verification of patient and specimen identification by at least two individuals before final reporting to the data base • Note codes attached to any data point with a detected deviation from normal due to procedural setup error, assay malfunction, equipment malfunction, or assay technical error • • Review of all final assay reports by the Cellular Immunology laboratory supervisor prior to entry into the data base. QC for each functional assay including phytohemagglutinin (PHA), pokeweed, mixed lymphocyte culture (MLC), and natural killer cell consisted of monitoring assay controls, duplicate sample reproducibility, and any trends in reagent reactivity. Assay precision was determined by calculating the CV of the quadruplicates for each variable tested. Also, a mean value of the CV for each assay was calculated. Individual CV's of 15 percent or less were the target values for the stimulated samples in the mitogen and natural killer cell assays. The Student's t-test was applied to duplicates to determine if there was a significant difference in sampling for the functional assays. Critical t-values at the 0.05 significance level were used to determin^ if duplicate sample results varied significantly. Grubbs' statistical test was used to identify any statistically significant outlier. This test was applied only to samples whose CV's were greater than 20 percent at a p-value of 0.01. The mitogen stimulation (PHA, pokeweed) effect was 6-6 �followed by daily evaluation of the radioactive counts in counts per minute (cpm) for each mitogen. When counts fell below expected values, suggesting that reagent deterioration had occurred, new aliquots were used. QC measures for the cell surface marker assays were calculation of T4+T8/T11 cell ratios, evaluation of flow cytometer computer outputs (cytograms and histograms), and duplicate sample testing. T +T8/T cellular ratios should approximate the value 1.0 for a normal population, validity of cytogram and histogram distributions generated by the flow cytometer was confirmed by the Cellular Immunology laboratory supervisor for each sample analyzed. The percent positive cells for each surface marker was determined in .the duplicates and viewed graphically using a microcomputer program. Any significant differences between duplicates were noted and follgwed for abnormal trends. On completion of this followup effort, the entire cellular immunology data base was reviewed by the Air Force team, laboratory staff, and consultants. Comments attached to the data points were also reviewed. Any data point that appeared unusual was reviewed and identified as an unexplained outlier. Unexplained outliers were deleted from the data base as errors of an unknown nature. This review was conducted without knowledge of exposure status. DATA MANAGEMENT QUALITY CONTROL Overviev of Quality Control Procedures The QC program for the data management activity consisted of multiple checks at all steps of the examination, data collection, and data processing cycle. Data QC procedures for data collection, conversion, and integration were developed before the clinical examinations began. Pretesting of all forms, procedures, and logistic arrangements was conducted 3 weeks before the examinations actually began. Additionally, during the first 2 months of the clinical examinations, all data collection activities were intensely scrutinized to detect and correct procedural deficiencies. QC activities also included automated QC techniques applied to laboratory data, clinical evaluations of all laboratory outliers, review of all physical examination findings by an independent diagnostician, and automated and manual data quality checking of hard copy against transcribed computer files for all questionnaire, physical examination, and medical coding data streams. Five interwoven layers of QC were instituted to ensure data integrity. Efforts focused on (1) data processing system design, (2) design and administration of all exams or questionnaires, (3) data completeness checks, (4) data validation techniques, and (5) quality control of medical records coding. In some cases, the QC procedures about to be described were implemented throughout the data management task rather than assigned to a particular activity. These comprehensive QC procedures will be mentioned where appropriate throughout the remainder of this section. 6-7 �Data Processing System Design For each data stream, standards were set to establish data element format (character or numeric), data element naming conventions, data element text labels, numeric codes for qualitative responses and results, QC range checks for continuous data elements, and QC validity checks for categorical data. A data dictionary provided detailed information on each data element. A systems integration approach was applied to the design and implementation of data collection procedures and techniques so that data emanating from the various study sources (physical examination, questionnaire, laboratory) were consistent in file format and structure. This was necessary to ensure that all data could be integrated into a single data base management system for analysis. Figure 6-1 provides an overview of the QC activities used in the data base management process. Forms and questionnaires were carefully designed to ensure that all required data elements would be collected according to the Study Protocol. The design of these instruments was such that they reflected the order in which the examination itself would be administered and provided for the sequential receding of information to streamline remaining data management activities. Completed medical records and questionnaires were converted from hard copy to machine-readable images using customized data-entry systems or state-of-the-art optical mark reading equipment. Verification procedures were performed to ensure that a uniquely identified participant record existed within each data file, and that the appropriate number of responses for each applicable field was provided. Data files were then verified against original data sheets and corrected as necessary. Data files were then subjected to validity checks. Any potentially conflicting results as well as any data values falling at the extremes of expected ranges were manually reviewed. Extreme values were reverified against the original raw data copies and either corrected or documented as valid results. Potentially conflicting results were returned to the examiners for review. These results were then documented as correctly recorded, corrected, or flagged for exclusion from analysis because of unresolvable examiner errors or omissions. Once the edits were completed and the data reverified, the "cleaned" files or tapes were transferred to the data analysis center for final inspection and integration into the study data base. For this QC measure, each data file was loaded into a Statistical Analysis System (SAS®) data set, and descriptive analyses were run. The validation, correction, transmission, and analysis QC procedures were repeated as necessary to ensure that all extreme or suspicious values had been validated. Design and Administration of Physical and Psychological Examination Forms As mentioned, the examination forms were designed to solicit all required data such that recording time was minimized, comprehension was enhanced, and data input could occur with a minimum of transcription errors. Optical Mark Recognition (OMR) technologies were selected to eliminate the risk of transcription errors and were applied to all psychological tests. 6-8 �Data QC Flow Chart Physical and Psychological Examinations / Laboratory Analysis 1I Onsite Review 1 Level 1 QC Contract Report ^^/~~^ / / / / Interview/ Questionnaire / Contract Report FIR Onsite Review Form QA Check Contract Report Data Cleaning Optical Scan Microprocessor Scan Univariate Analysis Contract Report Univariate Analysis Contract Report i_i Level 2 QC Merge/ Load Automated QC Final Contract Report Figure 6-1. Two Levels of Quality Control Applied to All Collected Data Prior to Statistical Analysis Univariate Analysis / / �Customized mark-sense forms were also developed and OMR technology was used to achieve these same objectives for segments of the physical examination and the self-administered questionnaires. The use of mark-sense forms allowed the creation of computerized data files directly from the raw data recorded on these forms. QC procedures for all data collection instruments began with a review of all forms immediately as they were completed. Any forms containing missing examination results were returned to the examining physician for completion before the participants left the site. Any questionable results or "hard-to-diagnose" conditions (such as heart sounds or peripheral pulses) were verified by the diagnostician at the outbriefing. All examination forms were signed by the examining physician, and the examiner identification number was coded in the data base so that interexaminer variation could be analyzed. Detailed QC records were maintained, which indicated the examining physician and the type of deficiency detected. Deficiency reports were reviewed by the study coordinator to detect any patterns of physician data entry error. A final level of QC audit was accomplished by Air Force statisticians, who conducted a detailed screening of the data and checked for errors. Data Completeness Checks Customized programming of the OMR allowed for the identification of those forms (and their corresponding data records) with missing responses, as well as those with multiple responses to questions that required a single response. The OMR scanner was programmed to reject forms that failed completeness and multiple response checks and to output a control code for each rejected form. The control code identified the location of the first three verification checks failed for a given form. Vhen a raw data form was rejected, the reason for the rejection was determined and the exact data element was corrected by comparing the rejected raw data form to the values recorded in the data record created by the scanner. A customized set of rejection and resolution codes was developed for the study to describe all the reasons for a form's rejection and any subsequent reasons for changing a data value. Various codes identified values recovered from light marks, missing marks explained by examiner comments, and missing comment flags resolved by the presence or absence of text in the comment areas. These codes ensured data completeness by accounting for all questionable or missing responses. (See examples of marksense forms in Figures 4-3 and 4-4.) Some of the rejected forms did not contain actual data errors but rather anomalies created in using mark-sense cards for data collection. For instance, incompletely erased responses and responses marked with too little carbon or graphite were incorrectly counted or missed, respectively, by the scanner. Examiners also tended to clearly mark responses for abnormal findings while bypassing or lightly marking responses for expected or desired findings. Failure of the form tp provide the correct number of expected responses always resulted in rejection. These technology-based errors were resolved, as were the anticipated, more traditional errors. The rejection code, data location code, resolution code, data inspector's initials, and correct data value were directly posted to a 6-10 �participant's data record. This innovative technique not only effectively maintained a comprehensive audit trail of all record manipulations, it also provided a mechanism for measuring the frequency of specific errors. Careful monitoring identified trends where individual data values were missed as well as the frequency with which individual examiners incorrectly marked their examination forms. Statistics were compiled on out-of-range results and data omissions that had been accepted in the previous QC audits. The results were monitored to detect trends, possible bias situations, and other data quality problems. This information was reviewed and relayed to examiners and internal auditors to assist in preventing or correcting chronic, but avoidable, problems. Data Validation Techniques QC activities also included data validation techniques. As mentioned earlier, data files were examined in a series of verification and validation procedures developed to check the results within each participant's record for logical consistency and abnormal findings. Any records noted to have ambiguous findings, incongruent observations, extreme results, or nonobvious errors or omissions were listed.and submitted for review to a physician. Again, clinical judgments were made by the auditing physician in assigning a validation code for each extreme or questionable data result. The validation codes allowed for indicating that data were deciphered from examiner comments or from related findings from another specialty area, or were accurately recorded and logically consistent with other findings for the participant. Data points that could not be definitively validated or recovered through clinical judgment and consultation with the original examiner were assigned codes noting missing or invalid data values. These unrecoverable data points were excluded from subsequent analysis. Medical Records Coding Quality Control Upon completion of the NORC data processing, all AFHS questionnaires were forwarded to SAIC for the medical coding of reported conditions. The International Classification of Diseases, 9th Revision, Clinical Modification (morbidity); International Classification of Diseases, 9th Revision (mortality); Systematized Nomenclature of Medicine (anatomic site); and American Hospital Formulary Service (medications) coding schemes were used, suitably modified. Each questionnaire was coded by two coders working independently. The results of the two coders were forwarded to the USAF for 100-percent QA/QC and final adjudication. The information from the physical examination was coded similarly. After the coding data were adjudicated, they were returned to SAIC for data entry. The coding sheets were batched, key entered, verified, and corrected. The corrections were also verified. The .key entry and verification functions were performed by.various operators. Five percent, or 100 records of each batch (whichever was larger), was randomly selected and subjected to manual reverification. An error rate of greater than 1 percent of this sample mandated reverification of the entire batch. In this final QA/QC check, the automated files were reviewed and compared to the hard copy by trained medical record coders, all of whom satisfied the minimum requirement of Accredited Record Technician or Registered Record Administrator eligibility. 6-11 �A manual tracking system was used to retrieve medical records. A chronological log was maintained to track participant requests for authorization to obtain medical record(s), receipt of the authorizations, requests for records from the provider, and receipt of the records from the provider. Identifying information in these logs included participant name, case number, date of action, condition(s) to be verified, dependent name (if appropriate), and type of medical provider (Federal/non-Federal). Due to the intricacies of obtaining medical records from Federal facilities, this task ultimately became the responsibility of the Air Force. STATISTICAL ANALYSIS QUALITY CONTROL Specific QC measures vere developed for activities falling within the statistical analysis task: construction of data bases for the statistical analysis of each clinical chapter, the statistical analysis itself, and the production of statistical reports to serve as the basis for the clinical chapters. Each specialized statistical data base was constructed by defining and locating each variable within the many subparts of the composite followup data base. Lists of variables and their data sources were submitted to the Air Force for approval. Although the data had been subjected to QC procedures during collection, statistical checks for outliers and other improbable values were conducted? anomalies identified by the statisticians were discussed with those responsible for the data collection, i.e., either NORC or SCRF. QA largely depended on regular communication and general agreement among statisticians. Several meetings and consultations among the Air Force team, the Principal Investigator, the SAIC statisticians, and the University of Chicago staff members were held in conjunction with the development of the data analysis plan. During the course of the analysis there were frequent telephone conversations. Any problems arising in the statistical analysis were resolved by team discussion. The software was checked by comparing results from analyses on the same variable by different programs (for example, BMDP*-LR [logistic regression] and BMDP®-4F [log-linear model] will give the same results for dichotomous variables when the program options are chosen properly). The statisticians frequently checked that the number of observations used in an analysis was correct, and peer review ensured that the program code was appropriate for the chosen procedure. The analyses were conducted in accordance with the data analysis plan which was reviewed extensively. Throughout the study, duplicate data bases were maintained by the USAF and SAIC. Upon completion of the analyses, SAIC delivered all analysis software and SAS data sets for each clinical area to the USAF for final review and archiving. All tables and statistical results were checked against the computer output from which they were derived, and all statistical statements in the text were checked for consistency with the results given in the tables. Additionally, drafts of chapters in the report were reviewed by the USAF and SAIC investigators, and the QRC. 6-12 �CHAPTER 6 REFERENCES 1. Bissell, A.F. 1969. CUSUM techniques for quality control. Appl. Stat. Vol. 18. 2. Lucas, J.M., and R. Crosier. 1982. Fast initial response for CUSUM quality control schemes: Give your CUSUM a headstart. Technometrics Vol. 24. 3. Grubbs, F.E. 1969. Procedures for detecting outlying observations in samples. Technometrics XI:1-21. 6-13 �CHAPTER 7 STATISTICAL METHODS . This chapter summarizes the key statistical elements of the study design, the statistical analysis issues, and the specific statistical methods used in the analysis. Additional details may be found in the USAF Study Protocol. The primary focus of the statistical analysis was a contrast of health status of the Ranch Hand and Comparison groups. Assessments were made of the proportions of participants with abnormal findings and of mean levels of key laboratory measurements. The analyses encompassed both simple contrasts between the two groups and more complex methods, in which adjustment was made for important covariates. In addition to these analyses, the possibility of an increasing response of medical problems with herbicide dose was explored, since if indeed there were an effect, more problems would be expected among the more heavily exposed. Although exact dosage information is not available, an exposure index was developed for the exposed population (the Ranch Hands) that approximates the potential herbicide exposure of each individual, incorporating information such as the occupation of the individual, his period of duty in the spraying operation, and the numbers of barrels per day of herbicide used during that period. Details on the exposure index are given in Chapter 8. Dose-response analyses were conducted for the Ranch Hands only, using this exposure index as a surrogate measure of dose. Interpretation of the results of the exposure index analyses, however, depends critically on the accuracy of the exposure index, which presently can be regarded as only fair. (Improved dosage information will be obtained for future studies from recently developed serum dioxin assay techniques.) Thus, the analyses of overall group differences between the Ranch Hands and the Comparisons are given primary emphasis, and the exposure index analyses merely supplement them. STATISTICAL STUDY DESIGN An overt herbicide effect would be characterized by more symptoms, signs, abnormal laboratory tests, syndromes, or diseases in the Ranch Hand group than in the Comparison group. If the disease(s) were fatal, increased mortality might also be observed. A subclinical herbicide effect would be detected as an increase in abnormal findings on the physical examination (particularly laboratory tests) that may or may not also be associated with symptom reporting or increased mortality. Thus, the basic objective of the statistical analysis is to test for differences between the Ranch Hand (exposed) group and the Comparison (nonexposed) group. 7-1 �In general, two types of data are used in the analysis. First, there are subjective data on symptoms reported by the participant in the questionnaire and in the review-of-systems section of the physical examination. Second, there are objective data, which include medical findings or signs identified during the physical examination, or by reviews of laboratory results, medical records, and death certificates. Symptoms reported by the study participants are subjective by definition, and are subject to influences that could result in erroneous conclusions. An association found between reported symptoms and herbicide exposure must be subjected to further confirmation, as the observations may result from over- or under-reporting bias and may not be indicative of a true herbicide effect. On the other hand, the medical findings data do not suffer from the same degree of participant influence. The medical findings and medical records review were conducted by highly trained individuals employed for the duration of the data collection and assessment phases of the study. They were held to stringent QC standards, as described in Chapter 6, to ensure that these data were as objective and accurate as possible. Incorporated in the study design is a feature that attempts to check for and correct symptom-reporting errors. A key component is a reported symptom verification process conducted by reviewing participant medical records and findings from the physical examination. In the retrospective morbidity portion of the study, the participant is questioned on past illnesses and medical conditions. With the participant's consent, an effort is made to obtain the medical records to verify the reported condition and, thus, to substantiate any unverified conditions. In addition, the study design includes verification of negative responses to determine unreported conditions. The medical record review process is time intensive and only a portion of the data was available for analysis in this study. Over-reporting was assessed by comparing the reported illness rates with the results of the physical examination and medical record review. Similarly, the assessment and correction of under-reporting requires the review of medical records to identify unreported illnesses. Obviously, this under-reporting assessment is restricted to conditions for which medical care was obtained or that were identifiable at the physical examination. STATISTICAL ISSUES In conducting the statistical analysis of the data in this study, there are a number of underlying issues. Except for bias, which is the topic of Chapter 5, these issues are discussed in this section. However, based upon the results of the bias analysis presented in Chapter 5, all statistical analyses in the clinical chapters use the contrast of Ranch Hands versus the total Comparison group. For the purposes of completeness and cross-reference to the Baseline report, identical analyses using the contrast of the Ranch Hands versus the Original Comparisons have been conducted, and these results are presented in the form of summary tables in each chapter appendix. 7-2 �Intervening Variables When comparing any two groups of individuals, the exact proportion of diseased individuals in each group is usually found to differ. The purpose of classical statistical hypothesis testing is to determine whether the observed difference in disease rates could be due to chance alone. If the observed difference is not attributable to chance, the two groups are considered representative of two truly different populations. If a statistically significant difference is found between the Ranch Hand group and the Comparison group, results from more rigorous statistical procedures must be examined and the medical context considered before the possibility of a causal relationship between disease and group (exposure) can be entertained. Alternatively, the absence of a statistically significant difference between groups does not exclude the possibility of a true causal relationship between exposure and disease. Thus, group associations, whether significant or not, should be examined with adjustment for other variables called intervening variables (explanatory variables, risk, factors, or covariates) that may account for, or mask, a true effect. For example, the two groups might differ with respect to age or racial composition, each of which may affect the outcome of the study. To protect against this, the technique of matching was used: The Ranch Hands and Comparisons were matched on age, race, and military occupation. Since it is not feasible to perfectly match a Comparison to an exposed individual with respect to all important explanatory variables, statistical procedures may be used to adjust for such explanatory variables so that valid interpretations can be made of apparent group differences. Thus, it was necessary to identify and collect data on suspected explanatory variables. Unfortunately, there is no way to ensure that all important intervening variables are taken into account. The best method that can be achieved is to incorporate all known covariates in the data collection and analysis. In most studies, covariates are variables measured prior to exposure. However, in the AFHS, except for the matching variables and historical data related to events prior to service in Southeast Asia, most covariate values were obtained at the Baseline or first followup interview and physical examination, which occurred 10 to 20 years following exposure. These covariates can generally be referred to as time-dependent covariates. They can elucidate the causal path between exposure and a particular disease; however, they are in a sense both dependent and independent variables, and therefore, analyses involving such covariates require careful interpretation. Besides covariates, both confounding variables and interactions must also be considered. A confounding variable is an intervening variable associated with both disease and exposure. (This is in contrast with a covariate that is associated only with disease.) Adjustments must be made for confounding variables to avoid a biased estimate of the group-disease relationship. An interaction exists when the effect of one variable varies across the levels of another variable. For example, the group difference might be large in one occupation, group and negligible in another. Incorporating interactions in the analysis allows for the identification of subpopulations at increased or decreased risk. 7-3 �Pover Conducting a statistical test using a Type I error, also called alpha level, of 0.05 (a =0.05) means that, on the average, in 5 cases out of 100, a false conclusion that an association (herbicide effect) exists would be made when in reality, there is no association. The other possible inference error (called a Type II error) is that of failing to detect an association when it actually exists. The probability of a Type II error (3) for a statistical test is 1 minus the power of the test. The power of the test is the probability that the test will reject the hypothesis of no herbicide effect when an effect does in fact exist. The power of a test depends on the group sample sizes, the disease prevalence rate, and the true group difference measured in terms of relative risk. Table 7-1 contains the approximate sample size required to detect specific relative risks with an approximate power of 0.8 (3 =0.2) using an alpha level of 0.05 for a two-sided test and assuming equal Ranch Hand and Comparison group sizes and unpaired analyses. Relative risk is the ratio of the disease prevalence rate of the Ranch Hand and Comparison groups. Conditions or diseases with comparison population prevalence rates and exposed group relative risks corresponding to those below the heavy black line on the table can be detected with an approximate 0.8 probability with the sample sizes used in this study. Table 7-2 provides the same information for continuous variables in terms of percentage mean shift and variability, assuming unpaired testing of a normally distributed variable and equal sample sizes. In the first followup of the AFHS, 1,016 Ranch Hands participated in the physical examination. In this size group, the chance of identifying zero cases of a disease with a prevalence of 1/500 or less is greater than 10 percent. Table 7-3 contains the probability of encountering no cases of disease states for cumulative prevalence rates of 1/200, 1/500, 1/1,000, 1/2,000, 1/5,000, and 1/10,000. Multiple Endpoints and Comparisons In developing the Protocol for the AFHS, previous animal and epidemiclogic studies, case reports, and veterans' concerns were reviewed to delineate the possible effects of exposure. The conclusion was reached that a comprehensive evaluation was needed due to the lack of an easily identifiable symptom complex in individual patients. Consequently, the morbidity study is very broad in scope, involving the collection and analysis of data related to general health indices as well as specific organ systems and clinical disease categories. The large number of endpoints under consideration presents a difficult problem in the assessment of Type I error rates. More than 150 dependent variables were tested, not to mention tests for interaction and multiple contrasts among the low, medium, and high exposure-level categories in the exposure index analyses. Furthermore, the dependent variables were correlated to varying degrees, and this makes it even more difficult to assess the attained significance levels. To allow for multiple endpoints, Bonferroni's inequality, which requires significance at the a /K, level where K. is the number of endpoints considered, may be used, but this procedure 7-4 �TABLE 7-1. Required SanpLe Sizes To Detect Group Differences in Two-Sanple Testing Assuming Equal Sanple Sizes* (Relative Risk Calculations) Occurrence Rate of Disease in Control Population Relative Risk (Multiplicative Factor of Occurrence Rate for Exposed Group) 1.25 1.50 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 ID75SO 2,822,082 783,901 235,164 78,384 43,544 29,391 21,944 17,415 14,393 12,243 10,640 B75QO 1,410,882 391,901 117,564 39,184 21,766 14,690 10,968 8,703 7,193 6,118 5,317 1 T^OO 281,922 78,301 23,484 7,824 4,344 2,930 2,187 1,735 1,433 1,218 1,058 1 300 140,802 39,101 11,724 3,904 2,166 1,460 1,089 863 713 606 526 1 KB 27,906 7,741 2,316 768 424 284 211 167 137 116 ICO 1 30 13,794 3,821 1,140 376 206 137 101 79 65 54 47 *This study has unequal sample sizes; therefore, the tabled values are understated. The similar table in the Baseline Morbidity Report, 24 February 1984, is in error because tabulated sample sizes were only one-half of their correct values. 7-5 �TABLE 7-2. Required Sample Sizes To Detect Group Differences in Two-Sample Testing Assuming Equal Sample Sizes* (Mean Shift Calculations) Mean shift Variability (CT/M) 0.05 0.10 0.25 0.50 0.75 0.5* 1,568 6,272 39,200 156,800 352,800 1.0* 392 1,568 9,800 39,200 88,200 1.5% 175 697 4,356 17,423 39,200 2.0% 98 392 2,450 9,800 22,050 2.5% 63 251 1,568 6,272 14,112 5.0% 16 63 392 1,568 3,528 7.5% 7 28 175 697 1,568 10.0% 4 16 98 392 882 *This study has unequal sample sizes; therefore, the tabled values are understated. The similar table in the Baseline Morbidity Report, 24 February 1984, is in error because tabulated sample sizes were only one-half of their correct values. 7-6 �TABLE 7-3. Probability of Zero Cases as a Function of Prevalence Disease Prevalence Probability of Finding Zero Cases in a Group of 1,016 Participants 1/10,000 1/5,000 1/2,000 1/1,000 1/500 1/200 0.903 0.816 0.602 0.362 0.131 0.006 becomes increasingly more conservative as the correlation among the endpoints increases. For the analysis results in this report, an alpha level of 0.05 was used for each dependent variable. In addition, group contrasts in strata defined by levels of a covariate involving in a group-by-covariate interaction were assessed by an alpha level of 0.05. The same was true for exposure level strata. In light of the multiple-endpoints problem, extreme caution in the interpretation of statistical results was required. A first consideration was the strength of the association in terms of the significance of the relative risk or difference in group means. All associations with p-values of 0.10 or less were examined and are described in this report. Then, careful consideration was given to the pattern of statistically significant results. Were only a few sporadic endpoints statistically significant, or was significance achieved on a number of endpoints indicating the same organ system failure? Were the significant results all in the same direction, and did they make biological and clinical sense? Did they confirm previous studies, or were they new findings? Paired Versus Unpaired Analyses Matching subjects in a study design on selected variables improves the comparability of the groups to be compared and, depending on the relationship of the matching variables to the study objective, the matching can be used explicitly in the analysis. In this study, the Comparison group was matched to the exposed group on age (to the nearest month of birth), race (Black, nonblack), and occupational category (officer-pilot, officer-navigator, officer-nonflyer, enlisted flyer, enlisted groundcrew). The matching was exact for occupational category, nearly exact for race (three mismatches occurred because of recording errors), and very close with respect to age (69% of the mortality population was matched to the nearest month of birth and more than 95% to the nearest year of birth). The general approach in this report, however, was to conduct unpaired analyses using all available data, based on stratification and/or covariate adjustment. In an unpaired analysis, the matching still serves to improve 7-7 �the comparability of the two groups, and precision is usually gained from the stratification and covariate adjustment. Mortality and Morbidity Data The AFHS incorporated both mortality and morbidity endpoints. The mortality data have been, and will continue to be, subjected to separate analysis. Interpretation of the morbidity analyses must be made in the light of the mortality results, particularly as the study continues and the number of deaths increases. Differential mortality in the two groups could obviously have an important impact on contrasts of physical examination findings in the surviving cohorts. This issue was examined in the analysis of selected diseases, for example, cancer. Outpoints The variables in this study were discrete, categorical, or continuous. Many served primarily as dependent variables, and when in the continuous form, powerful analyses were possible. In other settings, particularly when log-linear or logistic regression models were fitted, it is often necessary to dichotomize or discretize the continuous variables. Discretization, by establishing suitable nonoverlapping intervals or cutpoints, was often the result of a judgment requiring both statistical and clinical input. In general, cutpoint decisions considered the form of the variable, distribution of the variable, established values (e.g., cholesterol, normalabnormal, as specified by a given technique in a given laboratory), scientific values set by precedence (e.g., systolic and diastolic normal threshold 140/90), and error induction by another variable (e.g., use of the blood pressure threshold in obese-armed individuals). The approach to the selection of appropriate cutpoints was to select all cutpoints on a case-by-case basis and, where indicated, use the norms of the SCRF laboratory. Exclusions Due to medical considerations, certain subjects were excluded from the analyses of selected clinical categories. The exclusions were generally defined in the Baseline study and are identified in the clinical chapters of this report. Other exclusions were the result of missing data. OVERVIEW OF STATISTICAL PROCEDURES This section summarizes the basic statistical approach used in the data analysis of the first followup of the AFHS. The approach consisted of four parts: (1) preliminary analysis of the dependent variables and covariates to check for data anomalies and to obtain a general overview of the data, (2) core analyses to carefully determine any- possible effect of herbicide exposure, (3) analysis of the exposure index to investigate the dose-response relationship for the Ranch Hand group only, and (4) longitudinal analysis to examine changes over time. A summary of the statistical techniques utilized is provided in Table 7-4. This basic approach was utilized in the analyses for each clinical category. 7-8 �TABLE 7-4. Summary of Statistical Procedures Chi-Square Contingency Table Test The chi-square test of independence2 is calculated for a contingency table by the following formula: X2 = Z(f0-f9)2/f9 where the sum is taken over all cells of the contingency table and fo=observed frequency in a cell fa=expected frequency under the hypothesis of independence. Large values indicate deviations from the null 2 hypothesis and are tested for significance by comparing the calculated X to the tables of the chi-square distribution. Fisher's Exact Test Fisher's exact test is a randomization test of the hypothesis of independence for a 2x2 contingency table. This technique is useful for small samples and sparse cells. This is a permutation test based on the exact probability of observing the particular set of frequencies. General Linear Model Analysis The form of the general linear model1 for two independent variables is: Y - a + 01X1 + 32X2 + fl^Xj + e where Y = dependent variable (continuous) a = level of Y at Xx = 0 and X2 = 0, i.e., the intercept X1,X2 = measured value of the first and second independent variables, respectively, which may be continuous or discrete P^jSj = coefficient indicating linear association between Y and X a , Y and X2, respectively P12 = coefficient reflecting the linear interaction of Xt and X2 e = error term. This model assumes that the error terms are independent and normally distributed with a mean of 0 and a constant variance. Extension to multiple independent variables and interaction terms is immediate. 7-9 �TABLE 7-4. (continued) Summary of Statistical Procedures Linear regression, multiple regression, analysis of variance, and analysis of covariance are all examples of general linear model analysis. Kolmogorov-Smirnov Distribution Test The Kolmogorov-Smirnov (K-S) test is a nonparametric procedure which assesses differences between the distribution of two samples. Specifically, the K-S procedure tests the hypothesis that populations n, and Ji2 are identical and is designed to detect all possible deviations from this hypothesis. The assumptions of the K-S test are that the observations from the two samples are mutually independent and that both sets of observations are samples from the same distribution. Logistic Regression Analysis The logistic regression model2'4 enables a dichotomous dependent variable to be modeled in a regression framework with continuous and/or discrete independent variables. For two risk factors, such as group and age, the logistic regression model would be: logit P = <x+ (31X1 + 02X2 + 13^X^2 + e where P = probability of disease for an individual with risk factors X and X2 logit P = In (P/l-P), i.e., the log odds for disease X1 = first risk factor, e.g., group X2 = second risk factor, e.g., age. The parameters are interpreted as follows: a = log odds for the disease when both factors are at a 0 level 31 = coefficient indicating the group effect adjusted for age P2 = coefficient indicating the age effect adjusted for group 312 = coefficient indicating the interaction between group and age e = error term. In the absence of an interaction (|31? = 0), exp(p1) reflects the adjusted odds ratio for individuals in Group 1 (Xt = 1) relative to 7-10 �TABLE 7-4. (continued) Summary of Statistical Procedures Group 0 (Xx =0). If the probability of disease is small, the odds ratio will be approximately equal to the relative risk. Homogeneity of the odds ratios5 across different strata was assessed by the method of Breslow and Day. Throughout this report the adjusted odds ratios are referred to as adjusted relative risks. Correspondingly, in the absence of covariates (i.e., unadjusted analysis) the odds ratios are referred to as estimated relative risks. Proportional Odds Model The proportional odds model6 allows for the analysis of an ordered outcome variable. The model assumes that the odds of falling below a certain level rather than above it for individuals at different levels of an independent variable X are in constant ratio. For example, if the response takes one of the four values "excellent," "good," "fair," or "poor," and X is a simple indicator variable designating group (Ranch Hand versus Comparison), then the proportional odds model states that the odds for responding "excellent" versus "good," "fair," or "poor" in the Ranch Hand group are a multiple, exp(fJ), of the corresponding odds in the Comparison group. Likewise, the odds for responding "excellent" or "good" versus "fair" or "poor" in the Ranch Hand group are the same multiple, exp(P), of the corresponding odds in the Comparison group, as are the odds for responding "excellent," "good," or "fair" versus "poor" in the two groups. Thus, the model is appropriate whenever one frequency distribution is "shifted left" relative to another distribution. Incorporation of other variables into X allows the estimation of proportional odds ratios adjusted for covariates. Let the ordered response Y take values in the range 1 to K, and let JlA(X), i=l,...,K, denote the probability of responding at level i for an individual with covariate vector X. Let K.(X) be the odds that Y< j given X, i.e. , ~ n,(X) + rc (X) + ... + Ji.(X) K,(X) The proportional odds model specifies that ^(X) =• K.J exp(|3'X), for constant K^ 7-11 �TABLE 7-4. (continued) Summary of Statistical Procedures Thus the ratio of odds for individuals at covariate levels X1 and X2 is exp{p'(X1 - X2)} ^ ~2 and depends only on Xx - X2 and not on j . Log-linear Analysis Log-linear analysis2 is a statistical technique for analyzing crossclassified data or contingency tables. A saturated log-linear model for a three-way table is: 10 <Zijk> = U o U + U l(i) 13(ik) + U + U 2(j) + U 3(k, + U 12(lj, + U 23(jfk) + 123(ijk) where Zi Ut = expected cell count A . = specific one-factor effect U 12(i .. U = specific two-factor effect or interaction i23<ijk) = tnr ee-factor effect or interaction. The simplest models are obtained by including only the significant U-terms. Adjusted relative risks are derived from the estimated U-terms from an adequately fitting model. McNemar's Test McNemar's test4 effectively considers discordant pairs in which only the Ranch Hand or only the Comparison member in each pair experiences the abnormality. Using a chi-square approximation with continuity correction, the following statistic is used to test whether the off-diagonal entries are evenly divided: , (|b-c|-D2 b+c Where b and c are the number of pairs in which only the Ranch Hand is abnormal or only the Comparison is abnormal, respectively. This test is compared to a chi-squared distribution with one degree of freedom. 7-12 �TABLE 7-4. (continued) Summary of Statistical Procedures Test for Linear Trend For a kx2 contingency table in which the k groups fall into a natural order, Armitage developed a test for a linear trend in the proportions. Let PA denote the proportion of individuals in the ith row possessing some attribute (e.g., proportion of individuals with abnormal values at each of the three exposure level categories). A score, X., is assigned to each of the k levels of the row variable, and the regression coefficient, "& of Pi on XA is estimated. The regression coefficient is estimated in the usual way except that P1 is weighted by the sample size, n.. , in each row. provides a normal deviate for testing the null hypotheses of 0 = 0. 7-13 �Preliminary Analysis The preliminary analysis included the calculation of basic descriptive measures for the dependent and independent variables (covariates), for each group (Ranch Hand and Comparison). The descriptive measures included frequency distributions, histograms, mean, median, standard deviation, and range. These analyses provided an overview of each variable and the relationship of the Ranch Hand group to the Comparison group. In addition, the preliminary analysis provided insight for the construction of composite variables, the plausibility of normal/abnormal limits and cutpoints, and the choice of possible transformations to enhance the normality of the distribution of continuous dependent variables. Another purpose of the preliminary analysis was to examine the relationship between the covariates and the dependent variables and the relationships between and among the covariates. To accomplish this, cross tabulations of discrete variables were constructed and analyzed by the chi-square, or Fisher's exact test. For continuous variables, simple t-tests of group differences were done and product-moment correlation coefficients were computed. The preliminary analyses were accomplished with the use of the SAS*. Selected covariate tables are presented in the clinical chapters for illustration. Core Analysis The core analysis consisted of a series of steps taken to ascertain whether or not the data indicated a significant difference between the Ranch Hand and Comparison groups for each dependent variable. Both unadjusted and adjusted analyses were performed and are presented for each clinical chapter. Unadjusted analyses are simple contrasts between the Ranch Hand and Comparison groups of the mean values, or proportion with abnormal values, of each dependent variable, by t-tests, one-way analysis of variance, Fisher's exact test, or chi-square tests, as appropriate. Adjusted analyses take into account important covariates in the assessment of possible group differences, i.e., the covariates are included in the general linear, logistic regression, proportional odds models, or log-linear models. Continuous Dependent Variables When the dependent variable was continuous, the general linear models (GLM) procedure of SAS® was used to fit a model of the dependent variable in terms of the group indicator (Ranch Hand or Comparison) and appropriate covariates, and interactions between covariates. The covariates could be continuous or categorical variables. If necessary, the dependent variable was transformed prior to analysis by a transformation (e.g., logarithm) to enhance normality of its distribution. When a "best" model was fitted, according to the strategy outlined below, the test for significance of the group difference was then done on the adjusted group means, provided there were no significant interactions between the group indicator and any of the covariates. Group differences in the presence of interactions were assessed using stratification by different levels of the covariate(s) involved in the interaction or estimation of group differences at selected covariate levels using the best model identified. 7-14 �For some non-normally distributed dependent variables, the KolmogorovSmirnov (K-S) test of significant differences betveen the distributions of the variables in the two study groups was conducted. The K-S test is a nonparametrie test for the equality of two distributions designed to detect broad classes of alternatives. Categorical Dependent Variables Discrete dependent variables were analyzed by methods parallel to those used for continuous variables. For dichotomous variables, logistic regression was carried out by the program BMDP*-LR; for this analysis, the covariates could be either continuous or discrete. For polychotomous dependent variables, where the number of categories was three or more, log-linear modeling was performed by the use of the program BMDP®-AF, by incorporating the full (k)-factor interaction term involving the (k) covariates used in the model. For this type of analysis, all covariates had to be categorized. The models were fitted by the method of maximum likelihood. To make the results parallel to those obtained by logistic regression, i.e., because of the distinction between dependent and independent variables, the marginals were fixed in the model, effectively converting the log-linear model into a logit model. The significance of the relative risk for group was determined by examination of the appropriate model, as determined by the study, that includes all statistically significant effects and the group indicator or by examination of the significant interactions. Adjusted relative risks were derived from the coefficients of the appropriate model. Modeling Strategy In each clinical category, many covariates were considered for inclusion in the statistical models for adjusted group contrasts. The large number of such covariates and consequent interaction terms and the resulting difficulties of interpretation forced the adoption of a strategy for identifying a moderately simple model involving only significant effects. Interpretation of possible group differences was then made in the context of this simple model. A schematic representation of the generalized modeling strategy is provided in Appendix E. An initial model including all two-factor interactions and all threefactor interactions involving group was examined. Global tests at the 0.15 level, or individual tests at the 0.05 level, were used to screen out unnecessary three-factor interactions. A hierarchical stepwise deletion strategy was then used, eliminating effects with p>0.05 (except the main group effect) and retaining lower order effects if involved in higher order interactions, to result in the simplest model. Interactions between covariates, if significant, were retained as effects. The analysis was carried out by different statisticians, and there are necessarily subtle differences between them in presentation and approach. This, however, should not affect any of the final conclusions as to group differences. In some chapters, for instance, adjusted group means are presented, and in others the differences between the adjusted group means are 7-15 �presented. In each case, the same conclusion may be drawn since the statistic of relevance is the difference between the adjusted group mean and the associated p-value. Further, if an interaction of group with a continuous covariate was found, two equally valid methods were used to illustrate how the interaction was arising. One method was to categorize the continuous covariate and describe the group differences within each (covariate-defined) stratum. Another technique was to present group differences for several selected values of the covariatel Further, in the presence of small frequencies of abnormalities, exposure index analyses were occasionally carried out using only the main effects model (i.e., using group and all the covariates but not including interaction terms). It is recognized that, due to the large number of group-by-covariate interactions examined (up to 7 per dependent variable) for some 150 variables, some of the group-by-covariate interactions judged significant at the 0.05 level may be spurious, i.e., chance occurrences and not of biological relevance. This is analogous to the concept of Type I error for a two-sample adjusted contrast. When several covariates are used in an adjusted analysis of the group contrast for a single dependent variable, and each group-by-covariate interaction is tested at the 0.05 level, the chance of finding at least one that is statistically significant is, of course, greater than 0.05; this is assuming that there is no group effect or group-by-covariate interaction. How much greater depends on the interrelatedness of the covariates and their association with the dependent variable. For a study of this size, with many interrelated dependent variables being examined, it is not known how to estimate the number of group-bycovariate interactions that may be due to chance alone. However, this frequency clearly will be more than 5 percent. It is noted that this concept should be considered when significant group-by-covariate interactions are interpreted. Further, it is important that the size of the p-value associated with each group-by-covariate interaction be carefully weighed, as should be the pattern of the interaction findings for related dependent variables. EXPOSURE INDEX ANALYSES As described in Chapter 8, the exposure index was constructed to portray the level of dose of the herbicide for the Ranch Hand or exposed group only. Exposure index analyses were conducted on all dependent variables. The objective of the analyses was to determine if there was a difference in the levels of the dependent variable corresponding to the levels of the exposure index. The exposure index was trichotomized as high, medium, and low, separately, for each of the three occupational groups: officer, enlisted flyer, enlisted groundcrew. Thus, separate analyses were conducted for each occupational cohort. Discrete dependent variables were evaluated using log-linear and logistic regression models, treating exposure level as a categorical variable (by means of two indicator variables) and adjusting for covariates. For continuous dependent variables, a general linear model was fit, adjusting for covariates and using two indicator variables to designate exposure level. Contrasts between medium and low, and between high and low exposure levels, were also examined. 7-16 �LONGITUDINAL ANALYSES General Another objective of the AFHS is to observe the Ranch Hand population and the Comparison group carefully over time for the emergence, or deleterious rate change, of symptoms, signs, laboratory parameters, or frank disease. This followup objective is not without scientific and logistic challenge, considering mobile populations, problems of loss to study, changing laboratory methods and diagnostic criteria, and the diversity of many changing factors over a period encompassing numerous follovup examinations. The following sections describe the statistical procedures used for both continuous and categorical longitudinal data. Continuous Data A repeated measurements analysis of variance procedure10 was used to analyze the variables measured on a continuous scale. The model for the dependent variable (Y) measurement on the kth participant (i^) in the ith group (o^) at the jth time ( 3 ) is as follows: |. Y ijk = u + a£ + i^,., + 0j + 0 ^ + eijk 0 The sources of variation and associated degrees of freedom are given below: Source Degrees of Freedom* Group (Ranch Hand vs. Comparison) Subject/Group Time (Baseline vs. Followup) Group-by-Time (Subject-by-Time)/Group 1 2,108 1 1 2,108 *Based on 971 Ranch Hands and 1,139 Comparisons. The primary source of interest is the group-by-time interaction (o3ij). Vith measurements on each participant at only two times (Baseline and followup), a test on this interaction is equivalent to a test on the equality of mean differences (Baseline minus followup) between the Ranch Hand and Comparison groups. Care must be taken in the interpretation of the main effect, time ((3.) (i.e., overall Baseline mean versus overall followup mean). This effect is totally confounded with laboratory differences, and with over 2,000 participants, "significant differences".come easily. The source of variation due to group (a.) reflects a difference between the overall Ranch Hand and Comparison means (averaged over both times). This source should complement the group difference findings at Baseline and at 7-17 �followup, provided the group changes were consistent (no significant groupby-time interaction). All available participants were used at each Baseline and followup analysis, while only the participants with both measurements are included in the repeated measurement analysis. Covariates were not used in these analyses. Generally, time-independent (e.g., year of birth) and time-dependent (e.g., smoking) covariates can be used. Only the time-dependent covariates would affect the primary source of interest, namely the group-by-time interaction. Hence, all of the previously considered time-independent covariates would affect only the main group effect, a source not of primary interest since it is being considered in the separate cross-sectional analyses. Categorical Data Frequently, data were collected as normal-abnormal, or continuous measurements were discretized into this binomial response. For each Ranch Hand and Comparison group, a Baseline versus followup 2x2 (normal-abnormal) table of frequencies was prepared (paired data): Followup Ranch Hand Comparison Abnormal Normal Abnormal Normal Abnormal Abnormal Normal Normal Baseline As with the McNemar test, only the Abnormal-formal and Normal->Abnormal off-diagonal data were used in further contrasts. A conventional ) test was ( used to test the null hypothesis of a comparable change pattern for the two groups (unpaired data). Change Pattern Normal- AbnormalAbnormal Normal Ranch Hand Group Comparison This test is equivalent to 1 testing no group-by-time-by-endpoint interaction in a matched pair analysis. 7-18 �CHAPTER 7 REFERENCES 1. Neter, J., and W. Wasserman. 1974. Applied linear statistical models. Homewood, Illinois: Richard D. Irvin, Inc. 2. Bishop, Y.M.M., S.E. Feinberg, and P.W. Holland. 1975. Discrete multivariate analysis; Theory and practice. Cambridge: MIT Press. 3. Hollander, M., and D. Wolfe. 1973. Nonparametric statistical methods. New York: John Wiley & Sons. 4. Fleiss, J.L. 1981. Statistical methods for rates and proportions. 2d ed. New York: John Wiley & Sons. 5. Breslow, N.E., and N.E. Day. 1980. Statistical methods in cancer research. Volume I, The Analysis of Case-Control Studies, Lyon, International Agency for Research on Cancer (IARC Scientific Publications No. 32). 6. McCullagh, P. 1980. Regression models for ordinal data. J. Royal Stat. Soc. B42(2):109-142. 7. Armitage, P. 1955. Tests for linear trends in proportions and frequencies. Biometrics 11, 375-386. 8. Box, G.E.P., and D.R. Cox. 1964. An analysis of transformations. J. Royal Stat. Soc. 826:211. 9. Feinberg, S.E. 1981. The analysis of cross-classified data. Cambridge: MIT Press. 10. Winer, B.J. 1971. Statistical principles in experimental design. 2d ed. New York: McGraw Hill. 11. Breslow, N.E. 1982. Covariance adjustment of relative risk in matched studies. Biometrics 38:661-672. 7-19 �CHAPTER 8 EXPOSURE INDEX This chapter describes the development of the exposure index of the AFHS. Portions of this chapter are paraphrased from the Baseline Morbidity Report, 24 February 1984. An increased incidence of adverse health effects at higher levels of exposure represents a classic increasing dose-response relationship. The potential relationship of clinical endpoints with herbicide exposure can be tested using an estimate of exposure, hereinafter called an exposure index, for each member of the Ranch Hand cohort of the AFHS. However, due to a variety of biomedical mechanisms, there can be exceptions to the hypothesis of a consistently increasing dose-response relationship. An index of potential exposure to any of four TCDD-containing herbicides from fixed-wing spray missions was constructed for each Ranch Hand from the available historical data. The index serves as an estimate only, since the actual concentration of TCDD in the herbicides varied from lot to lot and individual assessments of actual body burden cannot be made. The four TCDDcontaining herbicides used in the development of the index are Herbicide Orange, Herbicide Purple, Herbicide Pink, and Herbicide Green. The exposure index was designed to correlate as closely as possible with exposure and is not an exact measure of actual individual exposures. Although the index contains errors when used to assess the exposure of a specific individual, it provides some degree of useful inference for groups of similarly exposed individuals. In summary, the exposure index in the AFHS is a surrogate indicator of TCDD exposure. The exposure index developed for the Baseline study and used in this report is defined in Table 8-1. The exposure index for the ith subject, denoted E , is defined as the product of the TCDD weighting factor, the gallons of TCDD-containing herbicide sprayed in the Republic of Vietnam theater during the tour of the ith subject, and the inverse of the number of men sharing the subject's duties during the tour of the ith subject. Each of these factors is described below. The TCDD weighting factor reflects the estimated relative concentration of TCDD in the herbicides sprayed. The estimated mean concentrations of TCDD in Herbicide Orange, Herbicide Purple, Herbicide Pink, and Herbicide Green are 2 parts per million (ppm), 33 ppm, 66 ppm, and 66 ppm, respectively. Archived samples of Herbicide Purple indicate a mean concentration of approximately 33 ppm, and samples of Herbicide Orange had a mean concentration of about 2 ppm. Since Herbicide Pink and Herbicide Green contained twice as much 2,4,5-T as Herbicide Purple, the mean concentration- of TCDD in these two herbicides was approximately 66 ppm. Based on procurement records and dissemination information, a combination of Herbicide Green, Herbicide 8-1 �TABLE 8-1. Algorithm for Exposure Index (Gallons of TCDDContaining Herbicide Weighting^ x Sprayed in the RVN Theater During the (Factor J Tour of the ith Subject, [TCDD E.. = ] Number of Men with Subject's Duties in the RVN Theater During the Tour of the ith Subject where E. = Exposure Index for the ith Subject TCDD Weighting Factor ( 24.0 if before 1 July 1965 I 1.0 if on or after 1 July 1965 Since prior to 1 July 1965 a combination of Herbicides Green, Pink, and Purple with a mean concentration of 48.0 ppm was sprayed, and after 1 July 1965 only Herbicide Orange with a mean concentration of 2 ppm was sprayed, the ratio is then 48?2 or 24:1. Gallons of TCDD-Containing'l Herbicide Sprayed in the RVN Theater During the Tour of the ith Subject [Number of Gallons of Herbicides Orange, ^Green, Pink, and Purple Expressed in Herbicide Orange Equivalent Gallons teased on Mean Concentration of TCDD Using the following: Herbicide Mean Concentration (ppm) of TCDD Green Orange Pink Purple 66 2 66 33 Number of Men with Subject's "| Duties in the RVN Theater During^ the Tour of the ith Subject ) Source: [Number of Personnel ^in the Same (Occupational Category Baseline Morbidity Report, 24 February 1984. 8-2 �Pink, and Herbicide Purple was sprayed between January 1962 and 1965. The estimated mean concentration of TCDD for this time was 48.0 ppm, using available data on the number of gallons procured and sprayed. The Herbs Tape and other data sources1 indicate that only Herbicide Orange was disseminated after 1 July 1965. Normalizing to Herbicide Orange, the weighting factor becomes 24.0 before 1 July 1965 and 1.0 after 1 July 1965. Using the Herbs Tape, Contemporary Historical Evaluation and Combat Operations (CHECO) Reports, and quarterly operations reports, a table of gallons of TCDD-containing herbicide sprayed for each month of the operation was constructed. Gallons of Herbicides Purple, Pink, and Green were converted to Herbicide Orange equivalent gallons based on the TCDD weighting factor of 24.0. This information is provided in Table F-l of Appendix F. The dates and occupational category of each Ranch Hand's tour(s) in the Republic of Vietnam were obtained by a manual review of military records. The study design specified five occupational categories: (1) officer-pilot, (2) officer-navigator, (3) officer-nonflying, (4) enlisted flyer, and (5) enlisted groundcrew. Based on the review of the records, the Ranch Hand manning for each occupational category by month was compiled. This information is also presented in Table F-l of Appendix F. A numeric exposure index reflecting the effective number of gallons of Herbicide Orange to which each individual was potentially exposed was computed. For the purpose of analysis, the values were categorized as high, medium, or low for each occupational category. Only three occupational categories were used. The three officer categories were combined into one since .pilots and navigators were exposed in the same manner and the officernonflying category, which included a relatively small number of participants, consisted of administrators whose exposure was considered to be essentially zero. The overall group of "nonexposed" Ranch Hands, estimated at approximately 2 percent of the Ranch Hand group, was analyzed in the low exposure category (see Table 8-2), conceivably leading to dilution of the exposure analyses and group contrasts. The exposure index categorizations developed for the Baseline study and used in this report are provided in Table 8-2, along with the frequencies of Ranch Hand participants by occupation and exposure level. The current exposure index is not specific to job and, therefore, may underestimate exposure for those individuals whose jobs required routine handling of herbicide. For example, maintenance schedules for the aircraft herbicide spray tank required that an emergency dump valve be periodically greased, requiring entry into the tank. The current exposure index cannot distinguish between men who received such exposure and men who did not. The extent to which individuals are misclassified by the current exposure index is not known, precluding bias calculations at this time. Because of the acknowledged imprecision of the exposure index, Air Force efforts are under way to develop.new perspectives of exposure. One effort is the construction of a new questionnaire for the 459 enlisted groundcrew personnel that may permit more accurate exposure analyses within this category. Another approach is the measurement of serum dioxin levels. 8-3 �TABLE 8-2. Exposure Index Categorization of 1,016 Compliant Ranch Hands Occupational Group Exposure Category Effective Herbicide Orange Gallons Corresponding to Exposure Category Number of Ranch Hand Participants in Exposure Category Officer Low Medium High <35,000 35,500-70,000 >70,000 127 130 123 Enlisted Flyer Low Medium High <50,000 50,500-85,000 >85,000 55 65 57 Enlisted Groundcrew Low Medium High <20,000 20,500-27,000 >27,000 154 163 142 Total 1,016 The Air Force currently is conducting a pilot study in conjunction with the laboratories of the Centers for Disease Control, Atlanta, Georgia, to determine levels of TCDD in serum and to establish the validity of exposure differential within the Ranch Hand and Comparison groups. This study is in accordance with the Study Protocol commitment to estimate dosage of TCDD as accurately as current technology permits. If successful, use of timeadjusted TCDD levels would permit more accurate exposure analyses within the Ranch Hand group. Perhaps of most importance, accurate TCDD levels within the Ranch Hand group could standardize exposure to a comparable baseline for all participants. Thus, the use of adjusted TCDD levels will place the exposure concepts on a firm scientific basis, and if herbicide effects exist, they can be discerned more accurately. 8-4 �CHAPTER 8 REFERENCES Young, A.L., J.A. Calcagni, C.E. Thalken, and J.W. Tremblay. 1978. The toxicology, environmental fate, and human risk of herbicide orange and its associated dioxin. Technical report OEHL-TR-78-92, USAF Occupational and Environmental Health Laboratory, Brooks AFB, Texas. 247 pp. 8-5 �CHAFFER 9 GENERAL HEALTH INTRODUCTION The effects of heavy, acute exposure to TCDD have been demonstrated in a number of different organ systems. It is plausible, therefore, that chronic low-dose exposure to TCDD might induce subtle, interrelated effects that are not organ-system specific, but are manifest only in general terms, or affect the state of "well-being." However, it is difficult to measure overall health objectively, and for this reason general health outcomes, as defined by this study, should be judged in context with other more specific clinical endpoints. (It should be noted that "general health" outcomes have not traditionally been considered in other dioxin morbidity studies.) Baseline Summary Results Five general health variables were included in the Baseline examination: self-perception of health, appearance of illness or distress, relative age, sedimentation rate, and percent body fat. In the analysis of the 1982 Baseline examination data, a statistically significant difference was found between the Ranch Hand and Comparison groups in self-perception of health, with a greater percentage of Ranch Hands reporting their health as fair or poor than Comparisons. This was true in both the younger and older age groups (p=0.017 and p=0.025 for individuals 40 or less and more than 40 years of age, respectively). The relative risk of the Ranch Hand group was also somewhat greater in the younger subgroup than in the older subgroup (1.8 and 1.4, respectively). Since only 9 of 1,811 individuals were reported by the examining physician as appearing ill or distressed, this designation was apparently reserved for only very ill or distressed individuals. Nevertheless, 8 of the 9 individuals were Ranch Hands, the difference being of borderline significance (p=0.056). Conversely, more Ranch Hands than Comparisons were reported by the examiners as appearing younger than their actual ages (4.9% versus 2.5%, p=0.029). No overall differences in percent body fat or sedimentation rate were found, although a significant interaction between age, group, and sedimentation rate was noted; younger exposed group members had fewer sedimentation rate abnormalities than did their Comparisons, whereas no difference was found in participants more than 40 years old. No statistically significant dose-response relationships were detected in the Ranch Hand group. Parameters of the 1985 General Health Assessment Variables of the Baseline examination (self-perception of health, appearance of illness or distress, relative age, sedimentation rate, and percent body fat) were analyzed for the third year followup effort. 9-1 �As an assessment of the general health status of each individual, three subjective measures were made as well as two more objective measures. During the health interview each study participant was asked, "Compared to other people your age, would you say that your health is excellent, good, fair, or poor?" This self-assessment of health is susceptible to varying degrees of conscious and subconscious bias. The examiner recorded the appearance of illness or distress (yes/no) and noted the appearance of the subject as younger than, older than, or the same as his stated age. To the degree that the examining physicians were kept blind to the study subject's group membership (Ranch Hand, Comparison), their assessments were less subject to bias. The two objective measures were percent body fat, calculated from the body mass index, and the erythrocyte sedimentation rate. Although both variables are rather indirect measures of the general state of health, they are accepted indicators of poor health. The adjusted statistical analyses below accounted for differences associated with age, race, and occupation. In the analysis of self-perception of health and sedimentation rate, adjustment was also made for personality score, determined from the Jenkins Activity Survey. This is a continuous variable derived by means of a discriminant-function equation based on items that best discriminate men judged to be Type A from those judged as Type B. Positive scores reflected the Type A direction and negative scores the Type B direction. Table G-l of Appendix G gives the distribution of the covariates in the Ranch Hand and Comparison groups. Age, race, and occupation were distributed similarly in the two groups (due to matching), and personality scores were also not significantly different. Aside from the subjective nature and potential bias in the self-reported perception of health, no specific issues related to assessment methodology require further comment. No individuals were excluded from analysis, except those with missing data. Chi-square tests and logistic regression models were applied to the categorical data. The sedimentation rate was normalized by logarithmic transformation. The proportional odds model was also used for ordinal data provided by the self-perception of health and relative age variables. Fisher's exact test was applied to the reporting of illness or distress by the examining physician because of the small number of cases who were classified as "ill." A two-sample t-test was used to assess differences in unadjusted group means, followed by multiple regression analysis to incorporate covariates, for percent body fat and sedimentation rate. RESULTS AND DISCUSSION Subjective Assessments Self-Perception of Health Each participant was asked to designate his health as excellent, good, fair, or poor. The frequency distributions of self-perception of health for each cohort are given in Table 9-1. 9-2 �TABLE 9-1, Unadjusted Analysis for Self-Perception of Health by Group Self-Perception of Health Excellent Group Good Fair Poor Number Percent Number Percent Number Percent Number Percent Total Ranch Hand 490 48 .2 434 42.7 74 7.3 18 1.8 1,016 Comparison 674 52 .1 525 40.6 81 6.3 13 1.0 1,293 P=0.14 The summarized data in Table 9-1 show that a higher percentage of Ranch Hands perceived their health to be fair or poor (9.1%) than the Comparisons (7.3%), although this difference was not statistically significant (Est. RR: 1.25, 95% C.I.: [0.95,1.64], p=0.14). Of considerable interest is that the percentage of both groups perceiving their health as only fair or poor was lower than that reported at the Baseline examination 3 years earlier (20.4% and 15.9% for Ranch Hands and Comparisons, respectively). This shift was the opposite of that expected from an aging effect. The data collection technique was an in-home interview in 1982 versus an onsite clinic interview in 1985, but this was not judged to be the likely cause of the improvement in health perceptions for the 3-year period. Whatever the cause, the effects were similar in both groups. A test of association between health perception (dichotomized as excellent/good and fair/poor) was performed with the covariates of age (born in or after 1942, born before 1942), race, occupation, and personality score (Jenkins score, trichotomized as low [less than -5], medium [between -5 and 5], and high [greater than 5]). These associations were examined both within the Ranch Hand and Comparison groups and pooled over the two groups. The findings were similar, and Table 9-2 shows the results after pooling. These results indicated a significant effect of age, with a higher percentage of the older cohort than the younger cohort reporting their health as fair or poor, as well as a significant effect of occupation, with the percentage of enlisted personnel reporting fair or poor health nearly twice that of the officers. No significant associations were noted for race or personality score. 9-3 �TABLE 9-2. Association Between Self-Perception of Health and Age, Race, Occupation, and Personality Score in the Combined Ranch Hand and Comparison Groups Self-Perception of Health Covariate Excellent/Good Fair/Poor Covariate Category Number Percent Number Percent Total p-Value Age Born >1942 903 Born <1942 1,220 94.0 90.5 58 128 6.0 9.5 961 1,348 0.003 Race Black Nonblack 130 1,993 90.9 92.0 13 173 9.1 8.0 143 2,166 0.76 819 94.8 45 5.2 864 Occupation Officer Enlisted Flyer Enlisted Groundcrew 347 89.7 40 10.3 387 957 90.4 101 9.6 1,058 Personality Score 827 716 573 92.2 91.2 92.6 70 69 46 7.8 8.8 7.4 897 785 619 Low Medium High 9-4 <0.001 0.61 �Adjusted analyses of self-perception of health were done by logistic regression using the covariates of age, race, occupation, and personality type. (Self-perception of health was dichotomized and the covariates categorized as in Table 9-2.) These analyses revealed statistically significant age and occupation effects, as well as a significant group-byoccupation interaction (p=0.015). Exponentiation of linear combinations of relevant regression coefficients generated adjusted relative risks for each occupational stratum. These summary data are presented in Table 9-3. TABLE 9-3. Adjusted Relative Risks of Self-Perception of Health by Occupation Adj. Relative Risk (95% C.I.) p-Value Officer 0.78 (0.42,1.46) 0.441 Enlisted Flyer 0.75 (0.38,1.46) 0.395 Enlisted Groundcrew 1.90 (1.25,2.88) 0.003 Occupation These analyses showed significant group differences in the selfperception of health for the enlisted groundcrew category but not for the officers or enlisted flyers. This is perhaps more clearly seen in Table 9-4, which gives the frequency distribution of self-perception of health stratified by occupation. Among officers and enlisted flyers, a lower percentage of Ranch Hands than Comparisons perceived their health as fair or poor. (These same Ranch Hands were also less likely to view their health as excellent.) In the enlisted groundcrew cohort, 12.7 percent of the Ranch Hands reported their health as fair or poor versus 7.2 percent of the Comparisons. Because the logistic model does not account for the ordinal nature of the self-perception of health variable, a proportional odds model for ordinal responses was also fit to the data in Tables 9-1 and 9-4. For the ordinal responses in Table 9-1, the proportional odds model yielded a statistically significant result (p=0.037), with poorer health estimated to be 1.18 times greater in the Ranch Hand group than in the Comparison group (952 C.I.: [1.01,1.39]). For the data in Table 9-4, a proportional,odds model fit to each occupational stratum (adjusting for age) yielded p-values of 0.65 for officers, 0.43 for enlisted flyers, and 0.031 for enlisted groundcrew. Thus, only the enlisted groundcrew category reached statistical significance, with adjusted proportional odds of 1.30 (95% C.I.: [1.02,1.64]). 9-5 �TABLE 9-4. Frequency of Self-Perception of Health by Occupation and Group Self-Perception of Health Excellent Occupation Good Fair Poor Number Percent Number Percent Number Percent Number Percent Total Officer Ranch Hand 238 Comparison 314 62.6 64.9 124 143 32.6 29.6 13 23 3.4 4.8 5 4 1.3 0.8 380 484 Enlisted Flyer Ranch Hand Comparison 67 94 37.8 44.8 94 92 53.1 43.8 13 19 7.3 9.0 3 5 1.7 2.4 177 210 Enlisted Groundcrew Ranch Hand Comparison 185 266 40.3 44.4 216 290 47.1 48.4 48 39 10.5 6.5 10 4 2.2 0.7 459 599 Similar results were obtained when the analyses were performed on the 1,016 Ranch Hands and 955 Original Comparisons completing the third-year health interview. These results are provided in Table G-2 of Appendix G. In the unadjusted analysis, the estimated relative risk for fair or poor health versus excellent or good health reached statistical significance (Est. RR: 1.43, 95% C.I.: [1.03,2.00], p=0.042). In the adjusted analysis, group membership, age, and occupation effects were all statistically significant with an adjusted relative risk of 1.48 (95% C.I.: [1.05,2.07]). The groupby-occupation interaction, however, did not reach statistical significance (p=0.23). Nevertheless, little difference was seen in the officers and enlisted flyers, whereas among the enlisted groundcrew, 12.7 percent of the Ranch Hands versus 7.4 percent of the Original Comparisons reported their health as fair or poor. Contrasts of the Ranch Hand and Original Comparison groups using the proportional odds model yielded only borderline significant results. For the unadjusted analysis applied to the overall data, the estimated proportional odds were 1.17 (95%, C.I.: [0,99,1.39], p«0.073). Stratifying by occupation and adjusting for age gave p-values of 0.76, 0.11, and 0.078 for the officers, enlisted flyers, and enlisted groundcrew, respectively. The adjusted proportional odds in the enlisted groundcrew cohort were 1.26 (95% C.I.: [0.97,1.62]). 9-6 �Appearance of Illness or Distress The recording of the appearance of acute ill health or physical distress at the examination was intended .to capture significant subjective health data that might (though not likely) escape corroboration by other physical examination or laboratory data. In particular, examining physicians were requested to affirm the presence of acute distress when the sign of hippocratic facies was present, a sign not easily feigned by participants. Very few participants were diagnosed as being acutely ill; these data are summarized in Table 9-5. TABLE 9-5. Unadjusted Analysis for Appearance of Acute Illness or Distress by Group Acute Illness or Distress Yes Group No Number Percent Number Percent Total Ranch Hand 4 0.4 1,010 99.6 1,014 Comparison 6 0.5 1,287 99.5 p-Value* 1,293 0.53 *Fisher's exact test, 1-sided. These data were too sparse to permit further meaningful analyses. Descriptively, it was noted that 9 of the 10 ill individuals were in the older age group; 9 of 10 were nonblack; and 2 were officers, 4 were enlisted flyers, and 4 were enlisted groundcrew. The 6 ill Comparison individuals were all Original Comparisons, as can be seen in Table G-3 of Appendix G. Further, these results were in substantial contrast to the Baseline findings that revealed a marginally significant excess (p=0.056) of acute distress among the Ranch Hands. Appearance of Relative Age The examining physicians scored each participant as appearing younger, older, or the same as his chronological age. These data are presented in Table 9-6. 9-7 �TABLE 9-6. Unadjusted Analysis for Appearance of Relative Age by Group Appearance of Relative Age Younger Group Number Percent Ranch Hand 16 1.6 Comparison 9 0.7 Same Number Older Percent Number Percent Total 957 94.3 42 4.1 1,015 1,233 95.4 51 3.9 p-Value 1,293 0.12 These frequency distributions showed that a slightly higher percentage of Ranch Hands than Comparisons appeared younger than their stated age, and almost equivalent percentages in both groups appeared older. Overall, there was no significant difference in the two distributions. The unadjusted findings in Table 9-6, however, did not confirm the significant tendency (p=0.029) at the 1982 Baseline examination for a higher percentage of the Ranch Hands than Comparisons to appear younger than their stated ages. Table 9-7 presents the association between each of the covariates and relative age (dichotomized as older looking versus the same or younger looking) after combining the Ranch Hand and Comparison groups. As noted from this table, age and race were not significantly associated with the appearance of relative age, whereas occupation did reveal a significant association, with about 6 percent of the enlisted personnel appearing older than their stated ages compared to 1 percent of the officers. An adjusted analysis using logistic regression with the covariates age, race, and occupation showed a significant effect due to occupation as well as a significant group-by-occupation interaction (p-0.038). Adjusted relative risks for each occupational stratum are given in Table 9-8. The adjusted relative risk was greater than 1 for the officers, i.e., the odds of appearing older were greater in the Ranch Hand group than in the Comparison group, but the relative risk was less than 1 for the enlisted flyers. However, the associated confidence intervals were rather broad and did not rule out a relative risk of 1 in each case. Again, because the logistic regression model does not account for the ordinal nature of the dependent variable, a proportional odds model was applied to the enlisted flyer cohort (data in the officer and enlisted groundcrew strata did not fit the model properly). The estimated proportional odds for the enlisted flyer cohort were nonsignificant (estimated odds: 0.49, 95% C.I.: [0.22,1.11], p=0.087). 9-8 �TABLE 9-7. Association Between Appearance of Relative Age and Age, Race, and Occupation in the Combined Ranch Hand and Comparison Groups Appearance of Relative Age Younger /Same Older Covariate Category Number Age Born >1942 Born <1942 914 1,301 95.2 96.5 46 47 4.8 3.5 960 1,348 0.14 Race Black Nonblack 138 2,077 96.5 95.9 5 88 3.5 4.1 143 2,165 0.91 Occupation Officer Enlisted Flyer Enlisted Groundcrew 855 362 99.0 93.5 9 25 1.0 6.5 864 387 998 94.4 59 5.6 Covariate Percent Number Percent Total p-Value 1,057 TABLE 9-8. Adjusted Relative Risks of Appearance of Relative Age by Occupation Adj. Relative Risk (95% C.I.) p-Value Officer 4.52 (0.94,21.9) 0.060 Enlisted Flyer 0.44 (0.23,1.27) 0.159 Enlisted Groundcrew 1.05 (0.62,1.78) 0.849 Occupation 9-9 <0.001 �A contrast of the Ranch Hand group with the Original Comparisons gave similar results, as shown in Table G-4 of Appendix G. Overall, there was little difference, but the group-by-occupation interaction was of borderline significance in the adjusted analysis (p=0.052). Differences were largely confined to the enlisted flyers, where fewer Ranch Hands than Comparisons appeared older than their stated ages (Adj. RR: 0.47, 95% C.I.: [0.20,1.12], p=0.089) (see Table G-5 of Appendix G). A proportional odds model applied to the enlisted flyer stratum gave adjusted proportional odds of 0.45 (95% C.I.: [0.20,1.02], p-0.055). Objective Assessments Two objective but nonspecific indicators of general health, the erythrocyte sedimentation rate and percent body fat, were analyzed in both discrete and continuous forms. Because the sedimentation rate was a highly skewed variable, it was normalized by logarithmic transformation for the continuous analyses. The sedimentation rate dichotomy was set at 20 mm/hr or less (normal) and greater than 20 mm/hr (abnormal) by the large-tube Westergren method. Percent body fat was based on height and weight obtained during the examination and was calculated according to the following formula: Percent Body Fat = (Weight[kg]/Height[m])(1.264) - 13.305. It is recognized that this formula will overstate the percent body fat for very muscular, large-boned men. Percent body fat was trichotomized into less than 10 percent (lean), 10 to 25 percent (normal), and greater than 25 percent (obese), consistent with the Baseline Report. Because of the sparseness of the lean category, it was often necessary to use a dichotomous variable of lean-normal versus obese. Erythrocyte Sedimentation Rate The unadjusted contrast of log sedimentation rate means revealed no significant group differences (mean±SE=l.620+0.026 in the Ranch Hand group versus 1.595±0.021 in the Comparison group, t=0.73, p=0.47). The geometric mean values were 5.05 and 4.93 for the Ranch Hand and Comparison groups, respectively. Tests of association of dichotomized sedimentation rate, with the covariates age, race, occupation, and personality score, pooled over both groups, were conducted; these summarized data are shown in Table 9-9. These results showed significant effects of age, with older individuals having a higher frequency of abnormal sedimentation rates than younger individuals, and a significant effect of personality score, with Type B individuals (low personality score) having more sedimentation rate abnormalities. The effect of occupation was of borderline significance (p=0.080), with a slightly higher percentage of abnormal values among the enlisted flyers than among officers or enlisted groundcrew. There was no evidence of any association between race and abnormal sedimentation rate. An analysis of the log sedimentation rate, adjusting for age, race, occupation, and personality score, detected significant effects for all of the covariates except race, as well as a significant age-by-personality score interaction. As in the unadjusted analysis, the adjusted analysis did not reveal any significant difference between the Ranch Hand and Comparison groups (p=0.412). 9-10 �TABLE 9-9. Association Between Sedimentation Rate and Age, Race, Occupation, and Personality Score in the Combined Ranch Hand and Comparison Groups Sedimentation Rate Abnormal >20mm/hr Normal <20mm/hr Covariate Number Percent Covariate Category Number Percent Total p-Value Age 941 Born >1942 Born <1942 1,263 97.9 93.7 20 85 2.1 6.3 961 1,348 <0.001 Race Black Nonblack 136 2,068 95.1 95.5 7 98 4.9 4.5 143 2,166 0.999 Occupation Officer 828 Enlisted 361 Flyer 1,015 Enlisted Groundcrew 95.8 93.3 36 26 4.2 6.7 864 387 0.080 95.9 43 4.1 1,058 94.0 96.6 96.1 54 27 24 6.0 3.4 3.9 897 785 619 Personality Score Low Medium High 843 758 595 9-11 0.026 �However, in the dichotoraous form, sedimentation rate abnormalities were significantly more prevalent in the Ranch Hands than Comparisons (Est. RR: 1.63, 95% C.I.i [1.12,2.38], p=0.013); these results are given in Table 9-10. Logistic regression analysis found significant effects for age and personality score, and the adjusted relative risk of 1.68 (95% C.I.: [1.13,2.49], p=0.011), was very similar to the estimated relative risk of 1.63. TABLE 9-10. Unadjusted Analysis for Sedimentation Rate by Group Sedimentation Rate Normal <20 mm/hr Group Number Percent Abnormal >20 mm/hr Number Percent Total Ranch Hand 957 94.2 59 5.8 1,016 Comparison 1,247 96.4 46 3.6 p-Value 1,293 0.013 The mean log sedimentation rate in the Original Comparisons was 1.636 plus or minus 0.025, not significantly different from the Ranch Hand mean (t=-0.45, p=0.65). The regression analysis yielded results very similar to those reported above, with little difference in the adjusted group means. Logistic regression analyses also gave similar results, with significantly more abnormalities in the Ranch Hand group (p=0.037). In summary, there was no difference between groups based upon mean values of the sedimentation rate, unadjusted or adjusted, but both unadjusted and adjusted discrete analyses snowed a significantly higher prevalence of sedimentation rate abnormalities in the Ranch Hand group. This finding was opposite to the Baseline findings in which Ranch Hands age 40 or less had significantly fewer sedimentation rate abnormalities than Comparisons, with no group difference in individuals over the age of 40. Percent Body Fat The mean percent body fat of Ranch Hands was significantly lower than that of Comparisons (21.10%±0.15.versus 21.54%±0.14, respectively; p=0.037). Because there were only a few values in the lean category (6 in the Ranch Hand group and 4 in the Comparison group), percent body fat was dichotomized into at most 25 percent (lean and normal) and more than 25 percent (obese) for tests of association between percent body fat and the covariates age, race, and occupation. The results are given in Table 9-11. 9-12 �TABLE 9-11. Association Between Percent Body Fat and Age, Race, and Occupation in the Combined Ranch Hand and Comparison Groups Percent Body Fat Lean/Normal <25% Obese >25% Covariate Covariate Category Age Born XL942 802 Born <1942 1,060 83.4 78.7 159 287 Race Black Nonblack 110 1,752 76.9 80.9 719 314 829 Occupation Officer Enlisted Flyer Enlisted Groundcrew Number Total p-Value 16.6 21.3 961 1,347 0.005 33 413 23.1 19.1 143 2,165 0.29 83.3 81.1 144 73 16.7 18.9 863 387 78.4 229 21.6 Percent Number Percent 0.023 1,058 These data demonstrated the significant effects of age, with a higher percentage of obesity in older men, and occupation, with a higher prevalence of obesity in enlisted personnel than in officers. Race was a noncontributory covariate. The covariate of smoking was unexplored. An adjusted analysis of percent body fat, with the same covariates, also showed the significant effects of age, occupation, and an age-by-occupation interaction. The adjusted results showed a small, but significantly lower mean level of body fat in the Ranch Hand group (adjusted difference=-0.443± 0.210, p=0.035). With percent body fat dichotomized into obese versus normal or lean, the percent obese was lower in the Ranch Hands than in the Comparisons (18.2% versus 20.2%), but the difference was not significant (Est. RR: 0.90, 95% C.I.: [0.71,1.08], p=0.25). Logistic regression analysis also failed to detect a significant group difference (Adj. RR: 0.87, 95% C.I.: [0.71,1.08], p-0.204). Analysis of percent body fat in the Ranch Hands and Original Comparisons gave somewhat different results. The overall difference in means was significant as before: 21.10 plus or minus 0.15 in the Ranch Hand group versus 21.58 plus or minus 0.16 in the Original Comparison group (t=-2.15, p=0.032). However, the regression analysis detected a statistically significant groupby-race interaction (p=0.041). The adjusted difference in mean percent "body fat (Ranch Hand versus Comparison) was greater in Black participants (-2.26%) 9-13 �than in nonblack participants (-0.34%). Of the Original Comparisons (Table G-7 of Appendix G), 20.4 percent were obese, greater than, but not significantly different from, the percent obese in the Ranch Hand group (p=0.230). Logistic regression analyses again detected significant age and occupation effects, but it detected no significant interaction between these variables. There was no strong evidence of a group-by-race interaction (models including all two-factor interactions gave a Z-value of 1.19 for the group-by-race interaction). The group effect was not statistically significant (Adj. RRs 0.87, 95% C.I.: [0.70,1.09], p=0.242). In summary, the unadjusted and adjusted tests of mean percent body fat showed a significantly lower value for Ranch Hands; correspondingly fewer Ranch Hands than Comparisons were obese, although this difference was not statistically significant. Few individuals were lean (less than 10 percent body fat). The 1982 Baseline examination found no difference in group means (p=0.67), or proportion of abnormalities (p=0.89). Further, analyses based solely upon the Original Comparison cohort found the difference in mean percent body fat between the Ranch Hand and Comparison groups to be greater in Blacks than nonblacks. EXPOSURE INDEX ANALYSES The exposure index, expressed in equivalent gallons of dioxin-containing herbicide potentially encountered by each Ranch Hand during his tour of duty in Vietnam, was categorized as low, medium, and high. Because it is not possible to assess the relative exposure between occupational groups, and since different cutoff values were used in the three occupational categories, separate analyses were performed within each occupational cohort. A detailed description of the exposure index is found in Chapter 8. Exposure analyses were performed on four of the five general health variables. Only four Ranch Hands were recorded as appearing ill or distressed (two were officers, both in the low-exposure category, and two were enlisted flyers, both in the high-exposure category). Further analysis was not done on this variable. Self-Perception of Health Table 9-12 presents dichotomized self-perception of health data by exposure level for the 1,016 Ranch Hands. While these unadjusted contrasts did not reach statistical significance within any of the occupational strata, the linear trend from low to high exposure in the officer cohort of the fair/poor category was of interest, and was subjected to further testing. Although the numbers were small at each exposure level, a test for linear trend led to a borderline significant increase of 2.5 plus or minus 1.3 percent per unit (step) increase in the exposure level category (p=0.064). Logistic regression analyses adjusted for age (dichotomized), race, and personality score (trichotomized) did not detect any significant exposure level effects. The only significant covariate effect found was for age in the enlisted groundcrew cohort. The adjusted relative risk for each occupational stratum is given in Table 9-13. 9-14 �TABLE 9-12. Unadjusted Exposure Index Analysis of Self-Perception of Health by Occupation Self-Perception of Health Excellent/Good Exposure Index Occupation Officer Number Percent Fair/Poor Number Percent Total p-Value* Low Medium High 124 124 114 97.6 95.4 92.7 3 6 9 2.4 4.6 7.3 127 130 123 0.17 Enlisted Flyer Low Medium High 51 59 51 92.7 90.8 89.5 4 6 6 7.3 9.2 10.5 55 65 57 0.83 Enlisted Groundcrew Low Medium High 134 146 121 87.0 89.6 85.2 20 17 21 13.0 10.4 14.8 154 163 142 0.51 *Chi-square tests, 2 d.f, TABLE 9-13. Adjusted Relative Risk of Self-Perception of Health by Occupation and Exposure Contrast Exposure Contrast Adj. Relative Risk (95% C.I.) p-Value Officer Medium vs. Low High vs. Low 2.00 (0.49,8.15) 2.93 (0.76,11.3) 0.334 0.119 Enlisted Flyer Medium vs. Low High vs. Low 1.30 (0.35,4.86) 1.50 (0.40,5.64) 0.700 0.549 Medium vs. Low High vs. Low 0.95 (0.47,1.92) 1.21 (0.62,2.35) 0.882 0.580 Occupation Enlisted Groundcrew 9-15 �Appearance of Relative Age The dichotomy of appearance of relative age was assessed for exposure effects in each occupational cohort. These unadjusted analyses, shown in Table 9-14, provided no evidence of a dose-response effect. As can be seen, the number of participants within each stratum appearing older than their stated ages was quite small. The adjusted analyses by logistic regression did not detect any significant exposure or covariate effects. TABLE 9-14. Unadjusted Exposure Index Analysis of Appearance of Relative Age by Occupation Relative Age Younger/Same Occupation Exposure Index Number Percent Older Number Percent Total p-Value* Officer Low Medium High 125 127 121 98.4 97.7 98.4 2 3 2 1.6 2.3 1.6 127 130 0.89 123 Enlisted Flyer Low Medium High 52 62 55 94.6 95.4 96.5 3 3 2 5.4 4.6 3.5 55 65 0.88 57 Enlisted Groundcrew Low Medium High 146 151 134 94.8 93.2 94.4 8 11 8 5.2 6.8 5.6 154 162 0.82 142 *Chi-square tests, 2 d.f. 9-16 �Erythrocyte Sedimentation Rate The sedimentation rate was analyzed both continuously on a logarithmic scale and dichotomously (normal, abnormal). One-way analyses of variance were performed on the sedimentation rate means categorized by occupation and exposure level. These tests showed no significant differences in the officer and the enlisted flyer strata (p=0.76, p=0.64, respectively). In the enlisted groundcrew stratum the means were marginally different, with the mean sedimentation rate increasing with increasing exposure level, but the differences were not statistically significant (p=0.12). When these data were adjusted by an analysis of covariance for age, the difference in mean sedimentation rates in the enlisted groundcrew was less noteworthy (p=0.33). Age was positively associated with the mean sedimentation rate in all three occupational strata (p<0.001, p=0.009, and p<0.001, respectively). The adjusted tests are reflected in Table 9-15 (means and confidence limits have been transformed back to the original scale). A categorical analysis of the sedimentation rate by exposure level for each occupational stratum was also conducted. Differing from the previous continuous analyses, the categorical contrasts revealed a significant exposure effect (p=0.027) in the enlisted flyer stratum, albeit with small numbers. These summarized data are shown in Table 9-16. Adjustment for age, race, and personality score revealed a significant high versus low exposure contrast in the enlisted flyer stratum. The adjusted analysis is fully shown in Table 9-17. TABLE 9-15. Adjusted Mean Sedimentation Rates by Occupation Exposure Index, Adjusted Mean, mm/hr (95% C.I.) Occupation Low Medium High p-Value Officer 5.40 (4.71 ,6.19) 4.78 (4.17 ,5.47) 4.69 (4.09,5.37) 0.31 Enlisted Flyer 5.10 (4.11 ,6.33) 6.00 (4.91 ,7.32) 5.00 (4.04,6. 19) 0.41 Enlisted Groundcrew 4.66 (4.10,5.29) 5.09 (4.49,5.77) 5.35 (4.69,6.12) 0.33 9-17 �TABLE 9-16. Unadjusted Exposure Index Analysis of Sedimentation Rate by Occupation Sedimentation Rate Abnormal >20mm/hr Normal <20mm/hr Exposure Index Number Percent Number Percent Total Low Medium High 117 125 119 92.1 96.2 95.9 10 5 5 7.9 3.8 4.1 127 130 123 Enlisted Flyer Low Medium High 53 62 48 96.4 95.4 84.2 2 3 9 3.6 4.6 15.8 55 65 57 0.027 Enlisted Groundcrew Low Medium High 142 156 136 92.2 95.7 95.8 12 7 6 7.8 4.3 4.2 154 163 142 0.290 Occupation Officer *Chi-square tests, 2 d.f. TABLE 9-17. Adjusted Relative Risk of Sedimentation Rate by Occupation and Exposure Contrast Occupation Exposure Contrast Adj. Relative Risk (95% C.I.) p-Value Officer Medium vs. Low High vs. Low 0.47 (0.16,1.41) 0.50 (0.17,1.52) 0.177 0.226 Enlisted Flyer Medium vs. Low . High vs. Low 1.28 (0.21,7.96) 4.97 (1.02,24.2) 0.790 0.047 Enlisted Groundcrew Medium vs. Low High vs. Low 0.76 (0.28,2.06) 0.54 (0.19,1.49) 0.592 0.234 9-18 p-Value* 0.27 �Percent Body Fat Exposure analyses of percent body fat were done using both linear models and logistic regression. One-way analyses of variance for means found no statistically significant exposure differences in the occupational cohorts. These statistics are presented in Table 9-18. TABLE 9-18. Unadjusted Means of Percent Body Fat by Occupation Exposure Index, Mean±SE Occupation Low Medium High p-Value Officer 20.99±0.36 21.11±0.41 21.26±0.36 0.88 Enlisted Flyer 20.65±0.55 21.26±0.77 21.59±0.77 0.65 Enlisted Groundcrew 20.91±0.42 21.43±0.41 20.79±0.44 0.53 Linear models including age, race, and two-factor exposure level-bycovariate interactions found no significant difference in the adjusted exposure level means for percent body fat. The effect of age was significant in the officer cohort (p=0.003), and of borderline significance in the enlisted groundcrew stratum (p=0.064). Race was nonsignificant throughout all the tests. The unadjusted categorical assessment of percent body fat, shown in Table 9-19, revealed no significant exposure effects. However, in the enlisted flyer stratum, a test for linear trend in the proportions gave a borderline significant result (p=0.054), with an estimated step increase of 6.8 plus or minus 3.6 percent per unit increase in exposure-level category. An adjusted analysis by logistic regression did not reveal significant exposure level effects but did detect significant effects of age in the officer and enlisted groundcrew categories. In summary, detailed exposure analyses were performed on four of five dependent variables used to assess general health status. Only a very few of the tests approached statistical significance (multiple comparisons notwithstanding); of these, three associations suggested a trend of adverse effects from low to high exposure; but only one was statistically significant, and there was no consistency across occupational strata (health perception in officers, p=0.064; sedimentation rate in enlisted flyers, p=0.027; and percent body fat in enlisted flyers, p=0.054). These results were relatively comparable to the negative exposure findings in the Baseline Report. 9-19 �TABLE 9-19. Unadjusted Exposure Index Analysis of Percent Body Fat by Occupation Percent Body Fat Obese <25% Lean/Normal <25% Occupation Officer Enlisted Flyer Enlisted Groundcrew Exposure Level Number Percent Number Percent Total p-Value* 104 110 100 81.9 84.6 81.3 23 20 23 18.1 15.4 18.7 127 130 123 0.76 Medium High 50 53 44 90.9 81.5 77.2 5 12 13 9.1 18.5 22.8 55 65 57 0.14 Low Medium High 126 131 113 81.8 80.4 79.6 28 32 29 18.2 19.6 20.4 154 163 142 0.88 Low Medium High L0¥ *Chi-square tests, 2 d.f. LONGITUDINAL ANALYSES Two variables, self-perception of health and sedimentation rate, were prescribed to assess the longitudinal differences between the 1982 Baseline examination and the 1985 followup examination. Both variables were analyzed in the discrete form. The four categories of perception of health were reduced to normal (excellent/good) and abnormal (fair/poor). The respective laboratory norms of 12 or less mm/hr and more than 12 mm/hr for the Baseline sedimentation rates, and 20 or less mm/hr and more than 20 mm/hr for the followup examination were used to categorize the sedimentation rate data into normal and abnormal groups. The off-diagonal data (normal to abnormal, abnormal to normal) from the two examinations were contrasted by group membership, a process equivalent to testing for a group-by-time-by-clinical endpoint interaction. The results of these tests, unadjusted for covariates, are given in Table 9-20. These analyses showed an equivalence of the change in self-perception of health in the two groups between examinations, but a highly significant group difference in the change in sedimentation rate abnormalities. The latter was explained by the fact that the Baseline examination determined a significant excess of sedimentation rate abnormalities in the Comparisons whereas at the followup examination, the Ranch Hands had a significantly higher proportion of abnormalities. Perhaps as a related fact, it is recognized that the sedimentation rate laboratory test procedure changed to a more sensitive one at the followup examination. 9-20 �TABLE 9-20. Longitudinal Analysis of Self-Perception of Health and Sedimentation Rate: A Contrast of Baseline and First Followup Examination Abnormalities Followup Examination Variable Group Ranch Hand Odds p-Value Baseline Examination Abnormal Normal Ratio (OR*) (ORRH vs. ORC) 0.21 62 27 127 750 Abnormal Normal 49 28 124 936 0.23 Ranch Hand Abnormal Normal 17 39 16 899 2.44 Comparison Sedimentation Rate Abnormal Normal Comparison SelfPerception of Health Abnormal Normal 14 27 37 1,061 0.73 0.84 0.002 Number Normal Baseline, Abnormal Followup *0dds Ratio: Number Abnormal Baseline, Normal Followup. SUMMARY AND CONCLUSIONS General physical health was evaluated by five measures, three of which were subjective (self-perception of health, appearance of distress, and appearance of relative age), and two of which were objective (percent body fat and sedimentation rate). Table 9-21 presents a summary of all the unadjusted and adjusted analyses of these five variables. The Ranch Hands rated their health as fair or poor more often than the Comparisons (9.1% versus 7.3%, respectively), but this difference was not significant by categorical testing. However, further analysis revealed a significant group-by-occupation interaction; differences were largely confined to the enlisted groundcrew category. Both the Ranch Hand and Comparison groups noticeably improved their perceptions of health from the 1982 Baseline examination. Only 10 individuals were reported as appearing acutely ill or distressed at the followup examination, 4 were Ranch Hands and 6 were Comparisons. This difference was not statistically significant and the data were insufficient for adjusted analyses. 9-21 �TABLE 9-21. Overall Summary Results of Unadjusted and Adjusted Analyses of General Health Variables Unadjusted Variable Categorical Self-Perception of Health NS Appearance of Illness/Disstress Mean Categorical Mean NS Appearance of Relative Age Adjusted — **** NS — **** Sedimentation Rate 0.013 NS 0.011 NS Percent Body Fat NS 0.037 NS 0.035 —Analysis not performed. ****Group-by-covariate interaction. "Analysis not possible due to sparse data. Appearance of relative age, as determined by the examining physician, showed 1.6 percent of the Ranch Hands appearing younger than their stated age, 94.3 percent appearing the same, and 4.1 percent appearing older (as contrasted to 0.7%, 95.4%, and 3.9%, respectively, in the Comparison group). There was a significant group-by-occupation interaction, but none of the estimated relative risks for the occupational categories was significant. This observation at the followup examination contrasted with the significant tendency at the Baseline for a higher percentage of Ranch Hands than Comparisons to appear younger than their stated ages. The geometric mean sedimentation rates (5.05 mm/hr Ranch Hand versus 4.93 mm/hr Comparison) did not differ significantly by group, either unadjusted or after adjustment for age, race, occupation, personality score, and an age-by-personality score interaction. However, in the dichotomous form, 5.8 percent of the Ranch Hands had sedimentation rate abnormalities as contrasted to 3.6 percent in the Comparison group. This difference was significant by both unadjusted and adjusted tests. Also, this finding was opposite to that of the Baseline examination, where it was noted that younger Comparisons had significantly elevated sedimentation rates. 9-22 �The mean percent body fat of the Ranch Hands was significantly lower than the Comparisons (21.10%±0.15, 21.54%±0.14, respectively, p=0.037), and was of nearly the same magnitude after adjustment for age, race, and occupation. However, both unadjusted and adjusted categorical tests did not reveal significant group differences, although the percent obese was lower in the Ranch Hands than in the Comparisons. No group differences in percent body fat were noted at the Baseline examination. Detailed exposure analyses were done on four general health variables (appearance of acute distress was too sparse for testing). Only one analysis demonstrated statistical significance, i.e., a positive association of sedimentation rate abnormalities with increasing exposure in the enlisted flyer cohort. Overall, no consistent pattern of exposure effects was discernible, and the exposure findings at the third-year followup were similar to the findings at Baseline. Longitudinal differences between the 1982 Baseline and the 1985 followup examination were assessed by analyses of two discrete variables, selfperception of health and sedimentation rate. Perceived health showed no significant group differences over time, but both the Ranch Hand and Comparison groups paradoxically reported symmetrical improvements in their perceptions over the 3-year period. The sedimentation rate analysis revealed a highly significant group difference (p=0.002), due to a reversal of findings between examinations, i.e., a significant detriment in the younger Comparisons at the Baseline versus a significant detriment in the Ranch Hands at the followup. The cause(s) and biological relevance of this observation are unclear. In conclusion, a nonspecific assessment of general physical health has shown relatively close similarity between the Ranch Hand and Comparison groups, with the Ranch Hands continuing to perceive their health more negatively than the Comparisons, having a slightly more favorable percent body-fat proportion, but a higher proportion of abnormal sedimentation rates that reflects a marked change since the Baseline examination. These findings must be placed in context with the organ and system-specific evaluations found in the succeeding chapters. 9-23 �CHAPTER 9 REFERENCES 1. Jenkins, C.D., R.H. Rosenman, and S.J. Zyzanski. 1974. Prediction of clinical coronary heart disease by a test for the coronary-prone behavior pattern. New Eng. J. Med. 290(23):1271-1275. 9-24 �CHAPTER 10 MALIGNANCY INTRODUCTION Cancer is a major suspect disease following exposure to chlorophenols, phenoxy herbicides, and dioxin. Both systemic cancer and skin cancer are key focal points of this study. The issue of military service related cancer in Vietnam veterans first arose in 1978-1979. Media presentations emphasized several early cancer deaths in several Army veterans, which were allegedly caused by exposure to Agent Orange. The media reinforced the causal allegations by citing animal studies, which demonstrated a carcinogenic effect, and a few human studies, which showed excessive cancer in specific occupational groups. So effective and sustained were the media presentations that today the public equates dioxin and Agent Orange exposure to cancer. In the larger context of environmental controversies, Young aptly described the Agent Orange issue as being at the crossroads of science and social concern. The scientific community has responded to the dioxin question by a massive research effort, which in concert with class action lawsuits, is expected to cost more than a billion dollars in the near future. The core of the overall research effort is basic and applied cancer research. Traditional animal-to-man extrapolation difficulties and interspecies variability have limited the direct applicability of much of the experimental work to the controversy. Major epidemiologic challenges have included: the ability to control/characterize bias; selection of suitable controls or reference groups; quality/quantity of exposure; misclassification of exposure; confounding exposure to known injurious chemicals; sample size and statistical power; number and selection of relevant risk factors; lack of antecedent disease or syndromes (other than chloracne); time to event (latency); rarity of the endpoint; and tumor type (carcinoma, sarcoma) differences found in many studies. For these reasons, there is no scientific consensus on the dioxin-cancer question. There is, however, a common thread, raising concern over soft tissue sarcomas (STS) and non-Hodgkin's lymphoma (NHL). Pertinent animal and human studies underscore the concern.over cancer. Numerous animal studies have been conducted to delineate the role of TCDD on tumor initiation, tumor promotion, mutagenesis, cocarcinogenesis, and DNA reactivity. The consensus of most research is that TCDD is only weakly mutagenic, does not covalently bind to DNA or cause it to initiate repair synthe|is, and behaves as a strong tumor promoter in already initiated cells. 10-1 �The oncogenic response to TCDD in animals has been repeatedly shown to depend upon animal species and strain, dose, age, sex, and route of administration. Conventional skin bioassays in mice produced mixed results in some studies ' but caused significant dermal fibrosarcomas in other studies using different strains of animals. In the presence of a strong carcinogen, TCDD induced skin papillomas in homozygous hairless mice (but not in the heterozygous strain), clearly supporting the promoter role of TCDD, a nongenetic mechanism judged to be related to receptor binding. Ingestion studies in several rat strains at doses of 0.07-0.1 yg/kg/day produced hepatocellular carcinomas, squamous cell carcinomas of the oropharynx and lung, and follicular cell thyroid adenomas. ' In two mouse strains, gavage doses of 0.07-0.3 yg/kg/day produced hepatocellular carcinomas and thyroid tumors. In the presence of partial hepatectomy and diethylnitrosamine, subcutaneous TCDD administration to rats resulted in hepatocellular carcinomas, demonstrating the promoter mechanism of TCDD. Based upon these and other studies, the International Agency for Research on Cancer (IARC) designated TCDD as carcinogenic in 1982. There are insufficient data to implicate 2,4-D and 2,4,5-T as carcinogens. The majority of animal studies have shown carcinomas rather than sarcomas, the tumor cited in some human studies. If TCDD oncogenicity in humans is to be supported, the differences in tumor types between animals and man requires explanation. In a series of publications beginning in 1974, commonly known as the "Swedish studies," extensive inquiry was made into occupational cancer following exposure to a variety of herbicides. Four related efforts using Swedish railroad workers found an increased cancer incidence mostly associated with non-TCDD herbicides. However, a case-control analysis of these data by other investigators suggested cancer promotion following phenoxy acid exposure. Prompted by a slight increase in STS in the railroad workers and clinical experience with a case series of STS, Hardell and coworkers launched an extensive second round of studies. ~ These efforts showed statistically significant increased risks for STS, Hodgkin's Disease (HD), and NHL. For exposure to phenoxy acids alone, the risk ratio ranged from 5.3 to 6.8 for STS in northern and southern Sweden, respectively, while a range of 3.3 to 6.6 was noted for exposure to chlorophenol alone. For malignant lymphoma (HD plus NHL), risk ratios of 8.4 and 4.8 were respectively demonstrated for chlorophenol and phenoxy acid exposures. An association of nasal and nasopharyngeal cancer to chlorophenol exposure (risk ratio, 6.7) was also detected, but other specifically focused studies of primary liver cancer and colon cancer were negative with respect to phenoxy acid or chlorophenol exposure. '2 The colon cancer study was conducted specifically to demonstrate a lack of respondent bias to "validate" previous questionnaire and interview methods used in the STS studies. From the outset, the Swedish studies have been criticized on methodologic issues, ~ prompting the primary authors, Axelson and Hardell, to respond with clarifications, new calculations, amplifying studies on additional cohorts, and studies on other cancers. 2'2 '2 ~ 1 The chief criticisms centered upon possible respondent and observational biases, 10-2 �selection of controls, confounding exposures, and degree of true exposure to phenoxy acids and chlorophenols. The authors answered these criticisms within the inherent constraints of the case-control methodology. Their efforts have been characterized as careful, clever, and properly stated, and have received favorable reviews. ' 3 Four small industrial mortality studies were conducted in the late 1970's and early 1980's. NIOSH investigators pooled the data from these studies and noted that three of the 105 deaths (2.9%) in these studies were due to STS as contrasted to an expected 0.07 percent in the U.S. general population. This study has been criticized for the hasty addition of possibly noncomparable industrial cohorts, and the lack of histologic confirmation of the STS cases. A subsequent case report added another STS case to the industrial studies, and two other reports revealed three unrelated STS cases also arising from the industrial sector. ' However, upon closer inspection, only two of the first four cases were confirmed as STS by an independent histologic review. Other review findings of the seven total cases were noteworthy: there was poor agreement on the histologic subtype of the soft tissue tumors, and because of a quirk in the International Classification of Diseases (ICD) System, wherein organ-specific sarcomas are coded separately from soft and connective tissue tumors (ICD 171), deathcertificaje based studies will underascertain STS by approximately 40 percent. ' This latter problem did not affect the Swedish studies. Other cancer studies throughout the world showed mixed support for the Swedish findings. An Italian case-control effort showed a weak association between ovarian mesothelial tumors and herbicide exposure, whereas a Finnish study of a small number of pesticide sprayers understandably did not detect any cases of STS or malignant lymphomas (ML).4 A study of more than 4,000 Danish phenoxy herbicide workers noted five STS cases (versus 1.8 expected) and seven ML cases (versus 5.4 expected). The author concluded that the STS observation supported the Swedish work and that the ML rate did not. One New Zealand case-control study showed a nonsignificant relative risk of 1.3 for STS among occupations consistent with phenoxy herbicide exposure, although a risk of 7.2 was noted for STS and potential chlorophenol exposure in tanneries. A related second cancer registry-based case-control study revealed significant excesses of agricultural and forestry occupations from ML cases and multiple myeloma cases (odds ratio 1.25). 8 In a similar but larger cancer registry study in.Sweden, there was no increased risk of STS (relative risk 0.9) in agricultural or forestry workers as contrasted to other industrial workers. Further, the STS risk was constant over time in spite of increased usage of phenoxy acid herbicides from 1947 to 1970. This Swedish study did not confirm or show a trend consistent with the earlier Hardell Swedish studies. A recent U.S. case-control study from the Kansas cancer registry has provided partial support for Hardell's observations. The Kansas study was very similar in methodology to the early Swedish studies and tried to avoid bias and misclassification. An overall risk of 1.6 was found for NHL in men exposed to herbicides, particularly 2,4-D. As the frequency of herbicide exposure increased to more than 20 days per year, the risk of NHL increased to sixfold vis-a-vis nonfarmers. For herbicide applicators, the risk for NHL 10-3 �was 8.0. A simultaneously published review of the Kansas work noted that this should shift scientific concern from STS to NHL. A population-based case-control study of STS and NHL in western Washington found no overall increased risk of these diseases associated with an occupational history of exposure to chlorophenols or phenoxy herbicides. However, risks of NHL were significantly elevated in the specific occupational categories of farmers, forestry herbicide applicators, and those individuals potentially exposed to phenoxy herbicides in any occupation for 15 years or more. An increased risk of NHL was also noted among those with occupational exposure to insecticides, organic solvents, lead, and welding fumes. A number of Vietnam veteran studies has attempted to determine whether veterans have experienced excessive mortality, particularly from cancer. Most of the studies used proportionate mortality ratio (PMR) methodology and equated Vietnam service with potential exposure to Agent Orange, a procedure of considerable imprecision (misclassification). These exposure allocation difficulties, coupled with the inherent methodological weaknesses of the PMR technique, have minimized the contribution of these studies to the overall cancer issue. As might be predicted by these problems, almost all of the veteran studies were negative for generic cancer associations, as well as for STS, HD, and NHL associations. As an example of the veteran studies, the Australian retrospective cohort mortality effort revealed an overall relative mortality ratio of 0.99, an overall cancer mortality ratio of 0.95, and nonsignificant statistical differences for STS, NHL, and HD. In a recent Vietnam experience study of STS using the case-control method, no significant association was found between military service in Vietnam and the subsequent occurrence of STS. No consistent pattern for other cancer types has emerged from the entire body of herbicide literature. None of the leukemias has been associated with exposure to Herbicide Orange nor any of its constituents. Two studies noted slight increases in gastric cancer ' and two others cited modest risks for lung cancer. ' A recent Swedish study reported slight excesses of rectal cancer in male workers and increased cervical cancer from the exposed female cohort. Overall, these and other observations have not been consistent with the expectation that dioxin, as a cancer promoter, should increase the occurrence of common "background" cancers. From another perspective, if clear-cut exposure to 2,4-D or dioxin is shown to cause an immunological deficiency (see Chapter 19), an expectation would be an excessive representation of B-cell tumors from the population of NHL cases. " An excess of B-cell neoplasms has, in fact, not been described in NHL cases from industrial or veteran cohorts to date. It is unlikely that the cancer question will be clearly resolved in the near future. Dioxin exposure in industry and agriculture has fallen precipitously since the 1970's, while exposures to 2,4-D and non-TCDD containing herbicides have continued. Veteran studies characterized by low or undocumented exposure to Agent Orange, and/or of small cohort size are unlikely to contribute substantive data for the evaluation of type-specific cancers although they may contribute to the resolution of the generic cancer issue. 10-4 �In summary, Swedish studies first noted an approximate sixfold risk of soft tissue sarcoma and malignant lymphoma in forestry workers exposed to both phenoxy acid herbicides (not containing the dioxin contaminant) and chlorophenols (containing dioxin). A large number of international studies were predominantly nonsupportive of the Swedish observations. Recent U.S. research on agricultural workers, however, provided some support for a nonHodgkin's lymphoma-phenoxy acid exposure association. The future scientific focus is expected to shift from dioxin herbicides to nondioxin herbicides and from soft tissue sarcomas to malignant lymphomas. Studies of other veteran populations will not likely contribute to the new emphasis, largely because of exposure uncertainties. Baseline Summary Results Cancer received major emphasis during the 1982 AFHS. The assessment of malignancy used data from both the in-home questionnaire and the review-ofsystems questionnaire obtained during the physical examination as well as data from the examination itself. All subjective data were verified by medical record reviews. In addition, tabulation of mortality count data from the Baseline Mortality Report was used in conjunction with cancer morbidity information. The overall results showed an equivalence of systemic cancer (p=0.46) in the two groups but significantly more nonmelanotic skin cancer (p=0.03) in the Ranch Hands. Of 50 reported systemic cancers from the Ranch Hand and Comparison groups, 28 (14 in each group) were verified by medical records and pathology reports. A visual inspection of anatomic sites showed a slight excess of genitourinary cancer and oropharyngeal cancer but a relative deficit of digestive system neoplasms in the Ranch Hands. A combined morbidity-mortality analysis derived from the initial 1:1 match (Ranch Hand to the C-l Comparison member) disclosed similar distributions. One case of soft tissue sarcoma and one case of Hodgkin's Disease were confirmed, both in the Comparison group. Exposure analyses for industrial chemicals and x rays were negative as were most of the herbicide exposure analyses in the Ranch Hand group. All of the exposure analyses were based upon very small numbers, and interactions were noted in several strata. Questionnaire data verified by medical record reviews revealed significantly more skin cancer in the Ranch Hands (relative odds 2.35). Basal cell carcinoma accounted for 83.9 percent of the reported skin cancers in both groups and was concentrated anatomically on the face, head, and neck. The few melanoma and squamous cell cancers were evenly distributed between the Ranch Hand and Comparison groups. All skin cancers occurred in nonblacks. Adjustments for occupational exposures (e.g., asbestos, degreasing chemicals) did not alter the increased rate of skin cancer in .the Ranch Hand group. Skin cancer in both groups was associated with exposure to industrial chemicals (p=0.03). Herbicide exposure analyses in the Ranch Hand group were essentially negative, although confounding was noted in many of the analyses. Outdoor occupations subsequent to military service as a covariate did not account for the significant skin cancer association. 10-5 �Parameters of the 1985 Malignancy Assessment The emphasis on cancer was increased during the first followup study in 1985. With the Baseline finding of excessive skin cancer in the Ranch Hands, and the lack of covariate data to refine that association, considerable attention was devoted to skin cancer. The questionnaire was altered to collect information on each geographic location in which a participant lived for more than 12 months in order to calculate a cumulative "lifetime" sun exposure index based on geographic latitude, since ultraviolet light exposure has been acknowledged as the primary cause of basal cell carcinoma. Detailed data on skin tannability, eye, skin, and hair color, and parental ethnicity were also obtained. In addition, emphasis at the dermatologic examination was shifted from acne/chloracne to skin cancer, and punch biopsies were sought for all suspected malignant lesions. The participants were asked to bring copies of their medical records to facilitate the verification of reported malignancies. Highly structured smoking data were collected for more detailed covariate adjustments, and Baseline questions on exposure to other carcinogens were repeated to gather interval data. No invasive procedures were used at the followup physical examination to detect evidence of systemic cancer. Thus, the dependent variables of the analyses below are Baseline analyses, but covariate analyses have been expanded and systemic cancers. The lifetime occurrence of cancer, as interval occurrence of skin and systemic cancers between the followup examinations, is analyzed. similar to the for both skin well as the Baseline and Minor numeric differences in various tables that follpw reflect missing data from the covariates. The statistical methods used throughout this chapter are Fisher's exact test, chi-square tests of association, and logistic regression models (BMDP®-LR) for adjusted group contrasts of neoplasm incidence rates. RESULTS AND DISCUSSION General Malignant and benign neoplasms, carcinomas in situ, and neoplasms of uncertain behavior or unspecified nature are studied in this chapter. The term "systemic" is used throughout to denote a nonskin neoplasm. The term "unspecified" is used to denote a neoplasm of uncertain behavior or unspecified nature. Neoplasm refers to any new and abnormal growth which may or may not be malignant. Malignant neoplasms (malignancies, cancer) are those neoplasms that are capable of invasion and metastasis. Questionnaire Data At the followup examination, participants provided information on cancer during the interval between examinations and participants who were new to the study gave their lifetime history. All reported neoplasms entered the medical records review process for verification. Only 11 Ranch Hands (1.1%) and 12 Comparisons (0.9%) reported neoplastic conditions which could not be 10-6 �substantiated (all of the skin); the group difference was nonsignificant (p-0.833). Physical Examination Data Some possible neoplastic conditions were discovered by the physicians at the physical examination. Many suspicious skin lesions were biopsied and the pathology determined. However, for some suspected skin neoplasms and all suspected systemic neoplasms, verification was not complete at the time of writing this report, and thus both verified and suspected neoplasms are described and analyzed. The term suspected is used throughout to denote those possible neoplastic conditions noted by the physicians at the followup examination for which the results of verification are not yet available. Consideration of suspected neoplasms was justifiable in particular for skin neoplasms, for which the biopsy confirmation rate is high. Statistical Analysis The statistical analysis is described in three sections. The first section presents unadjusted and adjusted analyses of skin and systemic neoplasm incidence in the Baseline-followup interval, and is referred to as interval analysis. In the second section, unadjusted and adjusted analyses of lifetime skin and systemic neoplasm incidence are analyzed for the followup participants, incorporating Baseline information. Since there were very few neoplasm occurrences before the SEA tours, this combined interval and Baseline analysis is referred to throughout as lifetime analysis. Lastly, the neoplasm history and mortality of the fully compliant Baseline participants subsequent to Baseline are described. All analyses are of the numbers of participants with (one or more) neoplasms, and not of the total number of neoplasms. The purpose of these three analyses is to present a comprehensive picture of the neoplasia history of the followup participants, and to provide some additional information on the neoplasia status of the Baseline participants subsequent to Baseline. There was a slight difference between the Baseline and followup cohorts. The interval and lifetime analyses pertain to neoplasm incidence among followup participants only. The third section pertains to Baseline participants only, describing their history of neoplasm incidence and mortality since Baseline. A fully combined morbidity-mortality analysis was not feasible for this report. Assuming a (two-sided) a -level of 0.05 and power 0.8, the sample sizes were sufficient to detect a relative risk of 2.56 when the Comparison neoplasm incidence rate is 1 percent, and a relative risk of 1.63 when the Comparison neoplasm incidence rate is 5 percent. For nonblacks only, the corresponding detectable relative risks are 2.63 and 1.65, respectively. All analyses of data from Ranch Hands and the Original Comparisons only are given in Appendix H. This appendix also contains other tabulations, such as covariate and interaction tables. 10-7 �Baseline-Follovup Interval Table 10-1 shows the Baseline-follovup interval neoplasm history for the followup participants. The interval began in January 1982 for participants new to the study, i.e., the 45 new Ranch Hands, the 71 new replacement Comparisons, and 83 newly compliant Original Comparisons. The total numbers of participants with verified neoplasms were 161/1,016 (15.8%) Ranch Hands and 170/1,293 (13.1%) Comparisons; the group difference was marginally significant (p=0.073). The relative frequencies of participants with verified plus suspected neoplasms, 17.4 percent of Ranch Hands and 16.2 percent of Comparisons, did not differ significantly between groups (p=0.466). Appendix Table H-l gives the numbers of participants with verified or suspected neoplasms and unadjusted analyses for the Ranch Hands and Original Comparisons in the Baseline-followup interval. Interval Skin Neoplasms Of Ranch Hands with verified neoplasms of all types (malignant, benign, and uncertain) 70.8 percent (114/161) had skin neoplasms; the corresponding percentage for the Comparisons was 68.2 percent (116/170). The difference in these proportions was not significant (p=0.634). When suspected neoplasms were included, the contrast was 70.1 percent (124/177) versus 67.6 percent (142/210), again not significant (p=0.660). No Blacks were found to have skin cancer, as anticipated since Blacks have a lower susceptibility to sun-induced skin cancer. Therefore, analysis of skin cancer was limited to nonblacks. Of Ranch Hands with skin neoplasms, 32.5 percent (37/114) had malignant neoplasms, as contrasted to 34.5 percent (40/116) of the Comparisons (p=0.781). When suspected malignant skin neoplasms were included, the contrast was 37.9 percent (47/124) versus 42.3 percent (60/142), and was not significant (p=0.531). For the remainder of this section, only malignant skin neoplasms are analyzed. The dependent variables examined were basal cell carcinomas, melanomas, squamous cell carcinomas, all skin cancers combined, and a group of skin cancers called sun exposure-related skin malignancies. The sun exposure-related skin malignancies were defined as basal cell carcinomas, melanomas, and malignant epithelial neoplasms not otherwise specified (NOS). The latter were included because they are frequently misdiagnosed basal cell carcinomas; three Ranch Hands had this diagnosis. Interval Malignant Skin Neoplasms Table 10-2 presents the numbers of participants with verified and suspected malignant skin neoplasms by cell type: basal cell carcinomas, squamous cell carcinomas, melanomas, all skin malignancies combined, and the sun exposure-related skin malignancies, together with the results of unadjusted group contrasts. For the sake of completeness, the total numbers of malignancies of each type are also given. The majority of the 10-8 �TABLE 10-1. Unadjusted Analyses of Followup Participants with Verified (Nonblacks and Blades) Group* Ranch Hand Site Skin Systanic All Neoplasm Behavior and Status Comparison , Number** Percent Number** Percent Malignant Verified Verified and Suspected Benign Verified Verified and Suspected Uncertain Behavior and Unspecified Nature: Verified Verified and Suspected Any Skin Neoplasm* Verified Verified and Suspected Total** p-Value*** 37 47 3.6 4.6 40 60 3.1 4.6 77 107 0.485 0.999 76 78 7.5 7.7 77 83 6.0 6.4 153 161 0.152 0.250 1 1 0.1 0.1 1 0.1 1 • 0.1 2 2 099 .9 0.999 114 124 11.2 12.2 116 142 9.0 11.0 230 266 0.030 0.393 8 12 08 . 1.2 7 12 0.5 0.9 15 24 0.603 0.680 42 48 4.1 4.7 50 61 3.9 4.7 92 109 0.749 0.999 6 6 0.6 06 . 7 11 0.5 0.9 13 17 0.999 0.625 55 65 5.4 6.4 61 80 4.7 6.2 116 145 0.445 0.863 Malignant, Benign, Uncertain Behavior, Unspecified Nature Verified 161 Verified and Suspected 177 15.8 17.4 170 210 13.1 16.2 331 387 0.073 046 .6 Malignant Verified ' Verified and Suspected Benign Verified Verified and Suspected Uncertain Behavior and Unspecified Nature: Verified Verified and Suspected Any Systemic Neoplasm3 Verified Verified and Suspected *Sample sizes: 1,016 Ranch Hands and 1,293 Comparisons. **Number of participants. ***Fisher's exact test. ^Participant has one or more malignant, benign, or unspecified skin neoplasms. Participant has one or more malignant, benign, or unspecified systemic neoplasms. °Participant has one or more malignant or benign skin or systemic neoplasms. 10-9 �TABLE 10-2. Unadjusted Analyses of Nonblack Follovup Participants vith Verified and Suspected Malignant Skin Neoplasms in the Baseline-Followup Interval by Cell Type and Group Group* Cell Type Status Statistic** Ranch Hand Comparison Est. Relative Risk (95% C.I.) p-Value Verified Number/% Total Neoplasms 29 42 3.0% 30 40 2.5% 1.23 (0.73,2.07) 0.429 Verified & Suspected Number/% Total Neoplasms 36 53 3.8% 48 63 4.0% 0.95 (0.61, 1.47) 0.824 Verified Number/% Total Neoplasms 4 6 0.4% 4 4 0.3% 1.27 (0.32,5.08) 0.738 Verified & Suspected Number/% Total Neoplasms 4 6 0.4% 5 5 0.4% 1.01 (0.27,3.78) Verified Number/% Total Neoplasms I 2 0.1% 3 3 0.3% 00, 0.42 ( . 4 4.06) 0.635 Verified & Suspected Number/% Total Neoplasms I 2 0.1% 6 7 0.5% 0.21 ( . 3 1.75) 0.142 00, All Malignant Verified Skin Neoplasms Verified & Suspected Number /% Total Neoplasms 37 56 3.9% 40 52 3.3% 1. 18 (0.75, 1.86) 0.486 Number/% Total Neoplasms 47 70 4.9% 60 81 5.0% 0.99 (0.67, 1.47) 0.999 Sun-Exposure Verified Related Malignant Neoplasms3 Verified & Suspected Number/% Total Neoplasms 32 47 3.4% 33 43 2.7% 1.24 (0.75,2.02) 0.447 Number/% Total Neoplasms 39 58 4.1% 53 71 4.4% 0.93 (0.61,1.42) 0.749 Basal Cell Carcinoma Squamous Cell Carcinoma o Melanoma *Number of participants—956 Ranch Hands and 1,210 Comparisons. **Number and percent of participants; total number of malignant neoplasms of specified cell type. a Basal cell carcinoma, melanoma, and malignant epithelial neoplasms NOS. 0.999 �participants with verified skin malignancies had basal cell carcinomas: 78.4 percent (29/37) Ranch Hands versus 75.0 percent (30/40) Comparisons; the difference between the groups was not significant (p=0.792). Unadjusted Analyses Table 10-2 shows that no significant group differences were found in the incidence rates of either verified or verified plus suspected malignant skin neoplasms. For verified basal cell carcinomas, the estimated relative risk of Ranch Hands versus Comparisons was 1.23 (95% C.I.: [0.73,2.07]) and was not significant (p=0.429). The estimated relative risk for verified squamous cell carcinoma, 1.27 (95% C.I.: [0.32,5.08]), was also not significant (p=0.738). The estimated relative risk for verified melanoma, 0.42 (95% C.I.: [0.04,4.06]), was also not significant (p=0.635). There were very few occurrences of melanoma (one Ranch Hand and three Comparisons) since this is a much rarer condition than other kinds of skin cancer. There were no significant differences between the groups for all verified malignant skin cancers combined (Est. RR: 1.18, 95% C.I.: [0.75,1.86], p=0.486) or for the category of sun exposure-related skin malignancies (Est. RR: 1.24, 95% C.I.: [0.75,2.02], p=0.447). When both verified and suspected malignant skin neoplasms were analyzed, the conclusions were similar, namely, there were no significant differences between the groups, and moreover, the estimated relative risks were closer to 1. No group differences were found in the parallel contrasts of Ranch Hands versus Original Comparisons (see Table H-2 of Appendix H). As shown in Table 10-3, additional analyses contrasted group differences in the anatomic location of basal cell carcinomas, melanomas, and sun exposure-related skin malignancies. Most occurrences of basal cell carcinoma and sun exposure-related skin malignancies were on the face, head, or neck, or the upper extremities. The relative frequency of occurrences of verified basal cell carcinomas at these combined sites was 89.7 percent for Ranch Hands and 80.0 percent for Comparisons of the total number of occurrences in each group, respectively. The group contrast (26/29 versus 24/30) was not significant (p=0.472). These combined sites accounted for 90.6 percent (29/32) of the sun exposure-related malignancies for Ranch Hands versus 72.7 percent (24/33) for Comparisons; this contrast was also not significant (p=0.108). The corresponding contrasts, when suspected malignant neoplasms were included with the verified malignant neoplasms, were also not significant. One Ranch Hand had verified melanoma of the face, and three Comparisons had verified melanoma on the trunk. Two other Comparisons had suspected melanoma, also on the trunk. The group contrast for melanomas on the trunk was not significant for verified conditions (p=0.260), but was marginally significant for verified plus suspected conditions (p=0.071), the detriment being in the Comparison group. Table 10-4 gives the frequencies of participants with face, head, and neck skin malignancies by group and occupation. Specifically, nonmelanoma malignant skin neoplasms and the sun exposure-related malignant skin neoplasms are listed by occupational category. For officers and enlisted groundcrew, the frequencies of participants with face, head, and neck malignant skin neoplasms (both malignant nonmelanoma and the malignant sun exposure-related skin neoplasms) did not differ significantly by group. However, the Ranch Hand enlisted flyers had a significantly higher frequency 10-11 �10-3. Malignant Skin Neoplasms in the Ra«*>liT*>-PnlLowup Interval by Anatomic Site and Group Basal (jell Carcinoma Sun-Exposure Related Malignancies Group* Group* Site Face, Head, Neck K Status** Ranch Hand Comparison p-Value Number/% Verified Verified & Suspected 24 2.5% 29 3.0% Upper extreai- Number/% ities Verified Verified & Suspected Melanoma Group* Ranch Hand Comparison p-Value Ranch Hand Comparison p-Value 034 .7 089 .9 27 2.8% 32 3.4% 23 1.9% 36 3.0% 0.194 0.622 1 0.1% 1 0.1% 0 0.0% 0 4 1 .4 .9 1 0.1% 0 9 9 5 0.5% 5 0.5% 3 0.3% 0.313 4 0.3% 0 5 0 .2 5 05 .% 5 0.5% 3 0.3% 5 04 .% 0.313 0.757 0 00 .% 0 00 .% 0 00 .% 0 00 .% a a Number/% Verified Verified & Suspected 2 0.2% 4 04 .% 6 0.5% 0 4 9 .7 11 0 9 0 2 0 .% .0 2 0.2% 4 04 .% 9 0.7% 14 1.2% 0.126 003 .9 0 0.0% 0 00 .% 3 0.3% 5 04 .% 020 .6 0.071 Lower Extrem- Number/% ities Verified Verified & Suspected 0 0.0% 0 0.0% 0 00 .% 0 00 .% 0 00 .% 0 00 .% 0 00 .% 0 00 .% 0 00 .% 0 0.0% 0 00 .% 0 0.0% a a Other Sites Number/% and Sites NOS Verified Verified & Suspected I 0.1% I 0.1% 2 0.2% 0 9 9 .9 2 0.2% 0 9 9 .9 1 0.1% 1 0.1% 2 0.2% 2 0.2% 099 .9 099 .9 0 0.0% 0 0.0% 0 00 .% 0 00 .% a a All Locations Number/% Verified Verified & Suspected 29 3.0% 36 3.8% 32 3.4% 39 4.1% 33 2.7% 53 4 4 .% 047 .4 079 .4 1 0.1% 1 0.1% 3 0.3% 6 0.5% 0.635 0.142 Trunk 23 1.9% 35 2.9% 30 2.5% 48 4 0 .% a a 049 .2 084 .2 *Number of participants — 956 Ranch Hands, 1,210 Comparisons. **Number and percent of participants. a No occurrences in either group. a a �TABLE 10-4. Unadjusted Analyses of Hnriblark FolloHup Rarticipants with Nomelanana Maligpant Skin Neoplasms and Sun-Exposure Related Skin Malignancies in the Basel Ine-RJloHUp Interval Occurring on the Face, Head, or Neck by Occupation Occupation Status Statistic n Group* Cotppyision Ranch H^ryi p-Value Number Percent Number Percent 477 373 Group* Conparison Ranch Hand p-Value Number Percent Number Percent 477 373 Verified Face, Head, Neck 14 Other Site 5 No Cancer 354 3.8 1.3 94.9 17 4 456 3.6 08 . 95.6 099 .9* 056 .1" 12 3 358 3.2 08 . 9. 60 13 5 459 2.7 1.1 96.2 068 .8* 099 .9" Verified & Suspected Face, Head, Neck 20 7 Other Site No Cancer 346 5.4 1.9 92.8 23 8 446 48 . 1.7 93.5 074 .5* 099 .9" 17 4 352 4.6 1.1 9. 44 20 9 448 4.2 1.9 93.9 086 .6* 048 .0" Officer n 167 167 193 193 Verified 8 Face, Head, Neck Other Site 1 No Cancer 158 Enlisted Flyer 4.8 06 . 94.6 3 0 190 1.6 00 . 98.5 0.121* 046 .5" 8 1 158 48 . 06 . 94.6 2 1 190 1.0 0.5 98.5 009 .4* 099 .9" 8 Verified & Suspected Face, Head, Neck 1 Other Site No Cancer 158 48 . 06 . 9. 46 5 1 187 2.6 06 . 96.9 036 .9* 099 .9" 8 1 158 48 . 06 . 94.6 4 2 187 2.1 10 . 96.9 028 .3* 099 .9" n 416 416 540 540 Verified 7 Face, Head, Neck Other Site 1 No Cancer 408 1.7 02 . 98.1 9 4 527 1.7 0.7 97.6 099 .9* 035 .9" 7 1 408 1.7 0.2 98.1 8 4 528 1.5 0.7 97.8 080 .0* 0.395" Enlisted Groundcrew Verified & Suspected Face, Head, Neck 7 Other Site 3 No Cancer 406 1.7 07 . 97.6 D 6 521 2.4 1.1 96.5 050 .0* 079 .3" 7 2 407 1.7 0.5 9. 78 12 6 522 2.2 1.1 96.7 064 .4* 047 .7" *Number and percent of participants. *Fisher's exact test for face, head, or neck versus no malignancy. Fisher's exact test for other site versus no malignancy. �of malignant sun exposure-related skin neoplasms than the corresponding Comparisons, 4.8 percent versus 1.0 percent (p«0.049). For nonmelanoma malignant skin neoplasms, the contrast was 4.8 percent versus 1.6 percent, but the difference was not significant (p=0.121). Inclusion of suspected malignant neoplasms with the verified malignant neoplasms reduced the significance of the difference between the groups for both the sun exposure-related skin malignancies and the nonmelanoma malignant skin neoplasms. Adjusted group contrasts of the incidence rate of basal cell carcinomas and malignant sun exposure-related skin neoplasms were done for verified and verified plus suspected conditions. Adjusted analyses were not carried out, however, for melanomas or squamous cell carcinomas because of the small frequencies. Covariates The covariates considered for the adjusted analyses of malignant skin neoplasm incidence, listed in Table 10-5, were the matching variables age and occupation; history of alcohol and cigarette use; host factors, comprising skin color, eye color, hair color, and ethnic background; reaction of skin to sun exposure; average lifetime residential latitude; and exposure to recognized carcinogens. Age was used as a continuous variable in the adjusted analyses, but was categorized for ease of presentation in the report. Eye color, hair color, and skin color were coded by the dermatologist at the physical examination. Hair color was determined by comparing the hair at the back of the neck with 17 numbered standardized hair samples and selecting the most closely matching hair sample. Similarly, skin color groupings from dark brown to pale peach were determined by comparing standardized flesh-colored squares against the skin of the inside upper arm. For the analysis, hair and skin colors were grouped as shown in Table 10-5. Each participant was assigned to one of four ethnic groups according to his responses to questions on race, as given in Table 10-5. (Blacks were omitted from the table because the analysis of malignant skin neoplasia was restricted to nonblacks.) These ethnic categories are approximate groupings in terms of susceptibility to sun-induced skin damage. The ethnic categories also generally correlate to skin color, a commonly known important risk factor for skin cancer. A lifetime residential history was obtained from all participants by a questionnaire. Residential history, relative to the equator, is a surrogate measure of sun exposure (but does not account for altitude or average sun-days at each location), an important risk factor for skin cancer. Each participant was asked to list all residences chronologically, citing both the city (or military installation) and the years of residence at each location since birth. Residences of less than 1 year were not sought because of the frequent short-term military travels of these cohorts. By standardized geographic atlases, the latitude (in degrees and minutes) of each residence was recorded. The Air Force subsequently checked all of the latitude determinations for accuracy. The average lifetime residential latitude of each participant was calculated by dividing the total degree-years (i.e., sum of latitude [degrees] times number of years lived there) from all residences by the total number of residential years listed. 10-14 �TABLE 10-5. Covariates for Analyses of Malignant Skin Neoplasms Category Covariate Age Born XL942, 1923-1941, <1922a Occupation Officer, Enlisted Flyer, Enlisted Groundcrew Lifetime Cigarette Smoking Pack-years: 0, >0-20, >20-40, >40 Lifetime Alcohol Consumption Drink-years: 0, >0-5, >5-30, >30-100, >100 Ethnic Background A, B, C, Db Skin Color Dark, medium, pale, dark peach, pale peach Hair Color Black, dark brown, light brown, blond, red Eye Color Brown, hazel, green, gray, bluec Reaction of Skin to Sun Exposure : (A.I) After first 30 minutes of summer sun (A.2) After >2 hours, after first exposure (A.3) After repeated sun exposures Burns, usually burns, burns mildly, rarely burns Burns painfully, burns, becomes red, no reaction Freckles with no tan, tans mildly, tans moderately, tans deep brown Sun-Reaction Index (Composite)*3 (1) Burns painfully (A.2) and/or freckles with no tan (A.3) (2) Burns (A.2) and/or tans mildly (A.3) (3) All other reactions Residential History (Average Latitude) Average latitude <37°, >37° Exposure to Carcinogens/Groups of Carcinogens Set I3 Asbestos Nonmedical X Rays Industrial Chemicals Herbicides Insecticides Degreasing Chemicals Yes, Yes, Yes, Yes, Yes, Yes, No No No No No No 10-15 �TABLE 10-5. (continued) Covariates for Analyses of Malignant Skin Neoplasms Covariate Set 2e Anthracene Arsenic Benzene Benzidene Chromates Coal Tar Creosote Aminodiphenyl Chlororaethyl Ether Mustard Gas Naphthylamine Cutting Oils Trichloroethylene Ultraviolet Light (not sun) Vinyl Chloride Composite Carcinogen Exposure Category Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, No No No No No No No No No No No No No No No Yes, if yes for exposure to any carcinogen in set 2, otherwise no. a Used as a continuous variable in adjusted analysis. b A B C D - English, Welsh, Scottish, Irish. Scandinavian, German, Polish, Russian, other Slavic, Jewish, French. Spanish, Italian, Greek. Mexican, American Indian, Asian. c Participant with one green eye and one brown eye is coded as green. Questionnaire data (see Appendix B). *AFHS Form 2 (see Appendix C). 10-16 �Recognizing that both total degree-years and average lifetime latitude could be covariates for malignant sun exposure-related skin neoplasms, average latitude was selected because of the high correlation of degree-years with chronological age, a separate risk factor already used in the analyses. Further, average residential latitude was believed to be a more stable measure in the presence of some lack of precision in the source data. In all analyses, the average residential latitude was used as a dichotomous variable (less than 37° N latitude, greater than or equal to 37° N latitude). A line across the United States at 37° N approximates a line from San Francisco, California, to Richmond, Virginia. Examination of the group distributions of the latitude variable suggest that it is a significant confounding variable. Specifically, 56.7 percent of the nonblack Ranch Hands had an average lifetime residential latitude greater than or equal to 37° N latitude versus 49.4 percent of the nonblack Comparisons (p=0.001). Although the average lifetime group residential latitudes appear similar (37.21° N latitude for the Ranch Hands, and 36.74° N latitude for the Comparisons), this difference is also highly significant (p=0.003), reflecting the substantial power of the analysis of continuous data. Participants reported their susceptibility to the effects of sunexposure damage by answering three questions about their skin reaction to sun: the reaction after the first 30 minutes of exposure to summer sun, the reaction after 2 or more hours of sun exposure after the first 30-minute exposure, and the reaction after repeated exposures (see questions 10-12 on page 71 of the questionnaire provided in Appendix B). Since these three responses are highly correlated, a composite sun-reaction variable for use in the adjusted analysis, called the sun-reaction index, was constructed from the last two questions (2-hour and repeated exposure reactions) after examination of the association between basal cell carcinoma incidence and the three skin reaction variables. The sun-reaction index had three categories. The first category corresponded to the most sensitive reaction on the last two questions, the second category corresponded to the next less sensitive .reaction on these two questions, and the third category comprised the remaining responses. Detailed questionnaire information on exposure to asbestos, nonmedical x rays, industrial chemicals, herbicides, insecticides, and degreasing chemicals was obtained from each participant. Self-reported information on exposure to 15 individual carcinogens was obtained at the physical examination. A composite carcinogen exposure variable was constructed from these responses on individual carcinogens: A participant had a positive score for this variable if he reported exposure to one or more of the 15 carcinogens, otherwise he had a negative score. Self-reported information on asbestos and radiation exposure was not used because this information was obtained in more detail from the questionnaire. The nonblack Ranch Hands differed significantly from the nonblack Comparisons in their exposure (yes/no) to nonmedical x rays (19.3% versus 25.6%, p<0.001). They also differed significantly from the Comparisons in their exposure to herbicides (94.1% versus 29.8%, p<0.001) and insecticides (70.2% versus 53.1%, p<0.001), possibly reflecting Vietnam experience. These variables were not used in the adjusted analysis. Further, there were significant or marginally significant group differences in the self-reported exposures to several individual carcinogens, in each instance relatively more (nonblack) Ranch Hands than Comparisons reported exposure: arsenic (2.7% 10-17 �versus 1.2%, p-0.016), naphthylamine (3.3% versus 1.7%, p=0.024), cutting oils (12.7% versus 8.7%, p=0.003), benzene (4.3% versus 2.7%, p=0.056), and benzidine (0.8% versus 0.3%, p=0.070). Results were similar when Blacks were included in the analysis. Covariate Associations Table 10-6 gives a summary of the chi-square tests of association between all covariates and the incidence of basal cell carcinomas and sun exposure-related malignancies. Details of these tests of association are provided in Appendix H, Table H-3. There was a significant increase in the incidence rate of verified basal cell carcinomas with increasing age (p=0.001). There was a significant difference in the incidence rate of basal cell carcinomas among occupation groups, with enlisted groundcrew having a lower incidence rate (1.8%) than officers (3.7%) and enlisted flyers (3.1%) (p=0.047). Since officers are, on the average, 5 years older than enlisted participants, this occupation effect may be due to some confounding with age. There was a higher incidence rate for average lifetime residential latitude less than 37° N versus greater than or equal to 37° N latitude (p=0.008). Furthermore, there was a strong difference for different levels of the sun-reaction index (p<0.001), and the three skin-reaction-to-sun variables (p<0.001 for all). Participants who tended to burn most had a lower rate (1.4%) than those with a milder reaction (6.0%), and a similar rate to those who tended to tan (1.9%) (an unexpected finding). There was a significant relationship between the incidence rate of basal cell carcinoma and total pack-years of lifetime smoking (p=0.023 for verifieds). This effect may also be due to confounding with age rather than to a primary smoking effect (see Table H-5 of Appendix H). No significant association was found between the incidence rate of verified basal cell carcinoma and lifetime drink-years. No significant associations were found with ethnic group, skin color, eye color, and hair color. However, when the ethnic group categories were dichotomized as Celtic or English versus other ethnic groups, the association was marginally significant (p=0.093). Skin color was dichotomized as dark peach or light peach versus other colors, and the association was significant (Est. RR: 3.00, 95% C.I.: [1.08,8.33], p=0.024). Hair color was dichotomized as blond or red versus other colors. The association of hair color with basal cell carcinoma incidence was not significant (p=0.384). Furthermore, no significant relationship was found between basal cell carcinoma incidence and the composite carcinogen-exposure variable (p=0.523) or the grouped or individual carcinogens. The associations between the covariates and the incidence of verified plus suspected basal cell carcinomas paralleled those for the verified basal cell carcinomas only, except that the difference in rates among ethnic groups was significant (p=0.046), hair color was significant (p=0.040), and a marginally significant positive relationship was found with nonmedical x-ray exposure (p=0.084) and herbicide.exposure (p=0.072). The difference among occupation groups, however, was more significant (p=0.003). 10-18 �TABLE 10-6. Summary of Associations Between Incidence Rates of Basal Cell Carcinoma and Sun Exposure-Related Skin Malignancies and the Covariates, in the Baseline-Followup Interval for Combined Follovup Ranch Hand and Comparison Nonblack Participants Covariate Sun Exposure-Related Skin Malignancies Basal Cell Carcinoma Verified & Verified & Suspected Verified Suspected Verified p-Value p-Value p-Value p-Value Age 0.001 <0.001 0.004 <0.001 Occupation 0,047 0.003 NS* 0.006 Lifetime Cigarette Smoking 0.023 0.005 0.012 0.007 Lifetime Alcohol Consumption NS NS NS NS Ethnic Background NS 0.046 NS 0.036 Skin Color NS** NS NS NS** Hair Color NS 0.040 NS NS* Eye Color NS NS NS NS Reaction of Skin to Sun Exposure: (Q.I) After first 30 minutes of summer sun (Q.2) After >2 hours, after first exposure (Q.3) After repeated sun exposures Sun-Reaction Index (Composite) Residential History (Average Latitude) Exposure to Carcinogens/Groups of Carcinogens Set la Asbestos Non-medical X Rays Industrial Chemicals Herbicides Insecticides Degreasing Chemicals 0.001 <0.001 <0.001 <0.001 <0.001 0.027 0.001 0.016 <0.001 0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.008 0.004 0.011 0.003 NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS 10-19 NS NS* NS NS* NS NS �TABLE 10-6. (continued) Summary of Associations Between Incidence Rates of Basal Cell Carcinoma and Sun Exposure-Related Skin Malignancies and the Covariates, in the Baseline-Followup Interval for Combined Follovup Ranch Hand and Comparison Nonblack Participants Covariate Set 2° Anthracene Arsenic Benzene Benzidene Chromates Coal Tar Creosote Aminodiphenyl Chloromethyl Ether Mustard Gas Naphthylamine Cutting Oils Trichloroethylene Ultraviolet Light (not sun) Vinyl Chloride Composite Carcinogen Exposure NS: Sun Exposure-Related Skin Malignancies Basal Cell Carcinoma Verified & Verified & Verified Suspected Verified Suspected p-Value p-Value p-Value p-Value NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS Not significant (p<0.10). NS*: Borderline significant (0.05<p<0.010). **Not significant when five categories of skin color examined; however, when dichotomized, p=0.024 for verified basal cell carcinoma and p=0.036 for verified and suspected sun exposure-related skin malignancies. a Questionnaire data. b AFHS Form 2. 10-20 �As expected, the relationships between the incidence of verified sun exposure-related skin malignancies and the covariates were similar to those just described for basal cell carcinomas (Table 10-6 and Table H-4 of Appendix H). For verified conditions, there was a strong increase in incidence rate with age (p=0.004), total lifetime smoking (p=0.012), average lifetime residential latitude (p=0.011), the reaction-to-sun exposure variables (p<0.001 for all), and the sun-reaction index (p<0.001), with similar strong associations for the verified plus suspected conditions. The difference among occupation groups was marginally significant (p=0.077) for verified conditions; this difference was significant (p=0.006) for verified plus suspected sun exposure-related skin malignancies (officers 5.9%, enlisted flyers 4.2%, enlisted groundcrew 2.8%). There was no association with the composite carcinogen-exposure variable, either for verified (p=0.879) or for verified plus suspected conditions (p=0.608). Table 10-6 shows no significant association between the incidence rate of verified sun exposure-related skin malignancies and ethnic group, hair color, skin color, or eye color. When suspected conditions were included, the ethnic group association was significant (p=0.036), and the association with hair color became borderline significant (p=0,051). There were higher incidence rates among those of Celtic or English background as opposed to other ethnic backgrounds, and among participants with blond or red hair as opposed to other colors (see Table H-4 of Appendix H). As in the analysis of basal cell carcinomas, the ethnic group, hair color, and skin color categories were collapsed, resulting in (for verified conditions): p«0.054 for those of Celtic or English backgrounds versus other ethnic backgrounds (Est. RR: 2.04, 95% C.I.: [1.00,4.17]) and p=0.031 for skin color peach versus not-peach (Est. RR: 2.61, 95% C.I.: [1.04,6.58]), but no significant association with hair color grouped as blond or red versus other (p=0.268) was found. Adjusted Analyses Because of the obvious interrelatedness among the host factors of hair color, skin color, eye color, ethnic background, and reaction of skin to sun, and because a smaller set of covariates was required for the adjusted analyses, a "main-effects" statistical model of basal cell carcinoma with the following covariates was used: age, occupation, total pack-years, lifetime drinking, ethnic background (dichotomized), hair color (blond or red versus other), eye color, skin color (peach tones versus other), the three skin-reaction-to-sun variables, average lifetime residential latitude (less than 37° N versus greater than or equal to 37° N), and the composite carcinogen exposure variable. The results of this analysis are given in Appendix H, Table H-5. The results showed that ethnic background, hair color, and the 30-minute skin-reaction-to-sun variable, while individually associated with basal cell carcinoma incidence, are relatively less important than the other host factors, namely skin color, and the 2-hour and repeatedexposure skin-reaction-to-sun variables, and were thus not included in the adjusted analyses. Total drink-years and the composite carcinogen exposure variable were not significant and thus were not used in the adjusted analyses. A parallel analysis was conducted in which the composite sunreaction index replaced all three skin-reaction-to-sun variables, and it was found that this substitution could be made without altering the relative contributions of the other covariates. For further reduction of the number of covariates, pack-years of smoking, although of interest (p=0.096), was 10-21 �also omitted. Thus, a reduced set of covariates for further analysis of the group contrasts was identified as age, occupation, skin color, average lifetime residential latitude, and the sun-reaction index. The results of adjusted analyses of group contrasts in the incidence rate of basal cell carcinoma and sun exposure-related skin malignancies are presented in Table 10-7. Parallel results for Ranch Hands contrasted with the Original Comparisons are given in Appendix H, Table H-6. A significant group-by-occupation interaction was found for verified interval basal cell carcinoma (p=0.044). Significant covariates were age (p=0.003), average residential latitude (p=0.003) and the sun-reaction index (p<0.001). The interaction was due to a significant difference in rates for enlisted flyers but not for officers or enlisted groundcrew: Ranch Hand enlisted flyers had a significantly (p=0.019) greater incidence rate of basal cell carcinomas than the corresponding Comparisons, 5.4 percent versus 1.0 percent (Adj. RR: 6.50, 95% C.I.: [1.36,31.01]) (see Appendix H, Table H-7). There was a significant group-by-sun-reaction index interaction in the analysis of verified plus suspected basal cell carcinomas (p=0.024); this was in part attributable to the absence of Ranch Hands who reported burning easily. The group frequencies for the three levels of this variable (burn easily, intermediate reaction, tan easily) were: Ranch Hands 0 (0%), 17 (8.9%), and 19 (2.7%), respectively, and Comparisons 4 (5.2%), 15 (5.7%), and 28 (3.2%), respectively. The incidence rate for Ranch Hands who had a moderate reaction to sun was (nonsignificantly) greater than that of the Comparisons. The details of this interaction are given in Appendix H, Table H-7. A skin color-by-age interaction (p=0.044) and average latitude (p=0.003) made significant contributions to the model. Results of the analyses for Original Comparisons were nonsignificant for verified conditions, although a marginally significant group-by-sun reaction interaction was found (p=0.051). The results for verified plus suspected conditions revealed a significant group-by-sun reaction index interaction (p=0.007) (see Table H-6 of Appendix H). Ranch Hands who had a moderate skin reaction to sun revealed a significantly greater incidence rate of verified basal cell neoplasms than corresponding Original Comparisons (Adj. RR: 2.81, 95% C.I.i [1.05,7.55], p=0.040) (Table H-8). This finding was marginally significant with the inclusion of suspected carcinomas (Adj. RR: 2.38, 95% C.I.: [0.98,5.76], p=0.055). The adjusted relative risk for the incidence rate of verified sun exposure-related skin malignancies was 1.37 (95% C.I.: [0.83,2.28]) and was not significant (p=0.221) (Table 10-7). Age (p<0.001), the sun-reaction index (p<0.001), and average lifetime residential latitude (p=0.008) contributed to the adjustment. No group difference was apparent when suspected malignancies were included. The adjusted relative risk was 1.05 (95% C.I.: [0.68,1.62], p=0.825), and the significant covariates were a skin color-bysun-reaction index interaction (p=0.028), a skin color-by-age interaction (p=0.028), and a skin color-by-residential latitude interaction (p=0.041). 10-22 �TABLE 10-7. Adjusted Analyses of Nonblack Follbwup Participants for Malignant Skin Neoplasm Incidence During the Baseline-Followup Interval Adj. Relative Risk (95% C.I.) Variable Status Basal Cell Carcinoma Verified **** **** AGE (p=0.003) LAT (p=0.003) SUNREAC (p<0.001) GRP*OCC (p=0.044) Verified & Suspected **** **** LAT (p=0.003) GRP*SUNREAC (p=0.024) SKIN*AGE (p=0.044) Sun-Exposure Verified Malignant Skin Neoplasms Verified & Suspected p-Value Covariate Remarks* 1.37 (0.83,2.28) 0.221 AGE (p<0.001) SUNREAC (p<0.001) LAT (p=0.008) 1.05 (0.68,1.62) 0.825 SKIN*SUNREAC (p=0.028) SKIN*AGE (p=0.028) SKIN*LAT (p=0.041) ^Abbreviations; LAT: average lifetime residential latitude SUNREAC: sun reaction index GRP: group OCC: occupation SKIN: skin color ****Group-by-covariate interaction—adjusted relative risk, confidence interval, and p-value not presented. 10-23 �Analysis of the Ranch Hands versus Original Comparisons contrasts found a significant group-by-skin color interaction for verified sun exposurerelated malignancies (p=0.036), and a significant group-by-sun reaction index interaction (p=0.030), similar to that found for basal cell carcinoma, for the verified plus suspected malignant neoplasms (see Appendix H, Tables H-6 and H-8, for details). The group-by-skin color interaction was due to a lower incidence rate for nonpeach Ranch Hands than Original Comparisons (Adj. RR: 0.20, 95% C.I.: [0.02,1.80], p=0.150), but a higher incidence rate for peach toned Ranch Hands than Original Comparisons (Adj. RR: 1.70, 95% C.I.: [0.95,3.04], p=0.073). The group-by-sun reaction index interaction (verified and suspected) was again due to Ranch Hands who react moderately to the sun having a higher incidence rate than similar Original Comparisons (Adj. RR: 2.74, 95% C.I.: [1.14,6.63], p=0.025). Interval Systemic Neoplasms As shown in Table 10-1, eight Ranch Hands (0.8%) and seven Comparisons (0.5%) had verified malignant systemic neoplasms in the interval between the Baseline and followup examinations. When suspected malignant systemic neoplasms were included, the numbers were 12 Ranch Hands (1.2%) and 12 Comparisons ( . % . The proportions of malignancies among the systemic neoplasms 09) of all types (malignant, benign, uncertain) were similar in the two groups: 14.5 percent (8/55) for Ranch Hands and 11.5 percent (7/61) for Comparisons (p=0.783). Inclusion of suspected conditions did not change the conclusion from this contrast: 18.5 percent (12/65) Ranch Hands versus 15.0 percent (12/80) Comparisons (p=0.656). For the remainder of this section, only malignant (verified and suspected) systemic neoplasms occurring in the Baseline to followup interval are analyzed. These occurrences were distinct from those reported at Baseline. No new metastatic systemic neoplasms were reported in the interval. Interval Malignant Systemic Neoplasms Table 10-8 shows the sites of the new malignant neoplasms reported by the eight Ranch Hands and seven Comparisons. Classification of malignancies was based on I CD-9 with special coding for tumor type as well as site, thus avoiding problems of underreporting of STS. Six Ranch Hands and five Comparisons had suspected systemic neoplasms in this interval (Table 10-9), making a total of 12 in each group, since 2 Ranch Hands with verified systemic neoplasms also had suspected systemic neoplasms. The frequencies were too small for indepth analysis of individual sites. Table 10-8 shows that two Ranch Hands had malignant neoplasms of the oral cavity and pharynx versus no Comparisons, and three Comparisons but no Ranch Hands had malignant neoplasms of the colon. For all digestive system malignancies (esophagus plus colon), there were four occurrences among Comparisons but none among Ranch Hands. The analyses that follow are based on the combination of all interval malignant systemic neoplasms regardless of specific site, both verified and verified plus suspected. Table H-9 of Appendix H lists the malignancy sites for the eight Ranch Hands and the six Original Comparisons in the Baseline-followup interval. 10-24 �TABLE 10-8. Summary of Follovup Participants with Verified Malignant Systemic Neoplasms in Baseline-Follovup Interval by Group Group Site Ranch Hand Comparison Total Oral Cavity and Pharynx 2a'b 0 2 Thyroid Gland 0 1 1 Esophagus 0 lc 1 Bronchus and Lung 1 0 1 Colon 0 3d'8 3 Kidney and Bladder 2 1 3 Prostate 1 1 2 Testicles 1 0 1 Connective and Other Soft Tissue 1 0 1 8 7 15 Total a lncludes one Ranch Hand with separate malignancies of tongue and epiglottis and also malignant neoplasm of bone. b lncludes one Ranch Hand with separate malignant neoplasms of tongue and oropharynx and secondary malignant neoplasm of other site. c Also has malignant neoplasm of bone. d lncludes one Comparison with secondary malignant neoplasms of liver and bone and bone marrow. 8 Includes one Comparison with secondary malignant neoplasm of liver. 10-25 �TABLE 10-9. Summary of Follovup Participants with Suspected Malignant Systemic Neoplasms at Physical Examination by Group Group Site Bronchus and Lung Rectum 4 a,b 0 c Comparison Total 2 6 1 Ranch Hand 1 Liver l 0 1 Prostate 0 1 1 Lymphatic and Hematopoietic Tissue ld 0 1 Unspecified Site 0 1 1 6 5 11 Total "includes one Ranch Hand with a suspected maglignant neoplasm of either lung, mediastinum, esophagus, or ill-defined site within digestive organs and peritoneum. Includes one Ranch Hand with a suspected secondary malignant neoplasm of lung. c Not specified as primary or secondary. Suspected as either Hodgkins disease, leukemia, or lymphoma. 10-26 �There is no parallel table for suspected malignant systemic neoplasms since the five Comparisons with suspected conditions in Table 10-9 are Original Comparisons. Unadjusted Analyses As shown in Table 10-10, the unadjusted group contrast for all verified malignant systemic neoplasms was not significant (p=0.603), with an estimated relative risk of 1.46 (95% C.I.: [0.53,4.03]). When suspected malignant neoplasms were included with the verified malignancies, the estimated relative risk was 1.28 (95% C.I.: [0.57,2.85]), and was also not significant (p=0.680). A parallel unadjusted analysis for Ranch Hands versus Original Comparisons gave similar nonsignificant results (Appendix Table H-10). Covariates The covariates considered for the adjusted analysis of all interval malignant systemic neoplasms combined were age, race, occupation, smoking and drinking history, exposure to the groups of carcinogens, exposure to the individual carcinogens, and the composite carcinogen exposure variable as listed in Table 10-5. The categories used for age, pack-years, and drinkyears were the same. Age was used as a continuous variable in the adjusted analyses but was categorized for ease of presentation in the report. No Blacks had verified systemic neoplasms, but in contrast to the skin cancer analysis, Blacks were retained in the analysis. Covaciate Associations Table 10-11 summarizes the results of chi-square tests of association between the incidence rate of all malignant systemic neoplasms combined and the covariates considered for use in the adjusted analyses. Details of the covariate relationships are given in Appendix H, Table H-ll. There was a significant increase in the incidence rate of all verified interval malignant systemic neoplasms with increasing age (p<0.001) and a marginally significant difference among occupations (p-0.056). The incidence rates for officers, enlisted flyers,, and enlisted groundcrew were 1.2 percent, 0.5 percent, and 0.3 percent, respectively. There was a marginally significant association with total lifetime alcohol consumption (p=0.082). The test for differences in incidence rates among pack-year levels of smoking was not significant (p=0.220), although an increasing trend was apparent. Some of the occupation effect may be attributable to confounding with age. There was a significant negative association with insecticide exposure for verified malignant systemic neoplasms (p=0.014). Table H-ll of Appendix H shows that there were a few significant or marginally significant positive associations with individual carcinogens: e.g., with naphthylamine (p=0.050), benzidine (p=0.088), and coal tar (p=0.079). However, in many instances the self-reported exposure frequencies were very small. 10-27 �TAKE 10-10. Unadjusted Analyses of Followup Participants with Verified and Suspected Malignant Systemic Neoplasms in the Baseline-Followup Interval by Group Status Verified Verified & Suspected Statistic Number of Participants/% Total Neoplasms Number of Participants/% Total Neoplasms Group Ranch Hand Comparison 8 0.8% 7 0.5% 12 12 0 9 .% 23 16 p-Value 1.46 ( . 3 4 0 ) 05,.3 0.603 1.28 ( . 7 2 8 ) 05,.5 060 .8 10 12 1.2% Est. Relative Risk ( 5 C.I.) 9% 10-28 �TABLE 10-11. Summary of Associations Between Incidence Rates of All Malignant Systemic Neoplasms Combined and the Covariates in the Baseline-Followup Interval for Combined Followup Ranch Hand and Comparison Groups Covariate Age Verified p-Value <0.001 Verified & Suspected p-Value 0.001 Race NS NS Occupation NS* NS Lifetime Cigarette Smoking NS NS Lifetime Alcohol Consumption NS* NS Exposure to Carcinogens/Groups of Carcinogens: Set la Asbestos Non-medical X Rays Industrial Chemicals Herbicides Insecticides Degreasing Chemicals Set 2b Anthracene Arsenic Benzene Benzidene Chromates Coal Tar Creosote Aminodiphenyl Chloromethyl Ether Mustard Gas Naphthylamine Cutting Oils Trichloroethylene Ultraviolet Light (not sun) Vinyl Chloride Composite Carcinogen Exposure NS NS NS NS 0.014 NS NS 0.049 NS NS NS* NS NS NS NS NS* NS NS* NS NS NS NS 0.050 NS NS NS NS NS NS* NS NS NS NS NS NS* 0.023 NS* 0.019 NS NS NS NS NS NS NS*; Borderline significant (0.05<p<0.10). NS: Not significant (p>0.10) a Questionnaire data. b AFHS Form 2. 10-29 �The covariate associations for verified plus suspected malignant systemic neoplasms were similar to those for verified only. The association with occupation was no longer significant (p=0.193), and there was a significant positive association with nonmedical x-ray exposure (p=0.049). There were some significant and marginally significant positive associations with individual carcinogens: with naphthylamine (p=0.019), chloromethyl ether (p=0.023), arsenic (p=0.069), mustard gas (p=0.090), and aminodiphenyl (p=0.061) (see Appendix H, Table H-ll). The covariates used for the adjusted group contrast of the incidence rate of all malignant systemic neoplasms were race, age (continuous), occupation, and pack-years. Adjusted Analyses The adjusted relative risks for all verified and verified plus suspected malignant systemic neoplasms are presented in Table 10-12. For verified malignant systemic neoplasms, there was no significant difference between groups (Adj. RR: 1.51, 95% C.I.: [0.54,4.22], p=0.434). Age made a significant contribution to the adjustment (p<0.001). Parallel results for Ranch Hands contrasted with Original Comparisons are given in Table H-12 of Appendix H. A significant group-by-occupation interaction was found in the adjusted analysis of verified plus suspected malignant systemic neoplasms (p=0.027). This was due to significantly more cases of malignant systemic neoplasms among Ranch Hand enlisted flyers than among corresponding Comparisons (4/175 [2 verified, 2 suspected] versus 0/209, Fisher's exact test=0.042), whereas the incidence rate for officers was lower (but not significantly) for Ranch Hands than for the corresponding Comparisons, and equivalent for the enlisted groundcrew (see Table H-13 of Appendix H). Age (p<0.001) and a race-by-packyear interaction (p=0.035) made significant contributions to the adjustment. Comparable results were found for the contrast of Ranch Hands with the Original Comparisons (see Tables H-12 and H-14 of Appendix H). Lifetime (Baseline and Interval) Data from the Baseline and followup examinations were merged to obtain records of the lifetime history of neoplasm incidence for those followup participants who participated at Baseline. New participants provided lifetime information at the followup examination. Neoplasms prior to service in Southeast Asia were excluded from all analyses. All data from the Baseline study have been verified, but as described in the previous section, the status of some suspected interval neoplasms remains unclear, and thus both verified and verified plus suspected neoplasms are described and analyzed in this section. Table 10-13 shows that 21.3 percent (216/1,016) of Ranch Hands and 16.2 percent (209/1,293) of Comparisons had skin or systemic neoplasms of some type (malignant, benign, and uncertain). The group difference in incidence rates was significant (p=0.002), with an estimated relative risk of 1.40 (95% C.I.i [1.13,1.73]). When suspected neoplasms were included, the contrast was less marked (22.7% [231] of Ranch Hands versus 19.3% [249] of Comparisons) but still statistically significant (p=0.044), with an estimated relative risk of 1.23 (95% C.I;: [1.01,1.51]). 10-30 �TABLE 10-12. Adjusted Analyses of Followup Participants for the Incidence of All Malignant Systemic Neoplasms During the Baseline-Followup Interval Variable Malignant Systemic Neoplasms Adj. Relative Risk (95% C.I.) 1.51 (0.54,4.22) p-Value Covariate Remarks 0.434 AGE (p<0.001) **** GRP*OCC (p=0.027) AGE (p<0.001) RACE*PACKYR (p=0.035) (Verified) Malignant Systemic Neoplasms (Verified & Suspected) **** ****Group-by-covariate interaction—adjusted relative risk, confidence interval, and p-value not presented. 10-31 �TABLE 10-13. Unadjusted Analyses of Follovup Participants with lifetime Occurrence of Verified and Suspected Neoplasms by Group (Nonblacks and Blacks) Group* Banch Hand Site Skin Systemic All Neoplasm Behavior and Status Comparison Number** Percent Number** Percent Malignant Verified 66 Verified and Suspected 75 Benign Verified 84 Verified and Suspected 86 Uncertain Behavior and Unspecified Nature: Verified 1 Verified and Suspected 1 Any Skin Neoplasm* Verified 150 Verified and Suspected 159 Total** p-Value*** 6.5 7.4 66 85 5.1 6.6 132 160 0.175 048 .5 8.3 8.5 79 85 6.1 6.6 163 171 009 .4 0.093 0.1 0.1 1 1 0.1 0.1 2 2 099 .9 099 .9 14.8 15.7 140 165 10.8 12.8 290 324 0.005 0.053 17 21 1.7 2.1 17 22 1.3 1.7 34 43 0.491 .0.538 51 57 5.0 5.6 64 75 5.0 5.8 115 132 0.999 0.857 15 15 1.5 1.5 14 18 1.1 1.4 29 33 0.453 0.862 81 91 8.0 9.0 87 106 6.7 8.2 168 197 0.259 0.548 Malignant, Benign, Uncertain Behavior, Unspecified Nature Verified 216 Verified and Suspected 231 21.3 22.7 209 249 16.2 19.3 425 480 0.002 004 .4 Malignant Verified Verified and Suspected Benign Verified Verified and Suspected Uncertain Behavior and Unspecified Nature: Verified Verified and Suspected Any Systemic Neoplasm13 Verified Verified and Suspected *Sample sizes: 1,016 Ranch Hands, 1,293 Comparisons. **Number of participants. ***Fisher's exact test. ^Participant has one or more malignant, benign, or unspecified skin neoplasm. Participant has one or more malignant, benign, or unspecified systemic neoplasm. c Participant has one or more malignant or benign skin or systemic neoplasm. 10-32 �Table H-15 of Appendix H is parallel to Table 10-13 for Ranch Hands and Original Comparisons only, Lifetime Skin Neoplasms As seen in Table 10-13, 69.4 percent (150/216) of Ranch Hands with neoplasms had skin cancer; the corresponding percentage for Comparisons was 67.0 percent (140/209). The group difference in these proportions was not significant (p=*0.604). This contrast, when suspected neoplasms were included, was 68.8 percent (159/231) versus 66.3 percent (165/249), which again was not significant (p=0.560). The overall percentage of Black and nonblack Ranch Hands with verified lifetime skin neoplasms of any type was 14.8 percent (150/1,016), versus 10.8 percent (140/1,293) for Comparisons. No Black followup participants had ever had skin neoplasms, nor did any Baseline Black participants. The overall percentage of nonblack Ranch Hands with skin neoplasms of any type was 15.7 percent (150/956) and was significantly (p=0.006) greater than that of the Comparisons with 11.6 percent (140/1,210). The estimated relative risk was 1.42 95% C.I.: [1.11,1.82]). When both verified and suspected neoplasms were in the analysis, the contrast was marginally significant (p=0.060): Ranch Hands 16.6 percent (159/956) versus Comparisons with 13.6 percent (165/1,210) (Estimated RR: 1.26, 95% C.I.: [1.00,1.60]). For the remainder of this subsection, only malignant skin neoplasms are examined. Furthermore, the analysis was restricted to nonblacks. The dependent variables examined were the same as those of the previous section (basal cell carcinoma, melanoma, squamous cell carcinoma, all malignant skin neoplasms combined and sun exposure-related skin malignancies). Lifetime Malignant Skin Neoplasms Table 10-14 presents the unadjusted analyses of the frequencies of nonblack participants in each group with lifetime occurrences of basal cell carcinoma, squamous cell carcinoma, melanoma, all malignant skin neoplasms, and the sun exposure-related skin malignancies. For completeness, the total number of malignancies of each type is also given. Table H-16 of Appendix H presents parallel analyses for Ranch Hands and Original Comparisons. Unadjusted Analyses There was a higher relative frequency (5.5%) of Ranch Hands who had basal cell carcinomas than of Comparisons (4.1%), but the difference was not significant (p=0.128). The estimated relative risk was 1.36 (95% C.I.: [0.92,2.02]). With the inclusion of suspected basal cell carcinoma, the estimated relative risk was also not significant (p=0.579). Of the 53 Ranch Hands with verified basal cell carcinomas, 17 (32.1%) had 2 or more occurrences. The corresponding number for the Comparisons was 14/50 (28.0%). The group contrast of the percentages with multiple basal cell carcinomas versus no basal cell carcinomas was not significant (17/920 versus 14/1,174, p=0.274), nor was the corresponding contrast when suspected basal cell carcinomas were included (19/916 versus 16/1,159, p=0.234). 10-33 �TABLE 10-14. Unadjusted Analyses of Nonblack Followup Participants vith Lifetime Occurrence of Verified and Suspected Malignant Skin Neoplasms by Cell Type and Group Group* Cell Type Status Statistic** Ranch Hand Comparison Est. Relative Risk (95% C.I.) p-Value Verified Number/% Total Neoplasms 53 77 5.5% 50 76 4.1% 1.36 ( .92,2. 02) 0.128 0 Verified & Suspected Number/% Total Neoplasms 59 88 6.2% 67 99 5.5% 1.12 ( .78,1. 61) 0.579 0 Verified Number/% Total Neoplasms 4 6 0.4% 6 7 0.5% 0.84 ( .24,3. 00) 0.999 0 Verified & Suspected Number/% Total Neoplasms 4 6 0.4% 7 8 0.6% 0.72 ( .21,2. 47) 0.764 0 Verified Number/% Total Neoplasms 5 6 0.5% 5 6 0.4% 1.27 ( .37,4. 39) 0.757 0 Verified & Suspected Number/% Total Neoplasms 5 6 0.5% 8 10 0.7% 0.79 ( .26,2. 42) 0.784 0 All Malignant Verified Skin Neoplasms Verified & Suspected Number/% Total Neoplasms 66 100 6.9% 66 100 5.5% 1.29 ( .90,1.83) 0.175 0 Number/% Total Neoplasms 75 114 7.9% 85 129 7.0% 0 1.13 ( .82,1. 56) 0.508 Sun-Exposure Verified Related Malignant Neoplasms3 Verified & Suspected Number/% Total Neoplasms 59 87 6.2% 55 83 4.6% 1.38 ( . 0 ,95,2. 02) 0.100 Number/% Total Neoplasms 65 98 6.8% 74 111 6.1% 1.12 ( . 0 ,79,1.58) 0.537 Basal Cell Carcinoma Squamous Cell Carcinoma Melanoma *Number of participants—956 Ranch Hands and 1,210 Comparisons. **Number and percent of participants; total number of malignant neoplasms of specified cell type. a Basal cell carcinoma, melanoma, and malignant epithelial neoplasms NOS. �The frequencies of participants who had squamous cell carcinoma were very small: 4 Ranch Hands (0.4%) and 6 Comparisons ( . % . The estimated 05) relative risk was 0.84 (95% C.I.: [0.24,3.00]), and the contrast was far from significant (p=0.999). Inclusion of suspected squamous cell carcinoma did not change this finding. The frequency of Ranch Hands who had melanoma, 5 (0.5%), was slightly greater than that of the Comparisons, 5 (0.4%), but the contrast was not significant (p=0.757); the estimated relative risk was 1.27 (95% C.I.: [0.37,4.39]). Inclusion of suspected melanoma inverted the relative risk to 0.79, which was again not significant. This analysis had little power due to small frequencies. For sun exposure-related skin malignancies, there was a higher percentage of Ranch Hands than Comparisons (6.2% versus 4.6%), but the contrast was only of borderline significance (p=0.100); the estimated relative risk was 1.38 (95% C.I.: [0.95,2.02]). When suspected sun exposure-related skin malignancies were included, the group difference was not significant (p=0.537), with estimated relative risk 1.12 (95% C.I.: [0.79,1.58]). As in the previous section, adjusted analyses were only carried out for basal cell carcinoma and the sun exposure-related skin malignancies. Covariates The same covariates as for the interval analysis (Table 10-5) were considered for the adjusted analysis of the lifetime incidence rates of basal cell carcinoma and sun exposure-related skin malignancies: age, occupation, history of cigarette smoking and alcohol consumption, the same host factors and average latitude, and exposure to the same recognized carcinogens. The covariates used for the adjusted analyses were the same as in the interval analysis, namely age, occupation, sun reaction index, average lifetime residential latitude, and skin color. Covariate Associations Table 10-15 presents details of the associations between the incidence rate of basal cell carcinoma and the following covariates: age; occupation; pack-years of smoking, lifetime drink-years; ethnic background, hair color, skin color, eye color; skin-reaction-to-sun variables, sun-reaction index; average residential latitude, and exposure to individual carcinogens and groups of carcinogens. For the incidence of verified basal cell carcinoma, the same associations were found as in the interval analysis, namely, an increasing incidence rate with increasing age (p<0.001), a significant difference among occupations (p=0.017; officers 6.4%, enlisted flyers 4.2%, enlisted groundcrew 3.6%), and significant associations with average lifetime residential latitude (p=0.026), all the skinr-react ion-to-sun variables (p<0.001 for all), the sun-reaction index (p<0.001), and increasing total pack-years (p=0.024). There was evidence of a higher incidence rate of basal cell carcinomas among the heavy drinkers, although the test for the difference among drinking categories was not significant. 10-35 �TABLE 10-15. Association Between Lifetime Incidence of Basal Cell Carcinoma and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Number* Percent p-Value 21 75 7 2.4 <0.001 6.3 8.1 24 91 11 2.7 7.6 12.6 <0.001 Officer Enlisted Flyer Enlisted Groundcrew 850 360 956 54 15 34 6.4 4.2 3.6 0.017 68 18 40 8.0 5.0 4.2 0.002 Total Lifetime Smoking (Pack- Years) 0 >0-20 >20-40 >40 616 998 391 157 32 36 21 14 5.2 3.6 5.4 8.9 0.024 37 43 31 15 6.0 4.3 7.9 9.6 0.010 Total Lifetime Alcohol Consumption (Drink- Years) 0 >0-5 >5-30 >30-100 >100 141 717 655 479 104 7 37 29 19 8 5.0 5.2 4.4 4.0 7.7 0.548 8 43 34 30 8 5.7 6.0 5.2 6.3 7.7 0.855 Ethnic Background* A 1,582 424 63 42 85 16 ,1 0 5.4 3.8 1.6 0.0 0.132 107 16 1 0 6.8 3.8 1.6 0.0 0.016 B C D Skin Color Dark Medium Pale Dark Peach Pale Peach 1 73 308 1,262 520 0 2 9 69 23 0.0 2.7 2.9 5.5 4.4 0.339 0 2 11 82 31 0.0 2.7 3.6 6.5 6.0 0.263 Born XL942 Born 1923-41 Born <1922 Occupation 03 Total Participants Number* Percent p-Value 882 1,197 87 Age o Verified and Suspected �TABLE 10-15. (continued) Association Between Lifetime Incidence of Basal Cell Carcinoma and the Covariates for Combined Followup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category -Eye Color Burns Painfully Burns Becomes Red No Reaction Skin Reaction Freckles, No Tan After Repeated Tans Mildly Exposure to Sun Tans Moderately Tans Deep Brown . 4.7 6.4 2.5 5.4 4.2 0.338 35 30 6 7 48 5.4 6.6 5.0 7.5 5.7 0.853 20 42 32 2 7 4.6 4.1 5.7 12.5 6.5 0.278 24 53 38 3 8 5.5 5.1 6.8 18.8 7.4 0.120 43 60 3.8 5.9 0.026 51 75 247 429 805 681 Skin Reaction Burns to First 30 Min. Usually Burns of Sun Exposure Burns Mildly Rarely Burns 30 29 3 5 36 1,136 1,022 >37° Residential History (Average <37° Latitude) Skin Reaction to >2 Hrs of Sun After First Exposure Number* Percent p-Value 439 1,038 563 16 108 Black Dark Brown Light Brown Red Blond o Number* Percent p-Value 645 455 119 93 850 Brown Hazel Green Grey Blue Hair Color Total Participants Verified and Suspected 21 36 29 16 8.5 <0.001 8.4 3.6 2.4 25 44 32 24 10.1 10.3 4.0 3.5 <0.001 120 338 1,043 663 9 31 42 21 7.5 <0.001 9.2 4.0 3.2 11 33 54 28 9.2 9.8 5.2 4.2 0.001 45 314 1,019 783 4 31 37 30 8.9 9.9 3.6 3.8 5 36 47 37 11.1 11.5 4.6 4.7 <0.001 <0.001 4.5 ' 7.3 0.006 �TABLE 10-15. (continued) Association Between Lifetime Incidence of Basal Cell Carcinoma and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified Covariate Covariate Category Sun Reaction Index Tends to Burn Mild Reaction Tends to Tan Exposures to Carcinogens Asbestos Nonmedical X Rays Total Participants Verified and Suspected Number* Percent p-Value 145 454 1,562 10 41 51 6.9 <0.001 9.0 3.3 Yes No 458 1,708 18 85 3.9 5.0 Yes No 494 1,672 29 74 Industrial Chemicals Yes No 1,126 1,040 Herbicides Yes No 1,261 905 Insecticides Yes No 1,313 853 Degreasing Chemicals Yes No Yes No Number* Percent p-Value 12 46 67 8.3 10.1 4.3 <0.001 0.389 25 101 5.5 5.9 0.822 5.9 4.4 0.187 37 89 7.5 5.3 0.080 49 54 4.4 5.2 0.365 60 66 5.3 6.4 0.314 65 38 5.2 4.2 0.357 81 45 6.4 5.0 0.164 69 34 5.3 4.0 0.181 82 44 6.3 5.2 0.303 1,261 905 60 43 4.8 4.8 0.999 72 54 5.7 6.0 0.852 489 1,653 21 80 4.3 4.8 0.716 24 100 4.9 6.1 0.379 CO 00 Composite Carcinogen Exposure ' �TABLE 10-15. (continued) Association Between Lifetime Incidence of Basal Cell Carcinoma and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified Verified and Suspected o i u> Total Participants Anthracene Yes No 2 2,161 0 103 0.0 4.8 0.999 0 126 0.0 5.8 0.999 Yes No 41 2,124 4 98 9.8 4.6 0.124 5 120 12.2 5.7 0.084 Benzene Exposure to Individual Carcinogens Covariate Category Arsenic Covariate Yes No 74 2,091 6 97 8.1 4.6 0.162 7 119 9.5 5.7 0.198 Benzidine Yes No 11 2,154 1 102 9.1 4.7 0.416 1 125 9.1 5.8 0.484 Chroma tes Yes No 84 2,079 4 97 4.8 4.7 0.999 5 119 6.0 5.7 0.812 Coal Tar Yes No 68 2,097 2 101 2.9 4.8 0.770 3 123 4.4 5.9 0.795 Creosote Yes No 159 2,007 9 94 5.7 4.7 0.560 10 116 6.3 5.8 0.726 Aminodiphenyl Yes No 6 2,157 0 102 0.0 4.7 0.999 0 125 0.0 5.8 0.999 Chloromethyl Ether Yes No 21 2,142 2 101 9.5 4.7 0.264 2 124 9.5 5.8 0.348 Mustard Gas Yes No 6 2,159 0 103 0.0 4.8 0.999 0 126 0.0 5.8 0.999 Number* Percent p-Value Number* Percent p-Value �TABLE 10-15. (continued) Association Between Lifetime Incidence of Basal Cell Carcinoma and the Covariates for Combined Follovup Ranch Hand and Comparison Nonblack Participants Verified and Suspected Verified Total Participants Naphthylamine Yes No 52 2,112 3 99 5.8 4.7 0.734 4 121 7.7 5.7 0.540 Cutting Oils Yes No 226 1,939 12 91 5.3 4.7 0.622 15 111 6.6 5.7 0.549 Yes No 184 1,979 5 98 2.7 5.0 0.207 7 119 3.8 6.0 0.253 Ultraviolet Light Yes No 44 2,120 5 98 11.4 4.6 0.055 5 121 11.4 5.7 0.179 Vinyl Chloride Exposure to Individual Carcinogens (continued) Covariate Category Trichloroethylene Covariate 31 Yes No 2,133 0 103 0.0 4.8 0.399 1 125 3.2 5.9 0.999 o Number* Percent p-Value *Number of participants with basal cell carcinomas * a Ethnic Background: A B C D - English, Welsh, Scottish, Irish Scandinavian, German, Polish, Russian, Other Slavic, Jewish, French Spanish, Italian, Greek. Mexican, American Indian, Asian • Number* Percent p-Value �There was a significant (p<0.001) association with the sun-reaction index. Participants with the most sensitive skin had a somewhat lower rate (6.9%) of verified basal cell carcinoma lifetime than the participants in the next most sensitive category (9.0%), although the difference was not as marked as in the Baseline-followup interval. However, the rate for those who tanned easily was much lower (3.3%) than for those who did not. A marginally significant positive association was found with self-reported exposure to non-sun ultraviolet light (p=0.055). The results were similar for the verified plus suspected basal cell carcinomas. There was a significant (p=0.016) difference among ethnic backgrounds, with participants with Celtic or English backgrounds having higher incidence rates than those with other backgrounds. Further, there were marginally significant positive associations in incidence rates with nonmedical x-ray exposure (p=0.080) and arsenic (p=0.084), a recognized skin carcinogen, but the association with ultraviolet light was not significant. The details of associations between the incidence rates of verified and suspected sun exposure-related skin malignancies and the covariates are given in Appendix H, Table H-17. The significant covariates for verified conditions were age (p<0.001), occupation (p=»0.009), total pack-years (p=0.021), average latitude (p=0.026), and sun-reaction index (p<0.001). The same pattern held for verified plus suspected sun exposure-related skin malignancies. There was a marginally significant positive association with ultraviolet light exposure (p=0.078) for the verified conditions only, and with herbicide exposure (p=0.076) for the verified plus suspected conditions. The covariates chosen for the adjusted analysis were age, occupation, skin color, average lifetime residential latitude and the sun-reaction index. Adjusted Analysis The results of adjusted analyses of group contrasts for lifetime skin malignancies are given in Table 10-16, There was significant evidence of a higher incidence rate of verified basal cell carcinoma in the Ranch Hand group as contrasted with the Comparisons (p=0.035). The adjusted relative risk was 1.56 (95% C.I.: [1.03,2.37]). A sun-reaction index-by-average latitude interaction (p=0.026), a skin color-by-sun-reaction index interaction (p<0.001), and an occupation-by-age interaction (p=0.047) made significant contributions to the model. The adjustment by average residential latitude, which is greater for Ranch Hands than Comparisons, contributed to a higher relative risk resulting from the adjusted analysis than from the unadjusted (see Table 10-14). When suspected basal cell carcinomas were . included in the analysis, a significant group-by-sun-reaction index interaction (p=0.040) was found. Age (p<0.001), a skin color-by-average residential latitude (p=0.024), and a skin color-by-sun-reaction index interaction (p<0.001) made significant contributions to the adjustment. This was due to a significant increase in basal cell carcinoma incidence for Ranch Hands with an intermediate skin reaction to sun over similar Comparisons (Adj. RR: 1.97, 95% C.I.: [1.04,3.73], p=0.038) (Appendix H, Table H-18). Similar results were found in the contrast of Ranch Hand versus Original Comparisons (Table H-19). Namely, for verified basal cell carcinoma, and for verified plus suspected basal cell carcinomas, significant group-by-sunreaction index interactions were found (p=0.010 and p=0.003, respectively [see Table H-20 for additional details on the interactions]). 10-41 �TABLE 10-16. Adjusted Analyses of Nonblack Followup Participants for Lifetime Malignant Skin Neoplasm Incidence Variable Status Adj. Relative Risk (95% C.I.) Basal Cell Carcinoma Verified 1.56 (1.03,2.37) 0.035 SKIN*SUNREAC (p<0.001) OCC*AGE (p=0.047) SUNREAC*LAT (p=0.026) **** **** AGE (p<0.001) GRP*SUNREAC (p=0.040) SKIN*LAT (p=0.024) SKIN*SUNREAC (P<0.001) 1.54 (1.04,2.29) 0.030 AGE (p<0.001) SKIN*LAT (p=0.016) SKIN*SUNREAC (p<0.001) 1.23 (0.86,1.77) 0.252 AGE (p<0.001) SKIN*LAT (p=0.013) SKIN*SUNREAC (p<0.001) Verified & Suspected Malignant Verified Sun-Exposure Skin Neoplasms Verified & Suspected p-Value Covariate Remarks ****Group-by-covariate interaction—adjusted relative risk, confidence interval, and p-value not presented. 10-42 �As shown in Table 10-16, there was a significantly higher incidence rate of sun exposure-related skin malignancies among Ranch Hands as contrasted with Comparisons (Adj. RR: 1.54, 95% C.I.: [1.04,2.29], p=0.030). Significant contributions were noted for age (p<0.001), a skin color-by-sunreaction index interaction (p<0.001), and an average latitude-by-skin color interaction (p=0.016). When suspected sun exposure-related skin malignancies were included in the analysis, the adjusted relative risk became 1.23 (95% C.I.: [0.86,1.77]) and was no longer significant (p=0.252). Age (p<0.001), a skin color-by-sun-reaction index interaction (p<0.001), and average latitude-by-skin color interaction (p=0.013) contributed significantly to the adjustment. When Ranch Hands were contrasted to Original Comparisons, significant group-by-sun reaction index interactions were found for verified, and verified plus suspected, sun-exposure related skin neoplasms (p=0.045,p=0.016, respectively). These interactions were due to significant relative risks for those participants with intermediate reactions of skin to sun, as was also found for basal cell carcinomas only (see Appendix Tables H-19 and H-20 for details). Lifetime Systemic Neoplasms Table 10-13 shows that 81 (8.0%) Ranch Hands and 87 (6.7%) Comparisons had a verified history of systemic neoplasms of any type (malignant, benign, or uncertain). The estimated relative risk was 1.20 (95% C.I.: [0.88,1.65]), and was not significant (p=0.259). With the inclusion of suspected systemic neoplasms, the frequencies were 9.0 percent (91/1,016) for Ranch Hands and 8.2 percent (106/1,293) for Comparisons, with an estimated relative risk of 1.10 (95% C.I.: [0.82,1.48]), and the contrast was also not significant (p=0.548). For Ranch Hands with systemic neoplasms of any type, the percentage with malignant neoplasms was 21.0 percent (17/81) and the corresponding rate for Comparisons was 19.5 percent (17/87), a nonsignificant group difference (p=0.849). Including suspected systemic malignancies, these frequencies were 23.1 percent (21/91) for Ranch Hands and 20.8 percent (22/106) for Comparisons. Again, the group difference was not significant (p=0.731). For the remainder of this section, only malignant systemic neoplasms are discussed. Lifetime Malignant Systemic Neoplasms Table 10-17 presents the frequencies of verified lifetime malignant systemic neoplasms by site. Three Ranch Hands versus no Comparisons had malignant neoplasms of the oral cavity and pharynx; these occurred at ages 45, 52, and 57. The group difference in incidence rate was marginally significant (p=0.085). No Ranch Hands but 5 Comparisons had malignant neoplasms of the colon; the group difference in incidence rate was also marginally significant (p=0.072). Three Ranch Hands but no Comparisons had testicular malignancies, but the.group difference in incidence rates was only marginally significant (p=0.085). These occurred at ages 35, 38, and 54. The suspected malignant neoplasms are listed in Table 10-9. Table H-21 of Appendix H gives a list of verified lifetime malignant systemic neoplasms for Ranch Hands and Original Comparisons. 10-43 �TABLE 10-17. Summary of Follovup Participants With Lifetime Incidence of Verified Malignant Systemic Neoplasms by Group Group Site Eye Oral Cavity and Pharynx Ranch Hand Comparison Total 1 0 1 0 3 3 a,b Larynx 0 1 1 Thyroid Gland 0 2 2 Esophagus 0 lc 1 Bronchus and Lung 2 0 1 Colon 0 Kidney and Bladder 4 3 7 Prostate 2 2 4 Testicles 3 0 3 Connective and Other Soft Tissue 1 1 2 Hodgkin's Disease 0 1 1 Ill-Defined Sites 1£ lg 2 Total 5d.e 17 17 5 34 a Includes one Ranch Hand with separate malignancies of tongue and epiglottis and also malignant neoplasm of bone. Includes one Ranch Hand with separate malignant neoplasms of tongue and oropharynx and secondary malignant neoplasm of other site. °Also has malignant neoplasm of bone. d lncudes one Comparison with secondary malignant neoplasms of liver and bone and bone marrow. 8 Includes one Comparison with secondary malignant neoplasm of liver. £ Malignant neoplasm of thorax. Malignant neoplasm of face, head, or neck. 9 10-44 �One Ranch Hand and one Comparison had neoplasms of connective and other soft tissue. The Comparison had a fibrosarcoma at age 28 (reported at Baseline) and the Ranch Hand participant had malignant fibrous histiocytoma at age 63 (reported at followup). Both of these conditions are classified as soft tissue sarcoma. Since soft tissue sarcoma and malignant neoplasms of the lymphatic system are of concern in this study, the occurrences of these malignancies are shown by group in Table 10-18. The occurrences of these four malignancies are too small to support further statistical analysis. TABLE 10-18. Summary of Followup Participants with Lifetime Soft Tissue Sarcoma, Leukemia or Lymphoma by Group Group Site Ranch Hand Comparison Verified Soft Tissue Sarcoma 1 1 Verified Hodgkin's Disease 0 1 Suspected Leukemia, Hodgkin's Disease, or non-Hodgkin's Lymphoma 1 0 Unadjusted Analysis Table 10-19 shows the results of unadjusted analyses of the frequencies of participants in each group with verified or verified plus suspected malignant systemic neoplasms combined. The estimated relative risk for all malignant systemic neoplasms was 1.28 (95% C.I.: 0.65,2.51) and was not significant (p=0.491). With the inclusion of suspected malignant, neoplasms, the estimated relative risk was 1.22 (952 C.I.: 0.67,2.23) and was also not significant (p=0.538). Similar nonsignificant results were found for Ranch Hands contrasted with Original Comparisons (see Table H-22 of Appendix H). Covariates The same covariates used for the interval history of malignant systemic neoplasms were used for the adjusted analysis of lifetime malignant systemic neoplasms, namely, age, race, occupation, history of cigarette smoking and alcohol consumption, and exposure to carcinogens. Total smoking and alcohol consumption were estimated up to the followup examination, and may be different if estimated only up to the year of diagnosis of a neoplasm (if any). Further, age at followup rather than age at diagnosis was used in the analysis. 10-45 �TAKE 10-19. Unadjusted Analyses of lifetime Incidence Bates of ALL Malignant Systemic Neoplasms Combined, by Group Group Status Statistic Verified Number of Participants/% Total Neoplasms Verified & Suspected Participants/% Total Neoplasms Ranch Hand Comparison Est. Relative Risk (95%C.I.) p-Value 17 1.7* 17 1.3* 1.28 (0.65, 2.51) 0 4 1 .9 25 22 Number of 21 2.1% 36 22. 1 7 .* 27 1.22 ( . 7 2 2 ) 06,.3 0.538 Covariate Associations Associations between the incidence rate of all malignant systemic neoplasms combined and the covariates are presented in Table 10-20. For verified malignant systemic neoplasms, strong associations were found with increasing age (p<0.001) and occupation (officers 2.3%, enlisted flyers 1.3%, and enlisted groundcrew 0.9%, p=0.028). These same associations were also found for verified plus suspected systemic malignancies. The association with smoking history was not significant, either for verified or for verified plus suspected malignancies. The incidence rate of all malignant systemic neoplasms increased marginally significantly (p=0.073) with increasing levels of total lifetime alcohol consumption. For verified plus suspected malignancies, the difference among drink-year categories was- also marginally significant (p=0.080). No significant association was found with the composite carcinogen exposure variable. A significant association was found between the incidence of verified malignant systemic neoplasms and naphthylamine (p=0.048). There was a significant positive association between the verified plus suspected conditions and naphthylamine (p=0.019), and a marginally significant association with chloromethyl ether (p=0.067). The covariates used for the adjusted analysis of the incidence of malignant systemic neoplasms were race, age (continuous), occupation, pack-years, drink-years, and the composite carcinogen-exposure variable, Adjusted Analysis Table 10-21 shows that, in the adjusted analysis of the group contrast in incidence of all systemic malignancies combined, there was a significant group-by-occupation interaction (p=0.023). This was due to a difference in rates for the enlisted flyers, 5.Ranch Hands versus 0 Comparisons (unadjusted p-value=0.019), whereas the incidence rates for officers and enlisted groundcrew did not differ significantly between groups (p=0.698 and 0.922, respectively) (Table H-23). Age made a significant'contribution to the adjustment. When suspected systemic malignancies were combined with the verified systemic malignancies, a group-by-occupation interaction (p=0.002) 10-46 �TABLE 1 - 0 02. Association Between Lifetime Incidence of All Malignant Systemic Neoplasms Combined and the Covariates for Combined Followup Ranch Hand and Comparison Participants Verified Total Participants Number* Percent Verified and Suspected Covariate Category p-Value Number* Percent p-Value Age Born >1942 Born 1923-41 Born <1922 961 1,261 87 4 24 6 0.4 1.9 6.9 <0.001 7 30 6 0.7 2.4 6.9 <0.001 Race . Nonblack Black 2,166 143 34 0 1.6 0.0 0.267 42 1 1.9 0.7 0.517 Occupation Officer Enlisted Flyer Enlisted Groundcrew 864 387 1,058 20 5 9 2.3 1.3 0.9 0.028 23 7 13 2.7 1.8 1.2 0.069 Total Lifetime Smoking (Pack- Years) 0 >0-20 >20-40 >40 658 1,081 406 158 6 15 9 4 0.9 1.4 2.2 2.5 0.237 8 20 11 4 1.2 1.9 2.7 2.5 0.324 Total Lifetime Alcohol Consumption (Drink- Years) 0 >0-5 >5-30 >30-100 >100 151 760 703 508 108 1 7 8 11 4 0.7 0.9 1.1 2.2 3.7 0.073 2 10 10 13 5 1.3 1.3 1.4 2.6 4.6 0.080 �TABLE 10-20. (continued) Association Between Lifetime Incidence of All Malignant Systemic Neoplasms Combined and the Covariates for Combined Follovup Ranch Hand and Comparison Participants Verified Total Participants Number* Percent Verified and Suspected p-Value Covariate Exposures to Carcinogens Asbestos 499 1,810 Yes No 5 29 1.0 1.6 0.405 7 36 1.4 2.0 0.459 Nonmedical X Rays 541 1,768 Yes No 9 25 1.7 1.4 0.684 14 29 2.6 1.6 0.150 Industrial Chemicals o Category 1,199 1,110 Yes No 14 20 1.2 1.8 0.229 20 23 1.7 2.1 0.539 Herbicides 1,339 970 Yes No 18 16 1.3 1.7 0.601 23 20 1.7 2.1 0.538 Insecticides 1,389 920 Yes No 17 17 1.2 1.9 0.223 23 20 1.7 2.2 0.432 Degreasing Chemicals 1,343 966 Yes No 18 16 1.3 1.7 0.600 26 17 1.9 1.8 0.876 519 1,762 Yes No 7 27 1.4 1.5 0.999 8 34 1.5 1.9 0.711 00 Composite Carcinogen Exposure Number* Percent p-Value �TABLE 10-20. (continued) Association Between Lifetime Incidence of All Malignant Systemic Neoplasms Combined and the Covariates for Combined Followup Ranch Band and Comparison Participants Verified and Suspected Verified Total Participants Number* Percent p-Value Number* Percent p-Value Covariate Exposure to Individual Carcinogens Anthracene 2 2,303 Yes No 0 34 0.0 1.5 0.999 0 43 0.0 1.9 0.999 Arsenic 42 2,266 Yes No 0 34 0.0 1.5 0.999 2 41 4.8 1.8 0.183 Benzene o i Category 83 2,225 Yes No 2 32 2.4 1.4 0.348 2 41 2.4 1.8 0.666 Benzidine 14 2,293 Yes No 1 33 7.1 1.4 0.188 1 41 7.1 1.8 0.227 Chroma tes 88 2,218 Yes No 2 32 2.3 1.4 0.375 2 41 2.3 1.9 0.679 Coal Tar 73 2,235 Yes No 2 32 2.7 1.4 0.292 2 41 2.7 1.8 0.397 Creosote 164 2,145 Yes No 2 32 1.2 1.5 0.999 4 39 2.4 1.8 0.543 Aminodiphenyl 6 2,300 Yes No 0 34 0.0 1.5 0.999 1 42 16.7 1.8 0.107 Chloromethyl Ether 23 2,282 Yes No 1 33 4.4 1.5 0.291 2 41 8.7 1.8 0.067 Mustard Gas 9 2,299 Yes No 0 34 0.0 1.5 0.999 1 42 11.1 1.8 0.156 �TABLE 10-20. (continued) Association Between Lifetime Incidence of All Malignant Systemic Neoplasms Combined and the Covariates for Combined Follovup Ranch Hand and Comparison Participants Verified Total Participants Covariate Category Exposure to Individual Carcinogens (continued) Naphthylamine 56 2,251 Yes No 3 31 5.4 1.4 0.048 4 39 7.1 1.7 0.019 Cutting Oils 243 2,065 Yes No 5 29 2.1 1.4 0.396 7 36 2.9 1.7 0.209 Trichloroethylene o I iji o Number* 200 2,106 Yes No 5 29 2.5 1.4 0.211 6 37 3.0 1.8 0.264 Ultraviolet Light 51 2,256 Yes No 1 33 2.0 1.5 0.535 1 42 2.0 1.9 0.621 Vinyl Chloride 33 2,273 Yes No 0 34 0.0 1.5 0.999 1 42 3.0 1.9 0.465 *Number of participants with malignant systemic neoplasms. Percent Verified and Suspected p-Value Number* Percent p-Value �TABLE 10-21. Adjusted Analyses for Lifetime Incidence of All Malignant Systemic Neoplasms Combined Variable Adj. Relative Risk (95% C.I.) p-Value Covariate Remarks Systemic Malignancies (Verified) **** **** GRP*OCC (p=0.023) AGE (p<0.001) Systemic Malignancies (Verified & Suspected) **** **** GRP*OCC (p=0.002) AGE (p<0.001) RACE*PACKYR (p=0.032) ****Group-by-covariate interaction—adjusted relative risk, confidence interval, and p-value not presented. was also found; this was also due to the high rates for the Ranch Hand enlisted flyers. Comparison of Baseline, Interval, and Lifetime Results Table 10-22 compares the unadjusted and adjusted contrasts from the Baseline report with those from the Baseline-followup interval and the whole post-SEA period, for the incidence of all verified malignant skin neoplasms combined, verified basal cell carcinomas, and all verified malignant systemic neoplasms combined. There were, of course, differences in the Baseline and followup cohorts, but there was a sufficiently large overlap to make such a comparative tabulation useful. Malignant Skin Neoplasms The significant relative risks for all malignant skin neoplasms seen at Baseline were not evident for the Baseline-followup interval. However, for lifetime basal cell carcinoma, a significant adjusted group contrast was found (p=0.035). The difference in the incidence rates of all skin neoplasms and in basal cell carcinomas only between the Ranch Hands and the Comparisons appears to have decreased over time, as evidenced by the fact that the interval estimated and adjusted relative risks were closer to 1 than those for the lifetime, i.e., interval plus Baseline period. Malignant Systemic Neoplasms The unadjusted group contrasts in incidence rates of all malignant systemic neoplasms combined were not significant for Baseline, for the Baseline-followup interval, or for lifetime (Baseline plus interval), nor was the adjusted group contrast for the Baseline-followup interval. The 10-51 �TABLE 10-22. Uhadjusted and Adjusted Analyses of the Incidence of ML Verified Malignant Skin and Systemic Neoplasms and Basal Cell Carcinoma: Baseline, Baseline-Followup Interval, and lifetime Occurrence Site Statistic All Malignant Number of Participants Skin with Neoplasms/Percent:0 Ranch Hand Neoplasms Comparison Baseline-Followup Interval Baseline* 35 25 Est. RR/p-Value 3.3% 2.0% (.7d 00) 1.62 Adj. RR/p-Value Basal Cell Carcinoma 31 21 Est. RR/p-Value Est. RR/p-Value Adj. RR/p-Value * 29 30 * 3.0% 2.5% 66 66 1.29 —* 53 50 6.9% 5.4% (.7) 0159 —* 5.5% 4.1% (.4) 007d 1.71 1.23 (.2) 0499 1.36 (.2) 0189 * **** **** 1.56 (.3) 005 8 7 0.8% 0.5% 17 17 * Adj. RR/p-Value All Malignant Number of Participants Systemic with Neoplasms/Percent :f Ranch Hand Neoplasms Comparison 3.0% 1.7% 3.9% 3.3% (.8) 0466 1.18 * —* Number of Participants with Neoplasms/Percent:0 Ranch Hand Comparison 37 40 Lifetime Occurrence 1.2% 0.9% 13 11 (.6d 04) 1.35 1.46 (.0) 0639 * 1.51 (.3) 044 * 1.28 **** 1.7% 1.3% (.9) 0419 **** 'Analysis not done a Baseline participants: 1 0 5 Ranch Hands, 1,224 Comparisons. ,4 b Followup participants: 1 0 6 Ranch Hands, 1,293 Comparisons. ,1 c Nonblacks only for followup participants ( 5 Ranch Hands, 1,210 Comparisons), both nonblacks and 96 Blacks for Baseline participants. d Chi-square test. "Fisher's exact test. f All participants. ****Grc>up-by-covariate interaction. 10-52 �estimated lifetime relative risk appears closer to 1 than for the two intervals separately, but the small number of occurrences and intervening mortality preclude more definitive statements. Baseline Participants This brief section summarizes the mortality and malignant neoplasm history of the fully compliant Baseline participants in the interval up to the followup examination. Mortality information up through the end of 1985 was considered. This discussion is directed to the question of whether competing mortality affected the preceding analysis of incident cancers among living participants. • Of the 1,045 Ranch Hands and 1,224 Comparisons who were fully compliant at Baseline, 971 Ranch Hands and 1,139 Comparisons returned to the followup examination. Table 10-23 presents the numbers of Baseline participants according to whether they completed the followup examination and whether they were alive at the end of 1985. TABLE 10-23. Fully Compliant Baseline Participants by Status at Followup Examination and Group Participated in Followup Examination Group Status Ranch Hand Comparison Total 2 1,137 5 2,105 Y e s Dead" Alive 3 968 No Dead Alive 9 65 15 70 24 135 1,045 1,224 2,269 Total a Died in 1985, but subsequent to participation in the examination. For the participants who did not return for the followup examination, Table 10-24 shows that 2 of the 9 deaths among Ranch Hands were due to malignant neoplasms, compared with 5 of the 15 deaths among the Comparisons. One Ranch Hand who died had a malignant skin neoplasm, but this was not the primary cause of death. Among the 65 Ranch Hands who did not return for the followup examination, 5 had verified malignant neoplasms at Baseline, including 1 systemic neoplasm (of the kidney), as contrasted with 2 among 70 Comparisons .who had verified malignant (skin) neoplasms. Thus, among the 74 Ranch Hands not returning for followup, there were 8 with incident or fatal neoplasms, as compared to 7 of 85 Comparisons; the group difference was not significant (p=0.788). 10-53 �TABLE 10-24. Fully Compliant Baseline Participants Who Did Not Participate in Followup Examination by Status and Group Group Status Ranch Hand Comparison Total Dead—Primary Cause of Death: Malignant Neoplasm 2a Other Causes 7C 5 10 7 17 Lost to Followup: Verified Malignant Neoplasm at Baseline No Malignant Neoplasm at Baseline 5d 60 2e 68 7 128 "Both with lung cancer. b Three with lung cancer, one with malignant neoplasm of intestine (location unspecified), one with malignant neoplasm of an ill-defined site (face, head, or neck). c lncludes one Ranch Hand with malignant skin neoplasm. Four with malignant skin neoplasms, one with malignant systemic neoplasm (kidney). 8 Two with malignant skin neoplasms. 10-54 �For the participants who did return for the followup examination, Table 10-25 gives the frequencies and percentages of the respective group totals according to neoplasm status at Baseline and at followup. Analysis showed that there was no significant group difference (p=0.115) in the pattern of neoplasm incidence at Baseline and/or at followup. The results of this section show approximate equivalence between the groups for the disease of cancer (fatal or nonfatal) since Baseline, and in the proportions of participants with malignancies at Baseline, followup, or both. EXPOSURE INDEX ANALYSES Unadjusted and adjusted exposure index analyses were conducted within each occupational cohort of the Ranch Hand group (see Chapter 8 for details on the exposure index). Interval and lifetime occurrences of basal cell carcinomas, sun-exposure related malignant skin neoplasms, and malignant systemic neoplasms were examined. As was done in the core analyses, verified conditions and verified plus suspected malignancies were each investigated. Blacks were excluded from all malignant skin neoplasm analyses. Group contrasts in incidence rates of malignant skin neoplasms were adjusted for the covariates of age, sun reaction index, and average residential latitude. Adjusted analyses for malignant systemic neoplasms accounted for the effects of age and race. For each dependent variable, exposure level frequencies and percentages are presented in Appendix Tables H-26 and H-27, for interval and lifetime, respectively, along with the results of the unadjusted analyses. Pearson's chi-square test was used to reflect overall exposure index differences, and Fisher's exact test was used to investigate medium versus low and high versus low exposure level contrasts. Results of the adjusted analyses are presented in Tables 10-26 and 10-27, for interval and lifetime, respectively. These results are presented in the context of a main effects model containing exposure index and all adjusting covariates. Several significant or marginally significant overall results were found. None was suggestive of a strictly increasing dose response effect; in fact, most showed decreasing incidence rates with increasing exposure. Among officers, in the unadjusted interval analysis, significant or marginally significant results were found among nonblacks for verified and suspected basal cell carcinomas (overall p=0.042), sun-exposure related malignant skin neoplasms (verified: overall p=0.096, verified plus suspected: overall p=0.021), and among Blacks and nonblacks for verified plus suspected malignant systemic neoplasms (overall p=0.081). These findings were primarily due to higher percentages of malignancies in the medium exposure level than in the high or low categories for each variable (see Appendix Table H-26 for frequencies). The corresponding adjusted analyses were nonsignificant for basal cell carcinoma (overall p=0.156), verified sun-exposure malignancies (overall p=0.272), and systemic malignant neoplasms (overall p=0.109). The adjusted results were marginally significant for verified plus suspected sun-exposure malignancies (overall p=0.095). 10-55 �TABLE 10-25. Fully Compliant Baseline Participants Also in Followup Examination by Malignant Neoplasm Status Group Malignant Neoplasm at Baseline Malignant Neoplasm at Followup Ranch Hand Number Percent Comparison Number Percent Total Yes 10 1.0 15 1.3 25 No 37 3.8 28 2.5 65 Yes 36 3.7 31 2.7 67 No 888* 93.5 1,953 Yes No Total 971 a 91.5 l,065a 1,139 Includes three Ranch Hands and two Comparisons who died after followup. 10-56 2,110 �TABLE 10-26. Adjusted Exposure Index Analysis for Followup Participants for occurrence of Malignant Neoplasms in the Baseline-Followup Interval Variable Occupation Officer Basal Cell3 Carcinoma (Verified Only) Enlisted Flyer Exposure Index Low Medium High Total* Total* Total* 124 54 138 Officer Basal Cella Carcinoma (Verified and Suspected) 61 121 51 Adj. Relative Risk ( 5 C.I.) 9% p-Value Overall M vs. L H vs. L 2.02 ( . 0 8 1 ) 05,.0 0.91 ( . 8 4 6 ) 01,.8 0.415 0.320 0.908 Overall H vs. L H vs. L 0.35 (0.05,2.20) 0.11 (0.01,1.10) 0.080 0.261 0.061 i Enlisted Groundcrew o i Ln 127 Contrast 124 Enlisted Flyer Enlisted Groundcrew 54 138 149 127 61 149 129 121 51 129 Overall M vs. L H vs. L 0.51 (0.07,3.53) 0.19 (0.02,2.14) 0.346 0.496 0.179 Overall M vs. L H vs. L 2.40 (0.73,7.88) 0.91 (0.22,3.76) 0.156 0.149 0.892 Overall M vs. L H vs. L 0.35 (0.05,2.20) 0.11 (0.01,1.10) 0.080 0.261 0.061 Overall M vs. L H vs. L 0.36 (0.06,2.25) 0.14 (0.01,1.44) 0.165 0.274 0.098 �TABLE 10-26. (continued) Adjusted Exposure Index Analysis for Follovup Participants for Occurrence of Malignant Neoplasms in the Baseline-Follovup Interval Occupation Variable Officer Low Total* 124 Exposure Index Medium High Total* Total* 127 121 Contrast Adj. Relative Risk (95% C.I.) p-Value Overall M vs. L H vs. L 2.38 (0.61,9.30) 0.95 (0.18,4.88) 0.272 0.214 0.949 Y a Sun-Exposure Related Malignancies (Verified Only) Enlisted Flyer 54 Enlisted Groundcrew Officer o m 00 Sun-Exposurea Related Malignancies (Verified and Suspected) 138 124 Enlisted Flyer Enlisted Groundcrew 54 138 61 149 127 60 149 51 129 121 51 129 Overall M vs. L H vs. L 0.35 ( . 5 2 2 ) 00,.0 0.11 (0.01,1.10) 0.080 0.261 0.061 Overall M vs. L H vs. L 0.83 (0.15,4.55) 0.50 (0.07,3.39) 0.767 0.826 0.481 Overall M vs. L H vs. L 2.68 (0.83,8.67) 0.93 (0.22,3.86) 0.095 0.100 0.921 Overall M vs. L H vs. L 0.35 ( . 5 2 2 ) 00,.0 0.11 (0.01,1.10) 0.080 0.261 0.061 Overall M vs. L H vs. L 0.59 (0.12,2.94) 0.36 (0.06,2.20) 0.514 0.519 0.268 �TABLE 10-26. (continued) Adjusted Exposure Index Analysis for Followup Participants for Occurrence of Malignant Neoplasms in the Baseline-Pollovup Interval Variable Occupation Officer Systemic15 Malignancies (Verified Only) Enlisted Flyer Exposure Index Low Medium High Total* Total* Total* 127 55 Enlisted Groundcrev 154 Officer 127 130 65 163 123 57 142 o Systemicb Malignancies (Verified and Suspected) Enlisted Flyer Enlisted Groundcrew 55 154 *Total number of participants. a Nonblacks only. b Blacks and nonblacks. —Analyses not done due to sparse cells. 130 65 163 123 57 142 Contrast Overall M vs. L H vs. L Overall M vs. L H vs. L Overall M vs. L H vs. L Overall M vs. L H vs. L Overall M vs. L H vs. L Overall M vs. L H vs. L Adj. Relative Risk (95% C.I.) p-Value 1.60 (0.15,17.22) — 0.365 0.696 — _—. — — — — — — — — 2.95 (0.31,27.73) — 0.25 (0.02,3.90) 0.38 ( . 3 4 9 ) 00,.0 — — 0.109 0.344 0.557 0.326 0.458 — — — �TABLE 10-27. Adjusted Exposure Index Analysis for Follovup Participants for Lifetime Occurrence of Malignant Neoplasms Variable Occupation Officer Basal Cell Carcinoma (Verified Only)a Enlisted Flyer Exposure Index Low Medium High Total* Total* Total* 124 54 Enlisted Groundcrew Officer o i Basal Cell Carcinoma (Verified and Suspected)3 138 124 Enlisted Flyer Enlisted Groundcrew 54 138 127 61 149 127 60 149 121 51 129 121 51 129 Contrast Adj. Relative Risk ( 5 C.I.) 9% p-Value Overall M vs. L H vs. L 1.33 (0.48,3.66) 1.27 (0.45,3.60) 0.841 0.580 0.647 Overall M vs. L H vs. L 0.23 (0.03,1.61) 0.08 ( . 1 0 7 ) 00,.8 0.024 0.141 0.030 Overall M vs. L H vs. L 1.10 (0.31,3.86) 0.87 (0.24,3.20) 0.937 0.881 0.832 Overall M vs. L H vs. L 1.49 (0.59,3.78) 1.22 (0.46,3.24) 0.699 0.404 0.694 Overall M vs. L H vs. L 0.23 (0.03,1.61) 0.08 ( . 1 0 7 ) 00,.8 0.024 0.141 0.030 Overall M vs. L H vs. L 0.89 (0.27,2.97) 0.71 (0.20,2,48) 0.860 0.849 0.589 �TABLE 10-27. (continued) Adjusted Exposure Index Analysis for Followup Participants for Lifetime Occurrence of Malignant Neoplasms Variable Occupation Officer Sun-Exposure Related Malignancies (Verified Only)3 Enlisted Flyer Low Total* 124 54 Enlisted Groundcrew 138 Officer 124 Exposure Index Medium High Total* Total* 127 61 149 121 51 129 o Sun-Exposure Related Malignancies (Verified and Suspected)3 Enlisted Flyer Enlisted Groundcrew 54 138 127 60 149 121 51 129 Contrast Adj. Relative Risk ( 5 C.I.) 9% p-Value Overall M vs. L H vs. L 1.19 (0.47,3.00) 0.99 (0.37,2.64) 0.906 0.717 0.980 Overall M vs. L H vs. L 0.42 (0.08,2.19) 0.09 (0.01,0.89) 0.045 0.300 0.039 Overall M vs. L H vs. L 1.35 (0.40,4.58) 0.88 (0.24,3.25) 0.785 0.627 0.850 Overall M vs. L H vs. L 1.33 (0.56,3.16) 0.97 (0.38,2.47) 0.722 0.518 0.952 Overall M vs. L H vs. L 0.42 (0.08,2.19) 0.09 (0.01,0.89) 0.045 0.300 0.039 Overall M vs. L H vs. L 1.10 (0.34,3.52) 0.72 (0.20,2.52) 0.785 0.879 0.603 �TABLE 10-27. (continued) Adjusted Exposure Index Analysis for Follovup Participants for Lifetime Occurrence of Malignant Neoplasms Variable - Occupation Officer Systemic Malignancies (Verified Only) Enlisted Flyer Exposure Index Low Medium High Total* Total* Total* 127 55 Enlisted Groundcrew 154 Officer 127 130 65 163 123 57 142 o t to Systemic Malignancies (Verified and Suspected) Enlisted Flyer Enlisted Groundcrew 55 154 *Total number of participants. a Nonblacks only. Blacks and nonblacks. —Analyses not done due to sparse cells. 130 65 163 123 57 142 Contrast Adj. Relative Risk (95% C.I.) p-Value Overall M vs. L H vs. L 1.11 (0.18,7.01) 1.49 (0.24,9.16) 0.902 0.911 0.669 Overall M vs. L H vs. L 0.86 (0.11,7.08) 0.46 (0.04,5.46) 0.806 0.892 0.540 Overall M vs. L H vs. L — — 0.073 — — Overall M vs. L H vs. L 1.69 (0.30,9.65) 1.47 (0.24,8.95) 0.829 0.554 0.679 Overall M vs. L H vs. L 0.51 (0.08,3.47) 0.54 (0.08,3.57) 0.741 0.494 0.527 Overall M vs. L H vs. L — —— 0.087 — — �For the interval analysis, enlisted flyers exhibited a marginally significant decreasing dose-response effect for verified basal cell carcinomas in both the unadjusted (p=0.073) and adjusted analyses (p=0.080). (All Ranch Hand enlisted flyer interval malignant skin neoplasms were verified basal cell carcinomas; thus, interval results for verified and verified plus suspected basal cell carcinoma and the corresponding sun-exposure related neoplasms were identical. Similarly, for lifetime analyses, verified and verified plus suspected analyses were the same). The percentages of participants with interval basal cell neoplasms were 11.1 percent, 3.3 percent, and 1.9 percent for the low, medium, and high exposure categories, respectively. The enlisted groundcrew exhibited a nonsignificant decreasing dose-response effect for basal cell carcinomas and sun-exposure related malignant neoplasms. In the adjusted lifetime analysis for enlisted flyers (Table 10-27), there were significant findings, similar to the interval analysis, namely a decreasing dose-response effect for basal cell carcinomas (overall p=0.024; Adj. RR (medium versus low]: 0.23, 95% C.I.: [0.03, 1.61], Adj. RR [high versus low]: 0.08, 95% C.I.: [0.01, 0.78]), and for sun-exposure related skin malignancies (overall p=»0.045; Adj. RR [medium versus low]: 0.42, 95% C.I.: [0.08, 2.19], Adj. RR [high versus low]: 0.09, 95% C.I.: [0.01, 0.89]). The percentages of participants with lifetime basal cell carcinomas were 13.0 percent, 3.3 percent, and 1.9 percent for the low, medium, and high exposure categories, respectively. The corresponding percentages for lifetime sun-exposure related skin malignancies were 13.0 percent, 4.9 percent, and 1.9 percent. For the enlisted groundcrew cohort, a marginally significant result was found for all systemic malignancies combined in the adjusted analyses (verified only: overall p=0.073; verified plus suspected: overall p=0.087). Of the four verified systemic malignancies, three were in the medium exposure category and one was from the high category. There was one additional suspected malignant neoplasm in the high exposure category. No significant results were found for officers in the lifetime analysis. DISCUSSION The statistical analyses of cancer endpoints in this chapter have carefully followed the prescribed boundaries of the SAIC analytic plan approved by the Air Force. Specific latency analyses of certain cancers associated with environmental exposures were not performed, nor were contrasts of cancer-specific incidence rates to SEER data judged appropriate. Further, embedded case control studies on selected cancers were not performed due to concern for bias. The statistical analyses focused on neoplasms occurring during the time interval between 1982 and 1985 (Baseline to followup). However, because these relatively young and healthy cohorts yielded small numbers of cancers in this short interval, and because of the intense scientific interest in malignant disease, the analysis went beyond the assessment of the incidence of malignant neoplasms in this interval. Lifetime (Baseline and followup data combined) analyses of malignant incident neoplasms were conducted. Cancers occurring prior to military duty in SEA were excluded. A full cancer mortality-morbidity analysis was not attempted but simple tabulations of cancer incidence and mortality of Baseline participants were made. Interval and lifetime analyses were expanded to include suspected cancers noted at followup. Further, grouped cancers that were not likely related were 10-63 �analyzed (all systemic cancers and malignant sun exposure-related skin neoplasms). These efforts, however, have introduced several subtle interpretive issues that should be noted, e.g., skin cancer rates are for nonblacks only, whereas systemic cancer rates are for all races; lifetime group rates are on only those attending the followup examination; and verified and suspected cancer categories included more cases but the data are less reliable. Further, contrasts of cancer rates, particularly skin cancer, between the Baseline results and followup results, or lifetime results, must account for the slight differences in the Baseline and followup cohorts, racial adjustment (Blacks were not omitted from skin cancer analyses at Baseline), skin cancer classification, the change in focus from the Original Comparisons to the total Comparison group, and whether the data were adjusted for covariates. Skin Cancer The emphasis on skin cancer at the followup examination was predicated upon the finding of a significant excess of such cancers at the Baseline examination, and the lack of risk factor data to conduct appropriate adjusted analyses. Because of shifting factors (cited above) between the examinations, a "direct look" at the skin cancer association is not straightforward. Figure 10-1 is presented as an aid to clarify the skin cancer observations over the two examinations. This diagram compares the Baseline and followup analyses. So that the unadjusted Baseline results could be contrasted to the followup results, the estimated relative risk of basal cell carcinoma among nonblack Ranch Hands versus all nonblack Comparisons (not just Originals) was calculated, using data in the Baseline Report. This unadjusted analysis gave a significant relative risk of 1.77 (p=0.049). These results could then be directly contrasted to the unadjusted followup results, which showed a narrowing of group differences over the 3-year interval (Est. RR: 1.23, p=0.429). (It is noted that this contrast compares skin cancer rates of approximately 23 years to 3 years at different levels of age risk.) The adjusted analysis revealed a significant group-by-occupation interaction, due to a significantly higher rate of basal cell carcinomas among Ranch Hand enlisted flyers than the corresponding Comparisons (Adj. RR; 6.50, p=0.019), but very similar rates in the two groups for officers and enlisted groundcrew were seen. The Baseline data were carefully merged (to avoid duplicate counts) with the followup data to assess the total lifetime incidence of basal cell carcinomas between groups. The addition of the nonsignificant followup results to the significant Baseline results produced a nonsignificant lifetime assessment (Est. RR: 1.36, p=0.128), as expected. However, when the lifetime data were adjusted for covariate effects, a significant result emerged (Adj. RR: 1.56, p=0.035), with Ranch Hands having a significant excess of lifetime basal cell carcinoma. A careful examination of the covariates showed that the variable of average residential lifetime latitude was most likely responsible for the significant adjusted results. The latitude variable was a significant confounding variable since it was associated with basal cell carcinomas and with average lifetime latitude which varied significantly by group. 10-64 �Baseline Results Followup Results Lifetime Results "Skin Cancer" Basal Cell Cancers Basal Cell Cancers (Unadj., Original Comparisons, All Races) NS (Unadj., Total Comparisons, Nonblacks) r [RR=1.23, p=0.429] New Analysis NS* (Unadj., Total Comparisons, All Races) I 1 Basal Cell Cancers NS (Unadj. Total "*" Comparisons, Nonblacks) [RR-1.36, p-0.128] **** (Adjusted for all Covariates, Total Comparisons, Nonblacks) (Adjusted for all Covariates, Total Comparisons, Nonblacks) [Group-by-Occ, p=0.044] [RR=1.56, p=0.035] (Unadj., Total Comparisons, Nonblacks) [RR=1.77, p=0.049] S: NS: NS*: »***. Significant (p < 0.05). Not significant (p > 0.10). Borderline significant (0.05 < p <, 0.10). Group-by-covariate interaction. Figure 10-1. Schematic Diagram of Unadjusted and Adjusted Skin Cancer Results, by Significance and Relative Risk, and by Examination Period (Time). 10-65 �Because of the significant confounding effect of the latitude variable, it was examined closely for misclassification or bias. An initial review of the residential history forms showed occasional discrepancies between total residential years and chronologic age. This was generally due to sporadic underreporting, and to the data collection instructions which required the citation only of residences of one year or longer. However, analyses showed fairly good concordance between reported residential years and chronologic age. No significant group difference was found for the inaccuracy of residential reporting (p=0.684), validating the use of all residential histories even though some were slightly imprecise. In the course of reviewing the covariate effects on basal cell carcinoma, the data suggested some unexpected associations. To sharpen these contrasts, adjusted risks were estimated at set levels of skin reaction to sun, skin color, average lifetime residential latitude, and age, relative to the lowest risk observed, i.e., Comparisons 40 years old (at Baseline) who have lived on average in northern latitudes and tan easily were arbitrarily assigned a risk of 1.00. These computed risks are given in Table 10-28. These results show uniform increased risks in the Ranch Hands over both the base level of one and the Comparisons in the same covariate strata. Further, in all strata, age, latitude, and skin color behave as expected. However, the sun-reaction index does not behave as expected since those who burn easily have lower relative risks than those who have an intermediate reaction to sun, although they do have higher relative risks than those who tan easily. This may represent avoidance of sun exposure or the use of sunblock by those individuals. Skin cancer, and particularly basal cell carcinoma, has been emphasized in this report because of the significant group differences detected at Baseline (and the theoretical link to TCDD causation), and the borderline significant adjusted results found for the lifetime rates. The results of the third-year followup analysis suggest that if group differences continue to narrow (where pX).15) at the fifth-year followup examination, the lifetime results would likely not be significant even with full adjustment. Systemic Cancer The analyses of systemic cancer for both the interval and lifetime periods have necessarily been limited by scant data. Cancer specific analyses, in particular, have not provided meaningful results because of low counts. However, some variation in tumor type was noted in the two groups: colon cancer (5 Comparisons, 0 Ranch Hands), testicular cancer (3 Ranch Hands, 0 Comparisons), and smoking related tumors of the oral cavity, pharynx, bronchus, and lung (5 Ranch Hands, 0 Comparisons). Testicular and smoking related tumors have not been associated with exposure to herbicides or TCDD. Table 10-18 cited counts of malignancies that have been associated to herbicides and dioxin exposure. Because of the relative rareness of the diseases soft tissue sarcoma (STS), Hodgkin's disease, and non-Hodgkin's lymphoma, lifetime rates were expected to be exceptionally low. Most of the covariate associations with systemic cancer were anticipated, but the change in significance for smoking (significant at Baseline, borderline significant for lifetime cancers) was not expected, particularly as the cancer cases increased during the interval. 10-66 �TABLE 10-28. Confuted Rides of Basal Cell Carcinoma by Group at Varying Levels of Four Risk Factors, Relative to Comparisons at law Risk* Covariate Categories Skin Reaction Average Lifetime to Sun Residential latitude Skin Color: Not Peach Age at Baseline Comparison Ranch Hand Skin Color; Peach Ranch Comparison Hand 10* .0* 2.99 1.48 4.43 1.55 4,62 2.30 6.85 40 60 1.63 4.87 2.42 7.23 2.52 7.53 3.74 11.18 >37°N 40 60 3.04 9.09 4.52 13.50 4.71 14.06 6.99 2.7 08 40 60 4.97 14.83 7.37 22.02 7.68 22.93 11.40 34.04 >37°N 40 60 2.02 6.04 3.00 8% , 3.13 9.33 4.64 13.86 <37°N Bums Easily 40 60 <37°N Intermediate Reaction >37°N <37°N Tans Easily 40 60 3.30 9.85 4.90 14.62 5.10 15.22 7.57 22.60 ^Computed from main effects model with latitude, skin reaction to sun, and skin color as covariates. **Base Category (Lowest Risk). 10-67 �All Cancers As previously noted, the interrelatedness of many of the analyzed cancer variables has created a compounding of statistical significance, and care should be taken in making inferences and final conclusions. An almost uniform dilutional effect was created by adding "suspected" cancers to the analyses, as there were more of this category in the Comparisons than in the Ranch Hands. The use of suspected neoplasms was deemed necessary in order to best describe the complete cancer findings, recognizing that confirmation of all suspected cases was difficult. Two patterns emerged from the analyses. All relative risks exceeded the value of one, except that of lifetime verified melanoma and verified or verified plus suspected squamous cell carcinoma. Some of the elevated risks were due to the relatedness of the variables as stated, but the relative risks for the unrelated variables skin cancer and systemic cancer both exceeded one. The joint consideration of both yielded a significant relative risk. The second pattern was of the group-by-covariate interactions observed for seven of the analyses; 3 of them involved the covariate of occupation and 4 involved skin reaction to sun. The three group-by-occupation interactions all showed a significant detriment to the Ranch Hand enlisted flying cohort. Further analyses of air crewmembers versus noncrewmembers revealed a significant risk of basal cell carcinoma for the Ranch Hand air crewmembers (RRs 1.94, p.O.049). Since enlisted Ranch Hand flyers in the interval exhibited more basal cell carcinomas (RR: 6.5, p=0.019) and more verified and suspected systemic cancers (4/175 RH with systemic neoplasms versus 0/209 Comparisons, p=0.042), there may be more reason to assume a biologic foundation than chance, although the reason is obscure. The four group-by-sun reaction index interactions all revealed a significant or marginally significant detriment to Ranch Hands who reacted mildly to the sun. In full context, the cancer observations cannot be viewed as disturbing at this time. The skin cancer group differences have narrowed over a 3-year period. An additional analytic observation on skin cancer is that inclusion or exclusion of only one or two cases was shown to alter the choice of the best statistical model, affecting the presence or absence of both covariates and group-by-covariate interactions, and also change the p-value of the adjusted group difference above or below the alpha level of 0.05. For systemic cancer, both groups are at the lower end of the expected ascending cancer curves, where numeric and tumor type fluctuations are expected. A recognized bench-mark for the latency of many cancers is 20 years, and this will not be achieved by most participants until the 5-year followup examination, 2 years from now. Cancer findings at that time will be the basis upon which firm conclusions can be made. SUMMARY AND CONCLUSIONS The cancer analysis focused on cancer occurrences in the Baselinefollowup interval, and also included analyses of the Baseline plus interval cancer history. A summary of the cancer findings is given in Table 10-29'. No significant unadjusted differences were found between nonblack Ranch Hands and Comparisons in the Interval (Baseline-Followup) incidence rates of basal cell carcinoma, melanoma, squamous cell carcinoma, all malignant skin cancers, sun-exposure related malignant neoplasms (comprising basal cell 10-68 �TABLE 10-29. Overall Summary Table: Unadjusted and Adjusted Analysis of Interval and Lifetime Skin and Systemic Cancer Incidence Cancer Type Baseline-Followup Interval UnadjustedAdjusted Lifetime (Baseline & Followup) UnadjustedAdjusted Malignant Skin Cancer (Nonblacks only) Verified Basal Cell Carcinoma NS **** NS Verified plus Suspected Basal Cell Carcinoma NS **** NS Verified Melanoma NS a NS Verified plus Suspected Melanoma NS a NS Verified Squamous Cell Carcinoma NS a NS Verified plus Suspected Squamous Cell Carcinoma NS a NS Verified Sun Exposure Skin Cancers NS NS NS* Verified plus Suspected Sun Exposure Skin Cancers NS NS NS All Verified Malignant Skin Cancers NS ~a NS Verified plus Suspected Malignant Skin Cancers NS NS Verified Skin Cancers of Any Type NS* S Verified plus Suspected Skin Cancers of Any Type NS NS* 10-69 **** NS �TABLE 10-29. Overall Summary Table: Unadjusted and Adjusted Analysis of Interval and Lifetime Skin and Systemic Cancer Incidence (continued) Cancer Type Baseline-Followup Interval UnadjustedAdjusted Lifetime (Baseline & Follovup) UnadjustedAdjusted Malignant Systemic Cancer (Blacks and Nonblacks) Verified Systemic Cancer NS NS NS **** Verified plus Suspected Systemic Cancer NS **** NS **** NS* —a S —a All Neoplasms (Blacks and Nonblacks) Any Type, Any Location" Verified NS: Not significant (p>0.10). ****Group-by-covariate Interaction. —aAnalysis not done. NS*: Borderline significant (0.05<p<0.10). Comprises malignant, benign, uncertain behavior. S: Significant (p<0.05). 10-70 �carcinoma, melanoma, and epithelial neoplasms NOS) or all malignant skin cancers as a group. The unadjusted group contrast of all skin neoplasms (comprising malignant and benign neoplasms, and neoplasms of uncertain behavior or unspecified nature) was marginally significant, with a higher rate among .Ranch Hands. When suspected malignant skin cancers (noted at Followup but not verified at the time of writing) were included in the analyses with the verified conditions, all the unadjusted group contrasts were nonsignificant. The covariates used for the adjusted analyses of basal cell carcinoma and the sun exposure related skin malignancies were age, occupation, skin color, reaction of skin to sun, and average latitude, all of which were highly associated with skin cancer incidence. Other host factors were related to skin cancer incidence, but not as strongly as those included in the analysis. A borderline association with smoking history was noted, and was determined to be partly an age effect. Analysis of the incidence of interval basal cell carcinoma revealed a significant group-by-occupation interaction, due to a significant group difference for enlisted flyers, but not for officers or enlisted groundcrew. Inclusion, of suspected basal cell carcinoma resulted in a group-by-sun reaction index interaction. This was due to Ranch Hands with an intermediate reaction to sun having a higher relative risk than the corresponding Comparisons. The adjusted group contrast of the incidence rates of verified sun-exposure related skin cancers was not significant; inclusion of suspected conditions did not alter this lack of significance. There was no significant group difference for Blacks and nonblacks in the unadjusted incidence rates of all interval verified malignant systemic neoplasms combined, nor was there a significant difference in the adjusted group rates. Analysis of the verified plus suspected interval systemic cancers showed a nonsignificant unadjusted group difference, but a group by occupation interaction was found in the adjusted analysis. This was due to a significant group difference of verified plus suspected systemic malignancies among the enlisted flyers with five occurrences among the Ranch Hands, but none among the Comparisons. Age and a race-by-packyear interaction were important adjusting factors. The Baseline and Followup data were combined for the assessment of lifetime incidence of cancer; occurrences of cancer prior to Vietnam were excluded. There were no significant unadjusted group differences in lifetime incidence rates among nonblacks for basal cell carcinoma, melanoma, squamous cell carcinoma, the sun exposure related skin cancers, or all malignant skin cancers combined. The unadjusted group contrast of all lifetime skin malignancies was significant, with a higher rate among Ranch Hands. Inclusion of suspected cancers with the verified cancers reduced the difference between the groups for all these malignant skin contrasts, except for the sun exposure related skin cancers, for which a marginally significant group difference was found. However, the contrast of all skin malignancies remained close to significance. Adjusted analysis of the incidence rates of lifetime basal cell carcinoma revealed a significantly higher incidence rate among Ranch Hands 10-71 �(Adj. RR: 1.56, p=0.035). Significant effects of an occupation-by-age interaction, a skin color-by-sun reaction index interaction, and a sun reaction index-by-average residential latitude interaction were seen. The adjustment resulted in a significant relative risk that, moreover, was higher than the unadjusted relative risk. Average residential latitude, associated with both group and skin cancer, and skin color, which was associated with the disease and marginally associated with group, played a major part in the change from the unadjusted analysis due to confounding. Inclusion of suspected basal cell carcinoma in the adjusted analysis resulted in a group by sun reaction index interaction, as was noted for the interval analysis. The adjusted group contrast in incidence rates of the sun-exposure related skin cancers was also significant (Adj. RR: 1.54, p=0.030), which is not surprising since the majority are basal cell carcinoma. Inclusion of the suspected conditions resulted in a non-significant group contrast. The unadjusted group contrasts of the incidence rates of all systemic cancers combined were not significant, both for verified and verified plus supected conditions. There was one new occurrence of a soft tissue sarcoma (Ranch Hand) and one suspected cancer of the lymphatic system (Ranch Hand), in addition to the one previously reported soft tissue sarcoma and one Hodgkin's disease in the Comparison group. Adjusted analysis of all lifetime malignant systemic neoplasms as a group, however, revealed a group by occupation interaction, due to a significantly higher rate for Ranch Hand enlisted flyers as contrasted to Comparisons. The same result was found for verified plus suspected systemic cancers. In conclusion, there were no adjusted or unadjusted differences between groups in basal cell carcinoma incidence in the Baseline-followup interval. At Baseline, a significantly higher rate of basal cell carcinoma was found for Ranch Hands when contrasted with Original Comparisons. When the Baseline data were combined with the interval data, adjusted analysis, but not the unadjusted analysis, revealed a significantly higher rate of basal cell carcinoma among the Ranch Hands than among all Comparisons. The relative risk of basal cell carcinoma appears to be declining over time. Relative risks of basal cell carcinoma and systemic cancer were found to be consistently larger than 1. Most of the skin cancers were basal cell carcinomas, upon which most of the skin cancer analysis focused, thus relative risks for sun-exposure related skin neoplasms and all malignant skin cancers as a group were very similar to those for basal cell carcinoma. The number of occurrences of systemic cancer was small, in part because the cohort is relatively young, and although the relative risks (lifetime and interval) are greater than 1, the difference between groups is not significant. Sufficient time may not have elapsed since Vietnam to enable a group difference in systemic neoplasms, if one exists, to be apparent. 10-72 �CHAPTER 10 REFERENCES 1. Young, A.L. 1981. Agent orange at the crossroads of science and social concern. Research report 2750-81, Air Command and Staff College, Air University, Maxwell AFB, Alabama. 2. Tschirley, F.H. 1986. Dioxin. Sci. Amer. 254:29-35. 3. Young, A.L., H.K. Rang, and B.M. Shepard. 1985. Rationale and descriptions of the federally sponsored epidemiologic research in the United States on the phenoxy herbicides and chlorinated dioxin contaminants. In Chlorinated dioxins and dibenzofurans in the total environment II; ed. L.H. Keith, C. Rappe, and G. Choudhary pp. 155-166.Stoneham, Massachusetts: Butterworth Publishers. 4. Poland, A. 1984. Reflections on the mechanism of action of halogenated aromatic hydrocarbons. In Banbury report 18; Biological mechanisms of dioxin action, ed. A. Poland and R.D. Kimbrough, pp. 109-117.Cold Spring Harbor, New York: Cold Spring Harbor Laboratory. 5. DiGiovanni, J.A., A. Viaje, D.L. Berry, et al. 1977. Tumor initiating ability of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and Aroclor 1254 in a two-stage system of mouse skin carcinogenesis. Bull. Environ. Contain. Toxicol. 18:522-557. 6. Berry, D.L., T.J. Slaga, J. DiGiovanni, et al. 1979. Studies with chlorinated dibenzo-p-dioxins, polybrominated biphenyls and polychlorinated biphenyls in a two-stage system of mouse skin tumorigenesis: Potent anticarcinogenic effects. Ann. N.Y. Acad. Sci. 320:405-414. 7. National Toxicology Program (NTP). 1982. Carcinogenesis bioassay of 2,3,7,8-tetrachlorodibenzo-p-dioxin (CAS no. 1746-01-6) in SwissWebster mice (dermal study). (Report 80-32, technical report series no. 201, NIH publication no. 82-1757, Research Triangle Park, North . Carolina. 8. Poland, A., D. Palen, and E. Glover. 1982. Tumour promotion by TCDD in skin of HRS/J hairless mice. Nature 300:271-273. 9. Kociba, R.J., D.G. Reyes, J.E. Beyer, R.M. Carreon, C.E. Wade, D.A. Dittenber, R.P. Kalnins, L.E. Frauson, C.N. Park, S.D. Barnard, R.A. Hummel, and C.G. Humiston. 1978. Results of a two-year chronic toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats. Toxicol. Appl. Pharmacol. 46:279-303. 10-73 �10. National Toxicology Program (NTP). 1982. Carcinogenesis bioassay of 2,3,7,8-tetrachlorodibenzo-p-dioxin (CAS no. 1746-01-6) in OsborneMendel rats and B6C3F1 mice (gavage study). (Report 80-31, technical report series no. 209, NIH publication no. 82-1765, Research Triangle Park, North Carolina. 11. Pitot, H.C., T. Goldsworthy, H.A. Campbell, and A. Poland. 1980. Quantitative evaluation of the promotion by 2,3,7,8-tetrachlorodibenzo-pdioxin of hepatocarcinogenesis from diethylnitrosamine. Cancer Res. 40:3616-3620. 12. Sundell, L., M. Rehn, and 0. Axelson. 1974. Exposure to herbicides— mortality and tumor incidence: An epidemiological investigation in Swedish railway workers. Lakartidningen 71:2466-2470. 13. Axelson, 0., and L. Sundell. 1974. Herbicide exposure, mortality, and tumor incidence. An epidemiological investigation on Swedish railroad workers. Work Environ. Health 11:21-28. 14. Axelson, 0., and L. Sundell. 1977. Phenoxy acids and cancer. Lakartidningen 74:2887-2888. 15. Axelson, 0., L. Sundell, K. Andersson, C. Edling, C. Hogstedt, and H. Kling. 1980. Herbicide exposure and tumor mortality; An updated epidemiological investigation on Swedish railroad workers. Scand. J. Work Environ. Health 6:73-79. 16. Huff, J.E., J.A. Moore, R. Saracci, and L. Tomatis. 1980. Long-term hazards of polychlorinated dibenzodioxins and polychlorinated dibenzofurans. Environ. Health Perspect. 36:221-240. 17. Hardell, L. 1977. Malignant mesenchymal tumors and exposure to phenoxy acids: A clinical observation. Lakartidningen 74:542-546. 18. Hardell, L. 1977. Soft-tissue sarcomas and exposure to phenoxyacetic acids and cancer. Lakartidningen 74:2735. 19. Hardell, L. 1979. Malignant lymphoma of histiocytic type and exposure to phenoxyacetate or chlorophenols. Lancet 1:55-56. 20. Hardell, L., M. Eriksson, P. Lenner, et al. 1981. Malignant lymphoma and exposure to chemicals, especially organic solvents, chlorophenols and phenoxy acids: A case control study. Br. J. Cancer 43:169-176. 21. Eriksson, M., L. Hardell, N.O. Berg, T. Moller, and 0. Axelson. 1981. Soft tissue sarcomas and exposure to chemical substances: A casereferent study. Br. J. Ind. Med. 38:27-33. 22. Hardell, L. 1981. Relation of soft-tissue sarcoma, malignant lymphoma and colon cancer to phenoxy acids, chlorophenols and other agents. Scand. J. Work Environ. Health. 7:119-130. 23. Hardell, L., B. Johansson, and 0. Axelson. 1982. Epidemiological study of nasal and nasopharyngeal cancer and their relation to phenoxy acid or chlorophenol exposure. Am. J. Indus. Med. 3:247-257. 10-74 �24. Hardell, L., N.O. Bengtsson, U. Jonsson, S. Eriksson, and L.G. Larsson. 1984. Aetiological aspects on primary liver cancer with special regard to alcohol, organic solvents and acute intermittent porphyria—an epidemiological investigation. Br. J. Cancer 50:389-397. 25. Hardell, L., and 0. Axelson. 1984. Phenoxyherbicides and other pesticides in the etiology of cancer: Some comments on the Swedish experience. Presented at Cancer Prevention—Strategies in the Workplace. University of California, San Francisco, December 1984. 26. Jannerfeldt, E. 1980. Epidemiological methodology and pesticide studies. Lakartidningen 77(12):1096. 27. (Editorial.) 1982. Phenoxy herbicides, trichlorophenols, and softtissue sarcomas. Lancet 1(8278):1051-1052. 28. Coggon, D., and E.D. Acheson. 1982. Do phenoxy herbicides cause cancer in man? Lancet 1(8278):1057-1059. 29. Axelson, 0. 1978. Aspects on confounding in occupational health epidemiology, Letter. Scand. J. Work Environ. Health 4:85-89. 30. Axelson, 0. 1980. Views on criticism of pesticide studies. Lakartidningen 77(12):1096-1099. 31. Axelson, 0. 1980. A note on observational bias in case-referent studies in occupational health epidemiology. Scand. J. Work Environ. Health 6:80-82. 32. Remington, R.D. 1980. Specific summary critique of five investigations related to concerns about agent orange. Congressional Record, August 6, 1980, pp. S 10911, S 10912. 33. U.S. Environmental Protection Agency. 1985. Health assessment document for polychlorinated dibenzo-p-dioxins. Final report, September 1985. 34. Ott, M.G., B.B. Holder, and R.D. Olson. 1980. A mortality analysis of employees engaged in the manufacture of 2,4,5-trichlorophenoxyacetic acid. J. Occup. Med. 22:47-50. 35. Cook, R.R., J.C. Townsend, M.G. Ott, and L.G. Silverstein. 1980. Mortality experience of employees exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). J. Occup. Med. 22:530-532. 36. Zack, J.A., and R.R. Suskind. 1980. The mortality experience of workers exposed to tetrachlorodibenzo-dioxin in a trichlorophenol process accident. J. Occup. Med. 22:11-44. 37. Zack, J.A., and W.R. Gaffey. 1983. A mortality study of workers employed at the Monsanto.Company Plant in Nitro, West Virginia. Environ. Sci. Res. 26:575. 38. Honchar, P.A., and W.E. Halperin. 1981. 2,4,5-T, trichlorophenol, and soft-tissue sarcoma. Lancet l(8214):268-269. 10-75 �39. Cook, R.R. 1981. Dioxin, chloracne, and soft-tissue sarcoma. 1(8220):618-619. Lancet 40. Moses, M., and I.J. Selikoff. 1981. Soft-tissue sarcomas, phenoxy herbicides, and chlorinated phenols. Lancet 1(8234):1370. 41. Johnson, F.E., M.A. Kugler, and S.M. Brown, 1981. Soft tissue sarcomas and chlorinated phenols. Lancet 2(8236):40. 42. Fingerhut, M.A., W.E. Halperin, P.A. Honchar, A.B. Smith, D.H. Groth, and W.O. Russell. 1984. An evaluation of reports of dioxin exposure and soft tissue sarcoma pathology in U.S. chemical workers. In Banbury report 18; Biological mechanisms of dioxin action, ed. A. Poland and R. D. Kimbrough, pp. 461-470.Cold Spring Harbor, New York: Cold Spring Harbor Laboratory. 43. Percy, C., E. Stanek, and L. Gloeckler. 1981. Accuracy of cancer death certificates and its effect on cancer mortality statistics. Am. J. Public Health 71(3):242-250. 44. Donna, A., P.G. Betta, F. Robutti, P. Crosignani, F. Berrino, D. Bellinger!. 1984. Ovarian mesothelial tumors and herbicides: case-control study. Carcinogenesis 5:941-942. A 45. Riihimaeki, V., S. Asp, E. Pukklala, and S. Hernberg. 1983. Mortality and cancer morbidity among chlorinated phenoxy acid applicators in Finland. Chemophere 12:779-784. 46. Lynge, E. 1985. A follow-up study of cancer incidence among workers in manufacture of phenoxy herbicides in Denmark. Br. J. Cancer 52:259-270. 47. Smith, A.H., N.E. Pearce, D.O. Fisher, et al. 1984. Soft-tissue sarcoma and exposure to phenoxyherbicides and chlorophenols in New Zealand. JNCI 73:1111-1117. 48. Pearce, N.E., A.H. Smith, and D.O. Fisher. 1985. Malignant lymphoma and multiple myeloma linked with agricultural occupations in a New Zealand cancer registry-based study. Am. J. Epidemic!. 121:225-237. 49. Wiklund, K., and L.E. Holm. 1986. Soft tissue sarcoma risk in Swedish agricultural and forestry workers. JNCI 76(2):229-234. 50. Hoar, S.K., A. Blair, F.F. Holmes, et al. 1986. Agricultural herbicide use and risk of lymphoma and soft-tissue sarcoma. JAMA 256:1141-1147. 51. Colton, T. 1986. Herbicide exposure and cancer. 256:1176-1178. Editorial. JAMA 52. Woods, J.S., L. Polissar, R.K. Severson, L.S. Heuser, and B.C. Kulander. 1987. Soft tissue sarcoma and non-Hodgkin's lymphoma in relating to phenoxy herbicide and chlorinated phenol exposure in western Washington. Preprint. To appear in JNCI, May 1987. 10-76 �53. Sarma, P.R., and G. Jacobs. 1981. Thoracic soft-tissue sarcoma in Vietnam veterans exposed to agent orange. New Engl. J. Med. 306(18):1109. 54. Greenwald, P., B. Kovasznay, D.N. Collins, and G. Therriault. 1984. Sarcomas of soft tissue after Vietnam service. JNCI 73:1107-1109. 55. Kogan, M.D., and R.W. Clapp. 1985. Mortality among Vietnam veterans in Massachusetts, 1972-1983. Boston, Massachusetts"Division of Health Statistics and Research, Massachusetts Department of Public Health. 56. Royal Commission on the Use and Effects of Chemical Agents on Australian Personnel in Vietnam. 1985. Cancer. Vol. 4 in Final Report, pp. VIII-129. Canberra, Australia: Australian Government Publishing Service. 57. Lawrence, C.E., A.A. Reilly, P. Quickenton, P. Greenwald, W.F. Page, A.J. Kuntz. 1985. Mortality patterns of New York State Vietnam veterans. Am. J. Public Health 75:277-279. 58. Wendt, A.S. 1985. Agent orange Iowa survey of Vietnam veterans. Report, Iowa State Department of Health, July 1985. Final 59. Holmes, A.P. 1986. West Virginia Vietnam-era veterans mortality study. Preliminary report, Health Statistics Center, West Virginia Department of Health, January 1986., 60. Anderson, H.A., L.P. Hanrahan, M. Jensen, D. Laurin, W. Yick, and P. Wirgman. 1986. Wisconsin Vietnam veteran mortality study. Final report, Wisconsin Department of Health and Social Services, March 1986. 61. Rang et al. 1986. Soft tissue sarcomas and military service in Vietnam: A case comparison group analysis of hospital patients. J_^ Occup. Med. 28:1215-1218. 62. Thiess, A.M., R. Frentzel-Beyme, and R. Link. 1982. Mortality study of persons exposed to dioxins in a trichlorophenol process accident that occurred in the BASF AG on November 17, 1953. Am. J. Ind. Med. 3:179-189. 63. Blair, A., D.J. Grauman, J.H. Lubin, et al. 1983. Lung cancer and other causes of death among licensed pesticide applicators. JNCI 71:31-37. 64. Barthel, E. 1981. Cancer risk in agricultural workers exposed to pesticides. Arch. Geschwulstforsch. 51(7):579-585. 65. Gatti, R.A., and R.A. Good. 1971. Occurrence of malignancy in immunodeficiency diseases: A literature review. Cancer 28:89-98. 66. Louie, S. and R.S. Schwartz. 1978. Immunodeficiency and the pathogenesis of lymphoma and leukemia. Semin. Hematol. 15:117-138. 10-77 �67. Di Carlo, E.F., J.B. Amberson, C.E. Metroka, P. Ballard, A. Moore, J.A. Mouradian. 1986. Malignant lymphomas and the acquired immunodeficiency syndrome. Arch. Pathol. Lab. Med. 110:1012-1016. 68. Lathrop, G.D., P.M. Moynahan, R.A. Albanese, and W.H. Wolfe. epidemiologic investigation of health effects in Air Force following exposure to herbicides—Baseline mortality study Brooks Air Force Base, Texas;Epidemiology Division, Data Division, USAF School of Aerospace Medicine. 1983. An personnel results. Sciences 69. Holman, C.D.J., and Armstrong, B.K. "Pigmentary Traits, Ethnic Origin, Benigh Nevi, and Family History as Risk Factors for Cutaneous Malignant Melanoma." Journal of the National Cancer Institute 72;257-266, 1984. 70. Scot to, J. and Fears, T.R. "Skin Cancer Epidemiology: Research Needs," National Cancer Institute Monograph 50:169-177, 1978. 10-78 �CHAPTER 11 NEUROLOGICAL ASSESSMENT INTRODUCTION Neurological signs and symptoms, as distinguished from overt diagnosable neurological disease, have been consistently associated with industrial exposure to chlorophenols, phenoxy herbicides, and TCDD. Thus, the neurological system comprises a major examination focal point in all dioxin morbidity studies. This report carefully separates central and peripheral neurological status from "neurobehavioral" parameters, which are discussed in Chapter 12, Psychological Assessment. Based on animal experiments, neurotoxicity can be attributed to the compounds 2,4-D and TCDD. For low to moderate doses, both central and peripheral acute effects occur but appear to be reversible. " The effects of 2,4-D are presumably due to disruption in the neuromuscular transport system of organic acid anions. A variety of 2,4-D experiments in several animal species generally shows a wide range of neural pathology including electroencephalographic (EEC) desynchronization, demyelination, myotonia, loss of coordination, and uncontrolled motor activity. No substantive data support the isolated neurotoxicity of 2,4,5-T. Numerous case reports following accidental human exposures or suicide attempts with 2,4-D have shown a remarkable neurologic parallel to the animal studies. ~ In particular, 2,4-D and TCDD have been implicated in a wide array of central neurological signs and symptoms, including headache, vomiting, dizziness, disorientation, sleep disturbance, stupor, memory loss, loss of coordination, and EEC abnormalities or alterations from a baseline tracing. ' ' Peripheral abnormalities have included demyelination, acute degeneration of ganglion cells, temporary paralysis, anesthesia, hyperesthesia, paresthesia, neuralgic pain, numbness, tingling, muscle pain, muscle fasciculations, depressed or absent deep tendon reflexes, weakness, decreased nerve conduction velocities, "polyneuritis," and limb fatigue. " These peripheral signs and symptoms in industrial workers have received the generic diagnostic label "neurasthenia." Both the number and severity of symptoms tended to aggregate in individuals with chloracne as contrasted to those without chloracne.11'16'17 In general, there is consistency between the various case reports of neurasthenia and results from uncontrolled clinical studies. Of particular relevance is the consistency in findings from studies of both industrial manufacturing and industrial accidents. This literature provides the clearcut conclusion that neurological.impairment is caused directly by exposure to 2,4-D and TCDD. Not answered satisfactorily in the literature, however, are the issues of complete reversibilty of observed signs and symptoms and the long-term impact on health and quality of life. 11-1 �Because of the conclusive evidence that two of three Agent Orange ingredients cause neurological "disease," it follows that significant exposure to Agent Orange could manifest neurologic signs, symptoms, or sequelae. In fact, over 10 percent of Vietnam veterans who enlisted in the VA^gent Orange Registry cited one or more symptoms of the neurasthenic complex. The VA Registry is a comprehensive listing, predominantly of veterans alleging health impairments due to Agent Orange exposure. The Registry does not purport to be a scientific effort upon which cause-and-effect relationships can be established. Nonetheless, some individuals believe that the symptom array in the VA Registry is so compatible with case reports and numerator-oriented clinical studies that the veterans must, in fact, have suffered adverse health effects from their Vietnam service and presumed exposure to Agent Orange. Others point to the intense media attention to "Agent Orange symptoms" during the formation of the Registry, and presume that the veterans' complaints are largely due to an "over-reporting" or compensation bias. Clearly, only well-controlled, well-conducted epidemiologic studies of veterans known to have been exposed to Agent Orange can answer the question of cause and effect for illnesses, including the specific question of whether single or multiple neurologic signs and symptoms are also attributable to these exposures. Baseline Summary Results The 1982 AFHS neurological assessment consisted of questionnaire, physical examination, and electromyographic data obtained by examiners and technicians who were blinded to the group identity of each participant. The physical examination required an average of 30 minutes to complete. Those few individuals with positive RPR tests, a screening serological test for syphilis, and those with peripheral edema were deleted from the statistical analyses. Covariates of reported alcohol usage, exposure to insecticides and industrial chemicals, and glucose intolerance (diabetes) were analyzed. Results of the questionnaire disclosed no significant group differences in reported neurological diseases. The physical examination did not reveal any statistically significant group differences in the function of all 12 cranial nerves, nor any effects due to the covariates of alcohol or diabetes. Peripheral nerve function was assessed by the quality of four reflexes (patellar, Achilles, biceps, and Babinski), muscle strength/bulk, and reaction to the stimuli of pin prick, light touch, and vibration. Other than a statistically significant increase (p»0.03) in Ranch Hand Babinski reflexes, significant group differences were not detected. The alcohol covariate demonstrated a marginal effect (p=0.07) on pin-prick reaction, while glucose intolerance showed a profound effect on the patellar and Achilles reflexes and reactions to light touch and vibration. Nerve conduction velocities were obtained on the ulnar nerve, above and below the elbow, and the peroneal nerve by highly standardized methods. The results for each segmental measurement were nearly identical in the Ranch Hand and Comparison groups. Conduction velocity showed highly significant inverse relationships to both alcohol (measured in drink-years) and glucose intolerance in almost all of the anatomic measurements. No group associations or interactions were detected with the covariates of industrial and degreasing chemicals and insecticides. 11-2 �No significant group differences were detected in four measures of central neurological function (tremor, finger-nose coordination, modified positive Romberg's sign, or abnormal gait). Alcohol usage was significantly associated with the presence of tremor, and glucose intolerance was highly correlated to abnormal balance and the presence of tremor. Of a total of 84 exposure index analyses on all of the dependent variables, 3 were statistically significant but were either nonlinear or biologically implausible. In summary, the detailed neurological examination and assessment did not reveal statistically significant increases in abnormalities in the Ranch Hands, nor were consistent dose-response relationships noted for herbicide exposure. The classical neurological effects of alcohol ingestion and diabetes were repeatedly observed in the neurological evaluations. Parameters of the 1985 Neurological Assessment The 1985 AFHS neurological examination deleted the measurements of nerve conduction velocities but otherwise repeated the format of the Baseline examination. The questionnaire maintained a historical focus of neurasthenia via five questions for the 1982-1985 interval. With this similarity in examination and questionnaire, the dependent variables of the analyses were almost identical to those of the Baseline study, however, the number of covariates was slightly increased. Diabetic status was trichotomized: Individuals reporting a history of diabetes (unverified) and individuals exhibiting glucose intolerance with postprandial glucose levels greater than or equal to 200 mg/dl were classified as diabetic, participants with glucose levels of at least 140 mg/dl but less than 200 mg/dl were classified as impaired, and participants with glucose levels less than 140 mg/dl were classified as normal. Race was included as a covariate, and lifetime alcohol use was updated on the basis of enhanced information from the 1985 questionnaire. The analyses were based on 1,016 Ranch Hands and 1,293 Comparisons. Individuals confirmed to be positive for syphilis by fluorescent treponemal antibody (FTA) testing were excluded from all analyses. Individuals with peripheral pitting or nonpitting edema were excluded only for the analyses of pin prick, light touch, and vibration. Numeric differences in the following tables are due to missing dependent variables or covariate data. The exclusions and missing covariate data are summarized in Table 11-1. The unadjusted analyses used chi-square or Fisher's exact test for frequency table analyses. Adjusted analyses were not performed where only sparse numbers of abnormalities were found. Logistic regression models were used in all adjusted analyses. Parallel analyses using Original Comparisons can be found in Appendix I, Tables 1-3 through 1-13. RESULTS AND DISCUSSION General Detailed neurological data were obtained on all participants by standard physical examination techniques. Four board-certified SCRF neurologists, all 11-3 �TABLE 11-1. Exclusions and Hissing Data for Neurological Assessment by Group Group Data Category Ranch Hand Comparison Total Lifetime Alcohol History (Drink-Years); Missing Data 39 40 79 Peripheral Edema (Exclusion Category for Pin Prick, Light Touch, and Ankle Vibration) 13 16 29 Diabetic Class (Missing Data) 0 4 4 Positive Syphilis Serology (RPR and FTA) Exclusion Category 0 1 1 blinded to the exposure status of the participants, conducted the examinations. Data were collected to assess three specific clinical areas: cranial nerve function, peripheral nerve function, and central nervous system (CNS) function. The analyses in this chapter are presented in the order of these functional areas. The unadjusted statistical analyses presented in this chapter are straightforward group contrasts of dichotomous (normal/abnormal) dependent variables using Fisher's exact test. Logistic regression models for adjusted analyses used the covariates of age (born in or after 1942, born between 1923 and 1941, born in or before 1922), race (Black, nonblack), occupation (OCC) (officer, enlisted flyer, enlisted groundcrew), diabetic class (DIAB) (normal, less than 140 mg/dl glucose? impaired, at least 140 mg/dl but less than 200 mg/dl glucose; diabetic, greater than or equal to 200 mg/dl glucose or past diabetic history), lifetime alcohol use (DRKYR) (total drink-years: 0, greater than 0 to 50, greater than 50), and unprotected exposure to insecticides (INS) (recorded as yes/no, excluding herbicide exposure). The models are "best-fit" following a step-down strategy beginning with all two-way interactions among the six covariates. Only variables with a substantial number of abnormalities were analyzed. Several summary indices were constructed for functionally related variables with low counts of abnormalities. A summary index was created for the cranial nerve function by combining the 15 cranial nerve parameters into a single index, which was classified as normal if all parameters were normal. Another cranial nerve function was created in a similar fashion, excluding neck range of motion due to the much higher percentage of abnormalities found for this variable relative to the other parameters. The four coordination parameters of the central nervous 11-4 �system were similarly combined to form a summary index. These constructed indices are presented more for the purpose of inspection than for inference making. Since the corneal reflex (as one measure of the trigeminal nerve function) contained no abnormalities for either group, no table is presented with this variable. The statistical power to detect a given relative risk in many of the subsequent analyses was somewhat limited. With the use of a two-sided cfr-level of 0.05 and power of 0.80, the sample sizes were sufficient to detect a 49 percent increase in the frequency of abnormal values for neck range of motion, a 69 percent increase for light touch but only a doubling for tremor, and an elevenfold increase for gag reflex. Power was generally poor in these analyses because of the extremely small number of abnormalities observed in both the Ranch Hand and Comparison groups. Questionnaire Data For the interval questionnaire, each participant was asked to update his health history for neurologic conditions occurring between 1982 and 1985. All affirmative histories were subjected to medical record verification, and appropriate ICD-9-CM coding. All verified neurological diseases were placed into six broad disease categories. These data are summarized in Table 11-2. TABLE 11-2. Unadjusted Analysis for Verified Neurological Disease by Group*—1982-1985 Group Abnormalities Ranch Hand Disease Category Inflammatory Diseases Hereditary and Degenerative Diseases Peripheral Disorders Disorders of the Eye Disorders of the Ear Other Disorders Comparison Number Percent Number Percent Total p-Value** 0 2 0.0 0.2 0 0 0.0 0.0 0 2 — 0.194 18 5 6 8 1.8 0.5 0.6 0.8 27 7 7 3 2.1 0.5 0.5 0.2 45 12 13 11 0.651 0.999 0.999 0.069 *Based on 1,016 Ranch Hands and 1,293 Comparisons; some participants may be classified in more than one category. **Fisher's exact test. 11-5 �All of these analyses were based on very small numbers of abnormalities, but none of the six general disease categories showed statistically significant differences between groups, although the marginal significance of the Other Disorders category is of interest. To determine whether lifetime differences in neurologic disease exist between the Ranch Hand and Comparison groups, verified followup data were combined with verified Baseline historical data. This tabulation is presented in Table 11-3. TABLE 11-3. Unadjusted Analysis for Verified Neurological Disease by Group*—Baseline and First Followup Studies Combined Group Abnormalities Ranch Hands Disease Category Comparisons Number Percent Number Percent Inflammatory Diseases Hereditary and Degenerative Diseases Peripheral Disorders Disorders of the Eye Disorders of the Ear Other Disorders Total p-Value** 0.660 0.999 3 2 0.3 0.2 2 3 0.2 0.2 5 5 23 2.3 1.6 2.4 1.5 2.9 1.8 2.2 1.1 61 39 0.361 . 16 24 15 38 23 53 29 0.889 29 14 0.747 0.453 *Based on 1,016 Ranch Hands and 1,293 Comparisons; some participants may be classified in more than one category. **Fisher's exact test. Like the followup data, the combined data revealed no statistically significant differences in any disease category. Also, there was no significant difference in patterns of disease for each group (p=0.721). Physical Examination Data Dependent Variable and Covariate Relationships: Cranial Nerve Function, Peripheral Nerve Status, and Central Nervous System Coordination Responses from both groups were combined and analyzed with the six covariates. In addition, current drinking (yes/no) and lifetime history of 11-6 �unprotected exposure to industrial and degreasing chemicals (yes/no) were also evaluated. Indices constructed from dependent variables from the cranial nerve function and central nervous system coordination processes were also included. A summary tabulation of covariate associations is shown in Table 11-4. The 10 variables in this table include variables from the peripheral nerve status and CNS process as well as the cranial nerve function and constitute the subset of variables for which adjusted analyses were performed. These results generally showed the profound association of classical risk factors for neurological deficits. Increases in the percentages of abnormalities for Achilles reflex, muscle status, neck range of motion, and the cranial nerve function index (which included neck range of motion) were associated with increases in age. Increasing percentages of abnormalities for pin prick and light touch were noted for increasing age from the young category (3.4% and 2.7% for pin prick and light touch, respectively) to the middle-aged category (8.1% and 4.7%, respectively), but a declining proportion of abnormalities was observed from the middle- to older-age categories (7.3% and 1.2%, respectively). No age effect was noted for gait, the CNS index, the cranial nerve index (neck range of motion excluded), and, surprisingly, for tremor. Race was not a significant covariate for any dependent variable. A significant occupational effect was observed for the CNS summary index (p=0.021, with both enlisted categories having a higher frequency of abnormalities [5.7% and 4.1% for enlisted flyers and enlisted groundcrew, respectively] than the officer category [2.6%]) and for the neck range of motion variable (p=0.010, with increasing proportions of abnormalities from the enlisted groundcrew [4.6%] to officers [7.5%] to enlisted flyers[8.0%]). Abnormalities in the Achilles tendon reflex were related to a graduated increase in drink-years of alcohol. For the variables of pin prick, light touch, muscle status, neck range of motion, and cranial nerve index (with neck range of motion included), the 0 drink-year category was related to a higher frequency of abnormalities than the greater than 0 to 50 drink-year category, which in turn was associated with a lower frequency of abnormalities than the greater than 50 drink-year category. For the current drinker (which was not used for modeling), the percentage of abnormalities for Achilles reflex and gait was significantly greater (p=0.007 and p=0.001 for Achilles reflex and gait, respectively) for current nondrinkers than for current drinkers. This relationship was reversed for the CNS summary index. For both the Achilles tendon reflex and the response to pin prick, the frequencies of abnormalities significantly increased from the diabetic classes of normal to impaired to diabetic (p<0.001 for both variables). For the variables of light touch, muscle status, gait, and CNS summary index, the associations with diabetic status were mixed: The normal diabetic class had a higher proportion of abnormalities than the impaired stratum which, in turn, had a lower proportion of abnormalities than the overtly diabetic class. Unexpectedly, the proportion of tremor abnormalities was highest for the normal diabetic class and became, successively lower in the impaired and diabetic strata (2.48%, 0.45%, and 0%, respectively). A higher proportion of pin prick abnormalities was associated with a history of unprotected exposure to insecticides (p=0.040; 6.94% for exposed versus 4.8% for unexposed). The other dependent variables were not 11-7 �TABLE Three Summary TiKyippg 3nd the Govariates in the Combined Ranch Band and Comparison Groups Covariate Exposure Degreasing Chemicals* NS 000 .5 NS <0.001 000 .4 NS NS 006 .2 NS** NS NS Race Occupation Total Drink-years Achilles Reflex O01 .0 NS NS 002 .2 007 .0 <0.001 Pin Prick 4.0 001 NS NS 004 .0 NS 007 .2 NS NS 006 .0 NS O01 .0 NS NS 001 .0 NS** O01 .0 NS 005 .2 Gait NS NS NS NS 001 .0 003 .3 NS NS NS CNS Index NS NS 001 .2 NS 002 .1 006 .1 NS NS NS Tremor NS NS NS NS NS 001 .1 NS NS NS Neck Range of Motion O01 .0 NS 000 .1 004 .1 NS NS** NS 009 .3 NS Cranial Nerve Function Index <0.001 NS NS** 0.032 NS NS NS NS** NS NS NS NS NS NS NS light Touch Muscle Status Cranial Nerve Function Index (Neck Range of Motion Excluded) NS: Not significant ( > . 0 . p01) * Variable not used in adjusted analyses. NS**: Borderline significant ( . 5 < p <0.10). 00 NS** Current Drinking* Industrial Chemicals* Age Dependent Variable Diabetic Class Insecticides NS NS** NS �significantly affected by the insecticide covariate. For most dependent variables, both Ranch Hands and Comparisons exposed to degreasing or industrial chemicals exhibited a smaller percentage of abnormalities than participants without exposure. Because the biologic basis of these findings is not readily apparent, these two variables were not used as adjusting covariates. Cranial Nerve Function All 12 cranial nerves were assessed as unilateral or bilateral; these unadjusted data are presented in Table 11-5. All bilateral assessments (e.g., right visual field, left visual field) were combined for the analyses; an abnormality consisted of a right and/or a left abnormality. The analysis of the 12 variables and two cranial nerve function summary indices did not reveal statistically significant group differences. Since no abnormalities are present for the variables of speech and tongue position in the Comparison group, the estimated relative risk for these variables was approximated by adding 0.5 to each cell. The low frequency of abnormal counts in all variables, except neck range of motion, contrasts with the 1982 Baseline findings, which found substantially more abnormalities. For example, ocular movement was recorded as abnormal in more than 30 percent of the participants at Baseline while only 0.7 percent of participants were found to be abnormal at followup. Because of the few abnormalities for all variables except neck range of motion, two summary indices of cranial nerve function were constructed. One indicated whether or not a participant is abnormal for any of the 15 variables, while the other was a composite for all except neck range of motion. The analyses of these indices are reflected in Table 11-5, and showed no statistically significant group differences, although the index excluding neck range of motion is of borderline significance. Speech and tongue position relative to midline were also of borderline significance, although the analysis was affected by sparse numbers of abnormalities. The constructed indices are presented more for the purpose of inspection than for inference making. Because of sparse numbers of abnormalities, adjusted analyses were performed only on the variable neck range of motion and the cranial nerve function summary indices, with and without neck range of motion data. The results of these analyses are given in Table 11-6. None of the results were statistically significant, although the cranial nerve function index, without neck range of motion, was marginally significant (p=0.061) when participants with missing drink-years were included. In the primary adjusted analysis for this variable, drink-years was included in a significant covariate interaction. However, an alternative model was also examined that included participants with missing drink-years due to the disparity in group response for these participants (4 out of 39 Ranch Hands abnormal, 0 out of 40 Comparisons abnormal). The results of these adjusted analyses are nearly identical to the unadjusted analyses (see Table 11-5). A borderline significant result of a group (GRP)-by-age interaction (p=0.0501) for neck range of motion existed, and an additional analysis stratifying by age is provided in Table 11-7. This table presents the results of interaction analyses from variables assessing the peripheral nerve status and central nervous system coordination process as well. 11-9 �TSBLE11-5. unadjusted Analyses for Cranial Nerve Function by (koup Group Ranch Hand Comparison Statistic Number Percent Number Percent I Olfactory n Abnormal Normal 1,016 10 1,006 1.0 99.0 1,292 10 1,282 0.8 99.2 1.27 ( . 3 3 0 ) 0.654 05,.7 n n Abnormal Normal 1,016 6 1,010 0.6 99.4 1,292 6 1,286 0.5 99.5 1.27 ( . 1 3 % 0.774 04,.) Light Reaction m n Oculomotor Abnormal Normal 1,015 8 1,007 0.8 99.2 1,289 9 1,280 0.7 99.3 1.13 ( . 3 2 9 ) 0.811 04,.4 1,016 6 1,010 0.6 99.4 1,292 10 1,282 0.8 99.2 0.76 ( . 8 2 1 ) 0 8 1 02,.0 .0 Ocular Movements m n Oculomotor Abnormal IV Normal Trochlear VI Abducens Facial Sensation V n Trigeminal Abnormal Normal 1,014 4 1,010 04 . 99.6 1,290 2 1,288 0.2 99.8 2.55 ( . 7 1 . 5 0 4 5 04,39) .1 Jaw Clench V n Trigeminal Abnormal Normal 1,016 2 1,014 0.2 99.8 1,292 2 1,290 0.2 99.8 1.27 ( . 8 9 0 ) 0.999 01,.5 n Abnormal Normal 1,016 7 1,009 0.7 99.3 1,292 4 1,288 0.3 99.7 2.23 ( . 7 7 4 ) 0.230 06,.1 n Abnormal Normal 1,015 7 1,008 0.7 99.3 1,292 7 1,285 0.5 99.5 1.28 ( . 5 3 6 ) 0.789 04,.5 n vm Acoustic . Abnormal Normal 1,015 2 1,013 0.2 99.8 1,292 1 1,291 0.1 99.9 2.55 ( . 3 2 . 5 0.586 02,81) Variable Cranial Nerve Smell Visual Fields Smile Optic vn Facial Palpebral Fissures Balance vn Facial 11-10 Est. Relative Risk ( 5 C.I.) p-Value 9% �TABLE 11-5. (continued) Unadjusted Analyses for Cranial Nerve Ruction by Group Group Ranch Hand Comparison Number Percent Number Percent Est. Relative Risk ( 5 C.I.) p-Value 9% Variable Cranial Nerve Gag Reflex IX n Abnormal Glossopharyngeal Normal 1,014 1 1,013 0.1 99.9 1,291 1 1,290 0.1 99.9 1.27 ( . 8 2 . 8 0.999 00,03) Speech X Vagus n Abnormal Normal 1,016 3 1,013 0.3 99.7 1,291 0 1,291 00 . 100.0 8.92 ( . 6 1 2 8 ) 0.085 04,7.9° Tongue Position Relative to Midline X Vagus n Abnormal Normal 1,015 3 1,012 0.3 99.7 1,292 0 1,292 00 . 100.0 8.94 ( . 6 173. 19)a 0.085 04, Palate and Uvula Movement XI Spinal Accessory n Abnormal Normal 1,014 2 1,012 0.2 99.8 1,291 1 1,290 0.1 99.9 2.55 ( . 3 2 . 6 0.586 02,81) Neck Range of Motion XH n Hypoglossal Abnormal Normal 1,016 61 955 6.0 94.0 1,292 84 1,208 6.5 93.5 0.92 ( . 5 1 2 ) 06,.9 0.666 Cranial Nerve Rnction Index n Abnormal Normal 1,003 96 907 9.6 90.4 1,275 115 1,160 9.0 91.0 1.07 ( . 0 1 4 ) 08,.2 0.663 Cranial Nerve Rjnction Index (Neck Range of Motion Excluded) n Abnormal Normal 1,003 42 961 4.2 95.8 1,275 35 1,240 2.7 97.3 1.55 ( . 8 2 4 ) 09,.4 0.062 Statistic "Estimated relative risk and 95% confidence interval calculated after adding 0.5 to each cell. 11-11 �TABLE 11-6. Adjusted Analyses for Selected Variables of Cranial Nerve Function by Group Ranch Band Variable Est. Relative .. Statistic Njnber Percent Number Percent Risk<95* C I ) p-Value Neck Range of Motion n 1,016 Abnormal 61 Normal 955 1,292 84 6.0 94.0 1,208 6.5 93.5 Cranial Nerve Function Index n 1,003 Abnormal % Normal 907 9.6 90.4 1,275 115 1,160 Covariate Remarks* 9.0 91.0 0.90 ( . 3 1 2 ) 0 5 1 06,.7 . 3 AGE(pO.OOl) GRP*AGE (nBrginal:pM).0501) 1.07 ( . 0 1 4 ) 0.666 08,.2 AGE(pO.OOl) Cranial n 964 1,232 08,.0 . 5 1.42 ( . 8 2 3 ) 0 1 3 DIAB*INS(p=0.022) OCC*EROR(p=0.011) 34 Nerve 38 3.9 Abnormal 2.8 Rnction Normal 926 96.1 1,198 97.2 OCC*DIAB(p=0.015) Index (Neck Alternative Model—Includes Missing Drink-Year Participants*fb Range of Motion n 1,271 1.56 ( . 8 2 4 ) 0 0 1 DIAB*INS(p=0.017) 09,.9 . 6 1,003 Excluded) Abnormal 42 4.2 34 2.7 OCC*DIAB(p=0.016) Normal 961 95.8 1,237 97.3 ^Abbreviations: GBP: group DIAB: diabetic class INS: insecticide exposure OCC: occupation EFKXR: drink-years 'lifetime alcohol consunption (total drink-years) not used as a covariate. b 79 missing drink-year participants: 4/39 Ranch Hands abnormal; 0/40 Comparisons abnormal. 11-12 �TftBLE 11-7. Sunmary "teble of Gtoup-by-Cbvariate Interactions for Neurological Variables Group Ranch Hands Variable Interaction Comparisons Adj. Relative Stratification Statistic Number Percent Number Percent Risk ( 5 C.I.) p-Value 9% n Abnormal Normal 412 10 402 2.4 97.6 549 5 0.9 3.03 ( . 2 9 0 ) 0.045 10,.0 544 99.1 n Abnormal Nornal 568 47 521 8.3 91.7 693 70 1 . 0.82 (0.55,1-21) 0.319 01 623 8 . 99 Bom <$1922 n Abnormal Normal 36 4 32 11.1 88.9 50 9 18.0 41 82.0 ( . 5 ( . 6 1.97) 0.361 05 01, Abnormal n Abnormal Normal 76 13 63 17.1 82.9 94 10 84 1.74 ( . 1 4.24) 0.223 07, n Abnormal Normal 105 1 104 1.0 99.0 n Abnormal Norual 822 45 777 5.5 9. 45 n Abnormal Normal 703 22 681 3.1 9. 69 683 8 1.2 2.60 ( . 5 5 9 ) 0.022 11,.0 675 9 . 88 Group-byInsecticides Exposure Not Exposed , n to Insecticide Abnormal Nonral 313 4 309 1.3 98.7 605 11 1.8 0 6 ( . 2 2 1 ) 0.532 .9 02,.9 594 98.2 Bom > 1942 Neck Range Group-byof Motion Age Pin Prick Group-byDiabetic Class Bom 1923-1941 Impaired Normal Exposed to Insecticides Tremor 11-13 10.6 89.4 174 16 9.2 0.09 ( . 1 0.69) 0.021 00, 158 90.8 1,005 53 5.3 952 94.7 1.02 ( . 8 1 5 ) 0.920 06,.4 �The stratified analysis for neck range of motion showed a higher proportion of younger Ranch Hands with neck range of motion abnormalities than younger Comparisons (p=0.045). Although not statistically significant, middle-aged and older Comparisons had higher proportions of abnormalities than did the Ranch Hands. Peripheral Nerve Status Peripheral nerve integrity was assessed by light pin prick, light touch (cotton sticks), visual inspection (and palpation, if indicated) of muscle mass, vibratory sensation as measured at the ankle with a tuning fork of 128 Hz, three deep tendon reflexes (patellar, Achilles, and biceps), and the Babinski reflex. The unadjusted analyses are given in Table 11-8. As noted previously, the analyses of pin prick, light touch, and vibratory sensation excluded the 29 participants with peripheral edema. These results showed that peripheral nerve function did not vary significantly by group. Adjusted analyses were performed by logistic regression on four peripheral nerve variables. The other variables had relatively sparse numbers of abnormalities. The covariates were age, race, occupation, drink-years of alcohol, diabetic class, and exposure to insecticides. These statistics are displayed in Table 11-9. For the variables light touch, muscle status, and the Achilles reflex, group differences were nonsignificant; the results were nearly identical to the unadjusted analyses. For the variable pin prick, however, a significant group-by-diabetic class interaction (p=0.003) was observed. This interaction was explored and the results are depicted in Table 11-7. As shown, the interaction suggests a difference, due to a lower proportion of abnormal pin-prick results in Ranch Hand impaired diabetics than in Comparisons (Adj. RRs 0.09,95% C.I.J [0.01,0.69], p=0.021), whereas both the abnormal and normal diabetic classes showed no significant group differences. Central Nervous System Coordination CNS coordination was evaluated clinically with four variables: hand tremor, rapid finger-to-nose coordination, one-foot standing balance (modified Romberg sign), and observation of gait for at least 10 steps. In addition, a constructed variable, the CNS summary index, was derived by summarizing abnormalities from all four CNS variables. The unadjusted analyses of these five variables are shown in Table 11-10. These results revealed no statistically significant group differences for the four primary CNS variables, although the borderline significance of tremor, with a higher proportion of abnormalities in the Ranch Hands, is interesting. The statistical power to detect a given relative risk was poor because of the small percentages of abnormalities. The CNS summary index was statistically significant, with Ranch Hands manifesting a higher proportion of abnormalities; this result should be interpreted with caution, however, si'nce this index was constructed after the data were examined. Three of the five variables with sufficient proportions of abnormalities were adjusted by six covariates, and these results are summarized in Table 11-11. 11-14 �TABLE 11-8. Unadjusted Analyses for Peripheral Nerve Function by Group Group Ranch Hand Variable Comparison Statistic Number Percent Number Percent Est . Relative Risk (95% C.I.) p-Value n Abnormal Normal 1,003 59 944 5.9 94. 1 1,276 80 1,196 6.3 93.7 0.93 (0.66,1 .32) 0.725 Light Touch n Abnormal Normal 1,003 38 965 3.8 96.2 1,276 47 1,229 3.7 96.3 1.03 (0.67,1 .59) 0.912 Muscle Status n Abnormal Normal 1,016 26 990 2.6 97.4 1,292 33 1,259 2.6 97.4 1.00 (0.60,1 .69) 0.999 Vibratory Sensation n 1,003 11 Abnormal Normal 992 1. 1 98.9 1,276 10 1,266 0.8 99.2 1.40 (0.59,3 .32) 0.510 Patellar Reflex n Abnormal Normal 1,016 1. 1 98.9 1,290 16 1,274 1.2 98.8 0.87 (0.40,1 .89) 0.846 Achilles Reflex n Abnormal Normal 1,009 5.7 94.3 1,284 75 1,209 5.8 94.2 0.98 (0.69,1 .40) 0.999 Biceps Reflex n Abnormal Normal 1,016 0.9 99.1 1,292 10 1,282 0.8 99.2 1.15 (0.46,2 .83) 0.819 n Abnormal Normal 1,011 0.4 99.6 1,287 5 1,282 0.4 99.6 1.02 (0.27,3 .80) 0.999 Pin Prick Babinski Reflex 11 1,005 58 951 9 1,007 4 1,007 11-15 �TfiBUBll-9. Adjusted Analyses for Selected Variables of Peripheral Nerve Ruction by Group Group Ranch Hand Comparison Statistic Number Percent ftnfcer Percent Pin Prick n Abnormal Normal 1,003 59 944 5.9 94.1 1,273 79 1,194 n Abnormal Normal 964 37 927 3.8 96.2 1,236 46 1,190 n Abnormal Normal 977 25 952 2.6 97.4 1,248 31 1,217 2.5 97.5 Achilles Reflex n Abnormal Normal 971 56 915 5.8 94.2 1,240 71 1,169 5.7 94.3 **** 3.7 96.3 Muscle Status p-Value 6.2 93.8 Light Touch Adj. Relative Risk (95XC.I.) **** Variable presented. Covariate Remarks GRP*DIAB(pd0.003) AGE(p<D.001) 1.02 (0.65,1.60) 0.921 OCC*RACE(p=0.013) AGE(p=0.043) EROR(p=0.031) 1.00 (0.57,1.75) 0.999 EROR*AGE(p=0.009) DIAB*INS(p4).039) 1.00 (0.69,1.45) 0.999 DRKXR*OCC(p=0.016) AGE(p<0.001) DIAB(p<D.001) interaction—adjusted relative risk, confidence interval, and p-value are not 11-16 �TABLE 11-10. Unadjusted Analyses for CNS Coordination Variables by Group Group Ranch Hand Variable Tremor Comparison Statistic Number Percent Number Percent Est. Relative Risk (95% C.I.) p- Value n Abnormal Normal 1,016 26 990 2.6 97.4 1,292 19 1,273 1.5 98.5 1.76 (0.97 ,3.20) 0.069 n Abnormal Normal 1,015 9 1,006 0.9 99.1 1,292 7 1,285 0.5 99.5 1.64 (0.61 ,4.43) 0.327 Romberg Sign n 1,015 2 Abnormal Normal 1,013 0.2 99.8 1,292 1 1,291 0.1 99.9 2.55 (0.23 ,28.15) 0.586 Gait n Abnormal Normal 1,016 20 996 2.0 98.0 1,290 16 1,274 1.2 98.8 1.60 (0.82 ,3.10) 0.178 n Abnormal Normal 1,015 48 967 4.7 95.3 1,290 39 1,251 3.0 97.0 1.59 (1.04 ,2.45) 0.036 Coordination CNS Summary Index 11-17 �TfiHE 11-11. Adjusted Analyses for Selected Variables of CMS Coordination by Group Group Ranch Hand "Variable Tremor Gait CNS Summary Index Comparison Adj. Relative Statistic Number Percent Number Percent Risk (95% C.I.) p-Value Covariate Remarks* n 1,016 Abnormal 26 Normal 990 2.6 97.4 1,288 19 1,269 1.5 98.5 1.70 ( . 3 3 0 ) 0 0 0 09,.9 .8 GRP*INS (marginal:p=0.055) DIAB(p=0.001) 977 20 957 2.0 98.0 1,246 15 1,231 1.2 98.8 1.74 ( . 8, . 7 0.110 08 3 4 ) DIAB(p=0.030) ERKXR*INS(p=0.047) n 1,015 Abnormal 48 967 Normal 4.7 95.3 1,286 38 1,248 3.0 97.0 1.57 (1.01,2.43) 0.042 DIAB(p=0.003) OCC(P=0.018) n Abnormal Normal These statistics were quite similar to the unadjusted tests, and showed borderline significance for tremor, nonsignificance for gait, and significance for the CNS summary index. The unexpected inverse relationship of tremor abnormalities to diabetic classification is again noted. The borderline group-by-insecticide interaction was investigated, and the results are given in Table 11-7. As shown, the relative risk for Ranch Hands exposed to insecticides was statistically significant (RR: 2.60, 95% C.I.: [1.15,2.90], p=0.022), whereas the relative risk for unexposed Ranch Hands was nonsignificant. This finding may have both an operational and biologic foundation, because records indicate that some Ranch Hands were exposed to the insecticide Malathion®, a cholinesterase inhibitor, during insecticide missions for malaria prevention. Comparisons, by definition, did not fly these missions. EXPOSURE INDEX ANALYSES Exposure index analyses were conducted within each occupation cohort of the Ranch Hand group to search fpr dose-response relationships (see Chapter 8 for details on the exposure index). All 27 variables and three summary indices were explored (unadjusted for any covariates) as with the unadjusted tests for group differences discussed previously in this chapter. These variables were investigated using Pearson's chi-square test and Fisher's exact 11-18 �test. Adjusted analyses were performed by logistic regression for the 10 variables (7 neurological parameters and 3 summary indices) for which adjusted analyses of group differences were previously examined. These analyses were accomplished, adjusted for age, diabetic class, insecticide exposure, and drink-years (all discretized), and any significant pairwise interactions between the exposure index and these covariates. Race was not included in adjusted analyses because of the absence of any race effect in the previous group difference analyses. Overall significance in the proportion of abnormalities among the exposure index levels of low, medium, and high was determined, as well as contrasts in the proportion of abnormalities between the medium and low exposure levels, and between the high and low exposure levels. Exclusions were made as described previously. Results of the adjusted analysis are presented in Table 11-12, and results for unadjusted analyses appear in Table 1-1 of Appendix I. Results from further study of exposure index-by-covariate interactions are given in Table 1-2 of Appendix I. Unadjusted analyses revealed borderline significant differences among exposure index levels for pin prick in enlisted groundcrew (p=0.052) and Achilles reflex in enlisted flyers (p=0.059). The data did not support an increase in the proportion of abnormalities with increasing exposure levels, however. Adjusted analyses yielded similar conclusions, in that significant or borderline significant results did not support an increase in the proportion of abnormalities with increasing exposure, and that very few significant results were observed. The pattern of abnormalities with the 10 variables was studied, and in no occupational strata was an increasing dose-response relationship evident. In fact, the high exposure level often had a smaller (although nonsignificant) proportion of abnormalities than the low and medium levels. Interactions were present for 5 of the 10 variables, and occurred primarily in the enlisted groundcrew stratum. A summary of these interactions is presented in Table 11-13. Meaningful interpretation of the interactions was difficult, due to the small numbers of abnormalities within a covariate strata. No significant adverse effects to participants with higher exposure levels were evident, however, in this analysis. In summary, no evidence of an increasing dose-response relationship at the followup examination was observed. No increase in prevalence rates was seen as exposure levels increased. These results essentially were in agreement with the findings of the Baseline Study. 11-19 �TABLE 11-12. Adjusted Exposure Index Analyses for Neurological Variables by Occupation Exposure Index Variable Occupation Low Total Medium Total High Total Contrast Adj. Relative Risk ( 5 C.I.) p-Value 9% Officer 127 120 Overall M vs. L H vs. L 096 .0 0.82 (0.31,2.18) 0.686 0.97 (0.37,2.56) 0.955 Enlisted Flyer 51 61 53 Overall M vs. L H vs. L 0.940 0.79 (0.20,3.20) 0.744 0.83 (0.21,3.31) 0.786 Enlisted Groundcrew 148 160 132 Overall M vs. L H vs. L 0.299 0.93 (0.27,3.21) 0.908 0.36 (0.09,1.51) 0.163 Officer Neck Range of Motion 125 120 127 119 Overall M vs. L H vs. L 0.551 0.63 (0.28,1.44) 0.277 0.78 (0.35,1.78) 0.560 Enlisted Flyer 51 60 53 Overall M vs. L H vs. L 0.808 1.00 (0.29,3.43) 0.999 0.68 (0.18,2.59) 0.569 145 158 131 Overall M vs. L H vs. L **() **! **() **! H» g Cranial Nerve Function Index Enlisted Groundcrew **() **! **() **! **() **! �TABLE 11-12. (continued) Adjusted Exposure Index Analyses for Neurological Variables by Occupation Exposure Index Variable Occupation Low Total Medium Total High Total Contrast Adj. Relative Risk (95% C.I.) p-Value 120 127 119 Overall M vs. L H vs. L 0.148 0.30 (0.08,1.22) 0.093 0.36 (0.09,1.45) 0.150 51 60 53 Overall M vs. L H vs. L 0.860 1.04 (0.13,8.27) 0.969 0.56 (0.05,6.58) 0.642 145 158 131 Overall M vs. L H vs. L 0.894 0.75 (0.23,2.45) 0.639 0.84 (0.25,2.76) 6.773 Officer 124 124 119 Overall M vs. L H vs. L 0.277 0.43 (0.13,1.38) 0.156 0.49 (0.17,1.43) 0.191 Enlisted Flyer 51 60 53 Overall M vs. L H vs. L 0.399 0.33 (0.05,2.35) 0.267 1.02 (0.23,4.60) 0.979 146 159 128 Overall M vs. L H vs. L 0.108 0.86 (0.32,2.34) 0.765 0.28 ( . 7 1 0 ) 0.062 00,.7 Officer Cranial Nerve Enlisted Function Flyer (Neck Range of Motion Excluded) Enlisted Groundcrew Pin Prick Enlisted Groundcrev �TABLE 11-12. (continued) Adjusted Exposure Index Analyses for Neurological Variables by Occupation Exposure Index Variable Occupation Low Total Medium Total High Total Contrast Adj. Relative Risk ( 5 C.I.) 9% p-Value Officer Light Touch 124 124 119 Overall M vs. L H vs. L 0.047 0.39 (0.11,1.40) 0.148 0.20 ( . 5 0 8 ) 0.027 00,.3 Enlisted Flyer 51 60 53 Overall M vs. L H vs. L ****(2) ****(2) ****(2) ****(2) ****(2) Enlisted Groundcrew 146 159 128 Overall M vs. L H vs. L 0.777 1.27 (0.34,4.80) 0.725 0.74 (0.16,3.35) 0.699 Officer ,_, i-1 125 127 120 Overall M vs. L H vs. L 0.105 0.15 (0.02,1.01) 0.051 0.57 (0.14,2.30) 0.433 Enlisted Flyer 51 61 53 Overall M vs. L H vs. L 0.979 0.90 (0.04,22.10) 0.946 0.74 (0.04,14.77) 0.841 148 160 132 Overall M vs. L H vs. L ****(3) ****(3) to Muscle Status Enlisted Groundcrew ****(3) ****(3) ****(3) �TABLE 11-12. (continued) Adjusted Exposure Index Analyses for Neurological Variables by Occupation Exposure Index Variable Occupation Low Total Medium Total High Total Contrast Adj . Relative Risk (95% C.I.) p-Value Officer Achilles Reflex 122 126 120 Overall M vs. L H vs. L 0.384 0.43 (0.13,1-46) 0.175 0.65 (0.21,1-99) 0.448 Enlisted Flyer 51 60 53 Overall M vs. L H vs. L 0.021 0.65 (0.16,2.76) 0.564 Overall M vs. L H vs. L **() **3 **() **3 ****(3) ****(3) **() **3 Enlisted Groundcrew 147 160 132 Officer 125 127 120 Overall M vs. L H vs. L 0.219 0.19 ( . 2 1 6 ) 0.132 00,.6 0.63 ( . 4 2 8 ) 0.548 01,.9 Enlisted Flyer 51 61 53 Overall M vs. L H vs. L 0.625 2.11 (0.19,23.39) 0.542 2.95 (0.29,30.43) 0.364 148 160 132 Overall M vs. L H vs. L 0.396 0.91 (0.22,3.66) 0 8 9 .8 0.28 (0.03,2.44) 0.248 to u> Tremor Enlisted Groundcrew �TABLE 11-12. (continued) Adjusted Exposure Index Analyses for Neurological Variables by Occupation Exposure Index Variable Occupation Low Total Medium Total High Total Contrast Adj . Relative Risk ( 5 C.I.) 9% p-Value Officer Gait 125 127 120 Overall M vs. L H vs. L 0.483 0.26 (0.02,3.25) 0.298 0.89 (0.12,6.76) 0.912 Enlisted Flyer 51 61 53 Overall M vs. L H vs. L 0.188 0.64 ( . 7 6 0 ) 0.693 00,.5 Enlisted Groundcrew ^ 148 160 132 Overall M vs. L fl vs. L 0.576 0.42 (0.07,2.51) 0.343 0.88 (0.19,3.99) 0.868 Officer M 125 127 120 Overall M vs. L H vs. L 0.123 0.22 (0.04,1.10) 0.066 0.57 (0.15,2.10) 0.399 Enlisted Flyer 51 60 53 Overall M vs. L H vs. L 0.930 1.21 (0.25,5.92) 0.818 0.90 (0.17,4.80) 0.899 148 160 132 Overall M vs. L H vs. L ****(2) ****(2) to -JS CNS Summary Index Enlisted Groundcrew **() **2 ****(2) ****(2) —No abnormal participants present in medium exposure index level for Achilles reflex (or high level for gait) in enlisted flyers. ****(l)Exposure index-by-diabetic class interaction—relative risk and p-value not presented. ****(2)Exposure index-by-insecticide exposure interaction—relative risk, confidence interval, and p-value not presented. ****(3)Exposure index-by-age interaction—relative risk, confidence interval, and p-value not presented. �TABLE 11-13. Summary of Exposure Index-by-Covariate Interactions for Neurological Variables p-Value Variable Occupation Covariate CNF Summary Index Enlisted Groundcrew Diabetic Class 0.045 Light Touch Enlisted Flyer Insecticide Exposure 0.026 Muscle Status Enlisted Groundcrew Age 0.026 Achilles Reflex Enlisted Groundcrew Age 0.014 CNS Summary Index Enlisted Groundcrew Insecticide Exposure 0.010 LONGITUDINAL ANALYSES Two variables, the modified Romberg sign and the Babinski reflex, were investigated to assess longitudinal differences between the 1982 Baseline examination and the 1985 followup examination. Both variables were classified as abnormal or normal. As shown in Table 11-14, 2x2 tables were constructed for each group for each variable. This table shows the number of participants who were abnormal at the Baseline examination and abnormal at the followup examination, abnormal at Baseline and normal at the followup, normal at Baseline and abnormal at the followup, and normal at both Baseline and the followup. The odds ratio is the ratio of the number of participants who were normal at Baseline and abnormal at the followup to the number of participants who were abnormal at Baseline and normal at the followup (the "off-diagonal" elements). The p-value was derived from Pearson's chi-square test of the hypothesis that there was comparable change in the two groups over time. These data showed no longitudinal difference in the change pattern in the Romberg sign in the two groups, but they did show a significant change in the Babinski reflex. In the Baseline examination, the Ranch Hands had a significantly greater proportion of reflex abnormalities than the Comparisons, but the followup examination showed approximately the same percentage of abnormality in both groups (Est. RR: 1.02, 95% C.I.s [0.27,3.80, p=0.999]). SUMMARY AND CONCLUSIONS Interval questionnaire data (1982 through 1985) on neurological illnesses, verified by medical records, revealed no significant group differences. These data were added to verified Baseline historical information to assess possible differences in the lifetime experience of neurological disease. Again, there was no significant difference between the Ranch Hand and Comparison groups. 11-25 �TABLE 11-14. Longitudinal Analysis of Romberg Sign and Babinski Reflex: A Contrast of Baseline and First Followup Examination Abnormalities Group Variable 1982 Baseline Exam 1985 Followup Exam Odds p-Value Abnormal Normal Ratio (OR)* (ORPH vs. OR^ K it C Ranch Hand Abnormal Normal 2 0 188 777 0 Comparison Abnormal Normal 0 1 250 886 0.004 Ranch Babinski Hand Reflex Comparison Abnormal Normal 1 3 7 953 0.43 Abnormal Normal 0 1 1,129 5.00 Romberg Sign 0.38 *0dds Ratio: 0.04 5 Number Normal Baseline, Abnormal Followup Number Abnormal Baseline, Normal Followup. A detailed neurological examination evaluated neurological integrity in three broad areas: cranial nerve function, peripheral nerve function, and central nervous system (CNS) coordination. The summary analytic results for all measurement variables comprising these three functional areas are presented in Table 11-15. Assessment of the 12 cranial nerves was based on the measurement of 14 variables. Two summary indices were constructed. Both the unadjusted and adjusted analyses did not disclose any statistically significant group differences, although two variables, speech and tongue position, were of borderline significance, with Ranch Hands faring worse than Comparisons. One of the two cranial nerve summary indices was marginally significant, again with the Ranch Hands at a slight detriment. The unadjusted and adjusted analyses of peripheral nerve function, as measured by eight variables (four reflexes, three sensory determinations, and muscle mass), did not reveal significant group differences. CNS coordination was evaluated by four measurements and a constructed summary variable. Hand tremor was found to be of borderline significance, with the Ranch Hands faring slightly worse than the Comparisons. The CNS summary index showed a significant detriment to the Ranch Hands. The exposure analyses for neurological variables with reasonable counts of abnormalities showed only occasional statistically significant results. No consistent pattern with increasing exposure was evident for any occupational category of the Ranch Hand group. 11-26 �TABLE 11-15. Overall Summary Results of Unadjusted and Adjusted Analyses of Neurological Variables Variable Unadjusted Adjusted Direction of Results** Questionnaire" Physical Examination Neurological Disease (Interval) Neurological Disease (History) NSb NS Cranial Nerve Function Smell Visual Fields NS NS Light Reaction Ocular Movements Facial Sensation Corneal Reflex NS NS NSc — Jaw Clench Smile NS NS Palpebral Fissures NS Balance Gag Reflex NS NS Speech Tongue Position Relative to Midline Palate and Uvula Movement Neck Range of Motion Cranial Nerve Function Index Cranial Nerve Function Index (excluding Neck Range of Motion) NS* RH>C NS* NS NS NS RH>C NS NS NS* NS* Peripheral Nerve Function Pin Prick Light Touch Muscle Status Vibratory Sensation Patellar Reflex Achilles Reflex Biceps Reflex Babinski Reflex NS NS NS NS NS NS NS NS 11-27 **** NS NS NS RH>C �TABLE 11-15. (continued) Overall Summary Results of Unadjusted and Adjusted Analyses of Neurological Variables Variable Unadjusted Adjusted Direction of Results** Central Nervous System Coordination Tremor Coordination Romberg Sign Gait CNS Summary Index NS* NS NS .NS 0.036 NS* RH>C — — NS 0.042 RH>C **RH>C: More abnormalities in Ranch Hand group than in Comparison group. "Disease categories include: inflammatory diseases, heriditary and degenerative diseases, peripheral disorders, disorders of the eye, disorders of the ear, and other disorders. NS:Not significant (p>0.10). No inflammatory diseases noted; borderline significant (p=0.069, RH>C) for other disorders; not significant for remaining categories. —Analysis not performed because of sparse number of abnormalities. c No abnormalities present. NS*Borderline significant (0.05<p<0.10)< Constructed variable. ****Group-by-covariate interaction. 11-28 �In a longitudinal analysis of the Romberg sign and the Babinski reflex, only the Babinski reflex revealed a significant difference between the Baseline and followup examination, with the Ranch Hands converting from significant adverse findings at Baseline to favorable nonsignificant findings at the followup examination. Overall, the followup examination findings are quite similar to the Baseline findings. However, several distinct patterns were evident from the analyses: (1) The followup examination detected substantially fewer abnormalities for almost all measurement variables, (2) the decrease in abnormalities was equivalent in both groups, (3) most of the covariate effects were classical, although exceptions were evident, (4) the adjusted analyses were uniformly similar to the unadjusted analyses, (5) the constructed summary variables were generally statistically significant, or of borderline significance (however some indices were created after the data were examined), and (6) although statistical significance at the pre-assigned a -level of 0.05 was not achieved for any of the measurement variables, abnormalities tended to cluster in the Ranch Hand group. Of the three group-by-covariate interactions in the adjusted analyses, only one, a borderline group-by-insecticide exposure interaction for hand tremor, where Ranch Hands exposed to insecticides had a marginally significant adverse effect, was of probable biologic (and operational) significance. In conclusion, none of the 27 neurological variables demonstrated a significant group difference, although several showed an aggregation of abnormalities in the Ranch Hand group, which merits continued surveillance. Historical reporting of neurologic disease was equal in both groups. The clinical sensitivity in detecting neurological deficits varied substantially between the Baseline and the followup examinations, but the number of statistically significant variables remained about the same. None of the exposure analyses revealed dose-response patterns in the Ranch Hand occupational categories. The longitudinal analyses disclosed a favorable reversal of significant Babinski reflex abnormalities at Baseline to nonsignificant findings at the followup examination for the Ranch Hands. The similarity in results between unadjusted and adjusted statistical tests is evidence of group equality for the traditionally important neurological covariates of age, alcohol, and diabetes. Of three group-by-covariate interactions in the adjusted analyses, only the Ranch Hand insecticide interaction with hand tremor was biologically plausible. 11-29 �CHAPTER 11 REFERENCES 1. Dougherty, J.A., G.E. Schulze, R.T. Taylor, and J. Blake. 1984. Behavioral toxicity of an agent orange component: 2,4-D. Oral presentation to the Veterans Administration Advisory Committee on Health-Related Effects of Herbicides, Washington, D.C., December 11, 1984. 2. Squibb, R.E., H.A. Tilson, and C.L. Mitchell. 1983. Neurobehavioral assessment of 2,4-dichlorophenoxyacetic acid (2,4-D) in rats. Neurobeh. Toxicol. Teratol. 5:331-335. 3. Desi, I., J. Sos, and I. Nikolits. 1962. New evidence concerning the nervous site of action of a chemical herbicide causing intoxication. Acta Physiol. 22:73-80. 4. Kim, C.S., L.A. O'Tuama, D. Mann, and C.R. Roe. 1983. Saturable cumulation of the anionic herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) by rabbit choroid plexus: Early developmental origin and interaction vith salicylates. J. Pharmacol. Exp. Ther. 225:699-704. 5. Goldstein, N.P., P.H. Jones, and J.R. Brown. 1959. Peripheral neuropathy after exposure to an ester of dichlorophenoxyacetic acid. JAMA 171(10):1306-1309. 6. Todd, R.L. 1962. A case of 2,4-D intoxication. J. Iowa Med. Soc. 52:663-664. 7. Berkley, M.C., and K.R. Magee. 1963. Neuropathy following exposure to a dimethylamine salt of 2,4-D. Arch. Int. Med. 111:133-134. 8. Berwick, P. 1970. 2,4-Dichlorophenoxyacetic acid poisoning in man. JAMA 214(6):1114-1117. 9. Wallis, W.E., A. Van Poznak, and F. Plum. 1970. Generalized muscular stiffness, fasciculations, and myokymia of peripheral nerve origin. Arch. Neurol. 22:430-439. 10. Park, J., I. Darrien, and L.F. Prescott. 1977. Pharmacokinetic studies in severe intoxication with 2,4-D and Mecoprop. Clin. Toxicol. 18:154-155. 11. Bauer, H., K.H. Schulz, and.U. Spiegelberg. 1961. Berufliche Vergif tungen bei der Herstellung von Chlorphenol-Verbindungen (Long-term hazards of polychlorinated dibenzodioxins and polychlorinated dibenzofurans). Arch. Gewerbephathol. Gewerbehyg. 18:538-555. Reported in IRAC (1978). 11-30 �12. Pazderova-Vejlupkova, J., M. Nemcova, J. Pickova, L. Jirasek, and E. Lukas. 1981. The development and prognosis of chronic intoxication by tetrachlorodibenzo-p-dioxin in men. Arch. Environ. Health 36:5-11. 13. Oliver, R.M. 1975. Toxic effects of 2,3,7,8-tetrachloro-dibenzo1,4-dioxin in laboratory workers. Br. J. Ind. Med. 32:49-53. 14. Boeri, E., B. Bordo, P. Crenna, et al. 1978. Preliminary results of a neurological investigaton of the population exposed to TCDD in the Seveso region. Riv. Pat. Nerv. Ment. 99:111-128. 15. Singer, R., M. Moses, J. Valciukas, R. Lilis, and I.J. Selikoff. 1982. Nerve conduction velocity studies of workers employed in the manufacture of phenoxy herbicides. Environ. Res. 29:297-311. 16. Moses, M., R. Lilis, K.D. Crow, J. Thornton, A. Fischbein, H.A. Anderson, and I.J. Selikoff. 1984. Health status of workers with past exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in the manufacture of 2,4,5-trichloro-phenoxyacetic acid: Comparison of findings with and without chloracne. Am. J. Ind. Med. 5:161-182. 17. Filippini, G., B. Bordo, P. Crenna, N. Massetto, M. Musicco, and R. Boeri. 1981. Relationship between clinical and electrophysiological findings and indicators of heavy exposure to 2,3,7,8-tetrachlorodibenzo-dioxin. Scand. J. Work Environ. Health 7:257-262. 18. Flicker, M.R., and A.L. Young. 1983. Evaluation of veterans for agent orange exposure. Presented at the Symposium on Chlorinated Dioxins and Dibenzofurans in the Total Environment given before the Division of Environmental Chemist-ry, American Chemical Society, Washington, D.C., September 1983. 11-31 �CHAPTER 12 PSYCHOLOGICAL ASSESSMENT INTRODUCTION Emotional illnesses or psychological abnormalities are not recognized as primary clinical endpoints following exposure to chlorophenols, phenoxy herbicides, and dioxin. "Neurobehavioral effects" occasionally ascribed to such exposures have been, in fact, predominantly neurological symptoms for which causation is not disputed (see Chapter 11). Higher CNS functioning, in terms of cognitive skills, personality, and reactivity, may be temporarily or permanently impaired depending on the exposure and the ability to measure accurately the psychological changes. Animal studies provide little insight into possible human psychological problems. Animal signs of lethargy, stupor, poor coordination, lack of feeding, and agitation have been observed in multiple studies involving many species. These signs have generally been attributed to the "wasting syndrome" or multi-organ toxicity, rather than primary CNS toxicity. A study of "behavioral" effects in rats following single and weekly doses of 2,4-D showed that the central effects of decreased coordination and lever-pressing behavior were transient and reversible. Further, no latent CNS impairment was detected after a d-amphetamine challenge. Human studies and case reports have occasionally noted psychological disorders or symptom complexes following exposure to herbicides and TCDD. Complaints included headache, anxiety, malaise, depression, abnormal anger, mood changes, sleep disturbances, decreased libido, and impotence. Scientific confirmation of these symptoms by psychological testing is difficult and exclusion of other plausible causes such as age, preexisting psychological abnormalities, or even motivation for compensation is often impossible. Most studies have merely recorded complaints and have not pursued their validation by indepth functional testing. Early studies of industrial chemical workers first provided the suggestion of psychological effects. Followup studies from the Nitro, West Virginia, accident in 1949, showed "nervousness," fatigue, irritability, cold intolerance, and decreased libido in many of the workers with chloracne, but most of these symptoms subsided over a 4-year period. ' Two followup studies in 1979, by different investigators of expanded (but slightly different) plant cohorts, noted reports of sexual dysfunction and decreased libido. ' One of these studies noted that these observations (and insomnia) were significantly increased in individuals with chloracne. Neither of these followup efforts conducted, neurobehavioral tests to validate the reported symptoms. Other industrially based studies reported symptoms of fatigue, decreased libido, impotence, sleep disturbances, ' 1reduced emotional responses, sensory deficits of smell, taste, and hearing, reading 12-1 �difficulties,9 memory loss,11 and emotional disorders.12'13 Symptoms of depression and anxiety have been associated with disfiguring chloracne. One study found a relationship between chloracne and hypomania as determined from the MMPI,1 and another noted that two of three chemists involved in the synthesis of TCDD developed marked personality changes. Although data interpretation problems exist, the Czechoslovakian 10-year followup study cited eight cases of severe dementia in exposed workers and reported that symptoms of anxiety and depression decreased over the followup period. A contemporary cross-sectional morbidity study of a mobile-home park, environmentally contaminated with dioxin, showed subclinical hepatic, hematologic, immunologic, and psychological changes in exposed residents. Significant abnormalities were recorded in the exposed group for the tension/ anxiety and anger/hostility scales of the profile of mood states (POMS) inventory, as well as the vocabulary subtest of the Wechsler adult intelligence scale (WAIS). However, functional testing by the Halstead-Reitan battery (HRB) did not reveal significant group differences. There was no way to differentiate between the primary effects of exposure and the secondary effects of media attention. In contrast to industrial cohorts, the study of chemically related psychological problems in veterans has proved more difficult because of the confounding effects of combat stress and the post-traumatic stress disorder (PTSD), and the uncertainty of exposure. Of almost 100,000 Vietnam veterans registered in the VA's Agent Orange Registry in 1983, 18 percent complained of "nervousness" and 10 percent cited personality disorders. A psychiatric review of 132 veterans included in the Registry, most of whom had been referred for treatment, disclosed a symptom hierarchy of sleep disorders (53%), mood depression (36%), suicidal thoughts (35%), and irritability (31%). Fifty-three percent of these veterans received the PTSD diagnosis. In 1980, the American Psychiatric Association established the term "post-traumatic stress disorder" to define a neurosis caused by extreme psychic trauma, e.g., natural disaster, war, imprisonment, or torture. PTSD comprises the symptoms of anxiety, "powder keg" anger, depression, irritability, restlessness, recurrent intrusive dreams, flashbacks, and sleeplessness. Quiescent PTSD may be acutely reactivated in some individuals by specific triggering events (e.g., visiting the Vietnam Memorial). The disorder is equally applicable to civilians following emotionally traumatic experiences. The onset of PTSD may immediately follow the traumatic event or it may occur years afterward. The older war terms shell shock, combat fatigue, and anxiety reaction generally referred to the more immediate symptoms following the trauma although components of PTSD are now recognized in veterans of earlier wars. The prevalence of PTSD in Vietnam veterans is unknown, and even the qualitative assessments of "common" or "rare" are debatable. 1 > 2 A 7-month incidence of legal and emotional maladjustments in returning Vietnam veterans occurred at the rate of 23 percent and did not differ significantly from comparable rates in nonveterans. Though a concise definition of PTSD exists, there is controversy as to the best means of diagnosis. Some workers prefer a full and thorough clinical interview while others favor empiric symptom scales. Clearly, each method serves a different, but highly related, purpose: clinical diagnosis in individuals versus an epidemiological/ statistical diagnosis in groups. 12-2 �Risk factors for the development of PTSD may include emotional predisposition, social/ethnic background, parental factors, race, and combat intensity ranging from slight involvement to atrocity behavior. ' ' Parallel conditions to PTSD (or perhaps unrecognized components of PTSD) encompass alcoholism, drug abuse, lawlessnes^ (arrests/felony convictions), personality disorders, and frank psychosis. ' ~ This chapter attempts to isolate any psychological disorders attributable to herbicide exposure. Baseline Summary Results Extensive psychological parameters were assessed on all participants during the 1982 Baseline questionnaire and physical examination. The expected high degree of concordance between education (college, high school) and military status (officer, enlisted) was observed and validated the sole use of education as a covariate representing socioeconomic status for most analyses. There were no questionnaire differences for past history of emotional or psychological illnesses between the Ranch Hand and Comparison groups. For the psychological indices of fatigue, anger, erosion, anxiety, and severity of depression (as determined by a modification of the Diagnostic Interview Schedule ), no group differences were detected among the college-educated Ranch Hands. However, for the high school-educated stratum, Ranch Hands demonstrated highly significant pathology for fatigue, anger, erosion, and anxiety. An unadjusted analysis of reported depression showed significantly more depression in the Ranch Hands, as did the isolation index adjusted for educational level. Exposure index analyses from the Ranch Hand questionnaire data did not suggest a relationship between exposure and psychological abnormality. At the time of the physical examination, additional self-reported data were collected with the Cornell Index and the MMPI. The CNS functional testing was conducted by a modified HRB, and intelligence was measured by the WAIS. The Cornell Index showed a significant increase in psychophysiologic symptoms in the high school-educated Ranch Hands. Six of 10 parameters of the Cornell Index were abnormal in the Ranch Hands (e.g., fear, startle, psychosomatic) as contrasted to the Original Comparisons, and all abnormal responses/parameters were inversely related to education to a statistically significant degree. MMPI results in the high school-educated participants, showed differences in the scales of denial, hypochondria, masculinity/ femininity, and mania/hypomania as contrasted to the college-educated group. Only the social introversion scale was significant in the college-educated participants. The effect of. education was influential (p<0.01) in all scales of the MMPI. Race was not a significant covariate. All self-reported data, including those from the in-home questionnaire, were not adjusted for possible group differences in PTSD or combat experience/intensity. Performance testing by the HRB showed no neuropsychiatric impairment in the Ranch Hands as contrasted to their overall self-administered MMPI and Cornell Index. In fact, Ranch Hand over-reporting was suggested in several parameters, but was not proved. The effect of education on the HalsteadReitan testing was profound (p<0.0001). WAIS intelligence scores revealed very close group similarities in the full-scale and verbal and performance 12-3 �scales. As expected, the intelligence quotient (IQ) of college graduates was significantly higher than the IQ of high-school graduates. Exposure index analyses of the HRB and WAIS data were negative and disclosed no patterns that suggested an herbicide effect. Parameters of the 1985 Psychological Assessment Two of the psychological tests (MMPI, HRB) conducted at the 1982 Baseline examination were repeated at the first followup examination in 1985. Repetitive testing was accomplished for purposes of clinical validation, establishment of comparable longitudinal parameters, and comparable covariate adjustments by concurrently derived PTSD and combat experience indices. Questions from the Diagnostic Interview Schedule were deleted from the followup questionnaire and were replaced by questions on combat experience in Vietnam. An updated history of mental and emotional disorders was obtained on all participants. A PTSD indicator was derived from a new MMPI subscale and was used for covariate adjustments of non-MMPI psychological data. The WAIS IQ assessment was deleted, but all parameters of the MMPI and HRB were retained. The Cornell Medical Index (CMI) was substituted for the Cornell Index in the 1985 psychological assessment. The dependent variables and covariates of the followup examination are similar to those analyzed at the Baseline. Longitudinal analyses of the MMPI scales of denial and depression consider the change of psychological test indices between groups. All statistical analyses are based on 1,016 Ranch Hands and 1,293 Comparisons. No individuals were excluded from the analysis of the psychological data for medical reasons. Sample size differences in the tables below reflect missing data from scale or battery test results, or from relevant covariates. The statistical tests use log-linear models, logistic regression models, Kolmogorov-Smirnov nonparametric tests, Fisher's exact test, and Pearson's chi-square test. Parallel analyses using Original Comparisons are in Tables J-8 through J-18 of Appendix J. RESULTS AND DISCUSSION Questionnaire Data At the followup interview, each participant was asked whether he had ever had a mental or emotional disorder. Whenever possible, the conditions were coded using ICD-9-CM. Reported disorders for which treatment was obtained were subsequently verified by reviews of medical records. Table 12-1 contains a tabulation of the distribution of these psychological illnesses, with information from the Baseline and followup studies combined. None of the types of illness categories showed statistically significant differences between groups; however, the "other neuroses" category is significant (p=0.037), with the Ranch Hands showing more adverse effects, when only Original Comparisons are used (see Table J-8 of Appendix J). 12-4 �TABLE 12-1. Unadjusted Analyses for Reported Psychological Illnesses by Group: Baseline and First Followup Studies Combined* Group Abnormalities Ranch Hand Type of Illness Psychoses Alcohol Dependence Anxiety Other Neuroses Comparison Number Percent 14 9 7 72 1.4 0.9 0.7 7.1 Number 9 8 13 74 Percent 0.7 0.6 1.0 5.7 Total 23 17 20 146 p-Value** 0.138 0.473 0.501 0.197 *Analyses based on 1,016 Ranch Hands and 1,293 Comparisons; some participants may have had more than one illness. **Fisher's exact test. Psychological Examination Data The MMPI is a self-administered test consisting of 566 questions on various aspects of behavior and personality. The results of the MMPI are numerical scores for 14 scales. The scales are anxiety (psychasthenia), consistency (F-scale), defensiveness (L-scale), denial (K-scale), depression, hypochondria, hysteria, mania/hypomania, masculinity/femininity, paranoia, psychopathic/deviate, schizophrenia, social introversion, and validity. The normal range of scores from 30 to 70 was used to categorize the results as normal or abnormal for all scales except validity. For validity (the number of unanswered questions) categories of 0 or greater than 0 were used. The test was administered to all 2,309 participants. A participant was considered nonresponsive in the MMPI if more than 30 questions (approximately 5%) were unanswered. Due to nonresponse, data on six participants, (two Ranch Hands and four Comparisons) were omitted from the analysis of all variables except validity. Thus, the MMPI analyses were based on 1,014 Ranch Hands and 1,289 Comparisons. The CMI is a self-administered instrument used to collect a substantial amount of medical and psychiatric data. The 195 questions of the CMI are partitioned into 18 sections (A to R) with the number of questions within a section ranging from 6 to 23. The analysis of the CMI was based on three scores: the total CMI score, an M-R subscore, and an A-H area subscore. The total CMI score is the number of affirmative responses on the entire questionnaire and is analyzed as a continuous variable. The M-R subscore, which deals with mood and feeling patterns, is a useful indicator of 12-5 �emotional ill-health. This subscore is the total number of affirmative responses to the 51 questions in sections M-R and is trichotomized as 0, 1 to 10, or greater than 10 for the analysis. The A-H area subscore is a measure of the scatter of complaints, indicating a diffuse medical problem, although other interpretations are possible. An abnormal A-H area subscore is defined as the number of sections (of A-H) with three or more affirmative responses. The A-H area subscore, which ranges from 0 to 8, is trichotomized as 0, 1 to 3, or 4 to 8 for the analysis. Consistent with the 5 percent nonresponse exclusion used for the MMPI, analysis of the total CMI score is based on scores with at least a 95 percent response rate or no more than 10 unanswered items from the total 195. M-R subscores are deleted from the analyses if three or more questions were unanswered from the 51 questions. For the A-H area subscore, participants who failed to answer all items were excluded from the analyses. Using these response criteria, analyses of the total CMI score are based on the scores of 1,000 Ranch Hands (16 deleted) and 1,268 Comparisons (25 deleted); the M-R subscore analyses use the results of 998 Ranch Hands (18 deleted) and 1,267 Comparisons (26 deleted); and the A-H area subscore analyses use 914 Ranch Hands (102 deleted) and 1,148 Comparisons (145 deleted). The HRB is a neuropsychological test that was administered to all participants to assess the functional integrity of the CNS. The battery consists of seven subtests: category (abstract recognition and analysis), total-time tactile performance, memory tactile performance, localization tactile performance, rhythm, speech, and finger tapping. In addition, other tests were performed (e.g., trailmaking, tests of recent memory) but do not contribute to the impairment index. For each participant who completed all seven s.ubtests, an impairment index, equal to the number of subtests in which the participant scored abnormally, is computed. This variable is dichotomized as normal (impairment index <3) or abnormal (impairment index X3). Twenty participants (10 in each group) refused or did not complete one or more of the seven subtests. Thus, the analyses of the HRB impairment index are based on data from 1,006 Ranch Hands and 1,283 Comparisons. Fisher's exact test was used to contrast the number of excluded participants between groups. A significant difference was not observed (p=0.654). The analyses of the psychological variables were adjusted for age (born in 1942 or after, born between 1923 and 1941, born in 1922 or before), race (Black, nonblack), education (high school, college), and drink-years (0, greater than 0 to 50, greater than 50). Education was dichotomized into high school and college categories, for purposes of analysis, from the classifications of (1) no high school diploma, (2) high school diploma, (3) attended college, and (4) college diploma. This variable was based on Baseline education levels, and participants with incomplete information were classified as high school educated. In addition, the analyses of the MMPI scales were adjusted for the combat index, a surrogate measure for PTSD. This index was constructed from 15 self-administered questions on combat experiences (see Appendix C, page C-15, AFHS Form 8). Associations of these 15 variables with PTSD, as measured from a subset of the MMPI questions, were examined, and responses to four questions showed statistically significant or marginally significant associations with PTSD. The four questions were (1) flew in aircraft that received battle damage, (2) had a close friend killed in action, (3) encountered mines or booby traps, and (4) wounded. An index, equal to the number of affirmative responses to these four questions, was computed and used as a trichotomized covariate (low, [0; n=708 (30.7%)], 12-6 �medium [1; n=814 (35.4%)], high [2-4; n= 781 (33.9%)], 6 missing participants, as with MMPI scales) for the analyses of the MMPI scales. While this index was associated with PTSD, it does not necessarily measure stress but does measure combat experience. The analyses of the CMI and HRB tests were adjusted for PTSD, based on the number of affirmative responses to a subset of 49 questions of the MMPI. For these analyses, PTSD was dichotomized as yes/no using greater than 30 affirmative responses2 as a positive indicator of PTSD. Sixteen participants (10 Ranch Hands, 6 Comparisons) were classified as having PTSD under this guideline. (Note that this indicator of PTSD was not used as a covariate for the analyses of MMPI scales, because the variable was based on the responses used in the calculation of the MMPI scores.) Current alcohol use (yes/no) and occupation were examined as potential covariates and are provided in the summary tables for inspection. Current alcohol use was highly correlated with drink-years, which better explained the dependent variables under study. Similarly, occupation was highly correlated with education (p<0.001). In this case, education was selected. Statistical Analysis Minnesota Multiphasic Personality Inventory (MMPI) The distributions of the Ranch Hand and Comparison groups for the 14 MMPI variables were contrasted using the Kolmogorov-Smirnov nonparametric tests and stratified by occupation (officer, enlisted flyer, enlisted groundcrew), for a total of 42 tests. Unadjusted analyses were performed using Fisher's exact test. Covariate analyses, using Fisher's exact or Pearson's chi-square test, were conducted for age, race, 'education, drinkyears, combat index, current alcohol use, and occupation. Logistic regression techniques were used to conduct the adjusted analyses. In the adjusted analyses, all covariates were used as discrete variables with the exception of age, which was used as a continuous variable. Current alcohol use and occupation were not used in the adjusted analysis. Using a two-sided a-level of 0.05, and with power of 0.80, the sample sizes are sufficient to detect a 38 percent increase in the rate of abnormal scores for depression, a 61 percent increase in the rate of abnormal scores for denial, and a 119 percent increase in the rate of abnormal scores for social introversion. Distributional Analyses The Kolmogorov-Smirnov tests identified no statistically significant differences between the Ranch Hand and Comparison distributions for the 14 MMPI variables at the 0.05 significance level for each occupational category. Only 2 of the 42 tests even approached significance, mania/ hypomania (Ranch Hand and Comparison officers, p=0.092) and psychopathic/ deviate (Ranch Hand and Comparison enlisted flyers, p=0.088). Results of the Kolmogorov-Smirnov tests are provided in Tables J-l to J-3 of Appendix J. It is noted that stratification by occupation reduced the sample size for each test and consequently decreased the power; that is, a larger maximum difference between the Ranch Hand and Comparison distributions is needed to show significance when the sample size is decreased, as is the case when stratification by occupation is performed. 12-7 �Unadjusted and Adjusted Analyses The unadjusted results, covariate tests of association, and adjusted results of the analyses for the 14 MMPI variables are summarized in Tables 12-2 to 12-4, respectively. Summary tables, which investigate interactions involving group, are provided in Table J-4 of Appendix J. The results of the tests of association for current alcohol use and occupation are presented in Table 12-3 for inspection, but are not discussed in the text since the measure of total drink-years was more appropriate for use in the analyses. Anxiety The unadjusted analysis showed no statistically significant difference in the anxiety scale between the Ranch Hands and the Comparisons (p=0.311). The tests of association with the covariates, using the pooled group categorical data, revealed statistically significant effects for age (p=0.010) and education (p<0.001). For age, 8.4 percent of the participants born in or after 1942 were scored as abnormal, as were 5.3 percent of those born from 1923 to 1941, and 4.6 percent of those born in or before 1922. The high school subgroup had a higher percentage (8.5%) of abnormalities than the college subgroup (4.4£). For the test of association, drink-years was marginally significant (p=0.058), based on the percent of abnormalities for 0, greater than 0 to 50, and greater than 50 drink-years: 10.0 percent, 5.9 percent, and 8.2 percent, respectively. In the adjusted analysis, there was no statistically significant difference between groups (p=0.512). In this analysis, education (EDUC) showed a statistically significant effect (p<0.001). The interaction, age-bycombat-index (CI), was also statistically significant (p=0.008). A group(GRP)-by-education interaction was marginally significant (p=0.057). Further investigation of this"interaction revealed an adjusted relative risk of 1.39 for the high school stratum and 0.68 for the college stratum. However, these relative risks were not significantly different from 1.00 (p=0.114, p=0.233, respectively). The exploration of this interaction is shown in Table J-4 of Appendix J. Consistency The unadjusted test of the MMPI consistency scale revealed no statistically significant difference between the Ranch Hand and Comparison groups (p-0.222). Based on the tests of association, education was statistically significant (p-0.010) with 3.9 percent abnormalities in the high school category and 2.0 percent abnormalities in the college category. In addition, the test of association with drink-years was statistically significant (p=0.021); the categories 0 and greater than 0 to 50 drink-years each had a percent abnormal frequency of 2.7, whereas there were 5.6 percent abnormalities in the greater than 50 drink-years category. In the adjusted analysis of the consistency scale, a group-by-education interaction was statistically significant (p=0.013). Further analysis of the interaction (shown in Table J-4 of Appendix J) revealed that the high school 12-8 �TABLE 12-2. Unadjusted Analyses for MMPI by Group Group Variable Anxiety Consistency to Defensiveness Denial Depression Hypochondria Hysteria Ranch Hand Statistic Number Percent Comparison Est. Relative Number Percent Risk (95% C.I.) p-Value n Abnormal Normal 1,014 73 941 7.2 92.8 1,289 79 1,210 6.1 93.9 1.19 (0.86,1.65) 0.311 n Abnormal Normal 1,014 36 3.6 978 . 96.4 1,289 34 1,255 2.6 97.4 1.36 ( . 4 2.19) 08, 0.222 n Abnormal Normal 1,014 23 991 2.3 97.7 1,289 35 1,254 2.7 97.3 0.83 (0.49,1.42) 0.592 n Abnormal Normal 1,014 17 997 1.7 98.3 1,289 58 1,231 4.5 95.5 0.36 (0.21,0.63) <0.001 n Abnormal Normal 1,014 114 900 11.2 88.8 1,289 126 1,163 9.8 90.2 1.17 (0.89,1.53) 0.272 n Abnormal Normal 1,014 119 895 11.7 88.3 1,289 129 1,160 10.0 90.0 1.20 (0.92,1.56) 0.198 n Abnormal Normal 1,014 123 891 12.1 87.9 1,289 125 1,164 9.7 90.3 1.29 (0.99,1.67) 0.067 �TABLE 12-2. (continued) Unadjusted Analyses for MHPI by Group Group Variable Ranch Hand Comparison Est. Relative Statistic Number Percent Number Percent Risk (95* C.I.) p-Value Mania/Hypomania n Abnormal Normal 1,014 63 951 6.2 93.8 1,289 88 1,201 6.8 93.2 0.90 (0.65,1.26) 0.611 Masculinity/ Femininity n Abnormal Normal 1,014 66 948 6.5 93.5 1,289 120 1,169 9.3 90.7 0.68 ( . 0 0.93) 05, 0.017 Paranoia n Abnormal Normal 1,014 31 983 3.1 96.9 1,289 28 1,261 2.2 97.8 1.42 (0.85,2.38) 0.187 Psychopathic/ Deviate n Abnormal Normal 1,014 120 894 11.8 88.2 1,289 149 1,140 11.6 88.4 1.03 ( . 0 1.33) 08, 0.845 Schizophrenia n Abnormal Normal 1,014 94 920 9.3 90.7 1,289 101 1,188 7.8 92.2 1.20 (0.90,1.61) 0.228 Social Introversion n Abnormal Normal 1,014 26 988 2.6 97.4 1,289 19 1,270 1.5 98.5 1.76 (0.97,3.20) 0.069 Validity n >0 0 1,016 224 792 22.0 78.0 1,293 271 1,022 21.0 79.0 1.07 (0.87,1.30) 0.540 NJ I �TABLE 12-3. Association Between MMPI Variables and the Covariates in the Combined Ranch Hand and Comparison Groups MMPI Scale Age Race Drink.- Combat Education Years Index Current** Alcohol Use Occupation** Anxiety 0.010 NS <0.001 NS* NS 0.001 <0.001 Consistency NS NS 0.010 0.021 NS NS <0.001 Defensiveness 0.028 0.025 0.001 <0.001 Denial 0.037 NS Depression NS NS Hypochondria 0.031 Hysteria <0.001 <0.001 NS* NS NS <0.001 0.002 NS NS <0.001 0.025 <0.001 0.041 0.044 <0.001 0.044 NS <0.001 0.006 NS 0.027 <0.001 Mania/Hypomania NS NS 0.011 0.001 NS 0.022 Masculinity/ Femininity 0.005 NS NS NS NS 0.005 Paranoia 0.022 NS NS NS NS NS* 0.014 Psychopathic/ Deviate NS 0.001 0.001 <0.001 NS NS* <0.001 Schizophrenia NS NS <0.001 NS Social Introversion 0.003 Validity NS <0.001 NS NS <0.001 NS 0.027 NS 0.014 NS NS* <0.001 NS* NS NS NS* <0.001 NS NS NS* NS NS - Not significant (p>0.10). *Borderline significant (0.05<p<0.10). **Not used in adjusted analyses. 12-11 NS �TABLE 12-4. Adjusted Analyses for MMPI by Group Group Adj. Relative Risk ( 5 C.I.) 9% p-Value Covariate Remarks* 1,285 1.12 (0.80,1.57) 0.512 EDUC (p<0.001) AGE*CI (p=0.008) GRP*EDUC (marginal: p=0.057) 974 1,246 **** **** 976 1,250 0.77 (0.45,1.33) 0.347 1,012 1,285 0.37 (0.21,0.66) <0.001 974 1,246 1.10 (0.84,1.45) 0.497 Hypochondria 1,012 1,285 1.12 (0.85,1.47) 0.431 AGE (p=0.002) RACE (p=0.026) EDUC (p<0.001) CI (p=0.043) Hysteria 1,014 1,289 1.27 (0.97,1.66) 0.077 AGE (p=0.003) EDUC (p<0.001) 974 1,246 0.80 (0.56,1.13) 0.203 DRKYR (p=0.006) AGE*CI (p=0.046) Variable Ranch Hand Total Comparison Total Anxiety 1,012 Consistency Defensiveness to h-1 NJ Denial Depression Mania/Hyporaania AGE (p=0.007) DRKYR (p=0.026) CI (p=0.041) GRP*EDUC (p=0.013) EDUC (p<0.001) DRKYR (p<0.001) EDUC*CI (p=0.044) EDUC (p<0.001) DRKYR (p=0.013) GRP*CI (marginal: p=0.055) �TABLE 12-4. (continued) Adjusted Analyses for MMPI by Group Group Variable Comparison Total Masculinity/ Femininity 1,014 Paranoia N5 Ranch Hand Total Adj. Relative Risk (95% C.I.) p-Value Covariate Remarks* 1,289 0.69 ( . 0 0 9 ) 05,.5 0.020 EDUC (p<0.001) RACE*AGE (p=0.008) 1,012 1,285 **** **** AGE*CI (p=0.003) GRP*AGE (p=0.036) Psychopathic/ Deviate 974 1,246 (0.79,1.36) 0.780 EDUC (p=0.011) AGE*CI (p=0.003) RACE*DRKYR (p=0.015) Schizophrenia 976 1,250 **** **** RACE*DRKYR (p=0.017) GRP*EDUC (p=0.010) Social Introversion 1,012 1,285 **** **** AGE (p=0.004) GRP*CI (p=0.037) Validity 1,014 1,289 **** **** AGE*CI (p=0.030) GRP*RACE (p=0.012) 1.04 *Abbreviations: EDUC: education Cl: combat index GRP: group DRKYR: drink-years of alcohol ****Group-by-covariate interaction — adjusted relative risk, confidence interval, and p-value are not presented. �Ranch Hand category had a marginally significantly higher percentage of abnormal participants (5.6%) than the high school Comparisons (2.9%) (p=0.051). The adjusted relative risk for the high school classification was 1.81 with 95 percent confidence bounds of 1.00 and 3.28. In contrast, the percentage of abnormalities in the Comparison college-educated stratum was higher than the corresponding Ranch Hand subgroup (2.6 percent, 1.4 percent, respectively), but the difference was not statistically significant (p=0.110). Age, drink-years (DRKYR), and combat index were also statistically significant (p=0.007, p«0.026, p=0.041, respectively) in the adjusted analyses. Defensiveness For the MMPI defensiveness scale, there was no significant difference between groups, based on the unadjusted analysis (p=0.592). The tests of association showed statistically significant differences for all variables except combat index, which was marginally different statistically. The percentage of abnormalities for the age categories (born in or after 1942, born between 1923 and 1941, and born in or before 1922) were 3.3, 1.8, and 4.6, respectively (p=0.028). There were 2.3 percent abnormalities for nonblacks as compared to 5.6 percent for Blacks (p=0.025). The percent abnormalities for the high school- and college-educated categories were 3.8 and 1.0, respectively (p<0.001). For the 0 drink-years category, there were 10.0 percent abnormalities; the percent abnormalities for the greater than 0 to 50 and greater than 50 drink-years were 2.4 and 0.6, respectively (p<0.001). For combat index, which was only marginally statistically significant (p=0.093), the percent abnormalities were 3.5 for the low, 2.1 for the medium, and 1.9 for the high categorizations. In the adjusted analysis, there was no significant difference between the Ranch Hand and Comparison groups (p=0.347). In this analysis, the covariates of education (p<0.001) and drink-years (p<0,001) were statistically significant. Denial Based on the unadjusted analysis, .there was a statistically significant difference between the two groups on the MMPI denial scale (p<0.001), with 4.5 percent abnormalities in the Comparison group as contrasted to only 1.7 percent in the Ranch Hand group. The estimated relative risk was 0.36 with a 95 percent confidence interval of 0.21 to 0.63. The tests of association found only age as a statistically significant covariate (p=0.037). Men born in or after 1942 and those born between 1923 and 1941 had 3.0 percent and 3.1 percent abnormalities, respectively, as compared to 8.0 percent abnormalities for those born in or before 1922. The adjusted analysis showed a statistically significant difference between groups (p<0.001). The adjusted relative risk estimate was 0.37 with 95 percent confidence bounds of 0.21 and 0.66. For this analysis, the education-by-combat index interaction was also statistically significant (p=0.044). 12-14 �Depression The unadjusted analysis of the depression scale revealed no statistically significant difference between the two groups (p=0.272). In the covariate tests of association, education and drink-years showed statistically significant effects (p<0.001, p=0.002, respectively). There was a higher percentage of abnormalities in the high school-educated category (13.1%) than in the college-educated category (7.2%). For drink-years, the highest rate of abnormality was in the highest category of alcohol use (15.8%), followed by the nondrinker with 10.7 percent abnormalities and the moderate category with 9.4 percent. In the adjusted analysis, there was no statistically significant difference between groups (p=0.497), but there was a marginally significant group-by-combat index interaction (p=0.055). This interaction was explored further and is shown in Table J-4 of Appendix J. The analysis of the groupby-combat index interaction revealed a marginal difference within the low (0) category of the combat index (p=0.055), but not within the medium and high categories. In contrasting the 192 Ranch Hands and the 490 Comparisons in the 0 category, there were 14.6 percent abnormalities in the Ranch Hand group versus 8.2 percent in the Comparisons (p=0.039). The adjusted relative risk for the 0 category of the combat index was 1.73 with a 95 percent confidence interval of 1.03 to 2.91. Education (p<0.001) and drink-years (p=0.013) also exhibited statistically significant effects in the adjusted analysis. Hypochondria There was no statistically significant difference for the MMPI hypochondria scale between the Ranch Hand and Comparison groups (p=0.198). In the covariate tests of association, all five variables were statistically significant. Of men born in or after 1942, 8.8 percent had abnormalities as compared to 12.2 percent and 12.6 percent of those born between 1923 and 1941 and in or before 1922, respectively (p=0.031). The rates of abnormalities for Blacks and nonblacks were 16.8 percent and 10.4 percent, respectively (p=0.025). There was a highly statistically significant difference for education (p<0.001) with the high school-educated category having 13.9 percent abnormalities and the college-educated category having 7.0 percent. There was also a statistically significant difference for drink-years (p-0.041). The lowest rate of abnormalities was in the greater than 0 to 50 drink-years category with 9.9 percent; the corresponding percentages for the 0 drink-year and greater than 50 drink-year categories were 12.7 and 14.3, respectively. The percent abnormalities in the low, medium, and high combat index categories were 9.8, 9.4, and 13.2, respectively (p=0.027). The adjusted analysis showed no significant difference between the Ranch Hand and Comparison groups (p=0.431). In this analysis, age (p=0.002), race (p=0.026), education (p<0.001), and combat index (p=0.043) were statistically significant covariates. 12-15 �Hysteria Based on the unadjusted analysis of the MMPI hysteria scale, the difference between the two groups approached statistical significance (p=0.067). The percent abnormalities were 12.1 and 9.7 for the Ranch Hand and Comparison groups, respectively. The estimated relative risk, was 1.29 with a 95 percent confidence interval of 0.99 to 1.67. The covariate tests of association showed that there were statistically significant differences for age (p=0.044), education (p<0.001), and drinkyears (p=0.006). There were 12.6 percent, 12.1 percent, and 8.9 percent abnormalities in the age categories born in or after 1942, born between 1923 and 1941, and born in or before 1922, respectively. The high school-educated category had a higher percentage of abnormalities (12.9%) than the collegeeducated category (8.2%). The drink-years category with the lowest percentage of abnormalities was greater than 0 to 50 with 9.6 percent? the 0 drink-years and the greater than 50 drink-years categories had 14.0 and 14.9 percent abnormalities, respectively. The adjusted analysis also approached significance (p=0.077). The adjusted relative risk was 1.27 with 95 percent confidence bounds of 0.97 and 1.66. Age and education were statistically significant covariates in the adjusted model (p=0.003, p<0.001, respectively). Drink-years was marginally significant (p=0.068) in the presence of other covariates, but was not included in the final adjusted model. Mania/Hypomania For the unadjusted analysis of the mania/hypomania scale of the MMPI, there was no statistical difference between the Ranch Hand and the Comparison groups (p=0.611). In the covariate tests of association, there were statistically significant differences for drink-years and combat index (p=0.011, and p=0.001, respectively). For the mania/hypomania scale, the 0 drink-years category had 6.7 percent abnormalities, the greater than 0 to 50 drink-years category had 5.8 percent, and the greater than 50 drink-years category contained 10.2 percent. The frequencies of abnormalities increased from the low to the high level of the combat index; the percentages were 5.0, 5.3, and 9.4, respectively. Based on the adjusted analysis, there was no statistically significant difference between the two groups (p=0.203). Drink-years was a significant covariate (p«0.006), as was the age-by-combat index interaction (p=0.046). Masculini ty/Feminini ty The masculinity/femininity scale of the MMPI measures the stereotype "macho" attitudes of the test subjects. There was a statistically significant group difference for this scale of the MMPI, unadjusted for covariates (p=0.017). There was a higher percentage of abnormalities in the Comparison group (9.3%) than in the Ranch Hand group (6.5%). The estimated relative risk was 0.68, and the 95 percent confidence interval was 0.50 to 0.93. 12-16 �There was a statistically significant difference detected for age (p=0.005) and for education (p<0.001), based on the pooled group data in the covariate tests of association. The highest rate of abnormalities was found in men born in or after 1942 (10.2%); whereas those born between 1923 and 1941 had 6.4 percent, and those born in or before 1922 had 8.0 percent. For education, the college-educated category showed an abnormal rate of 10.3 percent versus the high school category with 6.2 percent abnormalities. The adjusted analysis also showed a statistically significant difference between the two groups (p«0.020), with an adjusted relative risk of 0.69 (95% C.I.: [0.50,0.95]). Education and a race-by-age interaction were statistically significant in the adjusted analysis (p<0.001, p=0.008, respectively). These covariate associations follow expectations. Paranoia The unadjusted analysis of the MMPI paranoia scale did not reveal a statistically significant group difference (p=0.187). Based on the pooled group data, the covariate test of association for age was statistically significant (p=0.022). There was 3.6 percent abnormalities for men born in or after 1942, 2.0 percent for those born between 1923 and 1941, and no abnormalities for men born in or before 1922. The adjusted analysis revealed a significant group-by-age interaction (p=0.036). The age-by-combat index interaction was also statistically significant (p=0.003). The group interaction was examined by combining the participants born between 1923 and 1941 with those born in or before 1922, and basing the test on two age categories (born in or after 1942 and born before 1942), due to problems with 0 counts (see Table J-4 of Appendix J). The analysis showed a higher percentage of abnormal Ranch Hands than abnormal Comparisons for participants born before 1942 (2.7% and 1.2%, respectively; p=0.027). The relative risk estimate for this age category was 2.63 (95% C.I.: [1.11,6.20]). In contrast, for the stratum born in or after 1942, the frequencies of abnormalities were nearly the same in each group (3.7% for Ranch Hands, 3.5% for Comparisons; p=0.712). Psychopathic/Deviate No significant difference between the two groups was identified in the unadjusted analysis of this MMPI scale (p=0.845). In the covariate tests of association, there were statistically significant differences for race, education, and drink-years. There were 21.0 percent abnormalities for Blacks as compared to 11.1 percent for nonblacks (p=0.001). For education, there were 13.8 percent abnormalities in the high school-educated category and 9.1 percent in the college-educated category (p=0.001). The highest rate of abnormalities in the drink-year categories was 20.2 percent for the category of greater than 50 drink-years; the percent abnormalities for the 0 and greater than 0 to 50 categories were 11.3 and 10.1, respectively (p<0.001). Based on the adjusted analysis, there was no significant difference between the Ranch Hand and Comparison groups (p=0.780). In this analysis, education (p=0.011), the age-by-combat index interaction (p=0.003), and the 12-17 �race-by-drink-year interaction (p=0.015) were statistically significant adjusting variables. Schizophrenia The unadjusted tests showed no significant difference between the Ranch Hand and Comparison groups for the MMPI schizophrenia scale (p=0.228). Based on the pooled group data, the covariate tests of association revealed that education (p<0.001) and drink-years (p=0.014) had statistically significant effects. The high school-educated category had a statistically significant higher rate of abnormalities (11.0%) than the college-educated category (5.4%). For drink-years, the highest percent of abnormalities was in the greater than 50 drink-year category (12.6%), followed by the 0 drinkyear category with 8.7 percent, and the greater than 0 to 50 drink-year category, which had 7.7 percent abnormalities. In the adjusted analysis, the group-by-education interaction was significant (p=0.010) (see Table J-4 of Appendix J). The race-by-drink-year interaction was also statistically significant (p=0.017). Analysis of the high school and college strata showed a higher percentage of abnormal Ranch Hands than abnormal Comparisons in the high school classification (13.4% versus 9.5%, respectively; p=0.033). The relative risk estimate for high school participants was 1.51, with 95 percent confidence bounds of 1.05 and 2.16. The college-educated stratum revealed a nonsignificant group difference, but the Ranch Hands had a lower rate of schizophrenia abnormalities than the Comparison group (4.1% and 6.3%, respectively). Social Introversion Based on the unadjusted analysis, the difference between the two groups approached significance (p=0.069). The Ranch Hand group had 2.6 percent abnormalities as contrasted to 1.5 percent abnormalities in the Comparison group. The 95 percent confidence bounds on the estimated relative risk of 1.76 were 0.97 and 3.20. Age was the only statistically significant covariate (p=0.003). The participants who were born in or after 1942 had a higher percentage of abnormalities (3.1%) than either those born between 1923 and 1941 or those born in or before 1922; both of these latter age categories had a 1.1 percent frequency of abnormalities. Education was of marginal significance (p=0.099) with 2.4 percent of the high school-educated participants scored as abnormal as compared to 1.4 percent of the college-educated participants. The groupby-combat index interaction was statistically significant in the adjusted analysis (p=0.037) (see Table J-4 of Appendix J). The analysis of the group-by-combat index interaction showed a difference within the low (0) combat index category with the Ranch Hands having a significantly higher percentage of abnormalities than the Comparisons (5.6% and 1.2%, respectively; p=0.002). The adjusted relative risk for this combat index category was 4.86, with a 95 percent confidence interval of 1.77 to 13.36. The medium and high combat index strata showed no statistically significant group differences (p=0.478, p=0.677, respectively). In this adjusted model, age also had a significant effect (p=0.004). 12-18 �Validity For the MMPI validity scale, the unadjusted tests showed no significant difference between the Ranch Hand and Comparison groups (p=0.540). The covariate tests of association showed that Blacks had a significantly higher frequency of abnormalities (35.0%) than nonblacks (20.5%) (p<0.001). The adjusted analysis revealed a statistically significant groupby-race interaction (p=0.012). A covariate interaction, age-by-combat index, was also found to be statistically significant (p=0.030). Further investigation of the group interaction disclosed a higher percentage of Black Comparisons with scores greater than 0 than Black Ranch Hands (42.2%, 25.0%, respectively), with an adjusted relative risk of 0.46 (p=0.038, 95% C.I.: [0.22,0.96]). In contrast, the nonblack stratum revealed a slightly higher proportion of abnormalities in the Ranch Hands, with an adjusted relative risk of 1.20 (95% C.I.: [0.97,1.49], p=0.095) (see Table J-4 of Appendix J). Cornell Medical Index (CMI) Three variables derived from the CMI were analyzed: the total CMI, M-R subscore, and the A-H. area subscore. The total CMI was analyzed as a continuous variable, using a log (X+l) transformation, where X was the number of affirmative answers. Based on the Kolmogorov-Smirnov test, the distributions of the Ranch Hand and Comparison total CMI scores were contrasted. For this set of analyses, the data were stratified separately by the covariates of age, race, education, current alcohol use, and occupation. The unadjusted analysis of total CMI was based on the two-sample t-test. Analysis of variance and two-sample t-tests were used to analyze the covariates, and the adjusted analysis on the total CMI was based on analysis of covariance techniques, using SAS®-GLM. Age was analyzed as a continuous variable in the adjusted analysis. Using a two-sided ot-level of 0.05, and with power of 0.80, the sample sizes were sufficient to detect a 10.2 percent mean shift in the total CMI score relative to the mean observed in the Comparison group. Pearson's chi-square test was used to conduct the unadjusted analyses and the covariate tests of association of the M-R subscore and the A-H area subscore, which were trichotomized into low, medium, and high classes. The adjusted analyses of these two variables were conducted by log-linear techniques using BMDP*-4F. In all three CMI variables, a higher score is associated with a higher degree of abnormality. The results of the unadjusted analysis, covariate tests of association, and the adjusted analyses on the three CMI variables are summarized in Tables 12-5 to 12-7, respectively. As discussed for the MMPI variables, the results of the covariate tests of association for current alcohol use and for occupation are provided in the summary table for information only. 12-19 �TABLE 12-5. Unadjusted Analyses for the Cornell Medical Index (CHI) by Group Group Ranch Hand Comparison Variable Statistic Total CMI n Mean* 95% C.I.a 1,000 11.74 (11.17,12.35) 1,268 10.42 (9.95,10.90) n Number/% -0 (Low) 1-10 (Medium) >10 (High) 998 1,267 Est. Relative Risk (95% C.I.) M-R Subscore 538 408 52 53.9% 40.9% 5.2% 726 57 .3% 484 38 .2% 57 4.5% to I p-Value <0.001 Overall 0.252 Medium vs. Low 1.14 (0.96,1.35) High vs. Low 1.23 (0.83,1.82) 0.146 Overall 0.003 Medium vs. Low 1.33 (1.11, 1.60) High vs. Low 1.46 (1.08,1.98) 0.003 0.314 NJ O A-H Area Subscore n Number/% -0 (Low) 1-3 (Medium) 4-8 (High) 1,148 914 360 • 449 105 39.4% 49.1% 11.5% 537 46 .8% 504 43 .9% 107 9.3% 0.013 transformed from log (X+l) scale, where x was the number of questions answered "yes." —No relative risk given for Total CMI, which was analyzed as a continuous variable. �TABLE 12-6, Association Betveen CMI Variables and the Covariates in the Combined Ranch Hand and Comparison Groups CMI Variable Age Race Education DrinkYears PTSD Current* Alcohol Use Occupation* Total CMI <0.001 NS <0.001 <0.001 <0.001 <0.001 <0.001 M-R Subscore <0.001 0.022 <0.001 NS* <0.001 0.043 <0.001 A-H Area Subscore <0.001 NS <0.001 <0.001 <0.001 0.010 <0.001 NS: Not significant (p>0.10). NS*: Borderline significant (0.05<p<0.10). **Not used in adjusted analyses. 12-21 �TABLE 12-7. Adjusted Analyses for CMI Variables by Group Group Variable Statistic Ranch Hand Total CHI n Adj. Mean 95% C.I. 962 **** **** 1,229 **** **** Comparison M-R Subscore n 998 1,265 A-H Area Score n 881 1,113 to I to Adj. Relative Risk (95% C.I.) p-Value Covariate Remarks* **** PTSD (p<0.001) RACE*DRKYR (p=0.039) AGE*EDUC (p=0.005) GRP*EDUC (p=0.003) Overall Medium vs. Low: 1.14 (0.95,1.35) High vs. Low: 1.12 (0.74,1.70) 0.339 0.152 0.598 AGE (p<0.001) EDUC (p<0.001) PTSD (p<0.001) GRP*EDUC (marginal: p=0.067) Overall Medium vs. Low: 1.27 (1.06,1.53) High vs. Low: 1.24 (0.90,1.71) 0.040 0.011 0.190 AGE (p<0.001) EDUC (p<0.001) PTSD (p<0.001) DRKYR (p=0.014) *Additional Abbreviations; PTSD: Post-Traumatic Stress Disorder ****Group-by-covariate interaction—adjusted mean, confidence interval, and p-value not presented. No relative risk given for total CMI, which was analyzed as a continuous variable. �Distributional Analyses The Kolmogorov-Smirnov tests showed statistically significant differences between the Ranch Hand and Comparison distributions for the total CMI for one category for each of the covariates. For age, the distribution of Ranch Hands born in or after 1942 was statistically different from the corresponding distribution for the Comparisons (p<0.001). The distributions of the nonblack Ranch Hand and Comparison responses also differed significantly (p=0.003). The contrast of the high school-educated Ranch Hand and Comparison distributions revealed a statistically significant difference (p<0.001). The distributions for Ranch Hand and Comparison current drinkers were also statistically different (p=0.024). For occupation, the enlisted groundcrew distributions for Ranch Hands and Comparisons were statistically different (p=0.007). Except for the covariate age, all significant differences in distributions for each covariate were found in the category having the largest sample size. The results of the 12 Kolmogorov-Smirnov tests are summarized in Table J-5 of Appendix J. Unadjusted and Adjusted Analyses Total Cornell Medical Index Based on the unadjusted analysis, as depicted in Table 12-5, the total CMI means of the Ranch Hand and Comparison groups were statistically different (p<0.001). The mean, as transformed from the log (X+l) scale, of the 1,000 Ranch Hands was 11.74 as compared to 10.42 for the Comparisons. The covariate tests of association identified that age, education, drink-years, and PTSD were highly significant (p<0.001 for all). For age, the (transformed) means of the categories showed an increase; the means of those born in or after 1942, between 1923 and 1941, and in or before 1922 were 10.08, 11.49, and 14.53, respectively. The mean of the high schooleducated category (12.97) was statistically higher than the mean of the college-educated category (8.99). The mean of the greater than 50 drinkyears was 14.49 as compared to means of 10.37 and 10.34 for the 0 and greater than 0 to 50 drink-years, respectively. The mean of the participants with a positive measure of PTSD was 71.77, whereas 10.83 was the mean of those without a positive measure of PTSD. In the adjusted analysis, there was a significant group-by-education interaction (p-0.003). Further analysis of the interaction (see Table J-4 of Appendix J) snowed that the high school-educated Ranch Hands had a higher adjusted mean total CMI than the high school-educated Comparisons (p<0.001). No significant difference was seen in the college stratum. PTSD was a significant covariate (p<0.001). The covariate interactions, race-by-drinkyears and age-by-education, were also significant in the adjusted model (p=0.039, p=0.005, respectively). M-R Subscore The results of the unadjusted analysis on the M-R subscore, an indicator of emotional health, revealed no significant difference between groups (p-0.252). 12-23 �The covariate tests of association on the pooled group data showed that age (p<0.001), race (p=0.022), education (p<0.001), and PTSD (p<0.001) were statistically significant covariates. For age, participants born in or after 1942 had a higher percentage of scores greater than 0 when compared to the other categories. Blacks had a higher percentage of scores greater than 0 than nonblacks. For education, the college-educated category had a higher percentage of 0 scores. The M-R subscores were distributed differently for participants with and without PTSD. For example, 15 of 16 participants with PTSD had an M-R subscore greater than 10, whereas only 4.2 percent of the participants without PTSD had a similar score. Drink-years showed a marginally significant effect (p=0.054); the greater than 50 drink-year category exhibited the largest percentage of participants with scores greater than 0. No significant difference between the two groups was identified in the adjusted analysis. There was a marginally significant group-by-education interaction (p=0.067). Further investigation of this interaction (see Table J-4 of Appendix J) showed a significant difference for the high schooleducated stratum (p=0.030) but not for the college-educated stratum. This difference results from the contrast of the medium (1 to 10) and low (0) categories, with the Ranch Hands having a higher percentage of participants in the medium category for the M-R subscore than in the low category (Adj. RR: 1.37, 95% C.I.: [1.07,1.75], p»0.014). In this analysis, age, education, and PTSD were highly significant adjusting variables (p<0.001 for all). A-H Area Subscore Based on the unadjusted results, the A-H area subscore—an indicator of diffuse medical problems—revealed a significant difference between the Ranch Hand and Comparison groups (p=0.003). This was due to the increased percentage of Ranch Hands over Comparisons in both the medium (1 to 3) and the high (4 to 8) categories (p=0.003, p=0.013, respectively). The covariate tests on the A-H area subscore showed that age, education, drink-years, and PTSD were highly significant covariates (p<0.001 for all). Older participants (born in or before 1922) had the lowest percentage of 0 scores. The college-educated category had a higher percentage of 0 scores than the high school-educated category. For drink-years, the lowest percentage of 0 scores was in the greater than 50 drink-years category. Twelve of 16 participants with PTSD had scores of 4 to 8, as compared to. 9.7 percent of participants without PTSD. Results of the adjusted analysis were similar to the unadjusted analysis and indicated that the two groups were statistically different (p=0.040). The overall group difference was predominately due to an increased adjusted percentage of Ranch Hands over Comparisons in the medium (1 to 3) versus low (0) contrast (p=0.011). The adjusted relative risk for this contrast was 1.27 with 95 percent confidence bounds of 1.06 and 1.53. In the adjusted model, age, education, and PTSD were significant covariates (p<0.001 for all); drink-years was also statistically significant (p=0.014). Halstead-Reitan Battery (HUB) The unadjusted analysis of the impairment index, the one variable from the HRB, was performed by using Fisher's exact test. Fisher's exact test and 12-24 �Pearson's chi-square test were used to conduct the covariate tests of association. The adjusted analysis was based on logistic regression techniques using BMDP*-LR. The results of the analyses of the HRB impairment index are summarized in Table 12-8. The unadjusted contrast of the 1,006 Ranch Hand scores and the 1,283 Comparison scores for the HRB impairment index revealed no statistically significant group differences (p=0.533). The covariate tests of association showed that age, race, and education were highly significant covariates (p<0.001 for all), and drink-years also was statistically significant (p=0.002). For age, the highest percent frequency of abnormalities was in the category of participants born in or before 1922 (66.3%); the corresponding frequencies for the participants born between 1923 and 1941 and for those born in or after 1942 were 38.3 percent and 25.1 percent, respectively. Blacks had a significantly higher percentage of abnormal scores, with 57.1 percent as compared to 32.3 percent for nonblacks. The college-educated category had a 22.3 percent frequency of abnormalities versus 43.5 percent for the high school-educated category. With respect to drink-years, the highest percentage of abnormalities (41.2%) was for greater than 50 drink-years; the 0 drink-year and greater than 0 to 50 drink-year categories had 38.0 percent and 32.0 percent, respectively. There was no significant difference identified between the two groups based on the adjusted analysis (p=0.697). Age, race, and education were statistically significant covariates (p<0.001 for all). EXPOSURE INDEX ANALYSES Exposure index analyses were conducted within each occupational cohort of the Ranch Hand group (see Chapter 8 for details on the exposure index). All variables, except the total CMI, were investigated, (unadjusted for any covariates), using Pearson's chi-square test and Fisher's exact test. Analyses of the total CMI were accomplished by t-tests and analysis of variance and covariance techniques. A log transformation was used in both adjusted and unadjusted analyses, and participants with PTSD were deleted. Adjusted analyses were performed using logistic regression, incorporating the covariates of race, age, education, and drink-years, as well as any significant pairwise interactions between the exposure index and these covariates. Age was treated as a continuous variable in the analyses. For the MMPI variables, combat index was also included as a covariate. For the HRB impairment index, participants classified as having PTSD were deleted from the analysis. The M-R subscore and the A-H area subscore were collapsed into 2 categories for analysis: 0 and greater than 0. Participants with PTSD were also deleted from this analysis. Overall significance in the proportion of abnormalities among the exposure index levels of low, medium, and high was determined, as well as contrasts in the proportion of abnormalities between the medium and low exposure levels, and between the high and low exposure levels. Results of the adjusted analyses are presented in Table 12-9, and parallel results for unadjusted analyses are presented in Table J-6 of Appendix J. Results from further study of exposure index-by-covariate interactions are given in Table J-7 of Appendix J. 12-25 �TABLE 12-8. Summary Results for the Halstead-Reitan Battery Impairment Index Analyses Group Analysis Comparison Number Percent Est./Adj. Relative Risk (95% C.I.) p-Value Unadjusted Analysis to 1 eo Ranch Hand Statistic Number Percent n Abnormal Normal 1,283 427 856 1.06 (0.89,1.26) 1,006 348 658 34.6 65.4 33.3 66.7 0.533 N/A AGE (p<0.001) RACE (p<0.001) EDUC (p<0.001) DRKYR (p=0.002) PTSD (p=0.431) ALC (p=0.004) OCC (p<0.001) Covariate Tests of Association3 cr> Adjusted Analysis Covariate Remarks* n 1,006 1,283 1.04 (0.86,1.25) 0.697 AGE (p<0.001) RACE (p<0.001) EDUC (p<0.001) * Additional Abbreviations: ALC: current alcohol use (yes/no) OCC: occupation a Based on pooled group data; current alcohol use (ALC) and occupation (OCC) provided for information only. �TABLE 12-9. Adjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Officer Anxiety Statistic* n Low 125 Exposure Index Medium 126 61 Adj. Relative Risk (95* C.I.) p-Value High Contrast 120 Overall H vs. L H vs. L 2.46 (0.36,16.82) 2.43 (0.35,16.81) 0.562 0.358 0.367 Overall M vs. L H vs. L 0.44 (0.12,1.70) 0.28 ( . 05, 1.44) 0 0.215 0.235 0.127 Enlisted Flyer n ; 50 53 Enlisted Groundcrev n 148 160 131 Overall H vs. L H vs. L **() **! **() **! Officer n 125 126 120 Overall M vs. L H vs. L 1.10 (0.14,8.59) 0.274 0.925 Overall M vs. L H vs. L 0.39 0.30 (0.06,2.37) (0.03,2.93) 0.425 0.304 0.303 Overall M vs. L H vs. L 0.87 (0.32,2.34) 0.56 (0.18,1.67) 0.550 0.781 0.296 to Consistency Enlisted Flyer n Enlisted Groundcrev n 50 148 61 160 53 131 **() **! **() **! **() **! �TABLE 12-9. (continued) Adjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Statistic* Low Exposure Index Medium High Contrast 125 126 120 Overall H vs. L H vs. L n 50 61 53 Overall M vs. L H vs. L Enlisted Groundcrew n 148 160 131 Overall M vs. L ° H vs. L Officer n Officer Defensiveness NJ CO Denial Adj. Relative Risk (95% C.I.) p-Value n' Enlisted Flyer Enlisted Flyer n Enlisted Groundcrew n 125 50 148 126 61 160 120 53 131 0.518 0.613 0.17 (0.001,29.09) 0.503 1.37 (0.02,77.86) 0.878 0.79 (0.23,2.78) 1.31 (0.40,4.23) Overall M vs. L H vs. L ****(2) ****(2) Overall M vs. L H vs. L 1.03 (0.09,11.69) Overall M vs. L H vs. L 0.737 0.719 0.656 ****(2) ****(2) ****(2) 0.234 0.984 0.109 1.41 (0.18,11.09) 0.747 �TABLE 12-9. (continued) Adjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Officer Depression Statistic* n n Enlisted Groundcrev n Officer Hypochondria Enlisted Flyer n Enlisted Flyer n Enlisted Groundcrew n Low 125 50 148 125 50 148 Exposure Index Medium 126 61 160 126 61 160 High 120 53 131 120 53 131 Contrast Adj. Relative Risk ( 5 C.I.) p-Value 9% Overall H vs. L H vs. L 0.62 (0.20,1-88) 1.24 (0.46,3.33) 0.411 0.393 0.669 Overall H vs. L H vs. L 0.55 (0.18,1.67) 0.31 (0.09,1.10) 0.160 0.295 0.070 Overall M vs. L H vs. L **() **! **() **! **() **! **() **! **() **! Overall M vs. L H vs. L ****(3) ****(3) ****(3) ****(3) ****(3) Overall M vs. L H vs. L 0.33 (0.09,1.18) 0.74 (0.26,2.14) Overall M vs. L H vs. L **() **! **() **! 0.195 0.087 0.581 **() **! **() **! **() **! �TABLE 12-9. (continued) Adjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Officer Hysteria Statistic* n n Enlisted Groundcrew n Officer Mania/ Hypomania Enlisted Flyer n Enlisted Flyer n Enlisted Groundcrev n Low 125 50 148 125 50 148 Exposure Index Medium 126 61 160 126 61 160 High 120 53 131 120 53 131 Contrast Adjo Relative Risk ( 5 C.I.) p-Value 9% Overall M vs. L H vs. L ****(3) ****(3) Overall M vs. L H vs. L 0.55 (0.18,1.74) 0.41 (0.12,1.37) Overall M vs. L H vs. L **() **! **() **! **() **! **() **! **() **! Overall M vs. L H vs. L ****(4) ****(4) ****(4) ****(4) ****(4) Overall M vs. L H vs. L 2.51 (0.55,11.53) 1.66 (0.35,7.89) 0.474 0.236 0.527 Overall M vs. L H vs. L 0.97 (0.38,2.45) 0.61 (0.21,1.75) 0.597 0.945 0.356 ****(3) ****(3) ****(3) 0.306 0.312 0.148 �TABLE 12-9. (continued) Adjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Officer Statistic* n Low 125 Exposure Index Medium 126 High 120 Contrast Overall M vs. L H vs. L Adj. Relative Risk (95% C.I.) p-Value ****(3) ****(3) ****(3) ****(3) ****(3) 0.045 n 50 61 53 Overall M vs. L H vs. L n 148 160 131 Overall M vs. L H vs. L 0.50 (0.16,1.57) 0.75 (0.25,2.24) Officer Paranoia Enlisted Flyer Enlisted Groundcrew Masculinity/ Femininity n Overall M vs. L H vs. L ****(2) ****(2) ****(2) ****(2) ****(2) Overall M vs. L H vs. L ****(2) ****(2) ****(2) ****(2) ****(2) Enlisted Flyer n Enlisted n Groundcrew (a) 125 50 148 126 61 160 120 53 131 Overall M vs. L H vs. L 1.06 (0.31,3.66) 1.47 (0.44,4.92) 0.479 0.234 0.604 0.789 0.922 0.530 �TABLE 12-9. (continued) Adjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Officer Psychopathic/ Deviate Statistic* n n n Officer Schizophrenia Enlisted Flyer Enlisted Groundcrev to I OJ to Low n Enlisted Flyer n Enlisted Groundcrew n 125 . 50 148 125 50 148 Exposure Index Medium 126 61 160 126 61 160 High 120 53 131 120 53 131 Contrast Adj. Relative Risk (95% C.I.) p-Value Overall M vs. L H vs. L 1.01 (0.34,2.98) 1.78 (0.65,4.83) 0.427 0.985 0.259 Overall H vs. L H vs. L 1.20 (0.42,3.41) 0.79 (0.24,2.54) 0.759 0.731 0.689 Overall M vs. L H vs. L ****(3) ****(3) Overall M vs. L H vs. L 0.72 (0.18,2.97) 0.38 (0.07,2.12) 0.511 0.654 0.269 Overall M vs. L H vs. L 0.70 (0.21,2.35) 0.52 (0.14,1.97) 0.615 0.559 0.338 Overall M vs. L H vs. L 1.32 (0.66,2.61) 1.30 (0.64,2.64) 0.682 0.429 0.471 ****(3) ****(3) ****(3) �TABLE 12-9. (continued) Adjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Officer Statistic* n n n Officer Validity Enlisted Flyer Enlisted Groundcrev Social Introversion n Enlisted Flyer n Enlisted Groundcrew n Low 125 50 148 125 51 148 Exposure Index Medium 126 61 160 126 61 160 High 120 53 131 120 53 131 Contrast Adj. Relative Risk (95% C.I.) p-Value Overall M vs. L H vs. L 1.86 (0.16,21.91) 0.247 0.620 Overall M vs. L H vs. L 0.20 0.30 (0.01,4.85) (0.02,5.61) 0.521 0.321 0.418 Overall M vs. L H vs. L 0.47 (0.15,1.49) 0.87 (0.28,2.67) 0.394 0.199 0.805 Overall M vs. L H vs. L 0.97 (0.53,1.76) 0.48 (0.24,0.93) 0.049 0.920 0.031 Overall M vs. L H vs. L 0.67 (0.23,1.94) 1.26 (0.47,3.40) 0.479 0.459 0.649 Overall M vs. L H vs. L 1.22 (0.71,2.11) 1.22 (0.69,2.14) 0.718 0.470 0.499 �TABLE 12-9. (continued) Adjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Statistic* Exposure Index Medium Low High Contrast Adj. Relative Risk (95% C.I.) p-Value Overall M vs. L H vs. L ****(4) ****(4) ****(4) 51 48 61 n Adj. Mean ****(3,4) ****(3,4) ****(3,4) 95* C.I. ****(3,4) ****(3,4) ****(3,4) Overall M vs. L H vs. L ****(3,4) ****(3,4) ****(3,4) Enlisted Groundcrev 145 154 n 12.48 Adj. Mean(b) 13.67 95Z C.I.(b) (11.33, (10.30, 16.45) 15.09) 125 13.09 (10.81, 15.82) Overall M vs. L H vs. L Officer n 119 Overall M vs. L H vs. L 0.72 (0.41,1.28) 1.11 (0.64,1.93) 0.301 0.265 0.715 Overall M vs. L H vs. L ****(4) ****(4) **() **4 ****(4) ****(4) Officer Total CHI H-R Subscore 124 n Adj. Mean ****(4) 95X C.I. ****(4) Enlisted Flyer 123 Enlisted Flyer n 48 Enlisted Groundcrev n 146 124 ****(4) ****(4) 124 120 ****(4) ****(4) 61 51 152 127 Overall M vs. L H vs. L 0.608 0.319 0.655 0.82 (0.51,1.31) 0.73 (0.44,1.19) 0.427 0.403 0.201 �TABLE 12-9. (continued) Adjusted Exposure Index Analyses for Psychological Variables by Occupation Variable Occupation Officer A-H Area Subscore Statistic* n Enlisted Flyer n Enlisted Groundcrev n Officer n Low 111 45 129 Exposure Index Medium 109 5? 145 High Contrast 112 Overall M vs. L H vs. L 45 118 i-» NJ 1 U> 01 HRB Impairmerit Index Enlisted Flyer n Enlisted Groundcrev n 124 47 145 126 61 158 118 52 127 Overall M vs. L H vs. L Adj. Relative Risk (95* C.I.) p-Value 0.78 (0.44,1.37) 0.75 (0.43,1.31) ****(3,4) ****(3,4) 0.546 0.383 0.311 ****(3,4) ****(3,4) ****(3,4) Overall M vs. L H vs. L 0.84 ( . 0 1 4 ) 05,.0 0.92 (0.53,1.59) Overall M vs. L H vs. L 0.81 (0.43,1.53) 0.57 (0.29,1.12) 0.255 0.512 0.103 Overall H vs. L H vs. L 2.28 (0.96,5.44) 1.39 (0.58,3.37) 0.159 0.063 0.461 Overall M vs. L H vs. L * * ( ). **! **() **! 0.427 0.499 0.767 **() **! ****(!) ****(!) �TABLE 12-9. (continued) Adjusted Exposure Index Analyses for Psychological Variables by Occupation *n: represents total sample size for variable in given occupational stratum. (a): marginal exposure index by race interaction (p=0.055) — relative risk, confidence interval, and p-value presented, and additional information provided in interaction summaries. (b): converted from log (X-t-1) scale, where X was the number of questions answered yes. * * ( ) exposure index-by-race interaction — relative risk, confidence interval, and p-value not presented. **!: ****(2): exposure index-by-age interaction — relative risk, confidence interval, and p-value not presented. i-* tsj ^ ****(3): exposure index-by-education interaction — relative risk, confidence interval, and p-value not presented, a* **() * * 4 : exposure index-by-drink-year interaction — relative risk/adjusted mean, confidence interval, and p-value not presented. ****(3,4): exposure index-by-education and exposure index-by-drink-year interaction — relative risk/adjusted mean, confidence interval, and p-value not presented. : no relative risk given for Total CMI, which was analyzed as a continuous variable. �Unadjusted analyses revealed a borderline significant difference between the high and low exposure levels for masculinity/femininity in officers (Est. RR: 2.38, 95% C.I.: [0.94,6.06], p=0.075), and for the total CMI in officers (low means 7.99, high mean: 10.04, p=0.018; overall p-value: 0.049). These data supported an increase in the proportion of abnormalities with increasing exposure levels. Other significant or marginally significant results were associated with a decrease in the proportion of abnormalities with an increase in exposure level. The frequency of abnormalities for the different exposure index levels exhibited no graduated pattern across exposure levels. Within the officer stratum, five variables demonstrated an increasing dose-response relationship, although usually nonsignificant; however, four variables showed the opposite pattern, that is, a decreasing proportion of abnormalities with increasing exposure levels. Few significant results were observed in the adjusted analysis, as in the unadjusted analysis. The medium level of the HRB impairment index for enlisted flyers showed an increased relative risk over the low level (Adj. RR: 2.28, 95% C.I.: [0.96,5.44], p=0.063). Many exposure index-by-covariate interactions were present, however, which prevented a direct comparison. Interactions were present for 13 of the 18 variables, but no occupational stratum was predominant. A summary of these interactions is presented in Table 12-10. TABLE 12-10. Summary of Exposure Index-by-Covariate Interactions in Adjusted Analyses of Psychological Variables Variable Occupation Anxiety Denial Depression Hypochondria Hypochondria Hysteria Hysteria Mania/Hypomania Masculini ty/Femini ty Paranoia Paranoia Paranoia Psychopathic/Deviate Total CMI Total CMI Total CMI M-R Subscore A-H Area Subscore A-H Area Subscore HRB Impairment Index Enlisted Officer Enlisted Officer Enlisted Officer Enlisted Officer Officer Officer Enlisted Enlisted Enlisted Officer Enlisted Enlisted Enlisted Enlisted Enlisted Enlisted Covariate Groundcrew Groundcrew Groundcrew Groundcrew Flyer Groundcrew Groundcrew Flyer Flyer Flyer Flyer Flyer Groundcrew 12-37 Race Age Race Education Race Education Race Drink-Years Education Age Age Race Education Drink- Years Education Drink-Years Drink-Years Education Drink-Years . Race p-Value 0.020 0.048 0.050 0.005 0.033 0.018 0.007 0.015 0.018 0.044 0.004 0.055 (marginal) 0.040 0.034 0.027 0.021 0.042 0.009 0.004 0.031 �Significant or borderline significant results in these interactions, suggestive of a dose-response relationship (i.e., increasing abnormalities or more abnormal means as exposure increases), were as follows: (1) Hysteria in college-educated officers, overall p-value = 0.025; high versus low contrast (Adj. RR: 3.49, 95% C.I.: [1.17,10.32], p=0.024); increase in the proportion of abnormalities with increasing exposure levels. (2) Mania/Hypomania in officers with greater than 50 drink-years, high versus low contrast, p»0.067; analysis affected by sparse cell sizes, however. (3) Masculinity/Femininity in college-educated officers, medium versus low contrast (Adj. RR: 3.05, 95% C.I.: [1.01,9.08], p=0.048); increase in the proportion of abnormalities with increasing exposure levels. (4) Total CMI in high school-educated, nondrinking, enlisted flyers, medium versus low contrast, p=0.018. (5) Total CMI in college-educated, nondrinking, enlisted flyers, overall p-value =0.060; analysis affected by sparse cell sizes, however. (6) M-R subscore in nondrinking, enlisted flyers, overall p-value = 0.060; analysis affected by sparse cell sizes, however. (7) A-H area subscore in high school-educated, nondrinking, enlisted flyers, overall p-value = 0.007; analysis affected by sparse cell sizes, however. ( ) HRB impairment index in nonblack enlisted groundcrew, medium versus 8 low contrast (Adj. RR: 1.88, 95% C.I.: [1.09,3.25], p-0.024). All other significant interaction results were not consistent with a dose-response relationship. In summary, no consistent or strong patterns of increasing dose-response relationship were evident throughout the psychological exposure index analyses. LONGITUDINAL ANALYSES Two scales for the MMPI, depression and denial, were significantly different by group at Baseline and were investigated to assess the longitudinal differences between the 1982 Baseline examination and the 1985 followup examination. Both variables are scores and were classified as abnormal or normal according to criteria given previously. These variables have been stratified by education level. As shown in Table 12-11, 2x2 tables were constructed for each group for each variable. These tables show the number of participants who were abnormal at Baseline and abnormal at followup, abnormal at Baseline and normal at followup, normal at Baseline and abnormal at followup, and normal at both Baseline and followup examinations. 12-38 �TABLE 12-11. Longitudinal Analysis of Depression and Denial: A Contrast of Baseline and First Followup Examination Abnormalities Variable Education Group 1982 Baseline Exam 1985 Followup Exam Odds p-Value Ratio (OR)* (OR^ vs/ORc) Abnormal Ranch Hand Abnormal Normal 59 31 48 570 0.65 Comparison Abnormal Depression High School Normal 44 52 43 695 1.21 11 10 9 227 1.11 7 1.36 5 2.20 0.04 Normal College Ranch Band Abnormal Normal 0.73 Comparison Abnormal Normal High School Ranch Hand Abnormal Normal 2 11 690 Comparison Abnormal Normal Denial 15 . 11 276 6 32 10 786 Ranch Hand Abnormal 0 5 3 249 0.56 College Normal 3.20 1.67 0.32 Comparison Abnormal Normal 5 13 -WVM- THHn- Nurober Normal Baseline, Abnormal Followup Number Abnormal Baseline, Normal Followup 12-39 3 288 4.33 �The odds ratio given is the ratio of the number of participants who were normal at the Baseline and abnormal at the followup to the number of participants who were abnormal at the Baseline and normal at the followup (the "off-diagonal" elements). The changes in normal/abnormal status within each group are contrasted between the Ranch Hand and Comparison groups, and the p-value is derived-from Pearson's chi-square test of the hypothesis that the pattern of change in the two groups is the same. The data showed a significant difference (p=0.04) in the depression scores in the two groups between examinations for the high school-educated stratum: significantly more Comparisons developed depression in the interval. The percentage of Ranch Hands with abnormalities for depression decreased from the Baseline examination to the followup examination, in contrast to the Comparison group, which showed an increase in depression abnormalities. No significant difference in the pattern of change for depression was found in the college-educated stratum, nor were any significant differences observed for denial. DISCUSSION The MMPI is a comprehensive, self-administered questionnaire containing 566 questions that broadly assess behavior, personality, and validity and consistency indicators of the responses. The MMPI data are divided into 14 scales that are not mutually exclusive for specific questions. In this study, an additional MMPI scale for the characterization of PTSD is used to identify highly correlated combat experiences of the participants. Four combat questions were selected as a surrogate measure of PTSD, and an index of these questions is used as a covariate in all of the adjusted analyses of the MMPI subscales. Distributional testing for the 14 scales of the MMPI, stratified by occupation, yielded no significant differences or discernible patterns between the two groups. In contrast, both unadjusted and adjusted analyses showed significant group differences for the denial and masculinity/ femininity scales, with the Comparisons having higher proportions of abnormalities than the Ranch Hands. Also, borderline significant associations (0.05<p<0.10) were observed for the hysteria and social introversion scales, with the Ranch Hands having slightly higher proportions of abnormalities than the Comparisons. The discrepancy in results between Kolmogorov-Smirnov distributional testing and the refined statistical models was also noted in the 1984 Baseline Report. The unadjusted and adjusted results were completely comparable with respect to group differences when direct contrast was possible, i.e., when no group-by-covariate interactions were present. Of the seven group interactions noted in the adjusted analyses, three involved the covariate of education, with the high school-educated Ranch Hands faring worse than high school-educated Comparisons. Further, the high school strata usually exhibited a higher frequency of abnormalities than the college-educated strata. Overall education showed a profound effect either as a main effect or by an interaction with another covariate. The strong influence of education was also detected in the Baseline data. Analyses using only the Original Comparisons often showed stronger group differences than the analyses based upon the total Comparison group (see Tables J-13 to J-18 of Appendix J). 12-40 �A direct comparison of the MMPI results between the Baseline and followup examinations is hampered by the small change in cohorts and the difference in statistical models. In general, at the followup the Ranch Hands manifested more MMPI scale abnormalities than the Comparisons, as judged by the number of relative risks greater than one. However, the highly significant results for the denial scale, with the Comparisons having a higher proportion of abnormalities than the Ranch Hands, suggested that the Comparisons may be underreporting on all of the MMPI scales, and consequently more relative risks greater than one would be expected. A contrast of the adjusted Baseline MMPI results to the adjusted (and unadjusted results where interactions are noted in the adjusted tests) results of the followup suggest a relatively consistent pattern of narrowing group differences over time (e.g., hypochrondria, depression, hysteria, schizophrenia scales), either by a decrease in Ranch Hand abnormalities or an increase of Comparison abnormalities. This trend was also suggested in the longitudinal analysis of two scales (depression and denial) although only the "favorable" Ranch Hand change in depression for the high school stratum reached statistical significance. Overall, the followup MMPI data suggested a subtle, but consistent, decrease in reporting of concerns (or strength of concerns) in the Ranch Hands. Only 16 participants were identified as possibly having PTSD by the MMPI subscale. Further, only 4 of 15 combat experience questions manifested strong correlation to these possible PTSD cases. Most PTSD surveys have focused on U.S. Army ground personnel, obscuring direct comparisons to U.S. Air Force personnel because of inherent differences in combat experience, education, proportion of officers, and career motivation. The CMI revealed a significant group difference for the total score and the A-H area subscore, with the Ranch Hands exhibiting higher mean scores or higher frequencies of abnormal scores. There was no group difference for the M-R subscore. These results differed slightly from the distributional tests which showed one statistically significant stratum, where the Ranch Hand mean was greater than the Comparison mean, for each covariate (see Table J-5 of Appendix J). Because the Baseline CMI was in a different format, direct comparison of each psychological parameter to the followup CMI is not feasible. However, the Baseline CMI noted statistically significant group differences for 5 of 10 parameters, which is in approximate accord with the magnitude and direction of the results found at the followup examination. This analysis of the total CMI analyzed at followup has sufficient statistical power to detect a mean difference of one response out of 195 questions (0.5% difference, at power=0.8) between the groups. Education showed the same profound effect on the adjusted analyses as was noted at Baseline. The functional integrity of the CNS, as measured by the HRB impairment index, showed no significant group differences. There was similarity (Adj. RR: 1.04, 95% C.I.: [0.86,1.25], p«0.697) in results of the impairment index. As in the Baseline analysis, education was a major covariate in the followup examination; the additionally strong effects of age and race were also noted at the followup examination. Although valid differences exist between groups for some measures, there is no indication that these differences are manifest or confirmed by impaired CNS function, a reasonable medical expectation for chemically induced neurobahavioral pathology. Adjustment of the HRB results for PTSD (not feasible at the Baseline analysis) suggests that some group differences lack organic basis. 12-41 �SUMMARY AND CONCLUSIONS Questionnaire data (verified by medical record reviews) for the lifetime events of psychotic illness, alcohol dependence, anxiety, or other neuroses disclosed no significant differences between groups for these conditions. Analyses of the followup psychological examination emphasized 14 scales from the Minnesota Multiphasic Personality Inventory (MMPI), 3 parameters of the Cornell Medical Index (CMI), and the Halstead-Reitan Battery (HRB) impairment index. The similarity of the group distribution for the 14 MMPI variables, each stratified by the 3 occupational categories, was examined, and only 2 of the 42 tests approached statistical significance. The group distributions of the total CMI score were similarly contrasted, with separate analyses performed with stratification by the five covariates of age, race, occupation, education, and current drinking status. For one stratum of each of these covariates, a significant difference in the distribution of the Ranch Hand and Comparison scores was found. In all cases for the CMI, the Ranch Hand mean was greater than the Comparison mean. Distributional analyses using Original Comparisons generally reflected the same results as those involving the total Comparison group. Results of unadjusted and adjusted analyses on all of the 18 psychological variables are given in Table 12-12. The unadjusted analyses showed a significant difference for the MMPI scales of denial (p<0.001) and masculinity/femininity (p=0.017), the total CMI (p<0.001), and the Section A-H area subscore (p=0.003). A borderline significant difference was observed for the MMPI scales of hysteria (p=0.067) and social introversion (p=0.069). Comparisons had a greater percentage of abnormal scores for the denial and masculinity/femininity scales, whereas Ranch Hands showed adverse findings for the other four variables. The overall MMPI results have been interpreted in light of the significant increased denial in the Comparison group. The covariates age, education, drink-years, current alcohol use, and occupation had pronounced effects on the psychological variables, with a significant association or a borderline significant association with at least two-thirds of the 18 psychological variables. Many dependent variables in this chapter were affected by age in an expected pattern. Very few variables exhibited this pattern of consistency with drink-years. The intermediate category of greater than 0 to 50 drink-years often had the smallest proportion of abnormalities. The post-traumatic stress disorder (PTSD) variable, derived from a subset of the MMPI, was strongly associated with the CMI measures, but not with the HRB Impairment Index. Race and the Vietnam combat index (used for the MMPI subscales) had significant associations with a lesser amount of the psychological variables (6 of 18 variables and 3 of 14 variables, for race and combat index, respectively). The adjusted analyses were generally quite similar to the unadjusted analyses with respect to group differences, although a direct comparison of these analyses was often clouded by the presence of a substantial number of interactions (six group-by-covariate interactions were significant, and three interactions approached significance [0.05<p<0.10]). The MMPI scales of denial and masculinity/femininity were statistically significant in both the 12-42 �TABLE 12-12. Overall Summary Results of Adjusted and Unadjusted Analyses of Psychological Variables Variable Questionnaire; Psychological Illness Unadjusted Adjusted Direction of Results* NS Psychological Examination; MMPI Anxiety Consistency Defensiveness Denial Depression Hypochondria Hysteria Mania/Hypomania Masculini ty/Feminini ty Paranoia Psychopathic/Deviate Schizophrenia Social Introversion Validity NS NS NS <0.001 NS NS b NS*b NS 0.017 NS NS NS NS*b NS NS **** NS <0.001 NS NS NS*b NS 0.020 **** NS **** **** **** CMI Total CMI M-R Subscore A-H Area Subscore <0.001 NS 0.003 **** NS 0.040 HRB Impairment Index NS ORH RH>C ORH RH>C RH>C RH>C NS *RH>C - more abnormalities in Ranch Hands; ORH - more abnormalities in Comparisons. b lllnesses include psychosis, alcohol dependence, anxiety, and other neuroses. —Analysis not performed. NS: Not significant. NS*: Borderline significant (0.05<p<0.10). ****Interaction involving group. 12-43 �adjusted and unadjusted analyses, where Comparisons showed an adverse effect over Ranch Hands. The A-H area subscore of the CMI (suggesting diffuse medical problems) was also significant, where the Ranch Hands had higher mean scores than the Comparisons, suggesting the Ranch Hands had more illness. Education was often involved in significant group interactions with high school-educated Ranch Hands demonstrating a higher percentage of abnormal scores than high school-educated Comparisons. No group differences were observed in the college-educated stratum. The M-R subscore of the CMI, a broad indicator of emotional health, was not statistically different between the two groups. The HRB impairment index, a measure of central nervous system (CNS) functional integrity, did not differ significantly between the Ranch Hand and Comparison groups. Strong covariates in the adjusted analysis were age, race, and education. Because of alternate statistical models and slightly different psychological testing parameters, a direct contrast between the psychological results of the Baseline and followup examinations was not always possible. However, several broad patterns were observed: (1) the discordance between distributional tests and results from traditional statistical models of the MMPI variables was noted with data from both examinations; (2) there was a narrowing of group differences at the followup examination for most subjective variables, either by a decrease in Ranch Hand reporting, or by an increase in Comparison reporting; and (3) as at the Baseline, functional CNS testing, as measured by the HRB impairment index, showed no group differences, and did not support an organic basis for differences in self-reported symptomatology. The longitudinal analysis of two MMPI scales, depression and denial, showed a significant reversal of depression seen at Baseline in the high school-educated Ranch Hands. The determination of PTSD in both Air Force cohorts by a relatively new MMPI scale showed a prevalence rate of less than 1 percent. This low rate is strongly influenced by characteristics of the study population (e.g., age, education, and officer ratio). Unadjusted exposure index analyses did not reveal any patterns consistent with a dose-response relationship. For the adjusted exposure analyses, approximately one-third presented exposure interactions with the covariates of race, education, and age, but no consistent pattern could be identified. In conclusion, some test measures of psychological health (MMPI and CMI) did not show substantial adverse effects for either group. Significant test results were present in both groups or were noted in specific subgroups of a covariate.' Educational level, age, and alcohol use showed strong effects on the psychological scales and scores in this psychological assessment. There was a subtle but consistent trend for more favorable subjective test results at the followup examination for the Ranch Hands relative to the Comparisons. Testing of the CNS by the HRB demonstrated an almost identical prevalence of pathology in both groups. 12-44 �CHAPTER 12 REFERENCES 1. Peterson, R.E., M.D. Seefeld, B.J. Christian, C.L. Potter, C.K. Kelling, and R.E. Keesey. 1984. The wasting syndrome in 2,3,7,8-tetrachlorodibenzo-p-dioxin toxicity: Basic features and their interpretation. In Banbury report 18: Biological mechanisms of dioxin action, ed. A. Poland and R.D. Kimbrough, pp. 291-308. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory, 2. Dougherty, J.A., G.E. Schulze, R.T. Taylor, and J. Blake. 1984. Behavioral toxicity of an agent orange component: 2,4-D. Oral presentation to the Veterans Administration Advisory Committee on Health-Related Effects of Herbicides, Washington, D.C., December 11, 1984. 3. Ashe, W.F., and R.R. Suskind. 1949, 1950. Reports on chloracne cases, Monsanto Chemical Company, Nitro, West Virginia. In Report of the Kettering Laboratory, December 1949 and April 1950. 4. Suskind, R.R. July 1953. A clinical and environmental survey, Monsanto Chemical Company, Nitro, West Virginia. In Report of the Kettering Laboratory, July 1953. 5. Moses, M., R. Lilis, K.D. Crow, J. Thornton, A. Fischbein, H.A. Anderson, and I.J. Selikoff. 1984. Health status of workers with past exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in the manufacture of 2,4,5-trichloro-phenoxyacetic acid: Comparison of findings with and without chloracne. Am. J. Ind. Med. 5:161-182. 6. Suskind, R.R., and V.S. Hertzberg. 1984. Human health effects of 2,4,5-T and its toxic contaminants. JAMA 251:2372-2380. 7. Baader, E.W., and A.J. Bauer. 1951. Industrial intoxication due to pentachlorophenol. Ind. Med. Surg. 20:289-290. 8. Suskind, R.R. 1977. Chloracne and associated health problems in the manufacture of 2,4,5-T. Report to the Joint Conference, National Institute of Environmental Health Sciences and International Agency for Research on Cancer, World Health Organization, Lyon, France, January 1977. 9. Goldman, P.J. 1973. Schweist akute Chlorakne, eine Massenintoxikation durch 2,3,7,8-Tetrachlorpdibenzodioxin (Severe, acute chloracne, a mass intoxication due to 2,3,7,8-tetrachloridbenzo-dioxin). Per Hautarzt. 24(4):149-152. 12-45 �10. Vos, J.G., T.J. Sterringa, D. Zellenrath, H.J. Docter, and L.M. Daldkerup. 1977. TCDD accident at a chemical factory in the Netherlands. Report to the Joint Conference, National Institute of Environmental Health Sciences and International Agency for Research on Cancer, World Health Organization, Lyon, France, January 1977. 11. Telegina, K.A., and L.J. Bikbulatova. 1970. Affection of the follicular apparatus of the skin in workers employed in the production of the butyl ester of 2,4,5-T. Vestnik. Derm. Ven. 44:35-39. 12. Jirasek, L., J. Kalensky, K. Kubec, et al. 1974. Acne chlorina, porphyria cutanea tarda and other manifestations of general intoxication during the manufacture of herbicides, part 2. Czech. Dermatol. 49(3):145-157. 13. Pazderova-Vejlupkova, J., M. Nemcova, J. Pickova, L. Jirasek, and E. Lukas. 1981. The development and prognosis of chronic intoxication by tetrachlorodibenzo-p-dioxin in men. Arch. Environ. Health 36:5-11. 14. Poland, A.P., D. Smith, G. Metter, and P. Fossick. 1971. A health survey of workers in a 2,4-D and 2,4,5-T plant, with special attention to chloracne, porphyria cutanea tardas, and psychologic parameters. Arch. Environ. Health 22(3):316-327. 15. Oliver, R.M. 1975. Toxic effects of 2,3,7,8-tetrachloro-dibenzo-l, 4-dioxin in laboratory workers. Br. J. Ind. Med. 32:46-53. 16. Hoffman, R.E., P.A. Stehr-Green, K.B. Webb, G. Evans, A.P. Knutsen, W.F. Schramm, J.L. Staake, B.B. Gibson, and K.K. Steinberg. 1986. Health effects of long-term exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. JAMA 255:2031-2038. 17. Flicker, M.R., and A.L. Young. 1983. Evaluation of veterans for agent orange exposure. Presented at the Symposium on Chlorinated Dioxins and Dibenzofurans in the Total Environment, given before the Division of Environmental Chemistry, American Chemical Society, Washington, D.C, September 1983. 18. Blackburn, A.B. 1983. Review of the effects of agent orange: A psychiatric perspective on the controversy. Military Hed. 148:333-340. 19. Diagnostic and statistical manual of mental disorders, 3rd. ed. 1980. Washington, D.C.:AmericanPsychiatry Association. 20. Faltus, F.J., A.O. Sirota, J. Parsons, M. Daamen, and M.L. Schare. 1986. Exacerbation of post-traumatic stress disorder symptomatology in Vietnam veterans. Military Med. 151:648-649. 21. Van Putten, T., and J. Yager. 1984. Posttraumatic stress disorder. Arch. Gen. Psychiatry 41:411-413. 12-46 �22. Atkinson, R.M., R.G. Henderson, L.F. Sparr, and S. Deale. 1982. Assessment of Vietnam veterans for posttraumatic stress disorder in veterans disability claims. Am. J. Psychiatry 129:1118-1121. 23. Borus, J.F. 1974. Incidence of maladjustment in Vietnam returnees. Arch. Gen. Psychiatry 30:554-557. 24. Keane, T.M., R.F. Malloy, and J.A. Fairbank. 19B4. Empirical development of an MMPI subscale for the assessment of combat-related posttraumatic stress disorder. J. Consulting and Clinical Psychology 52:888-891. 25. Yager, T., R. Laufer, and M. Gallops. 1984. Some problems associated with war experience in men of the Vietnam generation. Arch. Gen. Psychiatry 41:327-333. 26. Laufer, R.S., M.S. Gallops, and E. Frey-Wouters. 1984. War stress and trauma: The Vietnam veteran experience. J. Health and Social Behavior 25:65-85. 27. Sierles, F.S., J.J. Chen, R.E. McFarland, and M.A. Taylor. 1983. Post-traumatic stress disorder and concurrent psychiatric illness: preliminary report. Am. J. Psychiatry 140:1177-1179. A 28. Robins, L.N., J.E. Helzer, K.S. Ratcliff, and W. Seyfried. 1982. Validity of the diagnostic interview schedule, version II: DSM-III diagnoses. Psychol. Med. 12:1855-1870. 29. Cornell University Medical College. 1949. Cornell medical index health questionnaire. Ithaca, New York: Cornell University. 12-47 �CHAPTER 13 GASTROINTESTINAL ASSESSMENT INTRODUCTION This system assessment centers on reported peptic ulcer and liver disease, and current hepatic function and porphyria as determined by comprehensive laboratory testing. The liver is a major target organ for single high-dose and continued low-dose exposure to chlorophenols and TCDD. Peptic/stomach ulcer disease and porphyria cutanea tarda (PCT) are suspected clinical endpoints following moderate- to high-level exposures. A variety of experimental animal studies ~ have demonstrated hepatic dysfunction and porphyria following a wide range of exposures to TCDD. The effects of exposure, as measured by enzymatic change, however, generally appear to be more related to species than to dose and route of administration. Gross organ pathology in the digestive system and associated clinical symptoms have been observed following TCDD oral administration to (or accidental ingestion by) animals. Pathological lesions have included gastric ulcers, metaplasia of the gastric mucosa, ileitus, hepatic hypertrophy and degeneration, hepatic parenchymal cell necrosis, and hepatic lipid accumulation. Scientific interest has centered on changes in hepatic enzymes following TCDD administration. Clearly, TCDD has proved to be an exceptional inducer of hepatic enzymes and mixed function oxidases, and a powerful inhibitor of other enzymes. Specifically, the induction of cytochrome P-450, a ferrocytochrome enzyme, by TCDD has been demonstrated in many species and most of their tissues. Further, marked increases in cytochrome P-450 have been implicated in the mechanism of hepatotoxicity, although other factors, such as genetic susceptibility via the Ah locus, iron levels, and lipid peroxidation (but not vitamin A), are also contributory. TCDD has also been shown to produce hepatic porphyria in animals by a reduction in uroporphyrinogen decarboxylase, possibly due to the activation of the P-450 enzyme. ' The porphyriogenic effect of TCDD has also been influencedl fby genetic susceptibility, iron levels, sex, and ambient tem2 perature. In correlation with some human studies, hexachlorobenzene was found to be more porphyriogenic than TCDD. Numerous morbidity studies, predominantly from the industrial sector, have noted significant abnormal liver function in exposed workers, with and without the presence of clinical hepatic disease. Abnormal liver function test results have been found for direct bilirubin, alkaline phosphatase, triglycerides, cholesterol, serum glutamic-oxaloacetic transaminase (SCOT), gamma-glutamyl transpeptidase (GGTP), urine d-glucaric acid, etc. ~ The consistent finding of elevated cholesterol levels may have predictive significance with respect to future heart disease (see Chapter 15), but at present there is no evidence for this. 13-1 �Contemporary studies have focused on two indirect measures of hepatic microsomal activity, GGTP and urine d-glucaric acid. In the study of the English industrial incident, several Seveso investigations, and the two studies of the Monsanto plant in Nitro, West Virginia, there was modest agreement in observing elevated GGTP and urine d-glucaric acid levels in exposed individuals.1 '19'2 ' Common to all studies was the observation that individuals with chloracne manifested significantly more abnormal liver function tests than exposed individuals without chloracne or unexposed individuals, suggesting a link to TCDD exposure. Several industrial studies have shown altered porphyrin excretion patterns (predominantly an increase in uroporphyrin) or clinical evidence of PCT, particularly in chronically exposed workers.2 " Individuals with low chronic exposure or high acute exposure (Seveso) have not shown these signs. Further, detailed reviews of the suspected association have identified the following scientific study design and interpretive problems: (1) multiple etiologies of PCT or abnormal porphyrin excretion patterns (chemical exposure, genetic makeup, alcohol consumption), (2) misdiagnosis of PCT, and (3) confounding of chemical exposures for the industrial cohorts. Some investigators believe that the PCT cases found in the early U.S. and European studies were more likely caused by exposure to chlorobenzenes than to TCDD. Overall, the evidence at present is inconclusive to establish a causal association between PCT and TCDD exposure. A recent industrial study based on questionnaire data has suggested an association of stomach/peptic ulcers with exposure to TCDD. This finding at the Monsanto plant differs from similar research using a slightly different cohort at the same plant which produced a negative conclusion on peptic ulcer disease. The gastric ulcer-TCDD association has not been reported in other cohort dioxin morbidity studies, but ulcer disease has generally not been a major research focus. The preliminary gastric ulcerTCDD association is fortified somewhat by studies that have shown significant gastric mucosal damage in monkeys following oral administration of TCDD. Baseline Summary Results The 1982 AFHS examination conducted an extensive evaluation of hepatic status by questionnaire, physical examination, and laboratory testing. The questionnaire elicited data on liver conditions, liver disease, and symptoms compatible with PCT, as well as detailed information on PCT risk factors (e.g., alcohol consumption, chemical exposures). The physical examination measured hepatomegaly when present and determined liver function and porphyrin patterns by a comprehensive battery of 12 laboratory tests. The questionnaire showed that Ranch Hands reported more miscellaneous liver conditions (verified by medical record reviews) and more skin changes compatible with PCT than their Comparisons. Although the PCT-reported data were statistically significant, no cases of PCT were diagnosed at examination in either cohort. The physical examination detected a twofold increase in hepatomegaly in the Ranch Hands, but the numbers were small and not statistically significant. Many of the laboratory test results demonstrated statistical interactions with the covariates. These interactions can be interpreted as being 13-2 �suggestive of an herbicide effect. Ranch Hands had slightly higher GGTP and lactic dehydrogenase (LDH) results and lower cholesterol levels; no differences were found for bilirubin or alkaline phosphatase levels. SCOT, serum glutamic-pyruvic transaminase (SGPT), and LDH results in the Ranch Hands interacted with the covariates alcohol, degreasing chemicals, and industrial chemicals differently than they did in the Comparisons. All of these two-factor interactions were statistically significant (p<0.05). There were no significant group differences in uroporphyrin, coproporphyrin, or d-aminolevulinic acid levels, nor did any test set support a diagnosis of PCT. Exposure analyses were essentially negative. The comprehensive hepatic evaluation did not reveal any consistent pattern of significant liver damage in the Ranch Hand group. Nevertheless, because of subtle profile differences in conjunction with questionnaire results and recent literature citations, the gastrointestinal system continues to be targeted for intensive examination throughout all phases of the followup effort. Parameters of the 1985 Gastrointestinal Assessment The 1985 AFHS examination continued the emphasis on hepatic function and expanded the porphyrin test battery to six assays. In addition, new components were added to the questionnaire to assess past and current diagnosed peptic ulcer disease, along with a series of screening questions to assess possible undiagnosed disease. Covariate data on aspirin usage, blood group, and family history of peptic ulcer were likewise obtained. Additional probes on intestinal parasites, gallbladder disease, and other liver conditions were also added. Because of the known profound effects of alcohol ingestion on hepatic function, a detailed alcohol consumption history was obtained by questionnaire. Thus, the dependent variables and covariates in the analyses below reflect a substantial enhancement over those assessed in the 1982 Baseline examination. Because of the effects of increased body temperature and past/current hepatitis B on some liver function tests, participants with a fever of 100 or more degrees Fahrenheit and/or a positive hepatitis B surface antigen (HBgAg) test were excluded from the analyses. Categorization of continuous clinical variables to dichotomous variables was largely accomplished by use of normal test values from the SCRF laboratory. Minor numeric differences in the tables that follow are due to an occasionally missing value. The analyses are generally based on 1,009 Ranch Hands and 1,289 total Comparisons after removal of the febrile and positive HBgAg participants. The statistical analyses relied largely on general linear models (SAS®-GLM), logistic regression techniques (BMDP®-LR), and log-linear models (BMDP®-4F). Parallel analyses using Original Comparisons are found in Tables K-7 to K-16 of Appendix K. RESULTS AND DISCUSSION This chapter, entitled "Evaluation of Hepatic Status" in the Baseline Report, incorporates the new elements of peptic ulcer disease and mortality from diseases of the digestive system; hence, the chapter name change to "Gastrointestinal Assessment." 13-3 �Because of the importance of gastrointestinal disorders, numerous historical and laboratory variables were chosen for evaluation. The analyses are reported in the following order: questionnaire data, mortality data, physical examination findings, laboratory results, exposure index analyses, and representative longitudinal analyses. Questionnaire Data: Liver Disorders At the followup examination, each participant was asked whether he had developed hepatitis, jaundice, cirrhosis, or other liver disorders during the interval 1982 to 1985. Affirmative responses were subsequently subject to verification by medical record reviews. Since the Baseline interview, eight Ranch Hands and five Comparisons cited a verified history of hepatitis (p=0.264); four Ranch Hands and five Comparisons reported a subsequently verified history of enlarged liver (p=0.999); one from each group noted a verified symptom of jaundice; one Ranch Hand cited a confirmed interval history of cirrhosis; and six Ranch Hands and six Comparisons gave verified histories of seven miscellaneous liver disorders (p=0.774). Table 13-1 presents the ICD code and descriptive diagnosis of the miscellaneous liver disorders by group. Because the number of respondents with new liver disorders was small and precluded meaningful analyses, the verified interval history was added to the verified Baseline history to assess possible lifetime differences for liver disease. These combined results are presented in Table 13-2. On the basis of combined data, the verified questionnaire responses for historic hepatitis, jaundice, cirrhosis, enlarged liver, and miscellaneous liver disorders did not vary significantly between the Ranch Hand and Comparison groups. The results for miscellaneous liver disorders differed from the Baseline findings. At Baseline, significantly more Ranch Hands than Original Comparisons had a verified liver disorder other than jaundice, hepatitis, or cirrhosis (13/1,045 versus 1/773; p=0.006). Subsequent to Baseline, the status of one additional Ranch Hand disorder and one more Original Comparison disorder was verified. Including these two new verified conditions with the data from replacement and shifted Comparisons, the group contrast at Baseline would have been of borderline significance (14/1,045 versus 7/1,224; p=0.077). Combining these Baseline data with the followup data resulted in nonsignificant lifetime results. However, the combined Baseline and interval analysis contrasting the Ranch Hands and the Original Comparisons was marginally significant (p=0.065) due to the contribution of the significant Baseline results. The verification status of reported liver symptoms and diseases is presented in Table 13-3. The data reflect the proportions of historic reporting that were verified by medical record reviews, and are contrasted by group for each variable. These data showed that the proportion of verified disease was not statistically significant between groups except for the category of enlarged liver which,showed a higher confirmation rate in the Comparison group. Thus, over-reporting or symptom/disease misclassification by the participants was not a function of group membership. 13-4 �TABLE 13-1. Number of Other Liver Conditions Reported by Study Participants at Followup by Group (Verified by Medical Record Review) ICD* Code (Meaning) Group Ranch H a n d C o m p a r i s o n 5713 (Alcoholic Liver Damage) 1 57420 (Calculus of Gallbladder without Mention of Cholecystitis) 0 7891 (Hepatomegaly) 1 1 7904/7905 (Enzyme Elevation) 3 0 7948 (Abnormal Liver Scan) 1 1 E9426 (Adverse Effect of Drug) 0 1 M81406 (Adenocarcinoma) 0 1 6 6 Total *ICD = International Classification of Diseases. 13-5 �TABLE 13-2. Unadjusted Analyses for Baseline and Interval History of Liver Disease by Group (Verified by Medical Record Review) Group Ranch Hand Disease Statistic Number Comparison Percent Number Est. Relative Percent Risk (95% C.I.) p-Value Hepatitis (Viral and Alcoholic) n Yes No 1,016 37 979 3.6 96.4 1,293 43 1,250 3.3 96.7 1.10 (0.70 ,1.72) 0.731 Jaundice n Yes No 1,016 20 996 2.0 98.0 1,293 28 1,265 2.2 97.8 0.91 (0.51 ,1.62) 0.771 n Yes No 1,016 3 1,013 0.5 99.5 1,293 . 2 1,291 0.2 99.8 1.91 (0.32 ,11.46) 0.660 Enlarged Liver n Yes No 1,016 17 999 1.7 98.3 1,293 24 1,269 1.9 98.1 0.90 (0.48 ,1.68) 0.874 Miscellaneous Liver Disorders n Yes No 1,016 17 999 1.7 98.3 1,293 13 1,280 1.0 99.0 1.68 Cirrhosis 13-6 (0.81 ,3.47) 0.195 �TABLE 13-3. Medical Record Verification of Reported Liver Symptoms and Diseases by Group (Baseline and Interval Questionnaires Combined) Group Variable Verification Status Ranch Hand Comparison 47 44 3 53 48 5 37 78.7 43 81.1 43 23 20 59 35 24 20 46.5 28 47.5 7 5 2 3 3 0 3 42.9 2 66.7 30 29 I 29 29 0 11 56.7 24 82.8 21 20 1 14 14 0 17 94.4 13 92.9 Hepatitis Number Reported Medical Records Reviewed Medical Records Pending or Not Released Number Verified Percent Verified Jaundice Cirrhosis Enlarged Liver Miscellaneous Liver Disorders Number Reported Medical Records Reviewed Medical Records Pending or Not Released Number Verified Percent Verified Number Reported Medical Records Reviewed Medical Records Pending or Not Released Number Verified Percent Verified Number Reported Medical Records Reviewed Medical Records Pending or Not Released Number Verified Percent Verified Number Reported Medical Records Reviewed Medical Records Pending or Not Released Number Verified Percent Verified 13-7 p-Value 0.806 0.999 0.999 0.047 0.627 �Peptic Ulcer Diseases The primary purpose of these analyses was to compare the ulcer disease experience of the Ranch Hand and Comparison groups. Since blood type has been reported to affect the incidence of peptic ulcer disease, blood type was used as a covariate in these analyses. The military medical and personnel records of the 2,309 study participants were reviewed to determine the blood type as recorded in these sources. The distribution of blood types in the two groups is shown in Table 13-4. TABLE 13-4. Unadjusted Analysis of Blood Type by Group Blood Type 0 Group A B AB Number Percent Number Percent Number Percent Number Percent Ranch Hand 378 45.4 334 40.1 87 10.5 33 4.0 Comparison 504 46.4 425 39.1 125 11.5 33 3.0 Total* 832 1,087 p=0. 60 *184 Ranch Hands and 206 Comparisons missing from blood type analysis. The blood type distribution was not significantly different in the two groups (p=0.60), and was similar to the distribution of blood types in the general U.S. white male population (p=0.57). Both physical examination diagnoses and questionnare responses to questions concerning ulcers were used as sources of data on the-occurrence of ulcer disease. A total of 58 participants was diagnosed as having ulcer disease at the time of the examination; however, 13 had to be deleted from the analyses of physical examination data and 15 from the analyses of questionnaires due to missing data on blood type. On questionnaires, 42 reported currently having ulcers and an additional 126 reported having had ulcers in the past. These data are summarized in Table 13-5. A three-factor log-linear analysis (group, ulcer, blood type) of data from the physical examination showed a significant three-factor interaction, with the Ranch Hand rate being higher in blood types AB and 0, and lower for types A and B (p=0.03). Stratified analyses of each blood type were conducted and did not reveal any statistically significant group differences. These data are shown in Table 13-6. 13-8 �TABLE 13-5. Frequency of Diagnosed and Reported Ulcer Disease by Group Group Ranch Hand Variable Statistic Number Percent Comparison Number Percent Total Diagnosed Disease (Physical Examination Data) n Yes No 832 19 813 2.3 97.7 1,087 26 1,061 2.4 97.6 1,919 45 1,874 Reported Disease (Questionnaire Data) n Current Past None 832 22 53 757 2.6 6.4 91.0 1,085 20 73 992 1.8 6.7 91.4 1,917 44 126 1,749 A three-factor log-linear analysis of questionnaire data was also performed. This analysis looked at current and past history of ulcer disease. No significant group differences or multifactor interactions were seen, with all p-values being greater than 0.10. These analyses demonstrated overall group equivalence within the Ranch Hand and Comparison groups with respect to blood type and present and past ulcer disease. Mortality Count Data Linkage of digestive system mortality to observed historic or examination morbidity has not been explored in this report; the linkage process, with the use of the Comparison replacement strategy, remains an open research issue. From a broader perspective, however, review of mortality count data in conjunction with current morbidity data may be useful in identifying disease pattern(s) with respect to group membership, organ-specific disease, and important covariates. For these purposes, the latest mortality count data (as of 31 December 1985) are summarized in Table 13-7. These data showed a large mortality contribution (approximately 50%)from liver disease in both groups and.a relative excess in Ranch Hands as contrasted to Comparisons. For malignant neoplasms, there was a relative excess in the Comparison group. There is also the suggestion that alcohol is an important risk factor. The relative excess of malignant neoplasms in the Comparison group is also striking. Overall, the slight excess of digestive system mortality in the Ranch Hands and the differences in distribution of 13-9 �TABLE 13-6. Unadjusted Analyses of Peptic Ulcer Disease by Blood Type by Group Group Ranch Hand Comparison Blood Type Statistic Number Percent Number Percent Est. Relative Risk (95% C.I.) p-Value 0 n Yes No 378 13 365 3.4 96.6 504 11 493 2.2 97.8 1.60 (0.70,3.60) 0.37 A n Yes No 334 4 330 1.2 98.8 425 12 413 2.8 97.2 0.42 (0.14,1.29) 0.21 B n Yes No 87 0 0.0 87 100.0 125 3 122 2.4 97.6 0.27a AB n Yes No 33 2 31 33 0 33 0.0 100.0 0.49a 6.1 93.9 —Estimated relative risk and confidence interval not calculated due to zero count in a cell. "Fisher's exact test. 13-10 �TABLE 13-7. Frequency of Digestive System Mortality by Group Deaths, by Group ICD Code Ranch Hand 1:5 Comparison 1 0 3 0 1 1 0 2 Total 2 6 5 1 1 0 1 15 8 Pancreatitis (5770) Alcoholic cirrhosis (5712) Nonalcoholic cirrhosis (5715) Nonalcoholic fatty liver (5718) Chronic liver disease (5728) Alcoholic liver disease (5711) Duodenal ulcer (5325) Malignant neoplasm (150-159) 31 deaths by cause in the two groups raise the issue of competing mortality. Interpretation of the analyses in this report of hepatic function and liver disease, with alcohol consumption taken into account, should be reviewed in the light of these mortality data. Physical Examination Data Gastrointestinal dysfunction was not a major focus of the physical examination except for a comprehensive biochemical profile of the liver. Consequently, only data on hepatomegaly were analyzed, and results of the analysis are shown in Table 13-8. The analysis showed a marginally significant excess (eight cases versus three) of hepatomegaly in the Ranch Hands (p=0.069). These results were in relative contrast to the Baseline examination findings of 1.56 percent and 0.78 percent in the Ranch Hand and Comparison groups, respectively (p=0.138), in the sense that fewer abnormalities were detected at the followup, although at both examinations the difference favored the Comparisons. The group data for hepatomegaly were pooled and compared to the covariates of age, race, occupation, current alcohol use (one or less drinks per day, more than one to four drinks per day, and more than four drinks per day), lifetime exposure to industrial chemicals, and lifetime exposure to degreasing chemicals. Only age and occupation showed significant associations with hepatomegaly (p=0.018, p=0.026, respectively). Because of sparse data, an adjusted analysis was not conducted. General Laboratory Examination Data As in the Baseline Report, the followup examination emphasized evaluation of laboratory data, particularly for hepatic function. Thus, this 13-11 �TABLE 13-8. Unadjusted Analysis of Enlarged Livers Diagnosed at Physical Examination by Group* Enlarged Liver Yes Group Number No Percent Number Percent Total p-Value 0.069 Ranch Hand 8 0.8 1,002 99.2 1,010 Comparison 3 0.2 1,287 99.8 1,290 *Excludes participants with positive HB a Ag. section reports on nine laboratory tests of hepatic function and on two tests reflecting porphyrin metabolism. Normal ranges for these 11 variables as determined by the SCRF and the Mayo Clinic Laboratories are presented in Table 13-9. Only values greater than the normal range were considered important in the assessment of dysfunction. Analyses of the nine hepatic variables were adjusted for the covariates of age, race, occupation (OCC), current alcohol use (ALC), days of exposure to industrial chemicals (1C), and days of exposure to degreasing chemicals (DC). For the two porphyrin analyses, blood urea nitrogen was used as a covariate. Because the hepatic test variables encompass acute to chronic effects, there was no "ideal" alcohol covariate (e.g., drink-years, current alcohol consumption in drinks per day). The covariate alcohol use was obtained from questionnaire data, centering on daily alcohol consumption (beer, wine, liquor) for those participants who reported drinking at least one drink in the 2 weeks preceding the examination. Thus, the alcohol covariate measures recent drinking intensity and may be more useful in" adjustment of acute variables (e.g., GGTP, SGPT) than variables related to chronic liver dysfunction (e.g., bilirubin determinations, alkaline phosphatase). Exposure to industrial chemicals and degreasing chemicals was measured in cumulative days of unprotected exposure, and was derived from the 1982 and 1985 questionnaires. These data, therefore, represent lifetime exposure. Exclusion categories consisted of fever (over 100 degrees Fahrenheit) and positive HBgAg tests, because of the known effects of these conditions on liver function tests. Three participants (two Ranch Hands, one Comparison) were excluded because of fever, and eight (five Ranch Hands, three Comparisons) because of a positive HBgAg test (seven positive, one missing). In addition, due to missing alcohol data, nine other individuals (six Ranch Hands, three Comparisons) were deleted from the analyses when current alcohol use was found to be a significant covariate. 13-12 �TABLE 13-9. Laboratory Norms for Nine Hepatic Function Variables and Two Porphyrin Determinations Variable SCOT SGPT GGTP Alkaline Phosphatase Total Bilirubin Direct Bilirubin LDH Cholesterol8 Triglycerides* Uroporphyrinb Coproporphyrin Unit SCRF Normal SCRF Abnormal U/L U/L U/L U/L rag/dl mg/dl U/L mg/dl mg/dl mg/24 hrs mg/24 hrs 27-47 3-36 15-85 50-136 <1.5 <0.36 100-190 <260 <320 <46 <96 >48 >37 >86 XL37 >1.5 20.37 >191 >261 >321 >47 >97 a SCRF provides age-dependent normal ranges; these values represent the maximum normal limits for those older than 40. b Performed at the Mayo Clinic. 13-13 �Statistical Analyses The nine dependent variables from the hepatic battery were subjected to three types of basic analyses: (1) a continuous dependent variable adjusted by continuous covariates (CC), (2) a continuous dependent variable adjusted by discrete covariates (CD), and (3) a discrete (categorical) dependent variable adjusted by discrete covariates (DD), except for current alcohol use, which was left as a continuous variable for model-fitting and power purposes. General linear models (SAS®) were used for the CC and CD analyses, and BMDP®-LR was used for the DD analyses. As noted in Chapter 7, Statistical Methods, all adjustments were carried out with the simplest model, including all significant covariates and twoand three-way interactions. The log transformation was used for the nine hepatic variables and for uroporphyrin, while a square root transformation was employed for the coproporphyrin variable. Since some direct bilirubin values were 0, the value 0.10 was added prior to log transformation. The sample sizes were sufficient to detect a 1.93-fold increase in the frequency of abnormal values for alkaline phosphatase and a 1.42-fold increase in the frequency of abnormal values for SGPT, using a (two-sided) a -level of .0.05 and power 0.80, Further, the sample sizes were sufficient to detect a 0.7 percent mean shift in alkaline phosphatase, a 1.8 percent mean shift in SGPT, and a 2.8 percent mean shift in uroporphyrin values. The results of the analyses on the 11 dependent variables are presented in the following summary tables (Tables 13-10 through 13-12), followed by descriptive narrative text. The summary tables are in the following logical order: unadjusted results, covariate tests of association, and adjusted results. Tables K-l and K-2 of Appendix K summarize interactions from the statistical analyses. All analytic information on any given variable can be obtained by scanning the summary tables. The following discussion condenses the key information on each dependent variable. Group-by-covariate interactions are narratively presented. The variables are organized in the same order as given in the tables. Serum Glutamic-Oxaloacetic Transaminase (SCOT) The unadjusted continuous (group means) and categorical (percent abnormalities) tests showed no statistically significant differences between groups (p=0.298 and p=0.999, respectively). Tests of association with the covariates using pooled group categorical data demonstrated the significant effect of race (a higher percentage of abnormalities in Blacks than nonblacks, 13.5% versus 7.6%; p<0.022) and current alcohol use (21.2% abnormal values associated with more than four drinks per day, 9.0% abnormals for more than one to four drinks per day, and 5.8% for one or less drinks per day; p<0.001). Similarly, the mean SGOT levels differed significantly between races (p<0.001) and by current alcohol use (p<0.001). The CC adjusted model showed no significant group differences (p=0.309). Significant covariates were race, an interaction of current alcohol use-bydegreasing chemicals, and an interaction of current alcohol use-by-age (all 13-14 �TABLE 13-10. Unadjusted Continuous and Categorical Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Group Group Variable Statistic Ranch Hand Comparison SCOT n Mean 95% C.I. Number/% Normal High 1,009 33.5 (32.8,34.1) 1,289 33.0 (32.5,33.5) n Mean 95% C.I. Number /% Normal High 1,009 21.6 (20.9, 22.3) n Mean 95% C.I. Number/% Normal High 1,009 32.8 (31.4,34.3) n Mean 95% C.I. Number/% Normal High 1,009 91.8 (90.4, 93.3) n Mean 95% C.I. Number /% Normal High 1,009 0.74 (0.73,0.76) SGPT GGTP Alkaline Phosphatase Total Bilirubin 929 92.1% 80 7.9% 872 86.4% 137 13.6% 919 91.1% 90 8.9% 953 94.5% 56 5.6% 982 97.3% 27 2.7% 13-15 Est. Relative Risk (95% C.I.) p-Value 0.298 1,187 92.1% 1.00 (0.74,1.36) 0.999 102 7.9% 1,289 22.5 (21.9,23.1) 0.051 1,102 85.5% 0.93 (0.73,1.17) 0.546 187 14.5% 1,289 32.4 (31.2,33.6) 0.632 1,172 90.9% 0.98 (0.74,1.31) 0.942 117 9.1% 1,289 89.3 (88.1,90.6) 0.009 1,236 95.9% 1.37 (0.93,2.01) 0.114 53 4.1% 1,289 0.75 (0.74,0.76) 0.576 1,250 97.0% 0.88 (0.54,1.45) 0.706 39 3.0% �TABLE 13-10. (continued) Unadjusted Continuous and Categorical Analyses for Hepatic Function Variables and Two Porphyrin Determinations by Group Group Variable Statistic Ranch Hand Comparison Direct Bilirubin n Mean 95% C.I. Number/% Normal High 1,009 0.18 (0.17,0.18) 1,289 0.18 (0.17,0.18) n Mean 95% C.I. Number /% Normal High 1,009 123.5 (122.2,124.8) n Mean 95% C.I. Number /% Normal High 1,009 214.3 (211.8,216.8) Est. Relative Risk (95% C.I.) p-Value LDH Cholesterol 971 38 999 10 96.2% 3.8% 99.0% 1.0% 863 85.5% 146 14.5% 0.981 1,246 96.7% 1.13 (0.73,1.77) 0.649 43 3.3% 1,289 123.9 (122.7 ,125.2) 0.655 1,272 98.7% 0.75 (0.34,1.64) 0.560 17 1.3% 1,289 215.0 (212.8 ,217.2) 0.688 1,082 83.9% 0.88 (0.70,1.11) 0.322 207 16.1% Triglycerides n Mean 95% C.I. Number/% Normal High 1,009 1,289 118.5 117.3 (113.8,123.3) (113.4 ,121.4) Uroporphyrin n Mean 95% C.I. 1,006 16.9 (16.2,17.7) 1,286 17.9 (17.3,18.6) 0.048 1,008 119.1 (116.2,122.0) 1,287 115.6 (113.0,118.2) 0.081 Coproporphyrin n Mean 95% C.I. 941 93.3% 68 6.7% 13-16 0.719 1,210 93.9% 1.11 (0.79,1.55) 0.549 79 6.1% �TABLE 13-11. Association Between Nine Hepatic Function Variables and Two Forphyrin Determinations and Six Covariates in the Combined Ranch Hand and Comparison Groups Analysis* Age Race SCOT C D NS NS SGPT C D <0.001 0.001 GGTP C D 0.012 NS Alkaline Phosphatase C D NS NS Total Bilirubin C D NS NS* NS <0.001 Direct Bilirubin C D NS NS LDH C D Cholesterol Industrial Degreasing Occupation Alcohol Chemicals Chemicals <0.001 0.022 Variable NS NS <0.001 <0.001 NS NS NS NS <0.001 <0.001 NS NS 0.017 NS 0.032 NS <0.001 <0.001 NS NS NS NS <0.001 0.003 <0.001a NS*a 0.011 NS 0.008 NS NS NS NS NS NS* 0.015 NS NS <0.001 NS NS NS NS NS <0.001 NS 0.006 NS NS NS NS NS NS NS C D <0.001 0.010 NS NS 0.002 0.008 <0.001 0.018 NS NS NS NS Triglycerides C D <0.001 <0.001 0.031 NS 0.013 NS 0.030 NS NS* NS* 0.019 NS Uroporphyrins C NS NS NS NS NS Coproporphyrins C NS NS <0.001 0.021 NS NS NS NS NS 0.003 NS NS Continuous (C)/Discrete (D). NS: Not significant (p>0.10) NS*: Borderline significant (p.05<p<0.10). a Wine consumption. 13-17 NS NS <0.001 <0.001 0.030 NS NS 0.010 NS* �TAHTK 13-12. Adjusted Continuous and Categorical Analyses for Hepatic Function Variables and T\*> Borphyrin Determinatdons by Group Group Variable Analysis Statistic Ranch Band Comparison CC SOOT CD CD OC p s CD CD CC QGIP CD CD Covariate Ranarks* n 1,003 Adj. Mean 34.8 95* C.I. (33.8,35.7) n 1,003 Adj. Mean **** 95% C.I. **** n 103 ,0 1,286 34.3 — 039 .0 (343.) 3.,53 1,286 **** _ **** **** ' 1,286 1 0 ( . 5 1 4 )0 8 8 . 3 07,.1 .6 n 1,286 ALC*DC(p=0.008), RAGE*DC(p=Q.015) 22.2 — 0.048 AGE*AKXP=0-001),r RACE*IC(p=0.017) (21.3, 23.3) ALD*DC(p=0.032), AG£*ALC(p=0.022) 1,286 — 0.029 OCC*AGE(p=0.026),, IC(p=0.049) 22.9 (21.1,24.8) 1,286 0.93 (0.73,1.18) 0.531 A(£*ALC(p=Q.004) 1,003 21.4 95% C.I. (20.4,22 •4) n 1,003 Adj. Mean 21.9 95% C.I. (20.2,23.8) n 1,003 Adj. Mean SGPT Adj. Relative Risk (95% C.I.) p-Value 1,286 n 103 ,0 Adj. Mean 37.5 37.0 0.575 95% C I ( 5 2 4 . ) ( 4 7 3 . ) . . 3.,01 3.,93 n 103 ,0 1,286 Adj. Mean 4 . 41 068 .6 4. 36 95% C I ( 0 0 4 . ) ( 9 6 4 . ) . . 4.,86 3.^79 n 1,286 1 0 ( . 4 1 3 )0 9 1 . 0 07,.4 . 7 1,003 ALC*DC(D<0.001) AGE*AlC(p<D.001) RACE(p«0.001) GRP*ALC(p=0.048) ALC*IC(p=0.008) DC(p=0.019), RACE(p<0.001) AGE*AI£(p<D.001) OCC*AIJC(D<0.001) RACE ( = . 2 ) p006 AGE*ALC(p<0.001) ,RACE*IC(p=0.011) ALC*DC(p<O.C01), ACE*TC(p=0.009) ACE*ALC(p=0.023), OCC*ALC(p=0.044) RACE(pO.COl) AGE*AIC(p<0.001), RACE(p=0.016) �TSHIE 13-12. Continued) feriahles and Ttao Porphyrin Determinations by Ckoup Group Variable Analysis Statistic Ranch Hand Comparison CO Alkaline Phosphatase CD ED OC Total Bilirubin CD DD OC Direct Bilirubin CD DD n 1,003 Adj. Mean 91.6 95% C.I. (89.4,93.9) n Adj. Mean 95% C.I. n 1,003 **** **** 1,003 1,285 89.1 (87.0,91.2) 1,285 Adj. Relative Risk (95% C.I.) p-Value — Covariate Remarks* AG£*IC(p=0.010), RACE*IC(p=0.007) 0.008 OCC(p<0.001), W3NE(p<0.001) GRP*IC(p=Q.011), #£*IC(p4).019) RACE*IC(p=0.002), OCC(p<0.001) WINE (p<0.001) 1.44 (0.97,2.13) 0.070 WINE*DC(p4).006), AGE*IC(p=0.005) RACE*IC(p=0.004), OCC*IC(p=0.016) _ 1,285 **** — A3G*DC(p=0.039) 0.599 RACE*ALC(p=0.007) RAC£*OCC(p=0.001) n 1,003 Adj. Mean 0.78 95% C.I. (0.75,0.81) 1,286 0.78 (0.75,0.81) n 1,003 Adj. Mean 0.83 95% C.I. (0.79,0.87) n 1,009 1,286 RAGE*ALC(p=0.004) — 0.598 OCC*ALC(p=0.034) 0.83 OOC*RACE(p=0.002) (0.80,0.87) 1,289 0.89 (0.54,1.47) 0.648 RACE(pO.OOl) n 1,003 Adj. Mean 0.18 95% C.I. (0.17,0.20) 1,286 0.18 (0.17,0.19) — 0.972 RACE*ALC(p=0.025) n 1,003 Adj. Mean 0.21 95% C.I. (0.19,0.22) n 1,003 1,286 0.20 (0.19,0.22) 1,286 — DC*IC(p=0.025), ALCMX:(p=0.012) 0.830 RACE*ALC(p=0.019), OCC*ALC(p=0.002) **** **** GRP*IC(p=0.012), RACE(p=0.014) ALC(p=0.026) �TAWR 13-12. (continued) Adjusted Continuous and Categorical Analyses for Hepatic Rncti.cn Variables and Two Borphyrin Deteodnaticns by Group Group Variable Analysis Statistic Ranch Hand Comparison CC LDH CD DD CC 8 Cholesterol CD ED CC Triglycerides CD ED n Adj. Mean 95% C I .. 1,003 Adj. Realtive Risk ( 5 C I ) p-Value 9% . . 1,286 «M Covariate Remarks* GRP*AGE(p=Q.018), OCC*IC(p=Q.014) RACE*IC(p=Q.024) AAAA n 1,003 1,286 RACE(p<D.001), AGE(p<D.001) Mj. Mean 130.0 130.5 — 0.671 DC(p=0.016) 95% C.I. (127.3,132.8) (127.8,133.1) n 1,009 1,289 0.75 (0.34,1.64) 0.560 n 103 ,0 126 ,8 Adj. Mean 295 1. 220.4 — 064 .0 95% C I ( 1 . , 2 . ) ( 1 . , 2 . ) .. 245247 255254 n 103 ,0 126 ,8 Adj. Mean 223.8 249 2. — 058 .4 95% C I ( 1 . , 3 . ) ( 1 . , 3 . ) . . 277201 289210 n 103 ,0 1 2 6 0 8 ( . 8 1 0 ) 0.1S1 ,8 . 5 06,.8 RACE*DC(p=O.021), RACE*OCC(p=0.005) I ( = . 4 ) ALC(p<0.001) Cp003, AG£(p<D.001) RACE*OCC(p=0.027), Al£(p<0.001) AGE(p<0.001) n 103 ,0 126 ,8 Adj. Mean **** **** — 95% C I .. **** **** n 103 ,0 126 ,8 Adj. Mean 125 1. 121 1. — 95% C I ( 0 . , 2 . ) ( 0 . , 2 . ) .. 137119 137112 n 109 ,0 129 ,8 **** GRP*AGE(p=0.015), AIC*DC(p=0.005) RACE*ALC(i>=0.031), OCC(p<D.001) **** RACE*AlJC(p=0.012), AGE(p=d0.029) OCC(p=0.039) OCC(p<D.001), RACE(p<0.001) 0 9 5 AGE(p<0.001), ALC(p=0.038) .0 **** GRP*OCC(p=0.027), RACE ( = . 2 ) p006 IC(p=0.038) �TABLE 13-12. (continued) Adjusted Continuous and Categorical Analyses for Hepatic Flncti.cn Variables and Two Borphyrin Determinations by Group Group Variable Analysis Statistic Ranch Band Comparison CC n Adj. Mean 1,000 **** rtCV Uroporphyrin **A* f t T y;X5 C.I. Coproporphyrin CC Jnlrrlr^Ij Adj. Relative Risk (95%C.I.) p-Value 1,283 **** ^«_ -* _t^t_l- * * * * n 1 0 2 , 0 1 2 4 , 8 Adj. Mean 1 9 3 1 5 7 1. 1 . 95% C I ( 1 . , 2 . ) ( 1 . , 1 . ) . . 164122 132182 "XfCfCX Covariate Remarks* GRP*EUN(p=0.015) DC*OOC(p=O.005) ALC(p=0.026) 0 0 5 BUN(p<0.001) .6 *Abbreviations; GRP: group OCC: occupation ALC: currait alcohol use WINE: vine consumption DC: exposure to degreasing chemicals 1C: exposure to industrial chemicals BIN: blood urea nitrogen — No relative risk or confidence interval given for continuous analyses. **** Group-by-covariate interaction—adjusted mean/relative risk, confidence interval, and p-value are not presented. �with p<0.001). The CD analysis revealed a significant group (GRP)-by-current alcohol use interaction (p=0.048), precluding a direct group contrast. Exploration of the interaction disclosed that the Ranch Hands had a significantly higher (p=0.010) mean SCOT for the more than one to four drinks per day category, whereas there were no significant group differences for the one or less drinks per day or more than four drinks per day categories (see Table K-l of Appendix K). Other significant covariate effects included d.egreasing chemicals (p=0.019), race.(p<0.001), and a. current alcohol use-by-industrial chemical (1C) interaction (p=0,008). The DD SCOT analysis showed no significant group differences (p=0.868). Covariates making significant contributions were race (p=0.026), an age-by-current alcohol use interaction (p<0.001), and an occupation (OCC)-by-current alcohol use interaction (p<0.001). Serum Glutamic-Pyruvic Transaminase (SGPT) The unadjusted categorical analysis was not significant (p=0.546), but the comparison of group means showed a borderline significant result, with the Comparisons having a higher mean SGPT than the Ranch Hands (p=0.051). Covariate associations with the pooled categorical Ranch Hand and Comparison group data showed an inverse relationship (p=0.001) between SGPT levels and age, with 17.1 percent abnormalities for those born in or after 1942, 12.3 percent for those born between 1923 and 1941, and 8.1 percent for those born in or before 1922.' The relationship with current alcohol use was also profound (p<0.001), with 23.4 percent abnormals noted for more than four drinks per day, 15.3 percent abnormals for more than one to four drinks per day, and 12.4 percent for one or less drinks per day. The direction and magnitude of the covariate effects of age and alcohol were quite similar for the tests of association with the mean SGPT level of both groups (p<0.001 for both covariates). No significant group interactions were detected in either the discrete or the continuous analyses. The CC-adjusted analysis yielded a significant group difference, with the Comparisons having a higher group mean than the Ranch Hands (p=0.048). The model was adjusted by the interactions of current alcohol use-by-degreasing chemicals (p=0.008), current alcohol use-by-age (p=0.001), race-by-degreasing chemicals (p=0.015), and race-by-industrial chemicals (p=0.017). The CD model also showed a significantly elevated mean SGPT in the Comparison group (p=0.029). The analysis was adjusted for exposure to industrial chemicals (p=0.049), and the interactions of age-byoccupation (p=0.026), age-by-current alcohol use (p=0.022), and current alcohol use-by-degreasing chemicals (p=0.032). A borderline significant interaction (p=0.0505) between group and current alcohol use was found, but because of modeling strategy, this interaction was not included in the final model. (This interaction is explored further in Table K-l in Appendix K, however.) The DD-adjusted analysis, like the unadjusted discrete analysis, disclosed a nonsignificant group difference (p=0.531). The model was adjusted for an age-by-current alcohol use interaction (p=0.004). Gamma-Glutamyl Transpeptidase (GGTP) The unadjusted contrasts of both mean levels of GGTP and the frequency of abnormalities showed no significant differences between the Ranch Hand and Comparison groups (p=0.632 and p=0.942, respectively). 13-22 �For discrete covariate associations, significance was noted for race, with 14.9 percent abnormals in Blacks and 8.6 percent for nonblacks (p=0.021), and current alcohol use, with 26.1 percent abnormals for more than four drinks per day, 10.5 percent for more than one to four drinks per day, and 6.2 percent for one or less drinks per day use (p<0.001). While the mean level of GGTP was similarly affected by race and current alcohol (p<0.001 for both covariates), it was also influenced by age (30.3 U/L for those born in or before 1922, 33.9 U/L for those born between 1923 and 1941, and 31.1 U/L for those born in or after 1942; p=0.012) and occupation (31.5 U/L for officers, 35.2 U/L for enlisted flyers, and 32.5 U/L for enlisted groundcrew; p=0.032). Each of the three adjusted analyses consistently produced nonsignificant group differences (CC: p=0.575; CD: p=0.668; DD: p=0.971). None of the three models was affected by a group-by-covariate interaction. The CC analysis was adjusted by four covariate interactions: age-by-current alcohol use (p<0.001), race-by-industrial chemicals (p=0.011), current alcohol useby-degreasing chemicals (p<0.001), and age-by-degreasing chemicals (p=0.009). The CD model was adjusted by race (p<0.001), by an age-by-current alcohol use interaction (p=0.023), and by an occupation-by-current alcohol use interaction (p=0.044). The DD analysis was adjusted by race (p=0.016) and by an age-by-current alcohol use interaction (p<0.001). Alkaline Phosphatase The analysis of group mean values showed a significantly higher (p=0.009) Ranch Hand mean (91.8 U/L) than that observed in the Comparison group (89.3 U/L). The unadjusted categorical analysis revealed a higher percentage of Ranch Hand abnormalities (5.6%) than Comparison abnormalities (4.1%), but this difference was not significant (Est. RR=1.37, 95% C.I.: [0.93,2.01], p=0.114). With pooled group data, significant covariate associations were found between the proportion of abnormal values and occupation (p=0.003), industrial chemicals (p=0.030), and marginally significant associations with wine consumption (p=0.056) and degreasing chemicals (p=0.091). The mean value of alkaline phosphatase depended significantly on all four of these covariates. The CC-adjusted analysis also showed a significantly higher mean value of alkaline phosphatase in the Ranch Hand group (p=0.008). The model was adjusted by the significant covariates of wine consumption (WINE) (p<0.001), occupation (p<0.001), and the interactions of age-by-industrial chemicals (p=0.010) and race-by-industrial chemicals (p=0.007). Wine consumption was used as a covariate instead of alcohol intensity since wine showed a very strong negative association with alkaline phosphatase. This effect masked a very weak positive association between beer or liquor consumption and alkaline phosphatase. In the CD model a significant group-by-industrial chemicals interaction was found (p=0.011). Specifically, in those individuals exposed to industrial chemicals, the Ranch Hands had a significantly higher mean value than the Comparisons (p<0.001), whereas in the unexposed stratum, the mean values were not significantly different between groups (p=0.973; see Table K-l of Appendix K). The CD analysis was also adjusted by wine consumption (p<0.001), occupation (p<0.001), and the interactions of age-by-industrial chemicals (p=0.019) and race-by-industrial chemicals (p=0.002). 13-23 �The DD model revealed a marginally significant group difference (Adj. RR: 1.44, 95% C.I.: [0.97,2.13], p=0.070) following adjustment by four significant interactions of wine-by-degreasing chemicals (p=0.006), age-byindustrial chemicals (p=0.005), race-by-ihdustrial chemicals (p=0.004), and occupation-by-industrial chemicals (p=0.016). Total Bilirubin Both the continuous and categorical unadjusted analyses found no significant differences in total bilirubin values between groups (p=0.576 and p=0.706, respectively). The covariate associations for both groups showed a significant effect of race (8.5% abnormal in Blacks versus 2.5% in nonblacks; p<0.001). Significant differences in mean total bilirubin levels were found between occupational groups (0.76 mg/dl for officers, 0.72 mg/dl for enlisted flyers, and 0.75 mg/dl for enlisted groundcrew; p=0.011), and with increasing levels of current alcohol use (0.80 for more than four drinks per day, 0.75 for more than one to four drinks per day, and 0.74 for one or less drinks per day; p=0.008). Further, increasing levels of total bilirubin were marginally associated with age (p=0,093). The CC model, adjusted for the interactions of age-by-degreasing chemicals (p=0.039), race-by-current alcohol use (p=0.007), and race-byoccupation (p=0.001), revealed no significant differences in total bilirubin means between groups (p=0.599). Similarly, the CD analysis found no difference between group means (p=0.598) after adjustment for the interactions of race-by-current alcohol use (p=0.004), occupation-by-current alcohol use (p=0.034), and occupation by race (p=0.002). The DD model, adjusted for race (p<0.001), also failed to detect significant group differences in the proportion of total bilirubin abnormalities (p=0.648). Direct Bilirubin Neither the continuous nor the categorical unadjusted tests disclosed significant differences between the Ranch Hand and Comparison groups (p=0.981 and p=0.649, respectively). A covariate association with the categorical data combined from both groups was noted for race, with 7.8 percent abnormalities found in Blacks as contrasted to 3.3 percent in nonblacks (p=0.015). There was a significant association between mean values of direct bilirubin and current alcohol use (0.21 mg/dl, 0.17 mg/dl, and 0.17 mg/dl for more than four drinks per day, more than one to four drinks per day, and one or less drinks per day, respectively; p<0.001) and a marginally significant difference due to race (0.20 mg/dl for Blacks versus 0.18 mg/dl for nonblacks; p=0.059). For both the CC and CD analyses, no significant group differences were found (p=0.972 and p=0.830, respectively). The CC model was adjusted for a race-by-current alcohol use interaction (p=0.025), and the CD model was adjusted for the significant interactions of race-by-current alcohol use (p=0.019), occupation-by-current alcohol use (p=0.002), current alcohol use-by-degreasing chemicals (p=0.012), and degreasing chemicals-by-industrial chemicals (p=0.025). The DD analysis revealed a group-by-industrial chemical 13-24 �exposure interaction (p=0.012). For participants exposed to industrial chemicals, the Ranch Hands had a higher proportion with abnormal values than the Comparisons (5.3% abnormal versus 2.9%, respectively; p=0.035), whereas there was no group difference for participants not exposed to industrial chemicals (p=0.144). Each stratum of the interaction was adjusted for race (p=0.014) and current alcohol use (p=0.026). The biological relevance of this interaction is unclear at this time. Lactic Dehydrogenase (LDH) No significant differences were found between the groups, either in the proportion of abnormal values (p=0.560) or in the mean levels of LDH (p=0.655). Significant effects for age (121.6 U/L, 124.6 U/L, 135.3 U/L for those born in or after 1942, between 1923 and 1941, and in or before 1922, respectively; p<0.001) and race (129.5 U/L for Blacks versus 123.4 U/L for nonblacks; p=0.006) were found in the tests of mean LDH levels. The CC analysis revealed a group-by-age interaction (p=0.018), although no significant adjusted group differences were found for any of the three age strata. The model was also adjusted for the significant interactions of occupation-by-exposure to industrial chemicals (p=0.014) and race by exposure to industrial chemicals (p=0.024). The CD model revealed no significant group differences after adjustment by age (p<0.001), race (p<0.001), and degreasing chemicals (p=0.016). Similarly, the DD analysis found no significant group differences, and no covariates made a significant contribution to the model. Cholesterol No significant differences were found between groups, either in the proportion of abnormal cholesterol levels (p=0.322) or in mean values of cholesterol (p=0.688) by unadjusted tests. However, in contrast, analysis of the Ranch Hand group versus the Original Comparisons (see Table K-9 of Appendix K) showed that the Comparisons had a significantly higher proportion of abnormal levels than the Ranch Hands (18.3% versus 14.5%, respectively; Est. RR: 0.76, 95% C.I.: [0.60,0.96], p=0.023). This observation was also found at Baseline. Significant covariate associations were noted between the proportion of participants with abnormally high cholesterol levels and age (12.7% for those born in or after 1942, 17.2% for those born between 1923 and 1941, and 18.4% for those born in or before 1922; p=0.010), occupation (14.9% for officers, 20.5% for enlisted flyers, and 13.9% for enlisted groundcrew; p=0.008), and current alcohol use (14.1% for one or less drinks per day, 16.4% for more than one to four drinks per day, and 21.7% for more than four drinks per day; p=0.018). For the associations between mean cholesterol levels and age, occupation, and current alcohol use, the significance of the covariate effects was greater than for the discrete analyses (p<0.001, p=0.002, and p<0.001, respectively). The CC results showed no significant group difference (p=0.604). The model was adjusted by age (p<0.001), current alcohol use (p<0.001), industrial chemical exposure (p=0.043), and the race-by-degreasing chemicals (p=0.021) and race-by-occupation (p=0.005) interactions. The CD analysis was negative for significant group differences (p=0.548). The analysis included the covariate contributions made by age (p<0.001), current alcohol use 13-25 �(p<0.001), and a race-by-occupation interaction (p=0.027). The DD analysis also showed no significant difference between groups for adjusted proportions of participants with abnormal cholesterol levels (p=0.181). Contributing covariates included age (p=0.029), occupation (p=0.039), and a race-bycurrent alcohol use interaction (p=0.012). In all of the discrete cholesterol analyses, the cutpoint of 260 mg/dl was used. Triglycerides In the unadjusted analyses, no significant differences in the proportion of participants with abnormal triglyceride levels or in mean values were found between the Ranch Hand and Comparison groups (p=0.549 and p=0.719, respectively). The covariate tests of association for percent abnormal triglycerides disclosed the significant effect of race (2.1% for Blacks and 6.7% for nonblacks; p=0.031) and a marginally significant association for industrial chemical exposure (p=0.073). For mean triglyceride levels, significant associations for age (p<0.001), race (p<0.001), occupation (p=0.013), current alcohol use (p=0.030), and degreasing chemicals (p=>0.019) were noted, in addition to a marginally significant association with exposure to industrial chemicals (p=0.077). The CC analysis revealed a significant group-by-age interaction (p=0.015), which showed a significantly elevated mean triglyceride level in Ranch Hands (p=0.039) born in or before 1922 as compared to similarly aged Comparisons (see Table K-l of Appendix K). There were no significant differences for the other two age strata. A significant adjusting covariate was occupation (p<0.001); in addition, the current alcohol use-by-degreasing chemicals (p=0.005) and race-by-current alcohol use (p»0.031) interactions were used for adjustment. The CD-adjusted analysis found no significant group differences (p=0.905). The model was adjusted by age (p<0.001), race (p<0.001), occupation (p<0.001), and current alcohol use (p=0.038). The DD analysis found a significant group-by-occupation interaction (p=0.027). Stratification by occupation revealed a significant increase in the proportion of abnormal triglyceride levels for Ranch Hand officers (Adj. RR? 1,77, 95% C.I.: [1.04,3.01], p=0.035) but no significant group differences were discerned for the enlisted flyer or enlisted groundcrew strata. The models were adjusted by race (p=0.026) and industrial chemical exposure (p=0.038). A cutpoint of 320 mg/dl was used to distinguish abnormal from normal. Uroporphyrin The uroporphyrin variable was analyzed only in the continuous form. The unadjusted analysis revealed a significant difference between group means (Comparisons 17.9 mg/24 hrs, Ranch Hands 16.9 mg/24 hrs; p=0.048). A CC model found a significant group-by-blood urea nitrogen (BUN) interaction (p=0.015; see Table K-l of Appendix K). To interpret the interaction, BUN was dichotomized at the median value of 14 mg/dl. Stratifying by BUN levels revealed a significantly greater (p<0.001) uroporphyrin mean for Comparisons than for Ranch Hands for BUN levels of 14 or less mg/dl and a nonsignificant but greater Ranch Hand mean for participants with BUN levels of 13-26 �more than 14 mg/dl. The stratified model was adjusted for current alcohol use (p=0.026) and the occupation-by-degreasing chemicals (p=0.005) interaction. Coproporphyrin As with the uroporphyrin variable, coproporphyrin was analyzed only as a continuous variable. The unadjusted analysis revealed a borderline significant difference in the mean coproporphyrin levels (119.1 mg/24 hrs for Ranch Hands and 115.6 mg/24 hrs for Comparisons; p=0.081). The covariate tests of association detected the significant effects of age (p»0.003) and current alcohol use (p=<0.001). A CC model, adjusted by BUN (p<0.001) and an age-by-current alcohol use interaction (p=0.003) revealed a borderline significant group difference (p=0.065) similar to the unadjusted analysis. The adjusted coproporphyrin means were 119.3 mg/24 hrs and 115.7 mg/24 hrs for the Ranch Hands and Comparisons, respectively. Discussion The results from the nine hepatic and two porphyrin analyses were not totally consistent with the Baseline findings. Several analytical reasons may possibly explain some of these differences, i.e., the adjusted analyses herein used the additional covariates of age, race, and occupation (the matching variables), and all two-way covariate interactions. However, as the Baseline data were not reanalyzed with the model process and total Comparison group used in this report, the contribution of analytic technique versus a true change in hepatic status is unknown. The Baseline Report noted a significantly lower mean cholesterol level in the Ranch Hands (opposite of an expected dioxin effect) and slight tendencies for higher GGTP and LDH values in the Ranch Hands. In this chapter, the analyses have shown equivalent group cholesterol levels, an increased SGFT mean in the Comparisons, an increased mean alkaline phosphatase in the Ranch Hands, an increased uroporphyrin mean in the Comparisons, and a borderline increased coproporphyrin mean in the Ranch Hands. The individual hepatic assay results were not suggestive of a pattern of significant hepatic damage in the Ranch Hands that might be supportive of an herbicide effect. Further, there was no consistent group-by-covariate interaction that suggests a detriment to a specific subcategory of the Ranch Hands. For those covariates used in both the Baseline study and this followup study, the direction and magnitude of their effects were relatively consistent between the studies. However, an unexpected association between wine drinking and alkaline phosphatase lacks a plausible explanation, particularly considering the inverse relationship, i.e., increasing alkaline phosphatase levels with decreasing wine consumption. These findings suggested the association between wine and alkaline phosphatase may be due to imprecision in data collection. None of the categorical (normal/abnormal categories) analyses was statistically significant, whereas all of the significant results were 13-27 �generated by the more powerful contrasts of continuously distributed hepatic data. Both porphyrin analyses showed group associations and are in distinct contrast to the otherwise largely negative hepatic findings. The significantly elevated uroporphyrin mean value in the Comparisons was directly opposite to that expected if dioxin-induced PCT were prevalent in the Ranch Hands. The primary biochemical defect in PCT is the reduced activity of uroporphyrinogen decarboxylase, an enzyme that metabolizes uroporphyrin. This defect leads to increased levels of uroporphyrin and coproporphyrin. Questionnaire-Laboratory Correlations; Porphyria Cutanea Tarda In the interval questionnaire all participants were asked whether their skin manifested "patches," excessive bruises, or sensitivity. These questions were deemed important in order to bound the maximum prevalence of cutaneous disorders compatible with a diagnosis of PCT. These historical data are given in Table 13-13. TABLE 13-13. Unadjusted Analysis for Interval History of Skin Bruises, Skin Patches, and Skin Sensitivity by Group Bruises, Patches, or Sensitivity Yes No Group Number Percent Ranch Hand 265 26.2 Comparison 260 20.2 Number Percent 746 1,029 Total p-Value 73.8 1,011 0.001 79.8 1,289 These data revealed that the Ranch Hands reported significantly more cutaneous symptoms (26.2%) than the Comparisons (20.2%). However, these data were substantially less than those reported at the Baseline in-home questionnaire, which also showed a statistically significant excess in the Ranch Hands. To determine if the skin histories might be related to PCT, the historic data were compared to the porphyrin test results. The abnormal/normal cutpoint of the coproporphyrin assays was reset to the 95th percentile because the normal range of the laboratory overclassified the proportion of abnormals. Table 13-14 gives the tabular display of both porphyrin test results by the reporting history of skin disorders. 13-28 �TABLE 13-14. Unadjusted Analyses for Porphyrin Abnormalities by Group and Skin Patch, Bruise, or Sensitivity Reported at Followup Questionnaire Group Abnormal Porphyrin Findings for a Participant Skin Patch, Bruise, or 0 1 2 Sensitivity Reported Number Percent Number Percent Number Percent Total p-Value* Both Groups Yes No 472 1,593 90.1 90.2 48 165 9.2 9.3 4 9 0.8 0.5 524 0.789 1,767 Ranch Hand Yes No 239 670 90.5 90.3 24 70 9.1 9.4 1 2 0.4 0.3 264 0.950 742 Comparison Yes No 233 923 89.6 90.1 24 95 9.2 9.3 3 7 1.2 0.7 260 0.742 1,025 *Chi-square test, 2 d.f. The data from both groups combined suggest that there is no relationship between a history of cutaneous disorders and porphyrin test positivity. The group-specific data in the table also show a lack of a statistically significant association between the reporting of skin patches, bruises, or sensitivity and the presence of an abnormal porphyrin test result. However, in both the Ranch Hand and Comparison groups, participants who had abnormal tests for both uroporphyrins and coproporphyrins were more likely to have reported cutaneous disorders than participants with normal findings for both tests. Consequently, the data were retabulated, focusing only upon uroporphyrin abnormalities (absolutely required for a diagnosis of PCT) and reporting of cutaneous disorders. These data are summarized in Table 13-15. These data suggest that the relative risk of increased uroporphyrin abnormalities for Ranch Hands is independent of whether or not a study participant reported skin patches, bruises, or sensitivities at the followup questionnaire (Breslow-Day test of homogeneity of odds ratio, p=0.791). In each instance (reported/not reported), the estimated relative risk was nonsignificant and less than 1, and in both the Ranch Hand group and the Comparison group there was a higher percentage of uroporphyrin abnormalities for participants who did not report skin patches, bruises, or sensitivity than for participants who did report these conditions. Thus, the sequential displays of Tables 13-13 through 13-15 show excessive reporting of PCT-like cutaneous symptoms in the Ranch Hand group that was not related to test abnormalities for both uroporphyrin and coproporphyrin abnormal test results, or for uroporphyrin abnormalities alone. These analyses were consistent with the clinical observation that 13-29 �TABLE 13-15. Unadjusted Analyses for Uroporphyrin Abnormalities by Group and Skin Patch, Bruise, or Sensitivity Reported at Followup Questionnaire Group Ranch Hand Comparison StratifiEst. Relative Variable cation Statistic Number Percent Number Percent Risk (95% C.I.) p-Value n Skin Patch, Abnormal Bruise, or Normal Sensitivity Reported 264 12 252 4.5 95 .5 260 12 248 4.6 95 .4 0.98 (0.43,2.23) 0.999 742 42 700 5.7 94.3 1,025 66 959 6.4 93.6 0.89 (0.62,1.28) 0.547 Uroporphyrin n Skin Patch, Abnormal Bruise, or Normal Sensitivity Not Reported only one differential diagnosis at the examination entertained the diagnosis of PCT. Based on all of these observations, PCT was a rare, if not nonexistent, condition in the Ranch Hands. EXPOSURE INDEX ANALYSES Both unadjusted and adjusted exposure index analyses were carried out for the nine laboratory tests of hepatic function and the two porphyrin metabolite tests. The porphyrin variables were analyzed only as continuous variables, while the others were analyzed both as continuous variables and discretized variables. Five covariates were included in the adjusted analyses: age, race, current alcohol use, exposure to degreasing chemicals (yes/no), and exposure to industrial chemicals (yes/no). Current alcohol use was treated as a continuous variable for all adjusted analyses, and age was treated as a continuous variable for the continuous adjusted analyses. Age was trichotomized (born in or after 1942, born between 1923 and 1941, and born in or before 1922) for the discrete adjusted analyses. In addition, the covariate BUN was used in the porphyrin analyses. For each variable, exposure level frequencies and percents are presented in Table K-3 of Appendix K along with the results of the unadjusted discrete 13-30 �analyses using Pearson's chi-square test to reflect overall exposure index differences and Fisher's exact test to investigate medium versus low and high versus low exposure level contrasts. Unadjusted means for each exposure level are presented in Table K-4 of Appendix K, along with the results of the unadjusted continuous analyses (using an F-test for an overall group assessment) and t-tests to examine medium versus low and high versus low exposure index contrasts. Results of the adjusted categorical and adjusted continuous analyses are presented in Tables 13-16 and 13-17, respectively. These results are presented in the context of a main effects model containing exposure index and all five covariates. Additional adjusted continuous analyses were conducted to examine pairwise interactions involving the exposure index and the covariates. Unadjusted and adjusted results for each variable are discussed in sequence. SGOT Within each occupation cohort, the low exposure level had the lowest percentage of abnormalities and the lowest mean. A marginally significant overall exposure level relationship (p=0.065) was found in the unadjusted discrete analysis for the enlisted groundcrew. This association was statistically significant in the adjusted analysis (p=0.023), exhibiting a doseresponse effect; medium versus low (Adj. RR: 2.14, 95% C.I.: [0.77,5.99], p=0,147) and high versus low (Adj. RRs 3.64, 95% C.I.: [1.36,9.72], p=0,010). A nonsignificant dose-response relationship was observed in the corresponding unadjusted and adjusted continuous analyses (p=0.418 and p=0.409, respectively), with unadjusted means of 32.9 U/L, 33.2 U/L, and 34.4 U/L for the low, medium, and high exposure levels, respectively. No significant results were found for enlisted flyers and officers. SGPT Within the enlisted groundcrew and enlisted flyer cohorts the low exposure level had the lowest percentage of abnormalities and the lowest mean value. This situation was reversed for the officers who exhibited the highest percentage of abnormal measurements and highest mean value in the low exposure categories. A significant overall result was found for enlisted flyers in the adjusted discrete analysis (p=0.036; medium versus low, Adj. RR: 6.55, 95% C.I.s [1.25,34.43], p=0.026); high versus low, Adj. RR: 4.29, 95% C.I.: [0.75,24.35], p=0.101). In the corresponding adjusted continuous analyses, a marginally significant dose-response relationship was observed (p=0.058) with adjusted means 18.1 U/L, 21.4 U/L, and 21.8 U/L for the low, medium, and high exposure levels, respectively. No significant results were found for officers or enlisted groundcrew. GGTP No significant or marginally significant results were found. A nonsignificant dose-response relationship was seen for enlisted flyers and officers in the continuous analyses but, conversely, a nonsignificant decreasing dose-response relationship was seen in the enlisted groundcrew. 13-31 �TABLE 13-16. Adjusted Categorical Exposure Index Analyses (Main Effects Model) Results for Hepatic Function Variables by Occupation Exposure Index Occupation Low Total Medium Total High Total Officer Variable 125 129 120 to 55 152 Officer i u> Enlisted Flyer Enlisted Groundcrew SCOT 125 Enlisted SGPT 65 160 129 55 57 140 120 65 Flyer Enlisted Groundcrew 152 160 140 Contrast Adj. Relative Risk (95% C.I.) p-Value Overall M vs. L H vs. L 1.60 (0.64,3.98) 1.02 (0.38,2.77) 0.508 0.312 0.963 Overall M vs. L H vs. L 7.79 (0.77,79.20) 5.38 (0.49,59.50) 0.108 0.083 0.170 Overall M vs. L H vs. L 2.14 (0.77,5.99) 3.64 (1.36,9.72) 0.023 0.147 0.010 Overall M vs. L H vs. L 0.97 (0.48,1.97) 0.77 (0.37,1.64) 0.768 0.933 0.504 57 M vs. L H vs. L Overall 6.55 (1.25,34.43) 4.29 (0.75,24.35) 0.026 0.101 Overall M vs. L H vs. L 1.53 (0.77,3.01) 1.18 (0.57,2.48) 0.457 0.223 0.655 0.036 �TABLE 13-16. (continued) Adjusted Categorical Exposure Index Analyses (Main Effects Model) Results for Hepatic Function Variables by Occupation Exposure Index Occupation Low Total Medium Total High Total Officer Variable 125 129 120 Enlisted Flyer 55 Enlisted Groundcrew 152 Officer GGTP 126 65 57 Contrast Adj. Relative Risk ( 5 C.I.) 9% p-Value Overall M vs. L H vs. L 1.02 ( . 8 2 7 ) 03,.2. 0.94 (0.35,2.54) 0.987 0.968 0.906 Overall M vs. L H vs. L 1.51 (0.41,5.65) 1.46 (0.37,5.78) 0.798 0.536 0.586 Overall M vs. L H vs. L 0.74 (0.34,1.64) 0.89 (0.40,1.97) 0.760 0.462 0.776 Overall M vs. L H vs. L 2.44 (0.65,9.05) 0.91 (0.19,4.36) 0.272 0.184 0.926 Overall M vs. L H vs. L 4.84 (0.52,44.80) 5.34 (0.58,49.06) 0.191 0.165 0.139 Overall M vs. L H vs. L 1.35 (0.50,3.59) 1.82 (0.72,4.59) 0.431 0.552 0.202 CO u> w Alkaline Phosphatase Enlisted Flyer 54 Enlisted Groundcrew 153 160 129 64 160 140 120 56 141 �TABLE 13-16. (continued) Adjusted Categorical Exposure Index Analyses (Main Effects Model) Results for Hepatic Function Variables by Occupation Exposure Index Total Bilirubin Occupation Low Total Medium Total High Total Officer Variable 125 129 120 65 57 152 160 140 Officer •Direct Bilirubin 54 Enlisted Groundcrew u> i Enlisted Flyer3 125 Enlisted Flyer 55 Enlisted Groundcrew 152 129 65 160 120 57 140 Contrast Adj . Relative Risk ( 5 C.I.) 9% p-Value Overall M vs. L H vs. L 0.67 (0.10,4.51) 1.10 (0.18,6.61) 0.851 0.677 0.915 — — — Overall M vs. L H vs. L 0.41 (0.10,1.65) 1.02 (0.32,3.23) 0.332 0.208 0.971 Overall M vs. L H vs. L 2.69 (0.46,15.82) 3.10 (0.56,17.25) 0.354 0.274 0.196 Overall M vs. L H vs. L 2.97 (0.48,18.38) 1.79 (0.24,13.43) 0.466 0.241 0.571 Overall M vs. L H vs. L 1.61 (0.43,6.06) 1.40 (0.36,5.51) 0.767 0.480 0.628 �TABLE 13-16. (continued) Adjusted Categorical Exposure Index Analyses (Main Effects Model) Results for Hepatic Function Variables by Occupation Exposure Index Occupation Low Total Medium Total High Total Officer Variable 125 129 120 55 152 Officer en Enlisted Flyer Enlisted Groundcrew Cholesterol 125 Enlisted Flyer 55 Enlisted Groundcrew Triglycerides 152 65 160 129 65 160 57 140 120 57 140 Contrast Adj. Relative Risk (95% C.I.) p-Value Overall M vs. L H vs. L 0.54 (0.27,1-09) 0.50 (0.25,1.03) 0.107 0.085 0.060 Overall M vs. L H vs. L 1.02 (0.38,2.73) 1.12 ( . 2 3 0 ) 04,.2 0.972 0.962 0.822 Overall M vs. L H vs. L 1.20 (0.57,2.55) 1.61 (0.78,3.30) 0.417 0.630 0.194 Overall M vs. L . H vs. L 0.97 ( . 8 2 4 ) 03,.5 1.35 (0.55,3.32) 0.721 0.946 0.514 Overall M vs. L H vs. L 2.66 (0.62,11.39) 2.06 ( . 4 9 6 ) 04,.0 0.379 0.189 0.358 Overall M vs. L H vs. L 0.44 (0.14,1.42) 0.60 (0.19,1.86) 0.363 0.173 0.375 *No analysis done since there were only three abnormal (one medium, two high)% �TABLE 13-17. Adjusted Continuous Exposure Index Analyses (Main Effects Model) for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index SGPT Low Medium High Contrast p-Value n Adj . Mean 125 33.6 129 34.7 120 33.8 Overall M vs. L H vs. L 0.718 0.450 0.904 Enlisted Flyer n Adj . Mean 55 30.3 65 32.8 57 32.7 Overall M vs. L H vs. L 0.276 0.144 0.184 n Adj . Mean 152 33.5 160 34.1 140 35.0 Overall M vs. L H vs. L 0.409 0.595 0.183 Officer ^ 0 0 i OO Statistic Enlisted Groundcrew SCOT Occupation Officer Variable n Adj . Mean 125 20.1 129 20.0 120 19.1 Overall M vs. L H vs. L 0.695 0.969 0.451 Enlisted Flyer n Adj, Mean 55 18.1 65 21.4 57 21.8 Overall M vs. L H vs. L 0.058 0.047 0.030 Enlisted Groundcrew n Adj . Mean 152 20.2 160 21.4 140 21.0 Overall M vs. L H vs. L 0.581 0.309 0.492 �TABLE 13-17. (continued) Adjusted Continuous Exposure Index Analyses (Main Effects Model) for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index Low Medium High Contrast p-Value n Ad j . Mean 125 30.9 129 32.2 120 32.4 Overall M vs. L H vs. L 0.828 0.611 0.580 Enlisted Flyer n Adj . Mean 55 36.6 65 42.6 57 44.6 Overall M vs. L H vs. L 0.286 0.230 0.132 n Adj . Mean 152 36.9 160 36.6 140 33.1 Overall M vs. L H vs. L 0.299 0.914 0.159 Officer ^ Statistic Enlisted Groundcrev GGTP Occupation Officer Variable n Adj . Mean 126 82.3 129 82.9 120 83.8 Overall M vs. L H vs. L 0.843 0.808 0.561 Enlisted Flyer n Adj . Mean 54 90.7 64 99.3 56 97.7 Overall M vs. L H vs. L 0.127 0.053 0.122 Enlisted Groundcrew n Adj . Mean 153 91.5 160 94.0 141 93.5 Overall M vs. L H vs. L 0.576 0.318 0.444 u> i u> Alkaline Phosphatase �TABLE 13-17. (continued) Adjusted Continuous Exposure Index Analyses (Main Effects Model) for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index Statistic Low Medium High Contrast p-Value n Adj . Mean 125 0.77 129 0.75 120 0.79 Overall M vs. L H vs. L 0.439 0.504 0.545 Enlisted Flyer n Adj . Mean 55 0.69 65 0.76 57 0.79 Overall M vs. L H vs. L 0.070 0.128 0.023 Enlisted Groundcrew n Adj . Mean 152 0.73 160 0.74 140 0.78 Overall M vs. L H vs. L 0.240 0.838 0.117 Officer. Total Bilirubin Occupation Officer Variable n Adj . Mean 125 0.20 129 0.19 120 0.21 Overall M vs. L H vs. L 0.567 0.517 0.689 Enlisted Flyer n Adj . Mean 55 0.18 65 0.19 57 0.19 Overall M vs. L H vs. L 0.724 0.471 0.498 Enlisted Groundcrew n Adj . Mean 152 0.17 160 0.19 140 0.18 Overall M vs. L H vs. L 0.550 0.277 0.670 u> Ui 00 Direct Bilirubin �TABLE 13-17. (continued) Adjusted Continuous Exposure Index Analyses (Main Effects Model) for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index Statistic Low Medium High Contrast p-Value n Ad j . Mean 125 134.0 129 131.3 120 128.9 Overall M vs. L H vs. L 0.232 0.373 008 .8 Enlisted Flyer n Adj . Mean 55 114.4 65 112.4 57 120.9 Overall M vs. L H vs. L 0.101 0.619 0.129 Enlisted Groundcrew n Adj . Mean 152 125.3 160 125.3 140 129.7 •Overall M vs. L H vs. L 0.092 0.997 0.055 Officer LDH Occupation Officer Variable n Adj . Mean 125 236.7 129 225.0 120 224.2 Overall M vs. L H vs. L 0.049 0.039 0.029 Enlisted Flyer n Adj . Mean 55 213.2 65 208.1 57 220.6 Overall M vs. L H vs. L 0.214 0.492 0.343 Enlisted Groundcrew n Adj . Mean 152 209.6 160 211.1 140 210.1 Overall M vs. L H vs. L 0.945 0.742 0.927 w u> Cholesterol �TABLE 13-17. (continued) Adjusted Continuous Exposure Index Analyses (Main Effects Model) for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index Statistic Low Medium High Contrast p-Value n Ad j . Mean 125 110.0 129 108.5 120 116.3 Overall M vs. L H vs. L 0.739 0.886 0.558 Enlisted Flyer n Ad j . Mean 55 112.5 65 111.2 57 113.7 Overall M vs. L H vs. L 0.981 0.919 0.927 Enlisted Groundcrew n Ad j . Mean 152 110.9 160 109.9 140 107.9 Overall M vs. L H vs. L 0.922 0.890 0.690 Officer Triglycerides Occupation Officer Variable n Ad j . Mean 125 17.49 129 16.69 120 17.45 Overall M vs. L H vs. L 0.856 0.621 0.977 Enlisted Flyer n Ad j . Mean 54 18.58 65 16.96 57 18.27 Overall M vs. L H vs. L 0.703 0.438 0.890 Enlisted Groundcrew n Adj . Mean 151 16.39 160 16.54 139 15.45 Overall M vs. L H vs. L 0.644 0.903 0.451 w i .• £» --- Uroporphyrin �TABLE 13-17. (continued) Adjusted Continuous Exposure Index Analyses (Main Effects Model) for Hepatic Function Variables and Two Porphyrin Determinations by Occupation Exposure Index Coproporphyrin Occupation Statistic Low Medium High Contrast p-Value Officer Variable n Ad j . Mean 125 127.65 129 128.84 120 130.26 Overall M vs. L H vs. L 0.901 0.833 0.649 Enlisted Flyer n Ad j . Mean 55 108.67 65 115.31 57 109.81 Overall M vs. L H vs. L 0.669 0.408 0.890 Enlisted Groundcrew n Ad j . Mean 151 115.28 160 115.71 140 122.88 Overall M vs. L H vs. L 0.325 0.935 0.177 �Alkaline Phosphatase For the enlisted groundcrew and the enlisted flyers, the lowest abnormal prevalence rate and lowest mean value were found in the low exposure category. A nonsignificant increasing dose-response relationship was seen within these occupations for the discrete analyses. In both unadjusted and adjusted continuous analyses, a marginally significant medium versus low contrast was found for enlisted flyers (p=0.086 and p=0.053, respectively), with unadjusted means of 88.9 U/L, 96.3 U/L, and 95.2 U/L for the low, medium, and high exposure levels, respectively. Total Bilirubin Discrete analyses revealed no significant findings; adjusted discrete analyses for enlisted flyers were not done due to sparse data. Continuous analyses revealed a significant overall effect (p=0.045, unadjusted) for enlisted flyers, which was marginally significant after adjustment (p=0.070). In both unadjusted and adjusted analyses, the high versus low mean contrast was significant (p=0.014 and p=0.023, respectively), with unadjusted means of 0.66 mg/dl, 0.73 mg/dl, and 0.76 mg/dl for the low, medium, and high exposure levels, respectively. Direct Bilirubin There were no significant exposure findings in either the continuous or discrete analyses, although within each occupational cohort, the lowest abnormal prevalence rate was found in the low exposure group. LDH The unadjusted discrete analyses revealed no significant or marginally significant results. No adjusted discrete analyses were done due to sparse data. The unadjusted continuous analyses for the enlisted groundcrew showed a significant overall relationship with the exposure index (p*0.031), with mean values of 123.1 U/L, 122.3 U/L, 127.9 U/L for the low, medium, and high exposure levels; the high versus low contrast was significant (p=0.037). After adjustment, the continuous analyses for enlisted groundcrew revealed marginally significant results (p=0.092, overall; p=0.055, high versus low). No significant or marginally significant results were seen for enlisted flyers or officers. Enlisted flyers and enlisted groundcrew had the largest mean values for their highest exposure category, which is reversed in the officers, who exhibited a nonsignificant decreasing dose-response relationship with exposure level. Cholesterol Significant or marginally significant results were found for officers in the direction of a decreasing dose-response relationship in both the adjusted continuous (overall p=0.049, medium versus low p=0.039, high versus low p=0.029) and adjusted discrete (medium versus low p=0.085, high versus low p=0.060) analyses. Neither of the enlisted cohorts demonstrated a similar decreasing response. 13-42 �Triglycerides No significant or marginally significant results were found. Uroporphyrins and Coproporphyrins No significant or marginally significant results were found. EXPOSURE INDEX ANALYSES Additional continuous analyses were done to examine pairwise interactions involving exposure level and the covariates. Ten exposure group-bycovariate interactions were found at p<0.05. All interactions were found in the enlisted flyer and enlisted groundcrew occupations. Eight of the interactions involved current alcohol consumption, one involved age, and one involved race. The interactions are summarized in Tables K-5 and K-6 of Appendix K. In Table K-5 of Appendix K, the slope of the continuous covariate with respect to the dependent variable is provided for each of the three exposure levels. Table K-6 of Appendix K presents the mean level of direct bilirubin for each of the three exposure levels by race. The interactions involving current alcohol consumption are mainly due to a nonsignificant dependent variable response to increasing alcohol consumption in the low exposure group in contrast to a significant positive response for the medium and high groups. The SCOT, SGPT, and GGTP interaction results for the enlisted groundcrew provide support for an interpretation of herbicide effect. In summary, the nine hepatic function variables and two porphyrin metabolite variables showed no conclusive evidence of a dose-response relationship at the followup examination. Five overall exposure group differences were found. Only two of these (SCOT for enlisted groundcrew, and total bilirubin for enlisted flyers) supported a dose-response relationship. LONGITUDINAL ANALYSES Three hepatic enzyme variables, SCOT, SGPT, and GGTP, were chosen for longitudinal analysis, spanning the spectrum of intermediate to acute effects. These test variables were chosen because both the Baseline and the followup assays were performed by the high-precision ACA 500® DuPont technology. The data from these three hepatic variables are arrayed in Table 13-18. The SGOT and SGPT data showed slight but uniform increases from the Baseline examination. These increases were proportionately the same for both the Ranch Hand and Comparison groups. These changes may reflect an aging effect or are due to laboratory variation. As indicated by the equality-ofdifference p-values, none of the three hepatic variables showed a statistically significant difference in the changes from Baseline to followup between groups. 13-43 �TABLE 13-18. Longitudinal Analyses for SCOT, SGPT, and GGTP: A Contrast of Baseline and First Follovup Examination Test Means Variable Means 1982 1985 Baseline Followup p-Value* (Equality of Difference) Group Total SCOT Ranch Hand Comparison 971 1,139 32.91 32.97 33.73 33.73 0.61 SGPT Ranch Hand Comparison 971 1,139 20.08 20.51 21.82 22.44 0.72 GGTP Ranch Hand Comparison 971 1,139 39.26 38.64 33.16 32.35 0.63 *Analyzed in log units. SUMMARY AND CONCLUSIONS The interval questionnaire revealed sparse reporting of liver disorders from 1982 to 1985 that was not significantly different between groups. Historical liver disease was verified by medical records, and these data were added to the verified Baseline history to assess possible lifetime differences. No significant differences were found. The medical record verification process showed that the historical data were generally correctly reported and classified between groups, except for the category of enlarged liver which showed a higher verification rate in the Comparison group. Digestive system mortality showed an overall nonsignificant excess in the Ranch Hands, but a relative nonsignificant excess of malignant neoplasms in the Comparisons. No differences were found for past or current peptic ulcer disease for the Ranch Hand and Comparison groups, adjusted for standard covariates as well as blood type. The physical examination disclosed a borderline significant increase of hepatomegaly in the Ranch Hand group. Emphasis was placed on nine laboratory test variables measuring liver function, i.e., serum glutamic-oxaloacetic transaminase (SCOT), serum glutamic-pyruvic transaminase (SGPT), gammaglutamyl transpeptidase (GGTP), alkaline phosphatase, total and direct bilirubin, lactic dehydrogenase (LDH), cholesterol, and triglycerides. In addition, uroporphyrin and coproporphyrin measurements were obtained to assess liver function and the likelihood of porphyria cutanea tarda (PCT). The nine hepatic variables were subjected to continuous and categorical statistical tests, and were adjusted for the covariates age, race, occupation, current alcohol consumption, and unprotected exposure to both industrial chemicals and degreasing chemicals. Final statistical models used only 13-44 �the significant covariates and two-way interactions for adjustment. The two porphyrin measurements were analyzed only in the continuous form. The overall summary results of the analyses of these 11 variables are given in Table 13-19. The results showed a significantly lower mean SGPT level, a greater mean alkaline phosphatase level, a lower mean uroporphyrin level for Ranch Hands as contrasted with Comparisons, and a marginally significant greater mean coproporphyrin level. Only in the instance of alkaline phosphatase did the discrete analysis approach statistical significance. No group differences were noted for SCOT, GGTP, total and direct bilirubin, LDH, cholesterol, or triglycerides. However, an analysis using only the Original Comparisons revealed a significantly greater mean cholesterol level in the Comparison group. A review of the covariate effects in the adjusted statistical models revealed that all covariates behaved as expected with the exception of alcohol consumption for the alkaline phosphatase analysis, which showed an inverse relationship with wine consumption. Exploration of group-by-group covariate interactions for alkaline phosphatase, direct bilirubin, triglycerides, SCOT, and uroporphyrins revealed significant group differences within specific covariate strata. In particular, Ranch Hands exposed to industrial chemicals had a significantly higher adjusted mean level of alkaline phosphatase and a significantly higher abnormal prevalence rate of direct bilirubin than similarly exposed Comparisons. For triglycerides, Ranch Hands born in or before 1922 had a significantly higher adjusted mean level than similar aged Comparisons, while Ranch Hand officers exhibited a significantly higher abnormal prevalence rate than Comparison officers. For SCOT, Ranch Hand moderate current drinkers (more than one to four drinks per day) had a significantly higher mean level than corresponding Comparisons. In the opposite direction, Comparisons with a mean BUN level less than or equal to 14 (median for all participants) were found to have a significantly higher adjusted mean uroporphyrin level than similar Ranch Hands. These results did not disclose any common pattern detrimental to the Ranch Hand group. These findings were generally consistent with the 1982 Baseline data, which disclosed a significantly increased mean cholesterol level in the Comparisons and nonsignificant Ranch Hand mean elevations for GGTP and LDH. Slight differences in analytic results are probably due to the use of more fully adjusted models used for the followup examination data. Overall, the followup examination laboratory data showed no adverse clinical or exposure patterns in either group. Further, the detection of significant mean shifts (still within normal range) by the continuous statistical tests, not mirrored by the categorical tests, suggests a circumstance of statistical power rather than findings of biological relevance. Of the five significant or marginally significant results that were found in the adjusted exposure index analyses, four exhibited a trend suggestive of an increasing dose-response relationship. In the enlisted flyer cohort, the percentages of SGPT abormalities were significantly different and increased from the low to the high exposure categories. The corresponding mean values were marginally significantly different among exposure levels. Also,'the mean levels of total bilirubin were marginally significantly different among exposure levels, increasing with exposure level. For enlisted groundcrew, the percentage of SCOT abnormalities significantly differed among 13-45 �TABLE 13-19. Overall Summary Results of Unadjusted and Adjusted Analyses of Nine Hepatic Function Variables and Two Porphyrin Metabolite Tests Unadjusted Mean Adjusted* Categorical Variable Mean CC Questionnaire Liver Disease (Lifetime History) Hepatitis Jaundice Cirrhosis Enlarged Liver Miscellaneous Liver Disorders Peptic Ulcer Disease NS Physical Examination Hepatomegaly Categorical CD DD Direction of Results** NS* NS NS NS NS NS NSa RH>C Laboratory Testing .SCOT SGPT GGTP Alkaline Phosphate Total Bilirubin Direct Bilirubin LDH Cholesterol Triglycerides Uroporphyrin Coproporphyrin NS NS* NS 0.009 NS NS NS NS NS 0.048 NS* NS NS NS NS NS NS NS NS NS NS 0.048 NS 0.008 NS NS **** NS **** **** NS* **** 0.029 NS **** NS NS NS NS NS NS NS NS NS* NS **** NS NS **** Questionnaire-Laboratory Correlation Skin Bruises, Patches,— 0.001 and Sensitivity *C: Continuous D: Discrete **RH>Cj more abnormalities, or higher mean value, in Ranch Hands, ORH: more abnormalities, or.higher mean value, in Comparisons. a Adjusted for blood type. NS: Not significant (p>0.10). NS*: Borderline significant (0.05<p<0.10). —Analysis not performed. ****Group-by-covariate interaction. 13-46 ORH RH>C ORH RH>C RH>C �exposure levels. Within the enlisted flyer cohort, all nine laboratory tests of hepatic function had the lowest percentage of abnormalities in the low exposure category; correspondingly, six of the nine mean levels were lowest for the low exposure category. Of the ten group-by-covariate interactions that were found, three (SCOT, SGPT, and GGTP) supported a dose-response relationship in the enlisted groundcrew cohort. Exploration of these interactions revealed a trend that showed an increasing association between current alcohol consumption and the dependent variables for increasing exposure levels. Longitudinal analyses for SCOT, SGPT, and GGTP disclosed no statistically significant group differences in the mean shifts from the Baseline to the followup examination. Interval reporting of PCT-like symptoms of skin patches, bruises, and sensitivity was significantly increased in the Ranch Hands (p=0.001). However, when these historic data were contrasted to both uroporphyrin and coproporphyrin abnormalities, no correlation was apparent, nor were there any significant group differences. Since an elevation in the uroporphyrin level is required for a diagnosis of PCT, the historic data were retabulated with only uroporphyrin abnormalities; again, no group differences were apparent, .and, in fact, uroporphyrin abnormalities in both groups were higher in those participants without a history of skin disorders than in those participants with such a history. The likelihood of bona fide PCT among study participants, and particularly among the Ranch Hands, appears to be remote. In conclusion, the followup examination disclosed more statistically significant findings for tests of liver function than the Baseline examination, but they were equally divided between the two groups and did not demonstrate clinical, statistical, or exposure patterns consistent with an herbicide-related effect on health. No evidence was found to suggest an increased likelihood of PCT among the Ranch Hand group. 13-47 �CHAPTER 13 REFERENCES 1. Kimbrough, R.D., C.D. Carter, J.A. Liddle, R.E. Cline, and P.E. Phillips. 1977. Epidemiology and pathology of a tetrachlorodibenzodioxin poisoning episode. Arch. Environ. Health 32(2):7-86. 2. McNulty, W.P. 1977. Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin for Rhesus monkeys: Brief report. Bull. Environ. Contam. Toxicol. 18(1):108-109. 3. Olson, J.R., M.A. Holscher, and R.A. Neal. 1980. Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the golden Syrian hamster. Toxicol. Appl. Pharmacol. 55:67-78. 4. Palmer, J.S., and R.D. Radeleff. 1964. The toxicologic effects of ' certain fungicides and herbicides on sheep and cattle. Ann. N.Y. Acad. Sci. 11:729-736. 5. Goldstein, J.A., P. Hickman, H. Bergman, and J.G. Vos. 1973. Hepatic porphyria induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin in the mouse. Res. Commun. Chem. Pathoi. Pharmacol. 6:919. 6. Madhukar, B.V., and F. Matsumura. 1981, Difference in the nature of induction of mixed-function oxidase systems of the rat liver among phenobarbital, DDT, 3-methylcholanthrene, and TCDD. Toxicol. Appl. Pharmacol. 61:110-118. 7. Kohli, K.K., and J.A. Goldstein. 1981. Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on hepatic and renal prostaglandin synthetase. Life Sci. 19:299-305. 8. Thunberg, T., and H. Hakansson. 1983. Vitamin A (retinol) status in the Gunn rat: The effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Arch. Toxicol. 53:225-234. 9. Goldstein, J.A., P. Hickman, and D.L. Jue. 1974. Experimental hepatic porphyria induced by polychlorinated biphenyls. Toxicol. Appl. Pharmacol. 27:437. 10. Sassa, S., H. De Verneuil, and A. Kappas. 1984. Inhibition of uroporphyrinogen decarboxylase activity in polyhalogenated aromatic hydrocarbon poisoning. In Banbury report 18: Biological mechanisms of dioxin action, ed. A.. Poland and R.D. Kimbrough, pp. 215-222. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory. 13-48 �11. Sweeney, G., D. Basford, B. Rowley, and G. Goddard. 1984. Mechanisms underlying the hepatotoxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin. In Banbury report 18: Biological mechanisms of dioxin action, ed. A. Poland and R.D. Kimbrough, pp. 255-239. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory. 12. Greig, J. 1984. Differences between skin and liver toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in mice. In Banbury report 18: Biological mechanisms of dioxin action, ed. A. Poland and R.D. Kimbrough, pp. 391-397. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory. 13. Goldmann, P.J. 1973. Schweist akute Chlorakne, eine Massenintoxikation durch 2,3,7,8-Tetrachlorodibenzodioxin (Severe, acute chloracne, a mass intoxication due to 2,3,7,8-tetrachlorodibenzo-dioxin). Per Hautarzt. 24(4):149-152. 14. Oliver, R.M. 1975. Toxic effects of 2,3,7,8-tetrachlorodibenzo 1,4-dioxin in laboratory workers. Br. J. Ind. Med. 32:49-53. 15. Reggiani, G. 1980. Acute human exposure to TCDD in Seveso, Italy. J^ Toxicol. Environ. Health 6:27-43. 16. Reggiani, G. 1979. Estimation of the TCDD toxic potential in the light of the Seveso accident. Arch. Toxicol. 2:291-302. 17. Suskind, R.R. 1978. Chloracne and associated health problems in the manufacture of 2,4,5-T. Report to the Joint Conference, National Institute of Environmental Health Sciences and International Agency for Research on Cancer, World Health Organization, Lyon, France, January 11, 1978. 7 pp. 18. May, G. 1982. Tetrachlorodibenzodioxin: A survey of subjects ten years after exposure. Br. J. Ind. Med. 39:128-135. 19. Ideo, G., G. Bellati, A. Bellobuono, A. Mocarelli, P. Marocchi, A. and P. Brambilla. 1982. Increased urinary d-glucaric acid excretion by children living in an area polluted with tetrachlorodibenzodioxin (TCDD). Clin. Chem. Acta. 120:273-283. 20. May, G. 1973. Chlorance from the accidental production of tetrachlorodibenzodioxin. Br. J. Ind. Med. 30:276-283. 21. Moses, M., R. Lilis, K.D. Crow, J. Thornton, A. Fischbein, H.A. Anderson, and I.J. Selikoff. 1984. Health status of workers with past exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in the manufacture of 2,4,5trichloro-phenoxyacetic acid: Comparison of findings with and without chloracne. Am. J. Ind. Med. 5:161-182. 22. Suskind, R.R., and V.S. Hertzberg. 1984. Human health effects of 2,4,5-T and its toxic contaminants. JAMA 251:2372-2380. 13-49 �23. Pazderova-Vejlupkova, J., M. Nemcova, J. Pickova, L. Jirasek, and E. Lukas. 1981. The development and prognosis of chronic intoxication by tetrachlorodibenzo-p-dioxin in men. Arch. Environ. Health 36:5-11. 24. Martin, J.V. 1984. Lipid abnormalities in workers exposed to dioxin. Br. J. Ind. Med. 41:254-256. 25. Hoffman, R.E., P.A. Stehr-Green, K.B. Webb, G. Evans, A.P. Knutsen, tf.F. Schramm, J.L. Staake, B.B. Gibson, and K.K. Steinberg. 1986. Health effects of long-term exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. JAMA 255:2031-2038. 26. Oliver, R.M. 1975. Toxic effects of 2,3,7,8-tetrachrloro-dibenzo-l, 4-dioxin in laboratory workers. Br. J. Ind. Med. 32:46-53. 27. Bleiberg, J., M. Wallen, R. Brodkin, and I.L. Applebaum. 1964. Industrially acquired porphyria. Arch. Dermatol. 89:793-797. 28. Jirasek, L., J. Kalensky, K. Kubec, et al. 1974. Acne chlorina, porphyria cutanea tarda and other manifestations of general intoxication during the manufacture of herbicides, part 2. Czech Dermatol. 49(3):145-157. 29. Poland, A.P., D. Smith, G. Metter, and P. Fossick. 1971. A health survey of workers in a 2,4-D and 2,4,5-T plant, with special attention to chloracne, porphyria cutanea tarda, and psychologic parameters. Arch. Environ. Health 22(3):316-327. 30. Peters, H.A., A. Gocmen, D.J. Cripps, G.T. Bryan, and I. Dogramaci. 1982. Epidemiology of hexachlorobenzene-induced porphyria in Turkey. Arch. Neurol. 39:744-749. 31. Rubenstein, E., and D.D. Federman, eds. 1986. Metabolism: The porphyrias. Chap. 9 in Scientific American Medicine. New York: Scientific American, Inc. 13-50 �CHAPTER 14 DERMATOLOGICAL EVALUATION INTRODUCTION The skin is a major target organ following heavy exposure to chlorophenols and dioxin and, therefore, is a primary focus of the AFHS clinical examination. Since the association between chlorinated chemicals and chloracne was first noted in 1957,1'2 a variety of animal experiments have shown the dermal sensitivity of rabbits, monkeys, and hairless mice to TCDD, 2,4,5-T (contaminated with TCDD), and other chlorinated dibenzo compounds, furans, or their brominated analogs.1"7 Chloracne is not associated with exposure to 2,4-D. Accidental exposure to waste oils containing TCDD has caused significant dermal symptoms, including loss of hair, ulceratiye dermatitis, and inflamed mucous membranes in horses, dogs, cats, and mice. ' Studies have suggested that the chloracnegens induce a series of pathological skin changes in target cells of the epithelial lining of sebaceous glands via the Ah receptor. Hyperkeratinization of these cells eventually leads to the formation of the comedone characteristic of aerie. In humans, development of the hallmark rash, chloracne, is generally acknowledged to represent substantial topical or systemic exposure to one or more chloracnegens. ' ' ' Acute fulminant chloracne is characterized by a maculopapular rash of active comedones, conforming to an eyeglass or facial butterfly(1distribution, often accompanied by chest, back, or eyelid lesions. The severity of the chloracne appears to be generally dose related, but may also depend on the route of administration, age, genetic predisposition, and/or the existence of acne vulgaris or other skin disorders. ' ' Occasionally, exposure, via contaminated clothes of an9 industrial worker, has been associated with chloracne in family members. Sequelae from severe chloracne include actinic elastosis, afpe scars, disfigurement, excessive hair growth, and Peyronie's disease. ' Severe chloracne is often accompanied by acute effects in other organ systems. In contrast, low to moderate exposure to chloracnegens generally produces mild chloracne with few, if any, attendant systemic signs and symptoms. The clinical diagnosis of acute chloracne is easier than the diagnosis of subacute and chronic chloracne. In the latter instances, a history of exposure to chloracnegens is essential in the diagnosis, particularly if the individual has experienced adolescent acne. Chronic chloracne has been clinically observed more than 30 years after onset, but a biopsy is often necessary to confirm these cases. Mild or transient cases of chloracne may be confused with persistent adolescent acne or other skin conditions. 14-1 �As noted in the AFHS Baseline Morbidity Report, over one-half of the veteran complaints in the Veterans Administration Herbicide Registry involved dermatological conditions, a fact sometimes alluded to as "evidence" of exposure to Agent Orange. In actuality, skin disease was a major medical problem among American troops serving in Vietnam. Forty-seven percent of the combat-days lost in the 9th Infantry Division from July 1968 to June 1969 were due to dermatological conditions. These diseases were directly related to the tropical climate and terrain. Only in rare cases has the Veterans Administration made a diagnosis of chloracne in a Vietnam combat veteran. The natural history of chloracne suggests that most cases should have been diagnosed while in Vietnam, but a dermatological survey failed to reveal any cases. Most recognized chloracne cases have been diagnosed in chemical plant workers or in victims of industrial accidents. Thousands of cases were recorded in the 1930-1940 era, and earlier descriptions of chloracne-like disease were found in 1897 to 1901. Industrial exposure to chloracnegens has been generally characterized as moderate-prolonged or severe-acute. In the setting of casual-sporadic exposure, as in the typical cases of the contaminated housing areas in Times Beach, Missouri, and the Quail Run Trailer Park, chloracne is virtually unknown. A number of dioxin morbidity studies have shown a clustering of abnormal laboratory tests in individuals with chloracne. ' ,24-27 Thig hag some investigators to believe that long-term sequelae to dioxin exposure will be found only in people with chloracne. Other investigators feel that this belief is not consistent with normal spectrum-of-illness concepts and that effects may occur in the absence of chloracne. Baseline Summary Results The 1982 Baseline clinical examination revealed an unexpected significant excess (p=0.03) of basal cell carcinoma in the Ranch Hand group. Risk factor data (e.g., sun exposure, host factors of tannability, complexion) were not collected in 1982. The 1982 examination focused on the diagnosis of chloracne both in historical terms by a detailed questionnaire and in contemporary terms via a comprehensive clinical assessment. The questionnaire data did not demonstrate anatomic, incidence, or onset-time patterns of acne in the Ranch Hand group that might support an inference of past chloracne, nor did the physical examination detect a single case. Fourteen biopsies from 11 participants also failed to document a chloracne diagnosis. A dermatology index (the number of clinically detected skin abnormalities per individual) was virtually identical between the Ranch Hand and Comparison groups, and was associated with the history of past acne in both groups. No exposure level associations were noted in any occupational category of the Ranch Hand group. The comprehensive dermatological assessment did not reveal evidence of past or current chloracne in the Ranch Hand group. Parameters of the 1985 Dermatological Evaluation Questionnaire data recaptured many of the acne parameters of the 1982 questionnaire, and the physical examination parameters were similar to the 14-2 �1982 Baseline examination. Particular emphasis was given to the diagnosis of basal cell carcinoma and to the collection of risk factor data, e.g., skin color, reaction to sun, ethnicity (see Chapter 10, Malignancy). Thus, the dependent variables and covariates of the analyses below closely approximated those previously conducted on the Baseline examination and questionnaire data. The adjusted statistical analyses were based on logistic regression (BMDP®-LR) and log-linear models (BMDP®-4F), and the unadjusted analyses primarily use Pearson's chi-square test and Fisher's exact test. In addition, an empiric Venn diagram was used to explore the potential of historic chloracne. Parallel analyses using only Original Comparisons are presented in Tables L-3 through L-ll of Appendix L. RESULTS AND DISCUSSION General Detailed dermatological data were obtained by standard physical examination techniques. Numeric differences in summary tables reflect missing dependent variable and undeterminable covariate information. One participant refused the dermatology examination; consequently, all skin disorder analyses were based on 2,308 participants. Data were collected on 22 skin disorders, which were in turn reduced to eight variables for analysis: comedones, acneiform lesions, acneiform scars, depigmentation, inclusion cysts, hyperpigmentation, other abnormalities, and the dermatology index. Descriptions of skin biopsies, which were also conducted at the physical examination, are given in this chapter. Followup questionnaire information regarding the presence, time, and location of acne was also analyzed. The analyses in this chapter first investigate questionnaire information on acne, and subsequent analyses center upon the eight skin disorder variables and the skin biopsies. Four covariates were included in this analysis: age, race, occupation, and presence of acne before duty in Southeast Asia. Age is used in its continuous form for all adjusted logistic regression analyses, but age is trichotomized (born in 1942 or after, born between 1923 and 1941, and born in 1922 or before) for presentation in summary tables and for use in dependent variable and covariate association analyses and log-linear models. Participants were categorized as either Black or nonblack. Occupation was divided into the three classifications of officer, enlisted flyer, and enlisted groundcrew. Sample size differences in subsequent adjusted analyses reflect missing dependent variable data or missing data on the presence of acne before duty in Southeast Asia. Questionnaire Data The acne status of each participant was determined by Baseline and followup questionnaire information. In particular, the occurrence of acne and the dates for acne occurrence have been determined and analyzed. Additionally, the analysis of the location of acne is presented for a subset of the participants who have had acne. Figure 14-1 below is a diagram explaining the occurrence of acne by time determination for the 2,309 participants, along with frequencies and an explanation of terms. 14-3 �Determination Presence of Acne All Acne in 1961 or Before (for Participants with Acne) Acne Reference to Beginning of First SEA Tour of Duty (for Participants with Acne Sometime after 1961) (138) (205) Yes to Acne — Reported acne on both/either Baseline and/or follow/up study. No to Acne — Never had acne. Pre-1961 Acne — Participants with acne who had last occurrence of acne in or before 1961. Post-1961 Acne — Participants with acne who had an occurrence of acne sometime after 1961. Undetermined — Time reference not determinable from date information available. Pre-SEA Acne — Participants with post-1961 acne who had all occurrences of acne before the start of first Southeast Asia (SEA) tour (as determined from military records). Post-SEA Acne — Participants with post-1961 acne who had all occurrences of acne after the start of first SEA tour. Pre- and Post-SEA Acne — Participants with post-1961 acne who had multiple occurrences, both before and after the start of first SEA tour, or a case of acne that began before the start of first SEA tour and-that ended after starting SEA tour. *: Analysis of location of acne performed for these participants. Figure 14-1. Occurrence of Acne by Time for First Followup Participants 14-4 �The distinction was made between pre-1961 and post-1961, since herbicide missions in Vietnam commenced in 1962. Responses of 2,309 participants indicated that 1,415 individuals never had acne, 379 had acne before 1961, 138 had acne after 1961 but before duty in SEA, 205 had acne both before and after duty in SEA, 146 had acne only after SEA duty, and 26 participants could not be specifically classified. Occurrence of Acne The reported occurrence of acne, as determined by Baseline and followup questionnaires, is displayed in Table 14-1. The analysis showed that the Ranch Hand group reported slightly more acne than the Comparison group, although the difference is nonsignificant (p=0.111). Analyses using Original Comparisons only showed a borderline significance (p=0.071) found in Table L-3 of Appendix L. The participants who responded "yes" to acne were categorized according to whether their acne occurred before or after 1961. The distribution of pre-1961 versus post-1961 acne is given in Table 14-2. TABLE 14-1. Unadjusted Analysis for Reported Historical Occurrence of Acne by Group Acne Yes Group Number Ranch Hand Comparison 412 482 Total 894 No Percent Number 40.6 37.3 604 811 1,415 14-5 Percent Total 59.4 62.7 1,016 1,293 2,309 Summary Statistics Est. RR: 1.15 952 C.I.: (0.97,1.36) p-Value: 0.111 �TABLE 14-2. Unadjusted Analysis for Reported Historical Occurrence of Acne Relative to 1961 by Group* Occurrence of Acne Post-1961 Group Number Ranch Hand Comparison 239 271 Total 510 Pre-1961 Percent Number 58.3 56.6 171 208 379 Percent Total 41.7 43.4 410 479 Summary Statistics Est. RR: (for post1961 cases): 1.07 95% C.I,: (0.82, 1.04) p-Value: 0.634 889 *Five participants deleted due to missing data at time of occurrence. As shown, no significant difference in the distribution of post-1961 versus pre-1961 acne existed between Ranch Hands and Comparisons (p=0.634). Cases of post-1961 acne were classified to SEA tour(s) of duty, as determined by military records. The distribution of post-1961 acne cases relative to SEA is shown in Table 14-3. This marginal significance (p=0.058) was due primarily to a larger percentage of Ranch Hands in the post-SEA category, as contrasted with the Comparisons (35.1% versus 25.3%). Duration of Acne The approximate duration of acne was examined among the three SEA categories by group using a two-factor analysis of variance. The calculation of acne duration for participants with multiple occurrences in overlapping time periods counted time periods only once. A square root transformation was used to normalize the duration data. Results from duration of acne analyses are given in Table 14-4. 14-6 �TABLE 14-3. Unadjusted Analysis for Reported Historical Occurrence of Acne Relative to SEA Tour of Duty for Post-1961 Acne by Group* Post-1961 Acne Pre-SEA Group Ranch Hand Comparison Total Post-SEA Pre- and Post-SEA Number Percent Number Percent Number 58 80 138 25.4 30.7 80 66 146 35.1 25.3 Percent Total 90 115 205 39.5 44.1 228 261 489 p-Value 0.058 *Twenty-one post-1961 participants with acne deleted due to missing data on time of occurrence. TABLE 14-4. Adjusted Analysis for Duration of Acne (in Years) for Post-1961 Acne by Group* Group Ranch Hand Comparison Total Total 219 252 471 Adjusted Mean** 95% C.I.** p-Value 8.18 7.49 (7.43,8.96) (6.82,8.19) 0.189 Covariate Remarks Time Reference to SEA (p<0.001) *Eighteen participants deleted due to missing data on time of occurrence. **Converted from square root scale. This adjusted analysis showed no significant effect due to group (p=0.189), but a highly significant effect due to SEA category (p<0.001), with the pre- and post-SEA category having higher mean durations than the pre-SEA or post-SEA categories, which were nearly identical. No interaction was present between group and SEA category (p=0.314). A categorical analysis was performed, in which duration was categorized into 5-year increments (five duration categories, the last being greater than 20 years). There was no significant difference between groups (pre-SEA, p=0.520; post-SEA, p=0.776; pre- and post-SEA', p=0.880). Location of Acne The location of acne for participants classified as post-SEA or pre- and post-SEA (351 participants) was analyzed. Spatial distribution of acne with 14-7 �primary emphasis on acne on the temples, around the eyes, or on the.ears was determined from the questionnaire; these data are presented in Figures 14-2 and 14-3. Figure 14-2 shows the distribution of acne for Ranch Hands and Comparisons, for post-SEA and pre- and post-SEA participants combined, whereas Figure 14-3 represents a similar distribution for only post-SEA participants. If more than one episode of acne occurred, cases involving the temples, eyes, or ears took precedence. Also, multiple-site involvement took precedence over single-site involvement. The Ranch Hand and Comparison Venn diagrams were contrasted by chisquare analysis of a 2x8 table, and no difference in the spatial distribution was noted for the combination of pre- and post-SEA and post-SEA groups (p=0.706), or for the analysis of only the post-SEA group (p=0.699). Sparse data cells were present in the analysis of both figures. Differences in spatial distributions were also not evident when the "other sites" classification was deleted (p=0.770 and p=0.664, respectively). If the intersection of the circles in these figures (i.e., temples, ears, and eyes) is contrasted with the rest of the locations of acne, no significant difference is seen (p=0.189 and p=0.627 for the combination of post-SEA and pre- and post-SEA groups and for only the post-SEA group, respectively). Physical Examination Data Twenty-two skin disorders were assessed at the dermatological examination (page C-9 of Appendix C). These disorders were combined into eight variables for analytic purposes. Comedones, acneiform lesions, acneiform scars, depigmentation, inclusion cysts, and hyperpigmentation were analyzed separately. The remaining 16 conditions were grouped to form a broad variable called "other abnormalities." Analysis of skin cancer is included in the malignancy chapter and will not be discussed here. Additionally, comedones, acneiform lesions, acneiform scars, and inclusion cysts were grouped to construct a dermatology index, which summed the number of abnormalities for these four conditions for each participant. Logistic regression techniques, with the use of BMDP®-LR, were utilized for adjusted analysis of all these variables except the dermatology index, which used BMDP®-4F. The sample sizes were sufficient to detect a 27-percent increase in the prevalence rate for comedones, a 30 percent increase in the prevalence rate for acneiform scars, and a 12 percent increase in the prevalence of at least one abnormality for the dermatology index, using a two-sided a -level of 0.05 with a power of 0.80. No cases of chloracne were chemically diagnosed. Preliminary Dependent Variables and Covariate Relationships The association of the eight skin disorder variables in both groups and the covariates of age (born in or after 1942, born between 1923 and 1941, born in or before 1922), race (Black or nonblack), occupation, and presence of preSEA acne (yes/no) was assessed using Pearson's Chi-square test and Fisher's exact .test. Table 14-5 is a summary of the associations of the dependent variables with these four covariates. Seven additional participants, who were initially classified as "undetermined," were reclassified as having acne before duty in SEA, based on data gathered by telephone. Nineteen participants were omitted from analyses involving presence of pre-SEA acne, because historical information on the date of onset of acne was not available. 14-8 �Ranch Hand 103 Other Sites n=169 Comparison 121 Other Sites n=181 Figure 14-2. Location of Post-SEA and Preand Post-SEA Acne by Group 14-9 �Ranch Hand 53 Other Sites n=80 Comparison 48 Other Sites n=66 Figure 14-3. Location of Post-SEA and Acne by Group 14-10 �TABLE 14-5. Association Between Dermatological Variables and Age, Race, Occupation, and Fre-SEA Acne in the Combined Ranch Hand and Comparison Groups Variable Comedones Acneiform Lesions Acneiform Scars Depigmentation Inclusion Cysts Hyperpigmentation Other Abnormalities Dermatology Index Age <0.001 <0.001 <0.001 NS NS NS <0.001 NS Race <0.001 NS* <0.001 0.009 NS <0.001 <0.001 NS Occupation <0.001 NS* <0.001 NS 0.036 <0.001 <0.001 0.010 Pre-SEA Acne NS <0.001 <0.001 NS NS 0.003 NS* <0.001 NS: Not significant (p>0.10) NS*; Borderline significant (0.05<p<0.10) effect with variable. Age had a significant effect on four of the variables. Prevalence rates for comedones and other abnormalities were highest for older participants (born in or before 1922). On the other hand, the prevalence of acneiform lesions and acneiform scars was higher in the younger participants (born in or after 1942). Nonblacks had a significantly higher prevalence of comedones and other abnormalities and a marginally significant increase (p=0.055) in acneiform lesions. Blacks had a significantly higher prevalence rate for acneiform scars, depigmentation, and hyperpigmentation. Occupation had a significant or marginally significant effect on seven of the eight variables, with either enlisted flyers or enlisted groundcrew generally having a higher percentage of abnormalities. Participants with pre-SEA acne had a significantly higher prevalence rate for acneiform lesions and acneiform scars, and a higher percentage with at least one abnormality in the dermatology index. Participants without acne pre-SEA had a significantly higher prevalence rate for hyperpigmentation, and a marginally significantly higher prevalence rate (p=0.084) for other abnormalities. Analyses of Individual Dependent Variables Comedones As reflected in Table 14-6, there was not a significant difference (p=0.361) between the proportion of participants with comedones in the Ranch Hand and Comparison groups, unadjusted for any covariates. 14-11 �TABLE 14-6. Unadjusted Analysis for Comedones by Group Comedones Absent Present Group Number Ranch Hand Comparison Percent Number 250 340 24.6 26.3 766 952 Percent 75.4 73.7 Total 1,016 1,292 Summary Statistics Est. RR: 0.91 95% C.I.: (0.76,1.10) p-Value: 0.361 Tests of association between the presence of comedones in both groups and the four covariates indicated that there was not a significant effect due to the presence of pre-SEA acne (p=0.355), but that there were significant effects due to occupation (p<0.001), age (p<0.001), and race (p<0.001). The proportion of participants with comedones increased with age (18.9% for participants born in or after 1942, 29.8% for participants born between 1923 and 1941, and 37.9% for participants born in or before 1922). Significantly more nonblacks had comedones than Blacks (26.5% versus 11.9%), and enlisted flyers had more than enlisted groundcrew or officers (34.4%, 24.8%, and 22.6%, respectively). An adjusted analysis of the proportion of participants with comedones was performed using logistic regression techniques. Results are presented in Table 14-7. TABLE 14-7. Adjusted Analysis for Comedones by Group Ranch Hand Total 1,007 Comparison Total Adjusted Relative Risk (95% C.I.) p-Value 1,282 0.89 (0.74,1.09) 0.260 14-12 Covariate Remarks Occupation (p<0.001) Presence of Pre-SEA Acne (p=0.038) Race-by-Age (p=0.046) �Again, no significant differences were found between groups (p=0.260). Occupation, pre-SEA acne, and a race-by-age interaction were significant (p<0.001, p=0.038, and p=0.046, respectively). Compared to Baseline findings, the percentage of participants with comedones increased in the Comparison group but decreased in the Ranch Hand group. Estimated and adjusted relative risks were both less than 1.0 in the followup study, while the estimated relative risk in the Baseline study was slightly greater than 1.0 (RR=1.05, with Original Comparisons used), but statistically nonsignificant. Acneiform Lesions As shown in Table 14-8, there was not a significant difference between the proportion of participants with acneiform lesions in the Ranch Hand and Comparison groups, unadjusted for any covariates (p=0.624). TABLE 14-8. Unadjusted Analysis for Acneiform Lesions by Group Acneiform Lesions Present Group Ranch Hand Comparison Number 188 228 Absent Percent Number 18.5 17.6 828 1,064 Percent 81.5 82.4 Total 1,016 1,292 Summary Statistics Est. RR: 1.06 95* C.I.i (0.86,1.31) p-Value: 0.624 Test.? of association between the presence of acneiform lesions in both groups and the four covariates revealed marginally significant effects due to race (p=0.055) and occupation (p=0.064), and significant effects for age (p<0.001) and presence of pre-SEA acne (p<0.001). Nonblacks had a marginally significantly higher proportion of participants with acneiform lesions than Blacks (18.4% versus 11.9%). The proportion of participants with lesions was greatest for enlisted groundcrew (20.1%), as compared to the other occupations (officers, 16.4%; enlisted flyers, 16.0%). The proportion of participants with acneiform lesions decreased with age (born in or after 1942, 23.0%; born between 1923 and 1941, 14.8%; born in or before 1922, 10.3%). A significantly higher proportion of participants with acne present before SEA had lesions (22.4%), as compared.with those not having acne before SEA (16.0%). An adjusted analysis of the proportion of participants with acneiform lesions was performed using logistic regression techniques. Results of this analysis are summarized in Table 14-9. 14-13 �TABLE 14-9. Adjusted Analysis for Acneiform Lesions by Group Ranch Hand Total 1,007 Comparison Total 1,282 Adjusted Relative Risk (95% C.I.) Covariate Remarks p-Value 1.08 (0.87,1.34) 0.512 Age (p<0.001) Race (p=0.014) Presence of Pre-SEA Acne (p=0.008) The results showed no significant differences between groups (p=0.512). Age (p<0.001), race (p=0.014), and presence of pre-SEA acne (p=0.008) were significant adjusting variables in this analysis. The Baseline and followup results for acneiform lesions were nearly identical with respect to group differences. Acneiform Scars Table 14-10 shows no significant difference between the proportion of participants with acneiform scars in the Ranch Hand and Comparison groups, unadjusted for any covariates (p«0.720). TABLE 14-10. Unadjusted Analysis for Acneiform Scars by Group Acneiform Scars Present Group Number Ranch Hand Comparison 150 183 Absent Percent Number 14.8 14.2 866 1,109 Percent 85.2 85.8 Total 1,016 1,292 Summary Statistics Est. RR: 1.05 95% C.I.: (0.83,1.33) p-Value: 0.720 Tests of association between the presence of acneiform scars in both groups and the covariates disclosed significant effects due to the four variables (p<0.001). As age increased, the proportion of participants with 14-14 �acneiform scars decreased (17.9% for participants born in or after 1942, 12.4% for participants born between 1923 and 1941, and 5.7% for participants born in or before 1922). Significantly more Blacks had scars than nonblacks (28.0% and 13.5%, respectively), and enlisted personnel had more than officers (enlisted groundcrew, 16.9%; enlisted flyers, 16.5%; and officers, 10.4%). The pre-SEA acne classification had a significantly higher proportion of participants with acneiform scars than the non pre-SEA acne classification. An adjusted analysis of the proportion of participants with acneiform scars was performed using logistic regression techniques. Results are given in Table 14-11. TABLE 14-11. Adjusted Analysis for Acneiform Scars by Group Ranch Hand Total 1,007 Comparison Total Adjusted Relative Risk. (95% C.I.) p-Value 1,282 1.07 (0.84,1.36) 0.584 Covariate Remarks Age (p=0.006) Race (p<0.001) Occupation (p=0.016) Presence of Pre-SEA Acne (p<0.001) No significant group differences were found (p=0.584). As in the covariate analysis with acneiform scars, significant effects in the adjusted analysis were observed due to all four covariates (age, p=0.006; race, p<0.001; occupation, p=0.0i6; presence of pre-SEA acne, p<0.001). The results for acneiform scars, as with the acneiform lesions, were quite similar in the followup and Baseline studies. Depigmentation Table 14-12 shows the contrast between the proportion of participants with depigmentation in the Ranch Hand and Comparison groups, unadjusted for any covariates. The proportion of participants with depigmentation was greater in the Comparison than in the Ranch Hand group; however, the difference between groups was nonsignificant (p=0.143). 14-15 �TABLE 14-12. Unadjusted Analysis for Depigmentation by Group Depigmentation Absent Present Group Ranch Hand Comparison Number Percent Number 102 155 10.0 12.0 Percent 914 1,137 90.0 88.0 Summary Statistics Total 1,016 1,292 Est. RR: 0.82 95% C.I.: (0.63,1.07) p- Value: 0.143 Tests of association between the presence of depigmentation in both groups and the four covariates determined a significant effect due to race (p=0.009), but shoved nonsignificant effects for age, occupation, and presence of pre-SEA acne. An adjusted analysis of the proportion of participants with depigmentation was performed using logistic regression techniques. The statistics are presented in Table 14-13. TABLE 14-13. Adjusted Analysis for Depigmentation by Group Ranch Hand Total 1,016 Comparison Total 1,292 Adjusted Relative Risk (95* C.I.) 0.82 (0.63,1.07) p-Value 0.144 Covariate Remarks Race (p=0.010) No significant difference was observed between groups (p=0.144). Race was the only significant covariate in this adjusted analysis (p=0.010). Depigmentation was not analyzed in the Baseline study. Inclusion Cysts As reflected in Table 14-14, there was not a significant difference between the proportion of participants with inclusion cysts in the Ranch Hand and Comparison groups, unadjusted for any covariates (p=0.303). 14-16 �TABLE 14-14. Unadjusted Analysis for Inclusion Cysts by Group Inclusion Cysts Present Group Ranch Hand Comparison Number Absent Percent Number 114 164 11.2 12.7 Percent 902 1,128 88.8 87.3 Total 1,016 1,292 Summary Statistics Est. RR: 0.87 95% C.I.: (0.67,1.12) p-Value: 0.303 Tests of association between the presence of inclusion cysts in both groups and the covariates of age, race, occupation, and presence of pre-SEA acne showed no significant effects due to age (p=0.437), race (p«0.506), or presence of pre-SEA acne (p-0.449). Occupation, however, exhibited a significant effect (p=0.036), with the enlisted flyer category having the highest proportion of participants with inclusion cysts (15.8% versus 11.9% and 10.8% for officers and enlisted groundcrew, respectively. An adjusted analysis of the proportion of participants with inclusion cysts was performed using logistic regression techniques. Results are presented in Table 14-15. TABLE 14-15. Adjusted Analysis for Inclusion Cysts by Group Ranch Hand Total 1,016 Comparison Total Adjusted Relative Risk (95% C.I.) p-Value Covariate Remarks 1,292 0.86 (0.67,1.12) 0.260 Occupation (p=0.041) No significant differences for inclusion cysts were found between the Ranch Hand and the Comparison groups (p=0.260). Occupation was the only significant covariate in this analysis (p=0.041). With reference to the Baseline study, the percentage of participants with inclusion cysts at the followup increased in the Comparison group, and 14-17 �decreased slightly in the Ranch Hand group. These differences could be due to changes in disease over time, different examiners, or changes in the cohorts examined. Both estimated and adjusted relative risks were less than one in the followup, while the estimated relative risk at the Baseline was slightly greater than one (RR=1.10 for Original Comparisons) but was not statistically significant. Hyperpigmentation Table 14-16 shows there was not a significant difference between the proportion of participants with hyperpigmentation in the Ranch Hand and Comparison groups, unadjusted for any covariates (p=0.762). TABLE 14-16. Unadjusted Analysis for Hyperpigmentation by Group Hyperpigmentation Present Group Ranch Hand Comparison Absent Number Percent Number 228 283 22.4 21.9 788 1,009 Percent 77.6 78.1 Total 1,016 1,292 Summary Statistics Est. RR: 1.03 95% C.I.: (0.85,1.26) p-Value: 0.762 Tests of association between the presence of hyperpigmentation in both groups and the four covariates revealed there was not a significant effect due to age (p=0.833), but that significant effects were due to race (p<0.001), occupation (p<0.001), and presence of pre-SEA acne (p=0.003). Blacks had a much higher prevalence of hyperpigmentation than nonblacks (53.1% for Blacks, 20.1% for nonblacks), and enlisted personnel had a higher prevalence of hyperpigmentation than officers (enlisted groundcrew, 25.5%; enlisted flyers, 23.5%; officers, 17.4%). The proportion of participants with hyperpigmentation was greater in the absence of pre-SEA acne (23.8%) than in the presence of pre-SEA acne (18.2%). An adjusted analysis of the proportion of participants with hyperpigmentation was performed using logistic regression techniques. Results are given in Table 14-17. 14-18 �TABLE 14-17. Adjusted Analysis for Hyperpigmentation by Group Ranch Hand Total 1,007 Comparison Total 1,282 Adjusted Relative Risk (95% C.I.) 1.04 (0.85,1.27) p-Value 0.720 Covariate Remarks Race (p<0.001) Occupation (p-0.009) Presence of Pre-SEA Acne (p=0.009) No significant group differences (p=0.720) were noted, although significant effects of race (p<0.001), occupation (p=0.009), and presence of pre-SEA acne (p=0.009) were evident. The proportion of participants with hyperpigmentation has increased since the Baseline study. Almost three times as many abnormalities were , found at the followup study (approximately 22% versus 8%). The relative risk estimate was closer to 1 in the followup study, but relative risks from both the Baseline and followup studies were not significantly different from 1. These differences could be due to disease or examination techniques. Other Abnormalities The study of other abnormalities encompassed a wide range of dermatological disorders. Included in this variable were the following abnormalities: (1) (2) (3) (4) (5) (6) (7) (8) Jaundice Spider Angiomata Palmar Erythema Suspected Melanoma Palmar Keratoses Actinic Keratoses Petechiae Ecchymoses (9) (10) (11) (12) (13) (14) (15) (16) Conjunctival Abnormality Oral Mucosal Abnormality Fingernail Abnormality Toenail Abnormality Dermatographia Cutis Rhomboidalis Suspected Basal Cell Carcinoma Suspected Squamous Cell Carcinoma With respect to the category "Other Abnormalities," a participant was considered normal only if he was.negative for all of these conditions. If one or more abnormalities existed, then the participant was considered abnormal. 14-19 �As reflected in Table 14-18, there was not a significant difference between the proportion of participants with other abnormalities in the Ranch Hand and Comparison groups, unadjusted for any covariates (p=0.349). TABLE 14-18. Unadjusted Analysis for Other Abnormalities by Group Other Abnormalities Abnormal Group Ranch Hand Comparison Number 608 748 Normal Percent Number 59.8 57.9 408 544 Percent 40.2 42.1 Total 1,016 1,292 Summary Statistics Est. RR: 1.08 95% C.I.: (0.92,1.28) p-Values 0.349 Tests of association between the presence of other abnormalities in both groups and the four covariates found a marginally significant effect due to the presence of pre-SEA acne (p=0.084), and significant effects due to age (p<0.001), occupation (p<0.001), and race (p<0.001). The proportion of participants with other abnormalities in the absence of pre-SEA acne (59.9%) was marginally significantly larger than the proportion of participants with other abnormalities who also had pre-SEA acne (56.1%). The proportion of participants with other abnormalities increased with age (with a low of 43.3% for participants born in or after 1942 to a high of 82.8% for participants born in or before 1922). Nonblacks had a significantly and substantially higher percentage of other abnormalities than Blacks (60.3% and 35.7%, respectively). Enlisted groundcrew had a lower proportion of abnormalities than officers or enlisted flyers (53.3%, 63.2%, and 63.8%, respectively). An adjusted analysis of the proportion of participants with other abnormalities was performed using logistic regression techniques. Results are presented in Table 14-19. Again, no significant difference was observed between groups (p=0.432). Age and race were significant covariates in this analysis (p<0.001 for both). In reference to the Baseline study, the percentage of participants with other abnormalities has increased in both the Comparison and the Ranch Hand groups. In the Baseline study, the estimated relative risk for Ranch Hands versus Original Comparisons was 0.77, significantly less than 1.00. The estimate of the relative risk has increased in the followup study to 1.08. The percentage of other abnormalities has increased from approximately 14 percent in the Baseline study to nearly 59 percent in the followup study. 14-20 �TABLE 14-19. Adjusted Analysis for Other Abnormalities by Group Ranch Hand Total 1,016 Comparison Total 1,292 Adjusted Relative Risk (95% C.I.) 1.07 (0.90,1.28) p-Value 0.432 Covariate Remarks Age (p<0.001) Race (p<0.001) Dermatology Index Four of the previously analyzed conditions (comedones, acneiform lesions, acneiform scars, and inclusion cysts) were used to construct a dermatology index. All four conditions are indicators of possible chloracne. The index was formulated by counting the number of abnormalities present in a participant for the four conditions. Consequently, the dermatology index ranged from 0 to 4, where 0 indicated that the participant had none of these abnormalities and 4 indicated that the participant had all of these abnormalities. Table 14-20 presents the number and the percent of participants with each of these five scores by group. A significant difference between the Ranch Hand and Comparison groups was not observed for this dermatology index, unadjusted for any covariates (p=0.576, 4 d.f.). Covariate main effect analyses found nonsignificant effects due to age (p=0.407) and race (p=0.558), but significant effects for occupation (p-0.010) and the presence of acne pre-SEA (p<0.001). These data are summarized in Table 14-21. By occupation, 55.8 percent of the officers had no abnormalities, whereas 50.8 percent of the enlisted groundcrew and 44.4 percent of the enlisted flyers had no abnormalities. The stratum corresponding to participants with pre-SEA acne present had a larger percentage of participants with at least one abnormality (see Table 14-21). An adjusted analysis of the five scores of the dermatology index was performed using log-linear modeling techniques. Significant effects were noted for occupation and an interaction between group and presence of pre-SEA acne (p=0.005, p=>0.041, respectively). Consequently, an analysis, stratifying by pre-SEA acne status, was performed, and the results are shown in Table 14-22. The adjusted relative risk for each of the index scores (1 to 4, separately, versus the 0 score), the 95 percent confidence interval, and the p-value for each contrast for each pre-SEA acne class are given in Table 14-23. 14-21 �TABLE 14-20. Unadjusted Analysis for the Dermatology Index by Group Dermatology Index Score 0 Group Ranch Hand Comparison 1 2 3 Number Percent Number Percent Number Percent Number 533 658 52.5 50.9 318 420 31.3 32.5 121 154 11.9 11.9 34 53 Overall p-Value (4 d.f.)=0.576 Contrast Est. Relative Risk (95% C.I.) p-Value* N9 to 1 vs. 0 2 vs. 0 3 vs. 0 4 vs. 0 *Fisher's exact test. 0.94 0.97 0.79 1.76 ( .78 ,1 .13) 0 ( .75 ,1 .26) 0 ( .51 ,1 .24) 0 ( .67 ,4.66) 0 0.480 0.840 0.317 0.327 Percent 3.3 4.1 Number Percent 10 7 1.0 0.5 Total 1,016 1,292 �TABLE 14-21. . o v - - r and Presence of Pre-SE&i^mt^ in the Combined Ranch —_ _ j vw^—jw 1 rbwi ani O"*nrwm<vn GCOUDS •»—-»~y-i» Dermatology Index Score* 0 1 2 3 4 Covariate Covariate Category Age Born XL942 501 52.2 2% 30.8 114 11.9 40 4.2 9 0.9 960 Bom 1923-1941 Born <L922 647 43 51.3 49.4 408 34 32.4 39.1 156 5 12.4 5.7 42 5 3.3 5.7 '8 0 0.6 0.0 1,261 87 83 1,108 58.0 51.2 38 700 26.6 32.3 17 258 11.9 11.9 4 83 2.8 3.8 1 16 0.7 0.8 143 2165 0.558 482 172 55.8 44.4 265 135 30.7 34.9 91 58 10.5 21 19 2.4 4.9 5 3 0.6 0.8 864 387 0.010 15.0 537 50.8 338 32.0 126 11.9 47 4.4 9 0.9 1,057 842 337 54.0 46.2 514 220 32.9 30.1 153 121 9.8 16.6 44 42 2.8 5.8 7 9 0.4 1.2 1,560 729 Race Number Percent Number Percent Number Percent Number Percent Number Percent Black Nonblack Total" p-VaW 0.407 Occupation Officer Bilisted Flyer Enlisted Groundcrev Presence of Pre-Sea Acne** No Yes 0.001 * Score denotes the number of abnormalities (for comedones, acneiform lesions, acneiform scars, and inclusion cysts) diagnosed. **Nineteen participants could not be classified. * One participant refused to take the dermatology examination. b Pearson's chi-square test. �TABLE 14-22. Adjusted Analysis for the Dermatology Index by SEA Acne Class and Group Dermatology Index Score* 1 0 Pre-SEA Acne Class ^ No pre-SEA -t- acne i Group Number Percent Number 2 Percent Number 3 ' Percent Number Percent 4 Number . Percent Total Ranch Hand Comparison 360 482 52.6 55.0 234 280 34.2 32.0 69 84 10.1 9.6 16 28 2.3 3.2 5 2 0.7 0.2 684 876 Ranch Hand Comparison 167 170 51.7 41.9 82 138 25.4 34.0 51 70 15.8 17.2 18 24 5.6 5.9 5 4 1.5 1.0 323 406 Pre-SEA acne - *Score denotes the number of abnormalities (comedones, acneiform lesions, acneiform scars, and inclusion cysts) diagnosed. �TABLE 14-23. Adjusted Relative Risks for Contrasts of Dermatology Index by Pre-SEA Class Pre-SEA Acne No Yes Contrast Adjusted Relative Risk 95% C.I. p-Value 1 abnormality vs. 0 abnormalities 2 vs. 0 3 vs. 0 4 vs. 0 1.12 1.10 0.77 3.09 (0.90,1-39) (0.77,1.55) (0.41,1.44) (0.65,14.62) 0.315 0.605 0.411 0.155 1 2 3 4 0.60 0.73 0.75 1.19 (0.42,0.85) (0.48,1.12) . (0.39,1.43) (0.33,4.38) 0.004 0.148 0.380 0.788 vs. 0 vs. 0 vs. 0 vs. 0 This analysis showed a significant difference between groups only when contrasting the proportion of participants with one abnormality (out of four) to the proportion of participants with no abnormalities for participants with pre-SEA acne (p=0.004). However, Comparisons were more likely to have one abnormality than the Ranch Hands, as is evidenced by the relative risk and confidence interval being less than 1. In contrast to the Baseline study, the percentage of participants with a score of 1 or more has increased at the followup examination for both the Ranch Hand and Comparison groups (8.1% for Ranch Hands, 12.1% for Comparisons). The estimated relative risks, when the dermatology index is condensed into two categories, were 1.11 for the Baseline examination and 0.94 for the followup examination. Biopsy Results Dermatologists were instructed to perform skin biopsies on any lesions they suspected of being malignant. Of the 40 biopsies collected from 35 participants, none was suggestive of chloracne. Histologic descriptions of these biopsies are presented in Table 14-24. With the exception of confirmed basal cell carcinoma, no single diagnostic category predominated. 14-25 �TABLE 14-24. Summary of Histologic Descriptions of Skin Biopsy by Group Group Ranch Hand Basal Cell Carcinoma Suspected Basal Cell Carcinoma Suspected Unspecified Carcinoma Unspecified Carcinoma Dermatofibroma Pigmented Nevus Dyschromia Keratoderma, Acquired Melanoacanthoma (Papilloma) Intradermal Nevus Junctional Nevus Cavernous Hemangioma Degenerative Skin Disorder Other Specified Disorders of Skin Local Infection of Skin Other Dermatoses Total Comparison 7 0 0 1 3 1 1 1 0 1 0 0 0 0 1 5 Histologic Description 4 3 1 0 0 2 0 1 1 0 1 1 1 1 0 2 21 Comments a,b b c d a c c 18 *0ne participant had a basal cell carcinoma at one site and an acquired keratoderma at another site. 'One participant had a basal cell carcinoma at one site and a suspected basal cell carcinoma at another site. : 0ne participant had a local infection of the skin, a suspected unspecified carcinoma, and a dermatosis at the same site. J 0ne participant had two cases of dyschromia at two different sites. 14-26 �EXPOSURE INDEX ANALYSES Exposure index analyses were conducted within each occupational cohort of the Ranch Hand group to search for dose-response relationships (see Chapter 8 for details on the exposure index). The dermatology index was collapsed into two categories, 0 and greater than 0. All eight dermatological variables were explored, unadjusted for any covariates, using Pearson's chi-square test and Fisher's exact test. Adjusted analyses were performed by logistic regression for these variables, using age, race, presence of pre-SEA acne, and any significant pairwise interactions between the exposure index and these covariates. Overall significance in the proportion of abnormalities among the exposure index levels of low, medium, and high was determined, as well as contrasts in the proportion of abnormalities between the medium and low exposure levels, and between the high and low exposure levels. Age was used as a continuous variable in the adjusted analyses. Results of the adjusted analyses for these eight variables are presented in Table 14-25, and counterpart results for unadjusted analyses are presented in Table L-l of Appendix L. Results from further investigation of exposure index by covariate interactions are given in Table L-2 of Appendix L. Significant or marginally significant results were present for some of these variables based on unadjusted analyses. A borderline significantly higher prevalence of comedones (Est. RR: 1.78, 95% C.I.: [0.95,3.35], p=0.084) for the contrast of medium exposure to low exposure was seen for officers. Marginally significant results for the contrast of high exposure to low exposure were also present for acneiform scars for officers (Est. RR: 2.38, 95% C.I.i [0.94,6.06], p=»0.075) and enlisted groundcrew (Est. RR: 1.82, 95% C.I.: [1.00,3.30], p-0.053), as well as for other abnormalities for officers (Est. RR: 1.66, 95% C.I.: [0.98,2.78], p=0.067). The data for these last three variable-occupation combinations supported an increase in the proportion of abnormalities from low to high exposure. Significant or marginally significant results were also observed for medium exposure versus low exposure in officers and enlisted groundcrew for depigmentation, and for high exposure versus low exposure in other abnormalities with enlisted flyers, but prevalence decreased as the exposure level increased in these cases. The frequency of abnormalities for the different exposure index levels exhibited no consistent pattern across occupations. However, within the officer and enlisted groundcrew occupations, most variables showed the low exposure level to have the lowest prevalence of abnormalities or the high exposure level to have the highest prevalence, whereas very few variables showed this pattern for enlisted flyers. Adjusted analyses revealed patterns similar to those of the unadjusted analyses. Results of the counterpart adjusted analyses to the situations described above are detailed below. (1) Comedones in officers, medium versus low: Adj. RR: 1.62, 95% C.I.: (0.83,3.15), p=0.154. 14-27 �TABLE 14-25. Adjusted Exposure Index Analysis for Dermatological Variables by Occupation Exposure Index Variable Occupation Low Total Medium Total High Total Contrast Adj . Relative Risk ( 5 C.I.) 9% p-Value Officer 122 Overall M vs. L H vs. L 0.334 1.62 (0.83,3.15) 0.154 1.44 (0.74,2.83) 0.283 Enlisted Flyer 55 65 56 Overall M vs. L H vs. L 0.413 0.65 (0.30,1.41) 0.276 0.61 (0.27,1.37) 0.234 152 162 140 Overall M vs. L H vs. L 0.878 0.94 (0.55,1.60) 0.808 1.08 (0.63,1.83) 0.782 Officer ^ *• 129 Enlisted Groundcrew Comedones 126 126 129 122 Overall M vs. L H vs. L 0.669 1.06 (0.52,2.15) 0.880 1.34 (0.67,2.66) 0.409 Enlisted Flyer 55 65 56 Overall M vs. L H vs. L 0.917 0.91 (0.32,2.60) 0.856 1.14 (0.39,3.35) 0.814 152 162 140 Overall M vs. L H vs. L 0.674 1.01 (0.58,1.75) 0.973 1.25 (0.71,2.20) 0.431 tsj OO Acneiform Lesions Enlisted Groundcrew �TABLE 14-25. (continued) Adjusted Exposure Index Analysis for Dermatological Variables by Occupation Exposure Index Variable Occupation Low Total Medium Total High Total Contrast Adj. Relative Risk ( 5 C.I.) 9% p-Value » Officer 126 129 122 Overall M vs. L H vs. L **() **! **() **! **() **! **() **! **() **! 55 65 56 Overall M vs. L H vs. L 0.363 0.82 (0.31,2.13) 0.682 0.47 (0.16,1.39) 0.174 152 162 140 Overall M vs. L H vs. L 0.068 1.22 (0.66,2.27) 0.519 2.00 (1.08,3.67) 0.026 Officer ^ *• " Enlisted Flyer Enlisted Groundcrewa Acneiform Scars 126 129 122 Overall M vs. L H vs. L 0.006 0.33 (0.11,0.98) 0.045 1.50 (0.69,3.25) 0.302 Enlisted Flyer 55 65 56 Overall M vs. L H vs. L 0.493 0.53 ( . 8 1 5 ) 0.245 01,.4 0.67 (0.24,1.90) 0.450 152 162 140 Overall M vs. L H vs. L ****(2) ****(2) to VO Depigmentation Enlisted Groundcrew **() **2 **() **2 **() **2 �TABLE 14-25. (continued) Adjusted Exposure Index Analysis for Dermatological Variables by Occupation Exposure Index Variable Occupation Low Total Medium Total High Total Contrast Adj. Relative Risk (95% C.I.) p-Value Officer13 Hyperpigmentation 122 Overall M vs. L H vs. L 0.221 2.05 ( .91,4.60) 0.082 0 1.32 ( .56,3.11) 0.532 0 Enlisted Flyer 55 65 56 Overall M vs. L H vs. L 0.881 1.24 ( .41,3.78) 0.707 0 1.33 ( .42,4.17) 0.630 0 152 162 140 Overall M vs. L H vs. L 0.916 0.91 ( .43,1.93) 0.806 0 1.07 ( .51,2.24) 0.856 0 Officer _, *u> o 129 Enlisted Groundcrew Inclusion Cysts 126 126 129 122 Overall M vs. L H vs. L 0.813 0.92 ( .47,1.79) 0.807 0 0. 0 ( .41,1.58) 0.525 8 0 Enlisted Flyer 55 65 56 Overall M vs. L H vs. L 0.656 0.71 ( .29,1.76) 0.465 0 1.04 ( .43,2.53) 0.930 0 152 162 140 Overall M vs. L H vs. L 0.365 1.20 (0.71,2.01) 0.494 0.81 ( . 6 1.41) 0.450 04, Enlisted Groundcrew �TABLE 14-25. (continued) Adjusted Exposure Index Analysis for Dermatological Variables by Occupation Exposure Index Variable Occupation Low Total Medium Total High Total Contrast Adj. Relative Risk ( 5 C.I.) p-Value 9% Officer 122 Overall M vs. L H vs. L 0.309 1.30 (0.75,2.24) 0.346 1.53 (0.88,2.65) 0.129 Enlisted Flyer 55 65 56 Overall M vs. L H vs. L 009 .4 0.66 (0.28,1.56) 0.341 0.35 ( . 4 0 8 ) 0.018 01,.3 152 162 140 Overall M vs. L H vs. L 0.764 0.85 (0.52,1.36) 0.489 0.87 (0.52,1.43) 0.580 Officer Dermatology Index 129 Enlisted Groundcrew Other Abnormalities 126 126 129 122 Overall M vs. L H vs. L **() **! **() **! Enlisted Flyer Enlisted Groundcrew **() **! **() **! **() **! 55 65 56 Overall M vs. L H vs. L 0.618 0.74 (0.36,1.54) 0.423 0.71 (0.33,1.51) 0.368 152 162 140 Overall M vs. L H vs. L 0.469 1.01 (0.65,1.59) 0.955 1.30 (0.82,2.06) 0.270 "Marginal exposure index-by-presence of pre-SEA acne interaction (p=0.056)—relative risk, confidence interval and p-value presented, and additional information provided in interaction summaries. * * ( ) Exposure index-by-presence of pre-SEA acne and exposure index-by-race interaction—relative risk, **!: confidence interval, and p-value not presented. * * ( ) Exposure index-by-presence of pre-SEA acne interaction—relative risk, confidence interval and **2: p-value not presented. �(2) Acneiforra scars in officers, high versus low: interaction present; direct contrast of adjusted and unadjusted analyses not possible. (3) Acneiform scars in enlisted groundcrew, high versus lows Adj. RR: 2.00, 95% C.I.'. (1.08,3.67), p=0.026; overall p-value=0.068, increase in the proportion of abnormalities with increasing exposure levels supported. (4) Other abnormalities in officers, high versus lows Adj. RR: 1.53, 95% C.l.t (0.88,2.65), p=0.129. Other adjusted analyses that showed significance or marginal significance exhibited a decreasing prevalence with increasing exposure level. All other adjusted analyses showed an interaction with covariates (described below) or nonsignificant results. Interactions were present for three of the eight variables and were observed for officers and enlisted groundcrew. A summary of these interactions is presented below in Table 14-26. TABLE 14-26. Summary of Exposure Index by Covariate Interactions Encountered in Adjusted Analysis of Dermatological Variables Occupation Covariate p-Value Acneiform Scars Officer Race 0.003 Acneiform Scars Officer Presence of Pre-SEA acne 0.003 Variable Acneiform Scars Enlisted Groundcrew Presence of Pre-SEA acne (marginal) 0.056 Depigmentation Enlisted Groundcrew Presence of Pre-SEA acne 0.035 Dermatology Index* Officer Race 0.026 Dermatology Index* Officer Presence of Pre-SEA acne 0.029 *Variable was collapsed into two categories, 0 and >0. 14-32 �As can be seen, all variables and occupations with interactions had a significant exposure index-by-presence of pre-SEA acne interaction or significant exposure index-by-race and exposure index-by-presence of pre-SEA acne interactions. Meaningful interpretation of many of the subsequent stratified analyses was hindered by small sample sizes, but two situations were of particular interest. For acneiform scars on officers, nonblack personnel without pre-SEA acne at low exposure had no participants with scars, whereas nonblack personnel exposed at the medium and high levels had 7.8 percent and 10.5 percent of participants with scars, respectively. Also, with acneiform scars for enlisted groundcrew, an increase in the prevalence of abnormalities for increasing levels of exposure was present for participants with pre-SEA acne, with an adjusted relative risk of 5.38 (95% C.I.: [1.45,19.96], p=0.012) for the contrast of high exposure versus low exposure. In summary, the results suggested the presence of an increasing doseresponse relationship in certain occupations for a few of the dermatological variables or within substrata of these variables, but no consistent pattern was evident throughout the dermatological exposure index evaluation. LONGITUDINAL ANALYSES The dermatology index was chosen to assess longitudinal differences between the 1982 Baseline examination and the 1985 followup examination. In testing for this difference, the dermatology index scores were collapsed into two categories: normal (dermatology index score of 0) and abnormal (dermatology index score greater than 0). As shown in Table 14-27, 2x2 tables were constructed for each group. These tables show the number of participants who were abnormal at the Baseline examination and abnormal at the followup, abnormal at Baseline and normal at followup, normal at Baseline and abnormal at followup, and normal at both Baseline and followup. The odds ratios given are the ratios of the number of participants who were normal at the Baseline and abnormal at the followup to the number of participants who were abnormal at the Baseline and normal at the followup (the "off-diagonal" elements). TABLE 14-27. Longitudinal Analysis of the Dermatology Index: A Contrast of Baseline and First Followup Examination Abnormalities 1982 1985 Followup Exam Group Baseline Exam Abnormal Normal Ranch Hand - Abnormal Normal 241 228 136 366 1.68 Comparison Abnormal Normal 283 283 136 437 2.08 Odds Ratio (OR)* p-Value (ORRH vs ORC) 0.15 *0dds Ratio: Number Normal Baseline, Abnormal Followup Number Abnormal Baseline, Normal Followup 14-33 �The changes in normal/abnormal status within each group were compared, and the p-value given was derived from Pearson's chi-square test of the hypothesis that the pattern of change in the two groups was the same. These results showed that the difference in the pattern is not significant (p=0.15). DISCUSSION The relative risks for all eight dermatological variables approached unity (none was statistically significant), an observation previously noted at the Baseline examination (except for the category "Other Abnormalities," which predominated in the Comparisons). More dermatological abnormalities were recorded at the followup (for six of the seven variables shared between the examinations) than at the Baseline—the increase in detection was slightly stronger in the Comparison group than in the Ranch Hand group. For example, in the category "Other Abnormalities," the reporting of skin lesions generally increased from about 14 percent to 59 percent. The overall difference between the two examinations probably reflects a combination of factors, e.g., changes in disease, chance, the addition of new participants, and possible differences in clinical practice between the two groups of dermatologists. The histologic categories of skin cancer (confirmed or suspected, any type), as examined by biopsies, showed a similarity between both groups. SUMMARY AND CONCLUSIONS Interval questionnaire data on the occurrence, time, and location of acne were analyzed to assess the possible historical diagnosis of chloracne. No significant difference was observed between groups for reported occurrence of acne, although the Ranch Hand cohort reported slightly more acne. The occurrence of acne relative to 1961 was comparable between groups. A marginally significant difference in the occurrence of post-1961 acne was found, with more Ranch Hands than Comparisons reporting acne strictly post-SEA. The duration of post-1961 acne was not significantly different between the two groups. For participants with post-SEA acne, the spatial eyeglass distribution of acne (suggesting chloracne) was observed to be similar for the Ranch Hand and Comparison groups, both for individual sites and the combination of acne on the eyelids, ears, and temples. This analysis suggested that the occurrence of skin disease compatible with chloracne was not different in the two groups. Analyses of the followup physical examination data, as with the Baseline examination, placed primary emphasis on six dermatologic disorders: comedones, acneiform lesions, acneiform scars, inclusion cysts, depigmentation, and hyperpigmentation. Secondary emphasis was given to 16 other minor conditions (generally not associated with chloracne) recorded at the physical examination. No significant findings occurred in any variable, as reflected in Table 14-28. 14-34 �TABLE 14-28. Overall Summary Results of Unadjusted and Adjusted Analyses of Questionnaire and Physical Examination Dermatological Variables Variable Unadjusted Adjusted Questionnaire Incidence of Acne Occurrence NS Relative to 1961 NS Relative to SEA (Post-1961 Cases) NS* Duration of Acne NS Location, of Acne NS — — — NS — Physical Examination Comedones NS NS Acneiforra Lesions NS NS Acneiform Scars NS NS Depigmentation NS NS Inclusion Cysts NS NS Hyperpigmentation NS NS NS NS NS **** Other Abnormalities . Dermatology Index NS: Not significant (p>0.10). — Analyses not performed. NS*: Borderline significant (0.05<p<0.10). ****Group-by-covariate interaction. 14-35 �No significant difference vas found for any of these variables in the unadjusted analyses. The variable consisting of the 16 secondary conditions, labeled "other abnormalities," had the largest difference in the prevalence of abnormalities for the Ranch Hand cohort over the Comparison group (Est. RR: 1.08, 95% C.I.: [0.92,1.28], p=0.349), but the difference was clearly nonsignificant. The covariate effects of age, race, occupation, and the presence of pre-SEA acne were often profound with respect to the recorded dermatologic conditions. The adjusted analyses closely mirrored the unadjusted analyses, with no significance noted between groups for any variable. Only one group-bycovariate interaction was observed in the adjusted analysis of the dermatology index, with a group-by-presence of pre-SEA acne interaction noted. However, further analysis of this interaction did not show an adverse effect in the Ranch Hand group. Exposure index analyses did support dose-response relationships for some of the variables in certain occupational strata, but did not reveal a strong pattern of results suggesting a relationship between skin disease and herbicide exposure. Overall, the followup examination results paralleled the Baseline findings. Although the followup examination detected more dermatologic abnormalities than those present at Baseline, slightly more abnormalities were found in the Comparisons, and most relative risks approached unity. The longitudinal analysis for the dermatology index showed no statistically significant differences between groups in the change in results from the Baseline to the followup examination. In conclusion, none of the questionnaire results disclosed an increased likelihood of past chloracne in the Ranch Hands. The physical examination did not diagnose a current case of chloracne. The dermatological data were similar between the Ranch Hand and Comparison groups, and the longitudinal analysis of the dermatology index suggested equivalence between the Baseline and followup examination results. 14-36 �CHAPTER 14 REFERENCES 1. Kimmig, J., and K.H. Schulz. 1957. Occupational acne due to chlorinated aromatic cyclic esters. Dermatologica 115:540. 2. Kimmig, J., and K.H. Schulz. 1957. Chlorinated aromatic cyclic ethers as the cause of so-called chloracne. Naturvissenschaften 44:337-338. \ 3. Jones, E.L., and H. Kizek. 1962. A technique for testing acnegenic potency in rabbits, aplied to potent acnegen, 2,3,7,8 tetrachlorodibenzo-p-dioxin. J. Invest. Dermatol. 9:511-517. 4. McConnell, E.E., J.A. Moore, and D.W. Dalgard. 1978. Toxicity of 2,3,7,8-tetrachlorodibenozo-p-dioxin in Rhesus monkeys (Macaca mulatta) following a single oral dose. Toxicol. Appl. Pharmacol. 43(1):175-187. 5. Kimbrough, R.D. 1980. Occupational exposure. No. 4 in Halogenated biphenyls, terphenyls, naphthalenes, dibenzodioxins, and related" roducts, ed. R.D. Kimbrough, p. 373. Topics in Environ. Health", Isevier/North Holland, Amsterdam. f 6. Young, A.L. 1980. The chlorinated dibenzo-p-dioxins. Chap. 5 in The science of 2,4,5-T and associated phenoxy herbicides, ed. R.L. Metcalf and W. Stumm, pp. 133-205. New York:Wiley-Interscience. 7. Knutson, J.C., and A. Poland. 1982. Response of murine epidermis to 2,3,7,8-tetrachlorodibenzo-p-dioxin: Interaction of the Ah and hr loci. Cell 30: 225-234. 8. Kay, J.H., R.J. Palazzolo, and J.C. Calandra. 1965. Subacute dermal toxicity of 2,4-D. Arch. Environ. Health 11:648-651. 9. Carter, C.D., R.D. Kimbrough, J.A. Liddle, R.E. Cline, M.M. Zack, W.F. Barthel, R.E. Koehler, and P.E. Phillips. 1975. Tetrachlorodibenzodioxin: An accidental poisoning episode in horse arenas. Science 188(4189):738-740. 10. Case, A.A. 1976. Tetrachlorodibenzodioxin (TCDD)—clinical aspects of poisoning. Clin. Toxicol. 9(6):963-967. 11. Greenlee, W.F., R. Osborne, L.G. Hudson, and W.A. Toscano. 1984. Studies on the mechanisms of toxicity of TCDD to human epidermis. In Banbury report 18: Biological mechanisms of dioxin action; ed. A. Poland and R.D. Kimbrough, Cold Spring Harbor, New York: Cold Spring Harbor Laboratory, pp. 365-372. 14-37 �12. Reggiani, G. 1980. Acute human exposure to TCDD in Seveso, Italy. Toxicol. Environ. Health 6:27-43. J^ 13. May, G. 1973. Chloracne from the accidental production of tetrachlorodibenzodioxin. Br. J. Ind. Med. 30:276-283. 14. Jirasek, L., J. Kalensky, and K. Kubec. 1973. Acne chlorina and porphyria cutanea tarda during the manufacture of herbicides, part 1. Czech. Dermatol. 48(5):306-315. 15. Bleiberg, J., M. Wallen, R. Brodkin, and I.L. Applebaum. 1964. Industrially acquired porphyria. Arch. Dermatol. 89:793-797. 16. Suskind, R.R,, and V.S. Hertzberg. 1984. Human health effects of 2,4,5-T and its toxic contaminants. JAMA 251:2372-2380. 17. Oliver, R.M. 1975. Toxic effect of 2,3,7,8-tetrachloro-dibenzo-l, 4-dioxin in laboratory workers. Br. J. Ind. Med. 32:46-53. 18. Crow, K.D. 1983. Significance of cutaneous lesions in the symptomatology of exposure to dioxins and other chloracnegens. In human and environmental risks of chlorinated dipxins and related compounds, R.E. Tucker, et al., eds.pp. 605-612, New York:Plenum Press. 19. Allen, A.M. 1977. Skin diseases in Vietnam, 1965-1972. Internal Medicine in Vietnam, Vol. 1, ed. A.J. Ognibene, p. 42. Washington, D.C.: Center of Military History, Government Printing Office. \ 20. Halprin, K.M. 1980. Chloracne recognition and its significance. Presented at the Second Continuing Education Conference on Herbicide Orange, Washington, D.C., May 28-30. 21. Crow, K.D. 1970. Chloracne. 56:79-90. Trans. St. John's Hosp. Dermatol. Soc. 22. Hoffman, R.E., P.A. Stehr-Green, K.B, Webb, G. Evans, A.P. Knutsen, W.F. Schramm, J.L. Staake, B.B. Gibson, and K.K. Steinberg. 1986. Health effects of long-term exposure to 2,3,7,8-tetrachlorodibenzo-p-dixoin. JAMA 255:2031-2038. 23. Stehr, P.A., G. Stein, H. Falk, et al. 1986. A pilot epidemiologic study of possible health effects associated with 2,3,7,8-tetrachlorodibenzo-p-dioxin contamination in Missouri. Arch. Environ. Health 41:16-22. 24. Jirasek, L., J. Kalensky, K. Kubec, et al. 1974. Acne chlorina, porphyria cutanea tarda and other manifestations of general intoxication during the manufacture of herbicides, part 2. Czech. Dermatol. 49(3):145-157. 25. Goldmann, P.J. 1973. Schweistakute Chlor-akne, eine Massenintoxikation durch 2,3,7,8-Tetrachlorodibenzodioxin (Severe acute chloracne, a mass intoxication due to 2,3,6,7-tetrachlorodibenzodioxin). Hautarzt 24(4):149-152. 14-38 �26. Okumura, H., and S. Katauki. 1969. A clinical study of oil disease (chlorinated biphenyl poisoning), particularly the internal medical signs. Fukuoka Acta Med. 60:440-446. 27. Moses, M., R. Lilis, K.D. Crow, J. Thornton, A. Fischbein, H.A. Anderson, and I.J. Selikoff. 1984. Health status of workers with past exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in the manufacture of 2,4,5-trichlorophenoxyacetic acid: Comparison of findings with and without chloracne. Am. J. Ind. Med. 5:161-182. 28. Lathrop, G.D. 1985. Assessments of a controversy: Agent orange and its association with dioxin, science assessment, toxicology forum. Given at the Institute of Pathobiology; Aspen, Colorado, July 1985. 14-39 �CHAPTER 15 CARDIOVASCULAR EVALUATION INTRODUCTION Cardiac disease and peripheral vascular disease are not classically recognized sequelae of exposure to phenoxy herbicides, chlorophenols, or dioxin. Most observational and experimental animal studies using 2,4-D, 2,4,5-T, or TCDD have not extensively commented on resulting cardiac abnormalities or dysfunction. The studies described below viewed the cardiac abnormalities as expected consequences of a moribund*state, and not as an indicator of primary cardiac toxicity to the putative chemical. Following oral administration of 2,4-D and 2,4,5-T, sheep and cattle developed cardiac hemorrhages. A lethal oral dose of TCDD in young Rhesus monkeys produced increased heart weights in another experiment.2 Horses and cats showed generalized vascular degeneration following exposure to soil contaminated with TCDD, and mice and guinea pigs fed high amounts of TCDD manifested low heart weights. A teratogenic experiment using 2,4,5-T in developing fish eggs showed graduated lethality and cardiovascular anomalies, which included enlarged veins and heart chambers. Another study using ventricular muscle strips from chick embryos exposed to PCB's (including TCDD) showed a marked decrease in contractility. This primary cardiotoxic response was presumably mediated by the Ah receptor, and was associated with increased prostaglandin synthesis. Human case reports, case series of individuals with chloracne, and epidemiological studies also confirmed that cardiac function is not a sensitive indicator of exposure to herbicides or TCDD. In three case reports of acute 2,4-D poisoning, cardiac dilation and cardiac arrest were observed in the one fatal case, while only transient nodal tachycardia was observed in one of the two nonfatal cases. ' Three laboratory technicians with chloracne, neurological symptoms, and hypercholesterolemia following sig- 10 nificant direct exposure to TCDD did not manifest any cardiac dysfunction, however, of 10 industrial workers with chloracne, 4 complained of heart palpitations and shortness of breath. In another two studies totaling 128 industrial workers, no excesses of cardiac complaints or findings were noted. 1? Furthermore, in two contemporary epidemiological studies using similar cohorts from the Nitro, West Virginia, plant, no significant cardiac impairments were detected in exposed workers. f l However, one study found significantly lower levels of high density lipoprotein (HDL) cholesterol in 1 individuals with chloracne as contrasted to individuals without chloracne. Two recent clinical-epidemiological pilot studies of residential areas in Missouri contaminated by TCDD did not disclose any significant cardiac disease in exposed residents,1 ' 8 although the Times Beach study noted a borderline association of diminished peripheral pulses in the exposed group (as did the AFHS Baseline study). 15-1 �Because the herbicide literature has not identified consistent cardiovascular findings that merited a specific clinical focus, this study has collected generalized data on past cardiac events by questionnaire and medical record reviews. Current cardiac and peripheral vascular status were measured by physical examination and laboratory procedures. Coronary heart disease (CHD) has been of general concern in this study because both male cohorts are largely within the high risk ages of 40 to 65. Since TCDD probably does not directly and permanently affect cardiovascular function, a theoretical question that arises is whether TCDD might have altered a cardiovascular disease risk factor that will exert a future adverse impact. There may be indirect evidence for such a possibility. Risk factors for CHD include age, sex, race, family history, past personal history, diabetes (all types), smoking, cholesterol (and cholesterol-HDL ratio), diet, blood pressure, body weight, exercise pattern, stress (personality type), and alcohol. ~ Of these risk factors, hypertension and cholesterol have received consistent attention in clinical and epidemiological evaluations. Hypertension, either at routine examination or via specific study, has*not been related to phenoxy herbicide or TCDD exposure. However, hypercholesterolemia has been 13 6 24 2 associated with repeatedly acute exposure to chlorophenols and dioxin. °'12' ' ' ' Baseline Summary Results The 1982 Baseline examination found no statistically significant differences between the Ranch Hand and Comparison groups in systolic or diastolic blood pressure, the frequency of abnormal electrocardiographs (ECG's), heart sound abnormalities, abnormal funduscopic findings, or carotid bruits. However, a statistically significant difference emerged in the frequency of abnormal peripheral pulses: 12.8 percent of the nonblack Ranch Hands exhibited absent or diminished peripheral pulses compared to 9.4 percent of the nonblack Original Comparisons (p=0.05). This difference was consistent across various pulse combinations and remained statistically significant when all Ranch Hands were contrasted with all Comparisons, adjusting for age, past smoking history, and cholesterol level. No statistically significant differences were found between the two groups-in the occurrence of reported or verified heart disease or heart attacks, although a significant group-by-heart disease-by-smoking interaction was noted in the older (40 or more years of age) subgroup, i.e., older Ranch Hands smoking more than 10 pack-years developed more heart disease than their Comparisons, whereas older Ranch Hands smoking less than 10 pack-years exhibited less heart disease. No significant dose-response relationships of any of the cardiovascular response variables with the exposure index were noted. Over 80 percent of reported cardiac conditions obtained from the study questionnaire were verified by a detailed review of medical records. There was also strong correlation between the past medical history of cardiac disease and the Baseline cardiovascular examination findings. However, the differences in peripheral pulse abnormalities primarily occurred in older individuals without a history of cardiovascular disease. These abnormalities, therefore, may be a precursor to more serious arterial disease or central dysfunction. 15-2 �. Finally, the veil-known risk factors of age, smoking, and cholesterol were found to be highly correlated with each other and with several of the cardiovascular response variables. Parameters of the 1985 Cardiovascular Examination The 1985 cardiovascular examination was very similar to the 1982 Baseline examination. Data collection was divided into three major categories: heart disease history, central cardiac function, and peripheral vascular function. Historical data were collected by a questionnaire administered at the examination site, covering the interval from 1982 through 1985. In addition, the review-of-systerns portion of the physical examination recorded the overall history of heart trouble and other serious illnesses. Medical records were sought on all individuals to verify the reported conditions and to determine the time of occurrence of major cardiac events. Each participant was classified as to whether or not he developed essential hypertension, and whether he developed heart disease or had an acute myocardial infarction since his tour of duty in Southeast Asia (SEA). These endpoints were analyzed along with all other dependent variables to assess the degree of correlation between the history of cardiovascular disease and present medical findings. In addition, mortality findings were combined with the cardiovascular disease histories to form additional endpoints. . Central cardiac function was assessed by the measurements of systolic blood pressure, heart sounds (by auscultation), and an ECG. Blood pressure was determined in a standardized manner (see section on Physical Examination Data), and all examiners.and diagnosticians were retrained on the detection of fourth heart sounds and the notation of innocent murmurs without recording them as abnormal heart sounds. ECG's were obtained after adherence to a 4-hour fast and abstinence from tobacco. Twelve-lead ECG's were recorded with a rhythm strip, and the following items were considered to be abnormal: right bundle branch block (RBBB), left bundle branch block (LBBB), nonspecific T-wave changes, bradycardia, tachycardia, arrhythmia, and other diagnoses (e.g., A-V block, evidence of a prior myocardial infarction). Evaluation of the peripheral vascular system was based on diastolic blood pressure, funduscopic examination, auscultation of the carotid arteries, and determination of the quality of five peripheral pulses. The presence of carotid bruits was recorded in both carotid arteries. The femoral, popliteal, dorsalis pedis, posterior tibial, and radial pulses were assessed both by manual palpation and Doppler techniques because of the significant group differences discovered at the Baseline examination. Doppler results were considered the "gold standard" for the pulse measurements, although sensitivity correlations were established with palpation results. Rate changes of abnormal pulses occurring since the Baseline examination were also examined. In addition to the above dependent variables, considerable analytical attention was directed to the cardiovascular risk factors of age, race, occupation (OCC), and updated values for smoking history (pack-years [PACKYR], and current smoking level [CSMOK]), alcohol history (drink-years [DRKYR], and current drinking level [ALC]), cholesterol (CHOL), HDL, cholesterol-HDL ratio (CHOL/HDL), percent body fat (%BFAT), personality score (PS), and differential cortisol response (DIFCORT). 15-3 �Individuals with a verified history of diabetes (or those with an elevated 2-hour postprandial glucose level) were excluded from all analyses except the morbidity-mortality analysis. In addition, individuals with peripheral edema were excluded from analyses of the manual peripheral pulses because of the difficulty of measuring the pulse in the presence of edema. Logistic regression models were used for dichotomous variables, and general linear models for continuous variables. All covariates except race and occupation were treated as continuous variables. Due to the large number of covariates, analyses were carried out as follows. Models adjusting only for age, race, and occupation were examined first, followed by models incorporating group (GRP)-by-age, group-by-race, and group-by-occupation interactions. Analyses were then performed, adjusting for (1) all covariates and (2) all covariates, but with only one variable selected from among each of the sets: pack-years of smoking, current smoking; cholesterol, HDL, cholesterol-HDL ratio; and drink-years of alcohol, current alcohol intensity. Selection of the covariate from each set was based on examination of the pairwise covariate-by-dependent variable associations and the coefficient from the fully adjusted model. Stepwise modeling was then conducted using all covariates, but with only one variable selected from each of the sets described above. Only group-bycovariate interactions were examined, as were the three-factor interactions of group-by-age-by-race, group-by-age-by-occupation, and group-by-race-byoccupation. "Best models" refer to the models including only the statistically significant covariate and interaction terms. Minor numeric disparities in the tables that follow reflect missing dependent variable or covariate data. Parallel analyses using Original Comparisons can be found in Tables M-12 through M-20 of Appendix M. Morbidity and mortality data on the full Ranch Hand cohort and an appropriate Comparison cohort were tabulated for four endpoints: (1) death (any cause) or verified nonfatal heart disease, (2) death (any cause) or verified nonfatal myocardial infarction, (3) fatal or nonfatal verified heart disease, and (4) fatal or nonfatal verified myocardial infarction or fatal heart disease. This analysis involved a number of assumptions, particularly with respect to missing histories in the noncompliant study subjects. RESULTS AND DISCUSSION Questionnaire Data; Reported and Verified Heart Disease For each participant, a cardiovascular disease history was obtained from both the questionnaire and physical examination review of systems history. The baseline and third-year followup data were merged to determine, for each participant completing the third-year followup examination, whether there was ever a reported history of cardiovascular disease following service in Vietnam. Reported conditions were verified by medical record reviews and classified according to the ICD-9-CM. The following three variables were analyzed in terms of both reported and verified events: Variables ICD-9CM Codes Essential Hypertension 401 Heart Disease (Excluding Essential 391, 393-398, 402, 404 Hypertension) 410-414, 415-417, 420-429 Acute Myocardial Infarction 410 15-4 �Table 15-1 gives the unadjusted analysis of reported and verified cardiovascular disease in the Ranch Hand and Comparison groups and the results of unadjusted group contrasts. Essential hypertension was reported in slightly over 25 percent of the participants, with rates not significantly different in the two groups (p=0.596). About 80 percent of these cases were verified, leaving similar rates of 20.7 and 20.2 percent in the Ranch Hand and Comparison groups, respectively, for verified essential hypertension. Reported heart disease was a little higher in the Ranch Hand group (28.1% vs. 26.1%) but the difference in the percentage of verified heart disease was of borderline significance (23.8% vs. 20.3%, p=0.054). The rates of reported and verified myocardial infarctions were about 2 percent and 1 percent, respectively, and not significantly different in the two groups. The associations between each of the covariates and the three verified cardiovascular endpoints are presented in Tables 15-2, 15-3, and 15-4. The tables containing the covariate associations with the reported cardiovascular diseases are included in Tables M-l through M-3 of Appendix M. All reported cardiac illnesses (verified and unverified) are included in these tables. Many of the classic risk factors were identified. Age, smoking, cholesterol and/or cholesterol-HDL ratio, percent body fat, differential cortisol, and alcohol use were significantly associated with reported and verified essential hypertension, although the smoking effect was in the opposite direction of that expected. Age, occupation, and the cholesterol-HDL ratio were significantly associated with reported and verified heart disease, with more disease found in officers than in enlisted personnel. Age, pack-years of smoking, cholesterol-HDL ratio, and drink-years of alcohol were significantly associated with reported and/or verified myocardial infarction (the smoking effect being in the expected direction). The results of logistic regression analyses adjusting for these variables are presented in Table 15-5. The results were similar to the unadjusted results, but the adjusted ralative risk for verified heart disease reached statistical significance (p=0.036). No significant group-bycovariate interactions were noted. Nearly identical results were obtained in the analysis of the Ranch Hands and Original Comparisons (see Tables M-12 and M-13 of Appendix M). Morbidity-Mortality Analysis Differential mortality in the two groups could introduce bias in the analysis of morbidity data. For the cardiovascular evaluation, morbidity and mortality data on all Ranch Hands (diabetics included) and the first Comparison of the randomly ordered set matched to the Ranch Hands were combined to estimate the frequency of four hierarchical cardiovascular endpoints. Because of competing mortality and possible misclassification of the cause of death, the endpoints of death (any cause) or verified nonfatal heart disease, and death (any cause) or verified nonfatal myocardial infarction were examined to assess group differences in the most extreme case (i.e., all deaths being associated with cardiovascular disease). The other two endpoints were limited to fatal or nonfatal verified heart disease, and fatal or nonfatal verified myocardial infarction or fatal heart disease. The analysis was based on 1,257 Ranch Hands and 1,253 Comparisons. The history of each individual from the end of his tour of duty in SEA to the present was reviewed. Histories of verified heart disease and myocardial 15-5 �TABLE 15-1. Unadjusted Analyses for Reported and Verified Heart Disease by Group Group Ranch Hand Variable Statistic Number Percent Comparison Number Percent Est. Relative Risk ( 5 C.I.) 9% p-Value Reported Essential Hypertension n Yes No 942 247 695 26.2 73.8 1,206 304 902 25.2 74.8 1.05 (0.87,1.28) 0.596 Verified Essential Hypertension n Yes No 942 195 747 20.7 79.3 1,206 244 962 20.2 79.8 1.03 (0.83,1.27) 0.787 Reported Heart Disease (Excluding Hypertension) n Yes No 942 265 677 28.1 71.9 1,206 315 891 26.1 73.9 1.11 (0.91,1.34) 0.298 Verified Heart Disease (Excluding Hypertension) n Yes No 942 224 718 23.8 76.2 1,206 245 961 20.3 79.7 1.22 ( . 0 1 5 ) 10,.0 0.054 Reported Myocardial Infarction n Yes No 942 20 922 2.1 97.9 1,206 22 1,184 1.8 98.2 1.17 (0.63,2.15) 0.617 Verified Myocardial Infarction n Yes No 942 9 933 1.0 99.0 1,206 13 1,193 1.1 98.9 0.88 (0.38,2.08) 0.779 Ul a, �TABLE 15-2. Association Between Verified Essential Hypertension and the Covariates in the Combined Ranch Hand and Comparison Groups Covariate Covariate Category Total Percent Abnormal p-Value Age Born >1942 Born <1942 934 1,214 17.2 22.9 0.001 Race Black Nonblack 126 2,022 25.4 20.1 0.191 Occupation Officer Enlisted Flyer Enlisted Groundcrev 807 354 987 21.2 20.1 20.0 0.798 Current Smoking 0 >0 - 20 >20 1,262 463 422 22.8 16.6 17.5 0.005 Pack-Years Smoking 0 >0 - 10 >10 512 760 869 24.8 17.9 20.0 0.010 Cholesterol <200 >200 - 230 >230 766 650 732 15.5 21.1 25.0 <0.001 <40 >40 - 50 >50 719 754 675 21.6 20.6 19.1 0.524 717 743 688 16.2 21.3 24.0 0.001 10 1,758 379 0.0 16.7 38.5 <0.001 HDL Cholesterol-HDL <4.2 Ratio >4.2 - <5.5 >5.5 Percent Body Fat <10 10 - 25 >25 Personality Score <-5 -5-5 >5 829 731 580 22.3 20.5 17.8 0.113 Differential CortjLsol <0.6 >0.6 - 4.0 >4.0 704 745 683 23.6 19.1 18.4 0.033 Current Alcohol Use (Drinks/Day) 0 >0 - 1 >1 592 809 738 21.4 17.2 23.2 0.011 Drink-Years Alcohol <1.25 >1.25 - 25 >25 691 719 666 21.1 18.4 22.8 0.116 15-7 �TABLE 15-3. Association Between Verified Heart Disease and the Covariates in the Combined Ranch Hand and Comparison Groups Covariate Covariate Category Total Percent Abnormal p-Value Age Born >1942 Born <1942 934 1,214 17.9 24.9 <0.001 Race Black Nonblack 126 2,022 23.0 21.8 0.826 Occupation Officer Enlisted Flyer Enlisted Groundcrev 807 354 987 24.8 20.9 19.8 0.034 Current Smoking 0 >0 - 20 >20 1,262 463 422 22.7 21.2 19.9 0.461 Pack-Years Smoking 0 >0 - 10 >10 512 760 869 23.2 20.9 21.9 0.617 Cholesterol <200 >200 - 230 >230 766 650 732 21.2 21.1 23.2 0.533 <40 >40 - 50 >50 719 754 675 21.7 20.4 23.6 0.357 CholesterolHDL Ratio <4.2 >4.2 - <5.5 >5.5 717 743 688 24.1 18.8 22.7 0.041 Percent Body Pat <10 10 - 25 >25 10 1,758 379 30.0 22.1 20.3 0.619 Personality Score <-5 -5-5 >5 829 731 580 21.4 20.9 24.0 0.369 Differential Cortisol <0.6 >0.6 - 4.0 >4.0 704 745 683 19.2 22.4 24.0 0.084 Current Alcohol Use Drinks/Day 0 <1 >0 - 1 592 809 738 23.0 22.5 20.3 0.441 Drink-Years Alcohol <1.25 >1.25 - 25 >25 691 719 666 21.4 22.0 21.8 0.968 HDL 15-8 �TABLE 15-4. Association Between Verified Myocardial Infarction and the Covariates in the Combined Ranch Hand and Comparison Groups Covariate Covariate Category Total Percent Abnormal Age Born XL942 Born <1942 934 • 1,214 0.2 1.6 0.002 Race Black Nonblack 126 2,022 0.0 1.1 0.471 Occupation Officer Enlisted Flyer Enlisted Groundcrev 807 354 987 0.9 1.4 1.0 0.697 p-Value Current Smoking 0 >0 - 20 >20 1,262 463 422 0.8 1.7 1.0 0.228 Pack-years Smoking 0 >0 - 10 >10 512 760 869 0.2 0.8 1.7 0.018 Cholesterol <200 >200 - 230 >230 766 650 732 0.5 0.9 1.6 0.095 <40 >40 - 50 >50 719 754 675 1.5 0.9 0.6 0.210 CholesterolHDL Ratio <4.2 >4.2 - <5.5 >5.5 717 743 688 0.4 0.9 1.7 0.046 Percent Body Fat <10 10 - 25 >25 10 1,758 379 0.0 1.0 0.8 0.872 Personality Score <-5 -5-5 >5 829 731 580 0.8 1.5 0.7 0.278 Differential Cortisol <0.6 >0.6 - 4.0 >4.0 704 745 683 1.0 1.1 1.0 0.989 Current Alcohol Use (Drinks/Day) 0 >0 - 1 >1 592 809 738 1.5 0.9 0.8 0.376 Drink-Years Alcohol SI. 25 >1.25 - 25 >25 691 719 666 1.3 0.4 1.4 0.143 HDL 15-9 �TABLE 15-5. Adjusted Analyses for Reported and Verified Heart Disease Adj. Relative Risk (95% C.I.) 'p-Value Reported Essential Hypertension 1.14 (0.93,1.41) 0.211 AGE (p<0.001), CSMOK (p=0.001),CHOL (p<0.001), %BFAT (p<0.001), ALC (p<0.001) Verified Essential Hypertension 1.11 (0.89,1-39) 0.347 AGE (p=0.021), CSMOK (p=0.021), CHOL (p<0.001), %BFAT (p<0.001), PS (p=0.039) Reported Heart Disease 1.12 (0.92,1.36) 0.258 AGE (p<0.001) Verified Heart Disease 1.25 (1.02,1.54) 0.036 AGE (p<0.001) Reported Myocardial Infarction 1.16 (0.60,2.23) 0.667 AGE (p<0.001),OCC (p=0.014), CHOL/HDL (p=.0.016) Verified Myocardial Infarction 0.93 (0.38,2.23) 0.865 AGE (p<0.001), CHOL/HDL (p=0.025) Variable ^Abbreviations; CSMOK: CHOLs %BFAT: ALC: PS: OCC: CHOL/HDL: Current smoking Cholesterol Percent body fat Current alcohol use (drinks/day) Personality score Occupation Cholesterol-HDL ratio 15-10 Covariate Remarks* �infarction for living individuals who were noncompliant at Baseline and at the followup were missing. For the living noncompliant individuals, the observed rate in the compliant individuals was used to estimate the number of nonfatal events among the noncompliant individuals for each cohort. It was assumed that there were no nonfatal cardiovascular events in the noncompliant individuals who died due to a cause other than cardiovascular system failure. The results are shown in Table M-4 of Appendix M. There was a total of 66 deaths in the Ranch Hand group and 77 in the group of Comparisons. The estimated percentage of Ranch Hands who died (any cause) or had a verified nonfatal history of heart disease was 27.4 as contrasted to 24.5 in the Comparisons. The rate of verified nonfatal myocardial infarctions was approximately 1 percent in each group. The estimated percentage of deaths (any cause) or verified nonfatal myocardial infarction was 6.4 percent in the Ranch Hands and 7.0 percent in the Comparisons. Only 5 of the 66 deaths in the Ranch Hands and 3 of the 77 deaths in the Comparisons either were from heart disease, or were individuals who had verified heart disease histories. The estimated percentage of fatal and nonfatal verified heart disease was 22.5 percent in the Ranch Hands and 18.6 percent in the Comparisons. Of the 66 deaths in the Ranch Hands only 1 individual died from cardiovascular disease or had a verified history of myocardial infarction as compared to 2 of the 77 deaths in the Comparisons. The estimated percentage of fatal or nonfatal verified myocardial infarction or fatal heart disease was 1.2 percent in the Ranch Hands and 1.0 percent in the Comparisons. These contrasts must be interpreted guardedly since they involve some unverifiable assumptions. Nevertheless, they are consistent with the morbidity findings presented in the chapter, and tend to show that the clinical cardiovascular disease spectrum is approximately equal in both groups. Physical Examination Data Central Cardiac Function Central cardiac function was assessed by the measurement of systolic blood pressure, heart sounds, and an ECG. Systolic blood pressure was determined by a standardized sphygmometer, at the appearance of the first sound with the ndndominant arm placed at heart level; the lowest value of three readings was recorded. Detection of abnormal heart sounds was conducted by standard auscultation with the participant placed in sitting, supine, and left lateral supine positions. Fourth heart sounds were assessed; murmurs were graded in intensity and location and were judged to be functional (normal) or organic (abnormal) in nature. Fourth heart sounds were scored as abnormal. ECG data were collected by a standardized 12-lead machine; approximately 95 percent of the clinical interpretations were performed by one cardiologist. All participants were asked to abstain from smoking for at least 4 hours prior to their ECG. 15-11 �Systolic Blood Pressure Systolic blood pressure was analyzed both as continuous and dichotomized variables (normal, 140 or less mm Hg; abnormal, more than 140 mm Hg). Combined distributional data from both groups revealed significant digit preference for values ending in zero (p<0.0001 for both systolic and diastolic readings), but standard statistical analyses were performed since the zero-digit peaks (e.g., 130, 140, 150 mm Hg) were relatively uniform and did not visually differ between the Ranch Hand and Comparison groups. Zero digit readings were recorded for 59.4 percent of the systolic blood pressures and 55.0 percent of the diastolic blood pressures. Table 15-6 gives the percentage of participants with abnormally high systolic values. The percent of abnormals was not significantly different from each other (p=0.529). Systolic blood pressure, analyzed as a continuous variable, had a mean of 118.96 mm Hg (95% C.I.: [118.06,119.86]) for the Ranch Hand group and a mean of 119.55 mm Hg (95% C.I.: [118.71,120.39]) for the Comparison group. These means were not significantly different (p=0.349). The means were also not significantly different when Original Comparisons were used (p=0.182). The association between each of the covariates (categorized into either two or three levels) and dichotomized systolic blood pressure in the combined Ranch Hand and Comparison groups is shown in Table 15-7. Age, cholesterol, percent body fat, personality score, and alcohol use (both current use and drink-years) were significantly associated with increased systolic pressure. These covariate effects were in the direction typically found in other studies,2 ' except for personality score where those participants with low scores (in the Type B direction) had the highest percentage of abnormal values. Adjustment of the categorical systolic blood pressure by the above covariates was performed by logistic regression analysis, and these results are presented in Table 15-8. As shown, there were no significant differences between the Ranch Hand and Comparison groups (p=0.920). Age, cholesterol, percent body fat, personality score, and current alcohol use all had statistically significant effects. An adjusted analysis of systolic blood pressure in the continuous form revealed a significant group (GRP)-by-ageby-race interaction (p=0.012) along with the significant main effects of current smoking (p<0.001), cholesterol (p<0.001), percent body fat (p<0.001), personality score (p<0.001), and current alcohol use (p=0.002). Exploration of the interaction revealed that among Blacks there was a group-by-age interaction (p=0.007), with a mean systolic pressure greater in the Ranch Hand group than in the Comparison group at the younger age levels, but lower at the older age levels. The estimated Ranch Hand-Comparison difference was 4.56 (± 3.30) mm Hg at the Baseline age of 35 and -16.01 (± 5.87) mm Hg at the Baseline age of 53 (see Table M-5 of Appendix M). In the nonblack cohort the group-by-age interaction was not significant (p=0.338), nor was there evidence of any overall group effect (p=0.356). In the analysis of the Ranch Hands and Original Comparisons, there were no statistically significant group differences, either unadjusted or adjusted for covariate effects (see Tables M-14 and M-15 of Appendix M). 15-12 �TABLE 15-6. Unadjusted Analyses for Central Cardiac Function By Group (Diabetics Excluded) Group Ranch Hand Statistic Systolic Blood Pressure n Abnormal Normal 942 60 882 6.4 93.6 1,205 85 1,120 7.1 92.9 0.90 (0.64, 1.26) 0.529 Heart Sounds n Abnormal Normal 941 31 910 3.3 96.7 1,206 32 1174 2.7 97.3 1.25 (0.76,2.06) 0.384 ECG (Overall) n Abnormal Normal 947 121 821 12.8 87.2 1,206 169 1,037 14.0 86.0 0.90 (0.70,1.16) 0.430 ECG: n' Abnormal Normal 942 5 937 0.5 99.5 1,206 9 1,197 0.7 99.3 0.71 (0.24,2.13) 0.542 n Abnormal Normal 942 0 942 0.0 100.0 1,206 0 1,206 0.0 100.0 n Abnormal Normal 942 85 857 9.0 91.0 1,206 107 1,099 8.9 91.1 i Percent Number Est. Relative Risk (95% C.I.) Variable Ul Number Comparison Percent p-Value u> ECG: RBBB LBBB ECG: Nonspecific T-¥ave Changes — 1.02 (0.76,1.37) 0.904 �TABLE 15-6. (continued) Unadjusted Analyses for Central Cardiac Function By Group (Diabetics Excluded) Group Ranch Hand n Abnormal Normal 942 45 897 4.8 95.2 1,206 56 1,150 4.6 95.4 n Abnormal Normal 942 0 942 0.0 100.0 1,206 0 1,206 0.0 100.0 n Abnormal Normal 942 31 911 3.3 96.7 1,206 41 1,165 3.4 96.6 0.97 (0.60,1.55) 0.889 n Abnormal Normal 942 97 845 10.3 89.7 1,206 132 1,074 11.0 89.0 0.93 (0.71,1.23) 0.631 ECG: Tachycardia ECG: Arrhythmia ECG: Other Diagnoses —No relative risk given, since no abnormals are present. Percent Est. Relative Risk (95% C.I.) Statistic ECG: Bradycardia I Percent Number Variable Ui Number Comparison 1.03 (0.69,1.54) p-Value 0.889 — �T6BUS15-7. Association Between Central Cardiac Flncticn Variables and the Govariates in the Gombinad Ranch Bawl and Comparison GEOUIIS (Diabetics Excluded) Current Pack-Years Race Occupation Snaking Smoking Percent Cholesterol- Body Personality Cholesterol HDL HDL Ratio Fat Score Differential Cortisol Current Alcohol Use (Drinks Drink-Years per Day) Alcohol Variable Systolic Blood Pressure 00.001 NS NS NS NS 00.001 NS NS 0.001 0.002 NS 0.018 00.001 Heart Sounds 0.005 NS NS* NS NS NS NS* 0.003 NS NS NS NS NS BOG (Overall) f t I K Age NS NS* NS 0.010 NS* NS* 0.016 00.001 NS NS NS NS N S NS NS NS BOG: RBBB 00.001 N S N S N S N S N S N S N S N S N S BOG: 00.001 Nonspecific T-Wave Changes NS NS NS 0.038 0.002 NS <D.001 00.001 NS NS NS 0.006 BOG: Bradycardia NS* NS 0.010 NS 0.007 NS* 0.002 00.001 NS NS NS NS, NS BOG: Arrhythmia NS NS 0.023 NS 0.028 NS NS NS NS NS NS NS NS BOG: Other Diagnoses 00.001 NS 0.011 NS* 0.023 NS NS NS NS NS NS 0.019 NS NS: Not significant (pX).10). NS*: Borderline significant (0.05 <p<D.10). �TABLE 15-8. Adjusted Analyses for Central Cardiac Function (Diabetics Excluded)* Variable Systolic Blood Pressure (Discrete) Systolic Blood Pressure (Continuous) Heart Sounds Adj. Relative Risk (95% C.I.) p-Value Covariate Remarks* 0.98 (0.69,1.40) 0.920 AGE (p<0.001) CHOL (p=0.004) 2BFAT (p<0.001) PS (p=0.002) ALC (p=0.020) **** **** GRP*RACE*AGE (p=0.012) CSMOK (p<0.001) CHOL (p<0.001) XBFAT (p<0.001) PS (p<0.001) ALC (p=0.002) 1.33 (0.80,2.24) 0.276 ECG (Overall) **** **** ECG: RBBB 0.72 (0.24,2.15) 0.555 AGE (p»0.008) ECG: Nonspecific ST-T-Wave Changes 1.12 (0.81,1.53) 0.497 AGE (p<0.001) RACE (p=0.005) CHOL (p=0.007) %BFAT (p<0.001) 15-16 AGE (p<0.001) RACE (p=0.003) CHOL/HDL (p=0.002) AGE (p<0.001) RACE (p=0.005) %BFAT (p<0.001) GRP*PACKYR (p=0.008) �TABLE 15-8. (continued) Adjusted Analyses for Central Cardiac Function (Diabetics Excluded)* Variable Adj. Relative Risk (95% C.I.) p-Value ECG: Bradycardia 1.08 (0.72,1-62) 0.726 ECG: Arrhythmia **** **** ECG: Other Diagnoses 0.92 (0.69,1.23) 0.575 Covariate Remarks** OCC (p=0.047) CHOL/HDL (p<0.001) AGE (p=0.001) OCC (p<0.001) GRP*PACKYR (p=0.018) GRP*%BFAT (p=0.038) AGE (p<0.001) RACE (p-0.015) CSMOK (p=0.039) *Some adjusted analyses did not explore effects of all covariates due to sparse number of abnormalities (see text). **Additional Abbreviations; GRP: group PACKYR: pack-years smoking ****Group-by-covariate interaction, relative risk/difference in group means, 95% confidence interval, and p-value not presented (see Table M-5 of Appendix M). 15-17 �Heart Sounds As shown in Table 15-6, the unadjusted frequency of abnormal heart sounds in the two groups was not significantly different (p=0.384). The covariate tests of association (Table 15-7) showed significant effects for age (p=0.005), cholesterol-HDL ratio (p=0.003), and a borderline association with occupation (p=0.069). Increased age (born before 1942)'had a frequency of 3.9 percent heart sound abnormalities as contrasted to 1.6 percent abnormalities in the younger age group (born in or after 1942). The cholesterol-HDL ratio (less than or equal to 4.2, between 4.2 and 5.5, and greater than or equal to 5.5) was positively associated with increasing frequencies of abnormal heart sounds (1.7, 2.6, and 4.7 percent, respectively). The observed frequencies of abnormal heart sounds were 3.8, 1.4, and 2.7 percent in the officers, enlisted flyers, and the enlisted groundcrew, respectively. The adjusted analysis (Table 15-8) did not detect any significant group differences (p=0.276). Age, race, and the cholesterol-HDL ratio were significant covariates (p<0.001, p=0.003, and p=0.002, respectively). No two- or three-way group interactions were noted. Similarly, nonsignificant results were found in the analyses of the Original Comparisons versus the Ranch Hands (see Table M-15 of Appendix M). Electrocardiograph Findings All EGG tracings were scored as normal or abnormal; specific abnormalities included RBBB, LBBB, nonspecific T-wave changes, bradycardia, tachycardia, arrhythmia, and other diagnoses. The unadjusted analysis of these variables (Table 15-6) showed no statistically significant differences in the overall ECG results, or any of the specific subcategories, between the Ranch Hand and Comparison groups. Two additional findings in the analysis were of interest: (1) the Ranch Hands had a uniformly lower number of ECG abnormalities than the Comparisons (though not statistically significant), and (2) the sum of the specific ECG findings exceeds the proportion of abnormalities scored on the overall ECG because some individuals accounted for two or more abnormalities. The associations between the covariates and the various ECG findings are presented in Table 15-7. Age was significantly associated with the overall ECG findings (p<0.001), nonspecific T-wave changes (p<0.001), and other ECG diagnoses (p<0.001), with more abnormalities found in the older age group. Occupation was significantly associated with bradycardia (p=0.010), arrhythmia (p=0.023), and other ECG findings (p=0.011). A higher percentage of officers than enlisted flyers or groundcrew had bradycardia, whereas enlisted flyers had the lowest proportion of arrhythmias, and enlisted groundcrew had the highest percentage. Officers and enlisted flyers had a higher percentage than the enlisted groundcrew cohort of other ECG findings. Pack-years of smoking was significantly associated with the overall ECG findings (p=0.010), T-wave changes (p=0.038), bradycardia (p=0.007), arrhythmia (p=0.028), and other ECG diagnoses (p=0.023). For the overall ECG findings, nonspecific T-wave changes, and arrhythmias, the moderate smoking group (greater than 0 to 10 pack-years) had the fewest abnormalities. 15-18 �Bradycardia was negatively associated with pack-years of smoking, with the highest frequency of abnormalities (7.2%) found in the 0 pack-years category versus the lowest proportion (3.6%) of abnormalities in the greater than 10 pack-years category. Cholesterol levels and/or the cholesterol-HDL ratio were positively associated with abnormalities in the overall ECG (p=0.016) and T-wave findings (p<0.001), but were negatively associated with bradycardia (p<0.001). Increased percent body fat was significantly associated with overall ECG abnormalities (p<0.001) and nonspecific T-wave changes (p<0.001). Drinkyears of alcohol was only associated with T-wave changes (p=0.006), with more abnormalities in the greater than 25 drink-years category than in the less than or equal to 1.25 drink-years category, but relatively fewer abnormalities in the more than 1.25 to 25 drink-years category. The covariate of current alcohol use was associated only with the category of other ECG diagnoses (p=0.019), but not in a consistent manner (individuals averaging less than one drink per day had more abnormalities than nondrinkers, but those averaging more than one drink per day had the lowest percentage of abnormalities). The covariates of race, current smoking, personality score, and differential cortisol level, however, did not significantly affect the variables of central cardiac function. Results from the adjusted logistic regression analyses are shown in Table 15-8, No significant group differences were detected for categorical RBBB, T-wave changes, bradycardia, and other ECG diagnoses. The covariates of age, race, percent body fat, pack-years of smoking, current smoking, cholesterol, and cholesterol-HDL ratio were significantly associated with one or more of the ECG variables. RBBB was adjusted only for age due to the small number of abnormalities. The adjusted analysis of the overall ECG findings revealed a significant group-by-pack-year interaction (p=0.008), and the analysis of the arrhythmia variable disclosed two significant interactions: a group-by-pack-year association (p=0.018) and a group-by-percent body fat association (p=0.038). All of these interactions are displayed in Table M-5 of Appendix M. In the case of the overall ECG findings, the adjusted relative risk among nonsmokers was significantly less than one (p=0.038), i.e., a lower risk for Ranch Hands than Comparisons. For heavy smokers (30 pack-years), the adjusted relative risk was 1.25 (95% C.I.: [0.89,1.76], p=0.197). For cardiac arrhythmias, exploration of the group-by-pack-year interaction at the approximate mean percent body fat of 21 percent showed a borderline significant relationship favoring the nonsmoking Ranch Hands (Adj. RR: 0.58, 95% C.I.: [0.30,1.10], p=0.093); heavy smoking Ranch Hands had a higher proportion of arrhythmias than heavy smoking Comparisons, but this association was not statistically significant (p=0.162). For the group-by-percent body fat interaction, 10 percent and 30 percent body fat levels were analyzed at the approximate median of 7 pack-years of smoking. The adjusted relative risk of "0.23 (95% C.I.: [0.07,0.78]) was statistically significant for the 10 percent body fat category (p=0.018), indicating a lower adjusted frequency of cardiac arrhythmias for nonobese Ranch Hands than for nonobese Comparisons, This situation was reversed for obese Ranch Hands, but the association was not statistically significant (RR: 1.88, 95% C.I.: [0.66,5.34], p=0.234). The adjusted analyses using the Original Comparisons were nearly identical to the analyses of the total Comparison group, including the three 15-19 �group interactions for overall EGG findings and cardiac arrhythmias described above. The analyses of the Original Comparison group are found in Tables M-15 and M-16 of Appendix M. Peripheral Vascular Function Peripheral vascular function was assessed by the diastolic blood pressure, funduscopic examination of small vessels, the presence or absence of carotid bruits,, and both manual palpation and Doppler bilateral measurements of the radial, femoral, popliteal, dorsalis pedis, and posterior tibial pulses. Individual peripheral pulses were combined to form overall indices of peripheral vascular status. Diastolic blood pressure was measured by the standard auscultatory technique, and was recorded at the pressure level corresponding to the disappearance of sound. The funduscopic examination was conducted with undilated pupils in a standard manner, with emphasis placed upon the detection of arterio-venous nicking, hemorrhages, exudate, and papilledema. Carotid bruits were assessed by standard bilateral auscultation; confirmation of bruits was not attempted by the Doppler technique. Manual pulse determinations were performed by the examining physician, independent of the Doppler measurements performed by qualified technicians. Tobacco abstinence for at least four hours was required for the Doppler examination, but not for the manual palpation. Only the physician diagnostician had access to both sets of pulse data. Diastolic Blood Pressure Diastolic blood pressure was analyzed as a continuous variable and as a dichotomized variable (normal value less than or equal to 90 mm Hg; abnormal value greater than 90 mm Hg). As with the systolic readings, a significant zero digit preference was noted for the diastolic blood pressure values. Table 15-9 arrays the results of the unadjusted categorical analyses. As shown, there are no statistically significant group differences for the proportions of diastolic abnormalities (p=0.999). Diastolic blood pressure, analyzed as a continuous variable, had a mean of 79.76 mm Hg (95% C.I.: [71.97, 80.35]) for the Ranch Hand group and a mean of 79.77 mm Hg (95% C.I.: [79.24, 80.30]) for the Comparison Group. These means were not significantly different (p=0.986). The means were also not significantly different when Original Comparisons were used (p=0.555). The tests of covariate association with diastolic blood pressure are given in Table 15-10. Cholesterol, cholesterol-HDL ratio, percent body fat, differential cortisol, and current alcohol use were significantly related to diastolic blood pressure (p<0.001, p=0.006, p<0.001, p=0.041, and p=O.OU, respectively). For increasing cholesterol, cholesterol-HDL ratio, and percent body fat, increases in proportions of abnormal diastolic blood pressure were obtained, whereas for increasing differential cortisol values, a decline in blood pressure abnormalities was found. Current alcohol use (drinks per day) revealed an inconsistent association with diastolic blood pressure abnormalities, with nondrinkers having a higher proportion of abnormalities than low-level drinkers, but a lower proportion of abnormalities than moderate drinkers (8.3, 6.4, and 10.6 percent abnormalities, respectively). The covariates of age, race, occupation, current smoking, pack-years of smoking, HDL, personality score, and drink-years of alcohol were not associated with diastolic blood pressure abnormalities. 15-20 �TABLE 15-9. Unadjusted Analyses for Peripheral Vascular Function by Group (Diabetics Excluded) Group Ranch Hand Variable Statistic Number Percent Comparison Number Percent Est. Relative Risk (95% C.I.) p-Value Diastolic Blood Pressure n Abnormal Normal 942 79 863 8.4 91.6 1,204 101 1,103 8.4 91.6 1.00 (0.74,1.36) 0.999 Funduscopic Examination n Abnormal Normal 941 7 934 0.7 99.3 1,206 6 1,200 0.5 99.5 1.50 (0.50,4.47) 0.472 Carotid Bruits n Abnormal Normal 941 7 934 0.7 99.3 1,205 7 1,198 0.6 99.4 1.28 (0.45,3.66) 0.646 Radial Pulses (Manual) n Abnormal Normal 929 4 925 0.4 99.6 1,191 8 1,183 0.7 99.3 0.64 (0.19,2.13) 0.465 Radial Pulses (Doppler) n Abnormal Normal 942 3 939 0.3 99.7 1,203 4 1,199 0.3 99.7 0.96 (0.21,4.30) 0.952 Femoral Pulses (Manual) n Abnormal Normal 929 20 909 2.2 97.8 1,191 25 1,166 2.1 97.9 1.03 (0.57,1.86) 0.932 ,_, In 1 £ • �TABLE 15-9. (continued) Unadjusted Analyses for Peripheral Vascular Function by Group (Diabetics Excluded) Group Ranch Hand Variable Statistic Comparison Number Percent Number Percent Est. Relative Risk (95% C.I.) p-Value Femoral Pulses (Doppler) Ln I n Abnormal Normal 942 6 936 0.6 99.4 1,205 4 1,201 0.3 99 .7 1.92 (0.54,6.82) 0.312 Popliteal Pulses (Manual) n Abnormal Normal 929 16 913 1.7 98.3 1,191 28 163 1, 2.4 97 .6 0.73 (0.39,1.35) 0.317 Popliteal Pulses (Doppler) n Abnormal Normal 942 10 932 1. 1 98.9 1,204 8 1,196 0.7 99 •3 1.60 (0.63,4.08) 0.322 Dorsalis n Pedis Pulses Abnormal (Manual) Normal 929 102 827 11.0 89.0 1,191 127 1,064 10 ,7 89 .3 1.03 (0.78,1.36) 0.818 n Dorsalis Pedis Pulses Abnormal (Doppler) Normal 938 228 710 24.3 75.7 1,202 274 928 22 .8 77 .2 1.09 (0.89,1.33) 0.412 Posterior n Tibial Pulses Abnormal (Manual) Normal 929 27 902 2.9 97.1 1,191 31 160 1, 2.6 97 .4 1.12 (0.66,1.89) 0.674 Posterior n Tibial Pulses Abnormal (Doppler) Normal 939 19 920 2.0 98.0 1,202 25 1,177 2.1 97 .9 0.97 (0.53,1.78) 0.928 �TABLE 15-9. (continued) Unadjusted Analyses for Peripheral Vascular Function by Group, (Diabetics Excluded) Group Ranch Hand Variable Statistic Number Percent Comparison Number Percent Est. Relative Risk (95% C.I.) p-Value Leg Pulses (Manual) 1_l n Abnormal Normal 929 131 798 14.1 85.9 1,191 177 1,014 14.9 85.1 0.94 (0.74,1.20) 0.624 Leg Pulses (Doppler) n Abnormal Normal 938 237 701 25.3 74.7 1,202 288 914 24.0 76.0 1.07 (0.88,1.31) 0.490 Peripheral Pulses (Manual) n Abnormal Normal 929 133 796 14.3 85.7 1,191 181 1,010 15.2 84.8 0.93 (0.73,1.18) 0.575 Peripheral Pulses (Doppler) n Abnormal Normal 938 239 699 25.5 74.5 1,202 290 912 24.1 75.9 1.08 (0.88,1.31) 0.472 All Pulses (Manual) n Abnormal Normal 929 133 796 14.3 85.7 1,191 182 1,009 15.3 84.7 0.93 (0.73,1.18) 0.535 All Pulses (Doppler) n Abnormal Normal 938 239 699 25.5 74.5 1,201 291 910 24.2 75.8 1.07 (0.88,1.30) 0.509 Ol i £ �TABLE 15-10. in the Combined Ranch Band and Comparison Groups (Diabetics Excluded) Variable Diastolic Blood Pressure NS NS Funduscopic Examination Carotid Bruits Ln Age Current Pack-Years Race Occupation inking Snaking Percent Cholesterol- Body Personality Differential Cortisol Cholesterol HDL HDL Ratio Fat Score Current Alcohol Use (Drinks Drink-Years per Day) Alcohol NS 0.006 O.C01 NS 0.041 0.014 NS NS 0.016 0.026 NS NS 0.004 NS NS* NS NS NS NS NS NS 0.021 NS NS NS 0.033 NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS* 0.002 O.001 0.001 NS NS NS <D.001 NS NS 0.001 0.006 NS* NS 0.019 NS NS NS NS 0.013 NS NS NS 0.001 NS NS 0.002 NS NS NS NS NS* 0.002 NS NS O.001 0.010 NS NS NS* NS NS NS NS NS NS NS NS 0.004 0.040 NS NS NS MS NS* NS NS NS* NS Radial Pulses NS (Manual) NS NS NS Radial Pulses NS (Doppler) NS NS Femoral Pulses (Manual) 40.001 NS Femoral Pulses (Doppler) NS* Popliteal Pulses (Manual) NS* Popliteal Pulses (Doppler) O.001 �TABLE 15-10. (contiiued) Association Between Peripheral Vascular Function feriables and the Covariates in tie Combined Band) Hand and Comparison Groins (Diabetics Excluded) Current Pack-Years Race Occupation Soaking Snaking Percent Current Alcohol Cholesterol- Body Personality Differential Use (Drinks Drink-Years Cholesterol BDL BEL Ratio Fat Score Cortisol per Day) Alcohol Variable Dorsal is Pedis Pulses (Manual) ^_t Age 0.018 NS NS* NS NS NS NS NS NS NS NS MS NS Dorsalis Pedis Pulses (Doppler) US* 0.007 NS NS MS MS MS MS MS MS NS NS Posterior 0.034 <0.001 NS <0.001 O.001 NS* NS NS* NS NS NS NS NS Posterior Tibial Pulses (Doppler) 0.003 NS* NS 0.021 NS NS NS 0.035 NS 0.028 NS* NS NS* Leg Pulses (Manual) 0.001 NS NS NS 0.031 NS 0.013 0.013 NS NS NS NS NS NS NS NS NS 0.001 Oi 1 KJ Ui Tibial Pulses (tonal) L e g Pulses (Doppler) Peripheral Pulses (Manual) 0.020 0.009 O.001 NS 0.012 NS N NS S N S N 0.028 NS S N S 0.013 N S 0.010 N N5 S N S NS NS NS �TWE 35-10. (ontiiuBd) in the Ocpihin^ R^vh Hand and Gwpiri.sm Grtmps (Diabetics RKrhpM) Current Pack-Years Race Occupation Sacking Smoking Percent Cholesterol- Body Personality Cholesterol HDL HDL Ratio Fat Score Variable Peripheral Pulses (Doppler) 0.037 0.015 0.019 NS NS N All Pulses (Manual) i Age O.C01NS NS NS 0.023 NS 0.013 All Pulses (Doppler) 0.032 0.014 0.023 NS NS NS NS 1 0 NS: Not significant (pX).10). NS*: Borderline significant (0.05<p<D.10). S N S N Differential Cortisol Current Alcohol Use (Drinks Drink-Years per Day) Alcohol S NS NS NS NS 0.012 NS NS NS NS NS NS NS NS NS NS NS S N �The adjusted categorical and continuous analyses are shown in Table 15-11. No significant group differences were found in the proportions of diastolic abnormalities (p=0.653), or in the difference of group mean values (p=0.299). The covariates of current smoking, cholesterol, percent body fat, and current alcohol use were statistically significant in both the categorical and continuous analyses; age was significant only in the analysis of group mean differences (p=0.005). No significant group-by-covariate interactions were found in either the logistic regression or general linear models. The adjusted analyses for the Original Comparisons were very similar to those described on the total Comparison group (see Table M-18 of Appendix M). Funduscopic Examination The funduscopic examination detected only 13 individuals with arteriovenous nicking (a sign of chronic blood pressure elevation) or vessel hemorrhages, 7 from the Ranch Hands and 6 from the Comparisons (p=0.472, Table 15-9). The covariate tests of association are given in Table 15-10.. Age, race, cholesterol/HDL ratio, percent body fat, and current drinking were statistically significant (p=0.004, 0.040, 0.016, 0.026, and 0.004, respectively). All funduscopic abnormalities were found in the older age group (born in or before 1942). Blacks had a higher proportion of abnormalities than nonblacks (2.4 percent versus 0.5 percent, respectively). The highest cholesterol-HDL category contained the highest proportion of funduscopic abnormalities; and increasing levels of percent body fat were associated with increases in proportions of abnormalities. Current alcohol consumption showed that nondrinkers had the highest proportion of abnormalities. The covariates of occupation, current smoking, pack-years of smoking, cholesterol, HDL, personality score, differential cortisol and drink-years of alcohol did not show significant effects. In the adjusted analysis by logistic regression (Table 15-11), there were no significant differences in funduscopic abnormalities between the Ranch Hand and Comparison groups (Adj. RR: 1.78; 95% C.I.: [0.56,5.62], p=0.322). Due to sparse data the model was adjusted only for the covariates of age, race, cholesterol-HDL ratio, percent body fat, and current alcohol consumption; and all were significant in the model. No group interactions were detected, and the results of the contrast of the Ranch Hand with the Original Comparison group were also nonsignificant (Table M-18 of Appendix M). Carotid Bruits The unadjusted group contrast of carotid bruits is displayed in Table 15-9. The proportions of bruits in both groups were similar (Est. RR: 1.28, 95% C.I.: [0.45,3.66], p=0.646). Overall, only 14 bruits were detected, 7 from each group, limiting the scope of the adjusted analyses. The covariate effects are given in Table 15-10. Age, current smoking, and cholesterol were of borderline statistical significance, whereas drinkyears of alcohol was significantly correlated with carotid bruits (p=0.021), with the greater than 25 drink-years category having the highest proportion. 15-27 �TABLE 15-11. Adjusted Analysis for Peripheral Vascular Function by Group (Diabetics Excluded)* Statistical/ Clinical Analysis Adj. Relative Risk (95* C.I.) p-Value Discrete 1.08 (0.78,1-48) 0.653 CSMOK (p-0. 040) CHOL (p<0.001) XBFAT (p<0.001) ALC (p-0. 008) Continuous 0.38 (-0.34, 1.11)' 0.299* AGE (p.O.005) CSMOK (p<0.001) CHOL (p<0.001) %BFAT (p<0.001) ALC (p.O.002) Funduscopic Examination 1.78 (0.56,5.62) 0.322 AGE (p<0.001) RACE (p<0.001) CHOL/HDL (p.O.017) XBFAT (p.O.037) ALC (p.O.038) Carotid Bruits 1.05 (0.35,3.16) 0.928 AGE.(p.O.024) DRKYR (p<0.001) Variable Diastolic Blood Pressure Covariate Remarks** Radial Pulses Manual Ooppler 0.64 (0.19,2.14) 0.96 (0.21,4.30)* 0.472h 0.952 AGE (p.O.040) Femoral Pulses Manual 1.21 (0.63,2.31) 0.562 AGE (p<0.001) CHOL/HDL (p-0.010) *BFAT (p<0.001) DIFCORT (p.O.002) Doppler 1.74 (0.48,6.31) 0.401 AGE (p.O.001) CSMOK (p.O.001) CHOL/HDL (p-0.042) Manual **** **** AGE (p-0.003) PACKYR (p«0.005) CHOL/HDL (p-0.Oil) GRP*RACE (p.O.038) Doppler 1.50 (0.58,3.91) 0.401 AGE (p<0.001) RACE (p.O.023) CSMOK (p<0.001) Manual **** **** AGE (p-0.004) DRKYR (p-0.038) GRP*OCC (p.O.046) Doppler 1.07 (0.87,1.31) 0.535 AGE (p.O.004) RACE (p.O,006) ZBFAT (p-0.003) Popliteal Pulses Dorsalis Pedis Pulses 15-28 �TABLE 15-11. (continued) Adjusted Analysis for Peripheral Vascular Function by Group (Diabetics Excluded)* Variable Statistical/ Clinical Analysis Adj. Relative Risk (95* C.I.) p-Value Covariate Remarks** AGE (p<0.001) RACE (p<0.001) PACKYR (p-0. 007) GRP*OCC (p-0. 01 7) 0.94 (0.50,1.77) 0.849 AGE (p<0.001) RACE (p.O.002) CSMOK (p.O.007) CHOL/HDL (p.O.015) Manual **** **** AGE (p<0.001) GRP*OCC (p.O.016) GRP**BFAT (p.O.034) 1.06 (0.87,1-30) 0.549 AGE (p-0.001) RACE (p.O.029) XBFAT (p-0.006) Manual **** **** AGE (p<0.001) XBFAT (p.O.018) GRP*OCC (p.O.033) 1.06 (0.87,1.30) 0.562 AGE (p.O.001) ZBFAT (p.O.006) Manual **** **** AGE (p<0.001) ZBFAT (p.O.022) GRP*OCC (p.O.036) Doppler All Pulses **** Ooppler Peripheral Pulses **** Doppler Leg Pulses Manual Doppler Posterior Tibial Pulses 1.06 (0.86,1.29) 0.603 AGE (p.O.001) ZBFAT (p.O.006) *Sonte adjusted analyses did not explore effects of all covariates due to sparse number of abnormalities (see text). **Additional Abbreviations: DRKYRsdrink-years of alcohol aDIFCORT: differential cortisol. "Difference in group means (Ranch Hand-Comparison) and associated p-value given, rather than relative risk, for continuous analysis of dependent variables. Unadjusted for any covariates—same results as for unadjusted analysis. ****Group-by-covariate interaction—relative risk, confidence interval, and p-value not presented (see Table M-6 of Appendix M). 15-29 �The adjusted analysis was performed with only the covariates of age and drink-years of alcohol due to the small number of detected bruits. The results (Table 15-11) demonstrate a lack of significant group differences (Adj. RR: 1.05, 95% C.I.: 0.35, 3.16], p=0.928). Both age and drink-years of alcohol were significant adjusting variables, but no significant group interactions were noted. The results of the Ranch Hand, Original Comparison group contrast was also nonsignificant (see Table M-18 of Appendix M). Peripheral Pulses Five peripheral pulses (radial, femoral, popliteal, dorsalis pedis, and posterior tibial) were analyzed using data assessments from both manual palpation and Doppler recordings. Palpation data from the examining physician were judged abnormal if the pulse was diminished or absent on either side. Assessment of the Doppler data was more complex and involved visual examination of the waveform morphology (pulsatility, systolic forward flow, and diastolic reverse flow) on analog strips and Polaroid® photographs, with careful comparison of the laterality of results. Confirmatory functional data (e.g., treadmill, segmental pressure readings) of abnormal pulses were not performed. The interpretation of each pulse was scored as normal, mild impairment, moderate impairment, severe impairment, or total occlusion (for the purpose of this analysis, all interpretations other than normal were considered abnormal). All Doppler measurements were conducted with a minimum of a 4-hour abstinence from smoking; compliance to the nonsmoking requirement was recorded by the Doppler technician. Besides analysis of each pulse as a distinct dependent variable, three pulse aggregates were prescribed for analysis in order to maintain continuity with the Baseline analysis. The rationale of the pulse aggregates was to localize pulse abnormalities in broad anatomic categories. The aggregates were: leg pulses (femoral, popliteal, dorsalis pedis, and posterior tibial); peripheral pulses (radial, femoral, popliteal, dorsalis pedis, and posterio tibial); and all pulses (peripheral pulses plus carotid pulses, the latter assessed by only manual techniques). Any one abnormal pulse in an aggregate constituted an abnormality for the overall category. The agreement of manual and Doppler assessments was-tested by McNemar's chi-square test using paired data when an individual was compliant to both examination procedures. The paired analyses for the radial, femoral, popliteal, dorsalis pedis, posterior tibial, leg, peripheral, and all pulses are displayed in Table 15-12. As shown, the two methods of pulse assessment differed profoundly (p<0.001) for the femoral, popliteal, dorsalis pedis, leg pulses, peripheral pulses, and all pulses, but only mildly (p=0.044) for the posterior tibial pulse; the methodology differences for the radial pulse were not significantly discordant (p=0.149). Further, as shown by the offdiagonal elements in the specific pulse tables, the manual palpation method classified more cases as abnormal for the femoral, popliteal, and posterior tibial pulses, whereas the Doppler technique detected more abnormalities for the dorsalis pedis pulse, and consequently, the three pulse aggregates. Overall, more credence is given to the Doppler results due to the more "objective" means of determining a pulse abnormality. The unadjusted analyses of all the pulses and pulse aggregates by manual and Doppler techniques (Table 15-9) showed that no statistically significant 15-30 �TABLE 15-12. Agreement Between Manual and Doppler Pulse Assessments (McNemar's )C Test) Radial Femoral DOPPLER 1*lormal DOPPLER Abnormal 2,102 3 9 Normal MANUAL Abnormal 3 Normal Abnormal Normal MANUAL Abnormal X2 = 2,08 p=0.149 Popliteal 2,072 3 38 6 X2 = 28.2 p<0.001 Dorsalis Pedis DOPPLER DOPPLER Normal Abnormal Normal Abnormal 2,067 8 35 Normal MANUAL Abnormal 8 Normal MANUAL Abnormal 341 71 155 X2 = 175.6 X2 - 15.7 p<0.001 Posterior Tibial 1,546 p<0.001 Leg DOPPLER Normal Abnormal Normal MANUAL Abnormal DOPPLER Normal Abnormal 2,035 23 40 16 Normal MANUAL Abnormal 1,462 346 135 170 X2 = 91.7 X2 = 4.1 p=0.044 p<0.001 All Peripheral DOPPLER Normal Abnormal Normal MANUAL Abnormal DOPPLER Normal Abnormal 1,455 347 139 172 Normal MANUAL Abnormal X2 = 88.2 p<0.001 1,454 347 138 173 X2 = 89.2 15-31 p<0.001 �group differences were detected for any pulse or pulse combination by either technique. The covariate tests of association for each pulse and pulse combination by technique are listed in Table 15-10. The following paragraphs describe the results shown in this table. Increased age (born before 1942) was significantly associated with a" higher proportion of pulse abnormalities for the femoral pulses (manual; p<0.001), popliteal pulses (Doppler; p=0.002), dorsalis pedis pulses (manual; p=0.018), posterior tibial pulses (manual, p=0.034; Doppler, p=0.003), leg pulses (manual, p=0.001; Doppler, p=0.020), peripheral pulses (manual, p<0.001; Doppler, p=0.037), and all pulses (manual, p<0.001, Doppler, p=0.032). Age was of borderline significance (0.050<p<0.100) for femoral pulses (Doppler), and for popliteal pulses (manual) and dorsalis pedis pulses (Doppler). Race was associated with dorsalis pedis pulses (Doppler, p=0.001), posterior tibial pulses (manual, p<0.001), leg pulses (Doppler, p=0.009), peripheral pulses (Doppler, p=0.015), and all pulses (Doppler, p»0.014), with Blacks having a lower proportion of abnormalities for the dorsalis pedis, leg, peripheral, and all pulses than nonblacks, but a higher proportion of abnormalities for the posterior tibial pulse. Race was of borderline significance for the Doppler-determined posterior tibial pulses. Occupation was significantly associated with abnormalities of the dorsalis pedis pulses (Doppler, p=0.007), leg pulses (Doppler, p=0.012), peripheral pulses (Doppler, p=0.019), and all pulses (Doppler, p=0.023), with officers uniformly having more abnormalities than enlisted flyers, who had more abnormalities than enlisted groundcrew. Current smoking (cigarettes per day) was significantly associated with increased abnormalities for the posterior tibial pulses (manual, p<0.001; Doppler, p=0.021), femoral pulses (Doppler, p=0.001), and the popliteal pulses (Doppler, p<0.001), despite the 4-hour abstinence prior to the Doppler examination. A relationship of increased smoking and increased abnormalities was only observed for the Doppler determination of the femoral and popliteal pulses. Pack-years of smoking was significantly related to increased abnormalities with popliteal pulses (manual, p=0.001; Doppler, p=0.010), posterior tibial pulses (manual, p<0.001), femoral pulses (Doppler, p=0.006), leg pulses (manual, p=0.031), peripheral pulses (manual, p=0.028), and all pulses (manual, p=0.023). Classical increasing associations were noted for the popliteal pulses (manual and Doppler), the posterior tibial pulses (manual), and the femoral pulses (Doppler). For the related variables involving cholesterol, the cholesterol-HDL ratio showed the most numerous and strongest associations with pulse abnormalities. The cholesterol-HDL ratio was significantly and positively associated with increases in manually determined radial, femoral, and popliteal pulse abnormalities (p=0.033, p<0.001, and p=0.002, respectively); however, other significant associations with all pulses and the leg and .peripheral pulse indices revealed an inconsistent pattern (p=0.012, p=0.013, and p=0.010, respectively). In addition, the ratio was significantly related to femoral and posterior tibial pulse abnormalities, as detected by the Doppler technique (p=0.019, p=0.035, respectively), but the relationships were not uniform from low to high values of the ratio. HDL was significantly associated with manually determined pulse abnormalities for femoral, leg, 15-32 �peripheral, and all pulses (p=0.002, p=0.013, p=0.013, p=0.013, respectively), but in all four cases, the mid-level category of HDL (greater than 40 to 50) was associated with the lowest proportion of abnormalities. Cholesterol showed only marginally significant associations with increased abnormalities of femoral pulse (manual and Doppler) and posterior tibial pulses (manual). Percent body fat was significantly associated with increases of femoral pulse abnormalities (manual, p=0.001); personality score was associated with posterior tibial deficits (Doppler; p=0.028; nonlinear pattern); and drinkyears of alcohol was related to femoral pulse abnormalities detected by both methods (manual, p<0.001; Doppler, p=0.013). Finally, in addition to numerous other marginally significant associations (e.g., drink-years and posterior tibial abnormalities, Doppler, p=0.083; drink-years and popliteal abnormalities, manual, p=0.085), differential cortisol showed a nonlinear association with posterior tibial pulse abnormalities (Doppler, p=0.074). The distribution of each of the covariates in the Ranch Hand and Comparison groups is presented in Table 15-13. As noted, the distributions of the three matching variables, age, race, and occupation, are nearly identical (p=0.987, p=0.745, and p=0.661, respectively). For current smoking, however, Ranch Hands smoke significantly more cigarettes per day (higher mean level) than the Comparisons (p=0.043) a finding also observed at Baseline. Additionally, the difference in mean percent body fat was of borderline significance (p=0.074), with a slightly higher average level in the Comparison group. The results of the adjusted analyses for the manual and Doppler pulse determinations are presented in Table 15-11. Due to the small number of abnormalities, manual radial pulses were adjusted only for age and the cholesterol-HDL ratio, and Doppler radial pulses were not adjusted for any covariates. Similarly, femoral Doppler pulses were adjusted only for age, current smoking, and the cholesterol-HDL ratio. Doppler popliteal pulses were adjusted only for main covariate effects, i.e., interactions were not examined. The adjusted analyses of all Doppler-determined pulse and pulse aggregate abnormalities did not disclose any significant differences between the Ranch Hand and Comparison groups. Age showed a consistent and profound effect in all of the adjusted Doppler analyses, whereas race, percent body • fat, and smoking were significantly influential in about half of the analyses, and the cholesterol-HDL ratio was significant for only two of the pulse variables. The effects of these four covariates were all in the expected (classical) direction. For the manual pulse readings, the adjusted results (Table 15-11) were decidedly different from the Doppler analyses, with all but the radial and femoral pulses involved in significant group-by-covariate interactions. There were no significant group differences for the radial and femoral pulses (p=0.472, p=0.562, respectively). For manually determined popliteal pulses, there was a significant group-by-race interaction (p=0.038), with Blacks having an adjusted relative risk of 6.74 (95% C.I.: [0.72,63.40], p=0.095) in contrast to nonblacks, who had an adjusted relative risk of 0.55 (95% C.I.: 0.28,1.12] p=0.099]). All significant group-by-covariate interactions are shown in Table M-6 of Appendix M. 15-33 �TABLE 15-13. Summary Statistics for Cardiovascular Covariates by Group Covariate Covariate Category Group Ranch Hand Comparison p-Value Percent Percent Black Nonblack 5.6 94.4 6.0 94.0 0.745 Officer Enlisted Flyer Enlisted Groundcrew 37.2 17.3 45.5 37.9 15.8 46.3 0.661 Mean ± SE Mean ± SE Age (At Baseline) 43.57±0.25 43.57±0.22 0.987 Current Smoking* 10.50±0.50 9.19±0.42 0.043 Pack-years Smoking 12.62±0.52 12.51±0.48 0.883 216.8±1.3 218.1±1.2 0.463 46.32±0.42 46.90±0.35 0.288 4.99±0.05 4.92±0.04 0.303 Percent Body Fat 20.85±0.16 21.23±0.14 0.074 Personality Score -1.11±0.30 -1.50±0.26 0.322 Differential Cortisol 2.31±0.13 2.46±0.12 0.398 Current Alcohol Use (Drinks per Day) 1.23±0.07 1.28±0.07 0.611 Drink-years Alcohol 25.62±1.44 22.91±0.96 0.117 Race Occupation Cholesterol HDL Cholesterol-HDL Ratio "Equivalent cigarettes/day. —Covariate not categorized for these results. 15-34 �For the dorsalis pedis, posterior tibial, leg, peripheral, and all pulses, significant group interactions with occupation were detected (p=0.046, p=0.017, p=0.016, p=0.033, and p=0.036, respectively). In all cases, the adjusted relative risk was less than one for the officers and greater than one for the enlisted flyers and groundcrew. In addition, the adjusted relative risk for enlisted flyers was consistently greater than the risk for the enlisted groundcrew. Statistically significant associations by pulse, by occupational category, were as follows: Posterior tibial pulses in enlisted flyer, p=0.032; leg pulses in officers (21% body fat level), p=0.026; peripheral pulses in officers, p=0.030; all pulses in officers, p=0.030. All other pulse-occupational strata contrasts were not statistically significant. As there was also a significant group by percent body fat interaction for leg pulses (p=0.034), each occupational category was analyzed by level of obesity (obese, percent body fat greater than 25 percent; nonobese, percent body fat equal to or less than 25 percent). For officers, the adjusted relative risks were less than one for both the obese (Adj. RR: 0.44, 95% C.I.: [0.17, 1.12], p=0.084) and the nonobese (Adj. RR: 0.66, 95% C.I.: [0.42,1.04], p=0.072). For enlisted flyer personnel, the adjusted relative risks were greater than one for both body fat categories, but were not statistically significant. The enlisted groundcrew manifested an adjusted relative risk of less than 1 for obese individuals (Adj. RR; 0.91, 95% C.I.: [0.39,2.10], p=0.818), and greater than 1 for nonobese individuals (Adj. RR: 1.20, 95% C.I.: [0.79,1.83], p«0.390), but also not statistically significant. The unadjusted analyses of the manual and Doppler pulse assessments (shown in Table M-17 of Appendix M), using the Original Comparisons, did not disclose any significant group differences. For the Doppler adjusted analyses, the results for the Ranch Hand versus Original Comparison contrasts were similar to those found in the Ranch Hand versus total Comparison group, i.e., no statistically significant group differences or group-by-covariate interactions. For the adjusted manual pulse determinations, however, the results differed somewhat from the contrast of the Ranch Hand versus total Comparison group in terms of the significant group-by-covariate interactions detected (see Tables M-18 and M-19 of Appendix M). As before, there were no statistically significant group differences for radial and femoral pulses. For popliteal pulses, however, there was a significant (p=0.048) group-byoccupation interaction, with an adjusted relative risk of less than one for the officers (p=0.219) and greater than one for the enlisted flyers, although not significantly so (p=0.165). For dorsalis pedis pulses, there were no significant group effects or interactions, but for posterior tibial pulses the results were similar to those found in the contrast of the Ranch Hands versus the total Comparison group analysis, i.e., a significant group-byoccupation interaction. For the three pulse aggregates, there were significant group-by-occupation and group-by-percent body fat interactions for the leg pulses (officers having a risk less than one; enlisted flyers and enlisted groundcrew having risks greater than one) and significant group-bypercent body fat interactions for peripheral pulses and all pulses (individuals with low percent body fat having adjusted relative risks greater than one, and obese individuals having an adjusted risk less than one). 15-35 �EXPOSURE INDEX ANALYSES Exposure index analyses were conducted for the Ranch Hand officer, enlisted flyer, and enlisted groundcrew cohorts separately to determine if any dose-response relationships could be identified. In many cases, the data were too sparse to permit statistical comparisons. Adjusted analyses included the exposure level and.only the main effects of age, race, packyears of smoking, cholesterol-HDL ratio, percent body fat, personality score, differential cortisol, and current drinks per day, whenever appropriate. (In several instances, the stepwise logistic modeling did not detect any statistically significant covariate effects. However, adjusted best model results may differ slightly from the unadjusted results due to the omission of individuals with missing covariate information from the adjusted analysis.) Reported and Verified Heart Disease Tabular results of adjusted exposure index analyses for reported and verified heart disease are presented in Table 15-14 (unadjusted exposure index analyses are in Table M-7 of Appendix M). There were no statistically significant differences for reported or verified essential hypertension or reported or verified myocardial infarction by exposure level. (The data on myocardial infarctions were quite sparse.) Results were also negative for reported and verified heart disease, except for the enlisted groundcrew cohort, where the percentage of individuals with reported or verified disease was lowest in the medium exposure category. Central Cardiac Function Table 15-15 gives the adjusted exposure results for systolic blood pressure (dichotomized), heart sounds, and ECG findings. The unadjusted exposure analyses are given in Table M-8 of Appendix M. The only exposure level effect reaching statistical significance was the medium versus low contrast for bradycardia in the enlisted groundcrew (p=0.048), where the adjusted relative risk was significantly less than one. There were borderline significant effects, with adjusted relative risks greater than one for systolic blood pressure (enlisted groundcrew, medium versus low exposure) and T-wave findings (enlisted flyers, medium versus low contrast). There were borderline significant effects, with relative risks less than one for T-wave findings in the enlisted groundcrew cohort, medium versus low exposure (unadjusted only), and high versus low contrast (adjusted only). The results for systolic blood pressure analyzed as a continuous variable showed no statistically significant exposure level effects, either unadjusted or adjusted for covariates. The adjusted medium versus low exposure level contrast was of borderline significance in the enlisted groundcrew (p=0.069). Age, percent body fat, and personality score were significant covariates in one or.more occupational strata. 15-36 �TABLE 15-14. Adjusted Exposure Index Analyses for Reported and Verified Heart Disease by Occupation Significant Covariates Officer Medium vs. Low High vs. Low 0.92 (0.50,1-70) 1.08 (0.58, 2.01) 0.795 0.810 AGE (p=0.016) %BFAT (p<0.001) Enlisted Flyer Medium vs. Low High vs. Low 0.84 (0.34,2.05) 1.34 (0.57,3.16) 0.704 0.509 %BFAT (p=0.002) Enlisted Groundcrew . i Occupation Reported Essential Hypertension OJ Medium vs. Low High vs. Low 1.37 (0.79,2.43) 1.26 (0.68,2.33) 0.289 0.459 ZBFAT (p<0.001) Officer Medium vs. Low High vs. Low 0.94 ( . 8 1 8 ) 04,.4 1.36 (0.70,2.65) 0.849 0.363 XBFAT (p<0.001) Enlisted Flyer Medium vs. Low High vs. Low 0.45 (0.15,1.33) 0.92 ( . 5 2 4 ) 03,.0 0.150 0.865 DIFCORT (p=0.026) Enlisted Groundcrew Medium vs. Low High vs. Low 1.47 (0.82,2.66) 1.33 (0.71,2.49) 0.201 0.379 %BFAT (p<0.001) Verified Essential Hypertension Contrast Adj. Relative Risk (95% C.I.) Variable p-Value 4 Reported Heart Disease Officer Medium vs. Low High vs. Low 0.79 (0.45,1.38) 0.69 (0.39,1.23) 0.407 0.204 AGE (p=0.011) Enlisted Flyer Medium vs. Low High vs. Low 1.30 ( . 8 2 9 ) 05,.4 0.66 (0.27,1.62) 0.529 0.368 NONE Enlisted Groundcrev Medium vs. Low High vs. Low 0.51 ( . 9 0 9 ) 02,.0 1.10 ( . 5 1 8 ) 06,.6 0.020 0.711 AGE (p=0.046) �TABLE 15-14. (continued) Adjusted Exposure Index Analyses for Reported and Verified Heart Disease by Occupation Adj. Relative Risk (95% C.I.) p-Value Medium vs. Low High vs. Low 0.75 (0.42,1.34) 0.73 (0.40,1.32) 0.332 0.298 AGE (p=0.007) Enlisted Flyer Medium vs. Low High vs. Low 1.11 (0.48,2.59) 0.57 (0.22,1.46) 0.803 0.242 NONE Enlisted Groundcrew Medium vs.. Low High vs. Low 0.38 (0.20,0.73) 0.95 (0.55,1.66) 004 .0 0.865 AGE (p=0.024) Officer Medium vs. Low High vs. Low 4.01 (0.43,37.2) 1.20 ( . 7 19.9) 00, 0.222 O.897 ALC (p=0.044) Enlisted Flyer Medium vs. Low High vs. Low Enlisted Groundcrew Medium vs. Low High vs. Low 0.86 (0.13,5.47) 0.79 (0.14,4.35) 0.873 0.787 AGE (p<0.001) Officer Medium vs. Low High vs. Low Variable Occupation Verified Heart Disease Officer Contrast Significant Covariates Oi i CO 00 Reported Myocardial Infarction • Verified Myocardial Infarction Enlisted Flyer Medium vs. Low High vs. Low — Medium vs. Low High vs. Low Enlisted Groundcrev — —Analysis not performed due to sparse cells. — — — — �TABLE 15-15. Adjusted Exposure Index Analyses for Central Cardiac Function Variables by Occupation Variable Occupation Contrast Adj. Relative Risk ( 5 C I ) 9% .. p-Value Significant Covariates Officer AGE (p=0.004) PS (p=0.010) Enlisted Flyer Medium vs. Low High vs. Low 0.94 (0.22,3.98) 0.74 (0.16,3.46) 0.936 0.697 NONE Medium vs. Low High vs. Low 2.76 (0.93,8.24) 1.97 (0.61,6.32) 0.069 0.254 AGE (p=0.041) %BFAT (p=0.006) Medium vs. Low High vs. Low 0.71 (0.16,3.16) 1.33 (0.33,5.40) 0.660 0.689 AGE (p=0.004) DIFCORT (p=0.009) Enlisted Flyer — — Enlisted Groundcrev Medium vs. Low High vs. Low 0.25 (0.05,1.41) 1.26 ( . 0 4 0 ) 04,.0 0.116 0.689 CHOL/HDL (p<0.001) Officer ECG 0.638 0.952 Officer Heart Sounds 0.78 (0.27,2.26) 1.03 ( . 5 3 0 ) 03,.8 Enlisted Groundcrew Systolic Blood Pressure Medium vs. Low High vs. Low Medium vs. Low High vs. Low 1.36 (0.63,2.97) 1.15 (0.50,2.62) 0.435 0.741 AGE (p=0.007) 2BFAT (p=0.009) Enlisted Flyer Medium vs. Low High vs. Low 1.54 (0.53,4.32) 0.86 (0.29,2.49) 0.412 0.779 AGE (p=0.007) %BFAT (p<0.001) Enlisted Groundcrew Medium vs. Low High vs. Low 0.66 (0.29,1.54) 0.76 (0.34,1.71) 0.342 0.516 AGE (p=0.001) PACKYR (p=0.036) — DIFCORT (p=0.038) �TABLE 15-15. (continued) Adjusted Exposure Index Analyses for Central Cardiac Function Variables by Occupation Variable Occupation Contrast Adj. Relative Risk (95% C.I.) p-Value Significant Covariates Officer Enlisted Flyer Medium vs. Low High vs. Low 3.10 (0.85,11.28) 1.64 (0.43,6.28) 0.085 0.472 AGE (p=0.032) £BFAT (p<0.001) Medium vs. Low High vs. Low 0.50 (0.19,1.28) 0.37 (0.14,1.02) 0.150 0.055 AGE (p=0.003) Medium vs. Low High vs. Low 1.06 (0.37,3.06) 1.10 (0.37,3.27) 0.912 0.865 RACE (p=0.035) Enlisted Flyer Medium vs. Low High vs. Low 5.09 (0.57,45.1) 2.04 (0.18,23.1) 0.144 0.569 NONE Medium vs. Low High vs. Low 0.21 ( . 4 0 9 ) 00,.8 0.50 (0.15,1.65) 0.048 0.254 NONE Officer Arrhythmia AGE (p=0.027) RACE (p=0.027) %BFAT (p=0.002) Enlisted Groundcrew Bradycardia 0.289 0.441 Officer Ln I 1.65 ( . 5 4 2 ) 06,.0 1.47 (0.55,3.92) Enlisted Groundcrew Nonspecific T-tfave Changes Medium vs. Low High vs. Low Medium vs. Low High vs. Low 0.21 ( . 4 1 1 ) 00,.4 0.17 (0.02,1.44) 0.070 0.105 AGE (p=0.011) Enlisted Flyer Medium vs. Low High vs. Low Enlisted Groundcrew Medium vs. Low High vs. Low — 0.63 (0.18,2.26) 0.99 (0.32,3.02) — 0.484 0.984 — PACKYR (p<0.001) �TABLE 15-15. (continued) Adjusted Exposure Index Analyses for Central Cardiac Function Variables by Occupation Variable Occupation Contrast Adj. Relative Risk (95% C.I.) p-Value Significant Covariates Officer 1.29 ( .61,2.72) 0 0.86 ( .37,1.96) 0 0.704 0.711 AGE (p=0.003) Enlisted Flyer Medium vs . Low High vs. Low 0.54 ( .17,1.78) 0 0 0.37 ( .11,1.32) 0.312 0.522 AGE (p=0.034) Enlisted Groundcrev Other Diagnoses Medium vs . Low High vs. Low Medium vs . Low High vs. Low 0 0.91 ( .36,2.29) 0.76 (0- 30,1.92) 0.841 0.562 AGE (p<0.001) RACE (p=0.030) —Analysis not performed due to sparse cells. �Peripheral Vascular System There were no significant dose-response effects for diastolic blood pressure (dichotomized), funduscopic abnormalities, or carotid bruits (Table 15-16). Analysis of diastolic blood pressure as a continuous variable also did not reveal any statistically significant exposure level effects. Significant covariates were percent body fat, personality type, cholesterol-HDL ratio, and current alcohol use. Exposure index analyses of the peripheral pulses did not detect any statistically significant exposure effects, either unadjusted (Tables M-9 and M-10 of Appendix M for the manual and Doppler pulse readings) or adjusted (Tables 15-17 and 15-18). Main-effect exposure analyses of 6 historical and verified heart disease variables, 10 central cardiac function variables, and 11 peripheral cardiac function variables (with both manual and Doppler results), showed no evidence of a dose-response relationship at the followup examination. Two statist;ically significant and several borderline significant exposure associations lacked a pattern of dose-response consistency, and appeared to be random in nature. Association of Cardiovascular Examination Findings With Verified Heart Disease The central and peripheral cardiovascular examination findings were analyzed together with the verified cardiovascular disease endpoints to determine the degree of correlation between the third-year followup examination and the past medical history. The results are shown in Table M-ll of Appendix M. There were highly significant associations between verified essential hypertension and systolic and diastolic blood pressures, ECG abnormalities, and abnormal fundi (p<0.001, <0.001, <0.001, 0.008, respectively). There was also a significant association between essential hypertension and abnormal heart sounds (p=0.036), as well as a borderline significant association between hypertension and carotid bruits (p=0.080). The frequency of verified essential hypertension, however, was not significantly different in those with and without peripheral pulse abnormalities (as determined by either the manual or Doppler technique). For verified heart disease, there was a negative association with diastolic blood pressure (p=0.043) and positive associations with ECG abnormalities, heart sounds, abnormal fundi, and abnormal peripheral pulses as determined by the Doppler technique (p<0.001, p=0.017, p=0.014, and p=0.007, respectively). Finally, there were significant positive associations between ECG and heart sound abnormalities (p<0.001 for both) and the occurrence of a verified myocardial infarction. The consistency between the examination findings and the past medical history provides support for the overall validity of the cardiovascular measurement systems, whether by selfreport, medical records, physician assessments, or objective determinations (e.g., ECG). 15-42 �TABLE 15-16. Adjusted Exposure Index Analyses for Diastolic Blood Pressure Funduscopic Abnormalities and Carotid Bruits by Occupation Variable Occupation Contrast Adj. Relative Risk (95% C.I.) p-Value Significant Covariates Officer 0.667 0.465 ZBFAT (p=0 .002) Enlisted Flyer Medium vs . Low 0.76 (0.19,2 .98) High vs . Low 1.06 (0.29,3 .84) 0.697 0.928 ZBFAT (P=0 . 0 ) 06 Enlisted Groundcrev Diastolic Blood Pressure Medium vs . Low 0.79 (0.27,2 .30) High vs . Low 1.44 (0.54,3 .86) Medium vs. Low High vs . Low 1.50 (0.63,3 .59) 1.24 (0.53,2 .86) 0.363 0.667 CHOL/HDL (p=0.037) ZBFAT (p<0.001) Officer u> 0.337a Funduscopic Enlisted Abnormalities Flyer Enlisted Groundcrew Officer Carotid Bruits 0.388* Enlisted Flyer Enlisted Groundcrew 'Overall analysis; sparse cells, chi-square test may not be valid. —Analysis not performed due to sparse cells. �TABLE 15-17. Adjusted Exposure Index Analyses for Peripheral Vascular System Manual Pulse Readings by Occupation Variable Occupation Contrast Officer Enlisted Flyer Enlisted Groundcrew Medium vs. Low High vs. Low 0.928 0.952 AGE (p=0.030) RACE (p=0.048) DIFCORT (p=0.048) Medium vs. Low High vs. Low Enlisted Groundcrew 0.91 (0.14,5.90) 1.05 (0.20,5.52) Medium vs. Low High vs. Low Enlisted Flyer RACE (p=0.006) XBFAT (p=0.032) PS (p=0.041) Medium vs. Low High vs. Low Officer 0.509 0.238 Medium vs. Low High vs. Low Medium vs. Low High vs. Low 0.36 (0.02,7.42) 3.20 (0.31,32.6) Medium vs. Low High vs. Low Enlisted Flyer Significant Covariates Medium vs. Low High vs. Low Officer p-Value Medium vs. Low High vs. Low Enlisted Groundcrew Radial Pulses Adj. Relative Risk ( 5 C.I.) 9% Ul Femoral Pulses Popliteal Pulses �TABLE 15-17. (continued) Adjusted Exposure Index Analyses for Peripheral Vascular System Manual Pulse Readings by Occupation Variable Occupation Contrast Adj. Relative Risk (95% C.I.) p-Value Significant Covariates Officer NONE Enlisted Flyer Medium vs. Low High vs . Low 1.53 (0.51,4.65) 1.39 (0.45,4.33) 0.447 0.569 NONE Medium vs. Low High vs. Low 0.61 (0.27,1.35) 0.95 ( . 5 2 0 ) 04,.2 0.222 0.897 NONE Medium vs. Low High vs. Low Enlisted Flyer Medium vs. Low High vs. Low 0.75 1.26 Enlisted Groundcrev Ul 0.764 0.171 Officer *- 0.88 (0.40,1-97) 0.52 (0.20,1-33) Enlisted Groundcrew Dorsalis Pedis Pulses Medium vs. Low High vs. Low Leg Pulses — — (0.13,4.20) (0.26,6.10) 0.741 0.772 RACE (p=0.027) Medium vs. Low High vs. Low 2.00 (0.48,8.33) 1.51 (0.37,6.17) 0.337 0.569 AGE (p=0.003) RACE (p<0.001) Officer Posterior Tibial Pulses — Medium vs. Low High vs. Low 0.96 (0.44,2.12) 0.84 (0.37,1.95) 0.920 0.697 NONE Enlisted Flyer Medium vs. Low High vs. Low 1.01 (0.37,2.75) 1.21 (0.45,3.27) 0.984 0.711 PS (p=0.034) Enlisted Groundcrew Medium vs. Low High vs. Low 0.69 (0.35,1.36) 0.89 (0.46,1.75) 0.285 0.741 NONE �TABLE 15-17. (continued) Adjusted Exposure Index Analyses for Peripheral Vascular System Manual Pulse Readings by Occupation Variable Occupation Contrast Adj. Relative Risk (95% C.I.) p-Value Significant Covariates Officer All Pulses NONE Enlisted Flyer Medium vs. Low High vs. Low 1.01 (0.37,2.75) 1.21 (0.45,3.27) 0.984 0.711 PS (p=0.034) Medium vs. Low High vs. Low 0.69 (0.35,1.36) 0.89 (0.46,1.75) 0.285 0.741 NONE Medium vs. Low High vs. Low 0.89 (0.41,1,94) 0.86 (0.42,1.75) 0.764 0.719 NONE Enlisted Flyer Medium vs. Low High vs. Low 1.01 (0.37,2.75) 1.21 (0.45,3.27) 0.984 0.711 PS (p=0.034) Enlisted Groundcrev I 0.764 0.719 Officer Ui 0.89 (0.41,1.94) 0.86 (0.42,1.75) Enlisted Groundcrev Peripheral Pulses Medium vs. Low High vs. Low Medium vs. Low High vs. Low 0.69 (0.35,1.36) 0.89 (0.46,1.75) 0.285 0.741 NONE —Analysis not performed due to sparse cells. �TABLE 15-18. Adjusted Exposure Index Analyses for Peripheral Vascular System Doppler Pulse Reading by Occupation Variable Occupation Contrast Adj. Relative Risk ( 5 C.I.) 9% p-Value Significant Covariates Officer 1.12 (0.63,1.97) 1.08 (0.60,1-96) 0.704 0.787 NONE Enlisted Flyer Medium vs. Low High vs. Low 1.30 (0.51,3.28) 1.43 (0.56,3.64) 0.575 0.447 NONE Enlisted Groundcrew Medium vs. Low High vs. Low 0.94 (0.53,1.65) 1.04 (0.58,1.87) 0.772 0.764 RACE (p=0.034) Officer Medium vs. Low High vs. Low 2.09 (0.38,11.6) 1.01 (0.13,7.58) 0.401 0.992 AGE (p=0.025) DIFCORT (p=0.026) Enlisted Flyer Medium vs. Low High vs. Low Enlisted Groundcrev Medium vs. Low High vs. Low Officer Dorsal is Pedis Pulses Medium vs. Low High vs. Low Medium vs. Low High vs. Low 1.26 (0.71,2.21) 1.19 (0.66,2.13) 0.430 0.562 NONE Enlisted Flyer Medium vs. Low High vs. Low 1.57 (0.63,3.90) 1.58 (0.63,3.98) 0.327 0.327 NONE Enlisted Groundcrew Medium vs. Low High vs. Low 0.94 (0.54,1.87) 1.01 (0.57,1.80) 0.818 0.772 NONE tn Posterior Tibial Pulses Leg Pulses — — — — — — �TABLE 15-18. (continued) Adjusted Exposure Index Analyses for Peripheral Vascular System Doppler Pulse Reading by Occupation Variable Occupation Contrast Adj. Relative Risk ( 5 C.I.) 9% p-Value Significant Covariates Officer NONE Enlisted Flyer Medium vs. Low High vs. Low 1.57 (0.63,3.90) 1.58 (0.63,3.98) 0.327 0.327 NONE Medium vs. Low High vs. Low 0 9 (0.52,1.57) .0 1.02 (0.58,1.79) 0.704 0.960 NONE Medium vs. Low High vs. Low 1.26 (0.71,2.21) 1.19 (0.66,2.13) 0.430 0.562 NONE Enlisted Flyer Medium vs. Low High vs. Low 1.57 (0.63,3.90) 1.58 (0.63,3.98) 0.327 0.327 NONE Enlisted Groundcrev *- 0.430 0.562 Officer Ul 1.26 (0.71,2.21) 1.19 (0.66,2.13) Enlisted Groundcrev Peripheral Pulses Medium vs. Low High vs. Low Medium vs. Low High vs. Low 0.90 (0.52,1.57) 1.02 (0.58,1.79) 0.704 0.960 NONE. 00 All Pulses —Analysis not performed due to sparse cells. �LONGITUDINAL ANALYSES Two cardiovascular variables, the index of all pulses (by palpation) and the overall ECG interpretation, were investigated to assess the longitudinal differences between the 1982 Baseline examination and the 1985 followup examination. Both variables are classified as abnormal or normal. As shown in Table 15-19, 2x2 tables were constructed for each group for each variable. These tables show the number of participants who were abnormal at Baseline and abnormal at followup, abnormal at Baseline and normal at followup, normal at Baseline and abnormal at followup, and normal at both Baseline and followup examinations. The odds ratio given is the ratio of the number of participants who were normal at the Baseline and abnormal at the followup to the number of participants who were abnormal at the Baseline and normal at the followup (the "off-diagonal" elements). The changes in normal/abnormal status within each group are contrasted between the Ranch Hand and Comparison groups, and the p-value is derived from Pearson's chi-square test of the hypothesis that the pattern of change in the two groups is the same. TABLE 15-19. Longitudinal Analyses of All Pulses Index and Overall ECG's: A Contrast of Baseline and First Followup Examination Abnormalities Variable Group 1982 Baseline Exam 1985 Followup Exam Odds* Ratio (OR) p-Value (ORRH vs. ORC) Abnormal Normal All Pulses Ranch Hand Abnormal (Manual) Normal 50 104 72 743 1.44 Comparison Abnormal Normal 40 153 63 880 2.43 Ranch Hand Abnormal Normal 86 43 192 650 0.22 Comparison Abnormal Normal 112 56 208 763 0.27 0.01 ECG (Overall) 0.42 Number Normal Baseline, Abnormal Followup *0dds Ratio: Number Abnormal Baseline, Normal Followup 15-49 • �The data showed a significant difference (p=0.01) in the pulse index in the two groups between examinations. The percentage of Ranch Hands and Comparisons with abnormalities for the pulse index increased from the Baseline examination to the followup examination; however, the Comparison group showed a larger increase in the proportion of pulse index abnormalities. The greater relative increase in the Comparisons caused the significant result. No significant group differences were detected between examinations for overall ECG abnormalities (p=0.42). DISCUSSION In general, the foregoing analyses on a wide range of cardiovascular variables, have shown a lack of significant differences between the Ranch Hands and the Comparisons. The sole exception was the finding of increased verified heart disease in the Ranch Hands versus the Comparisons (24% and 20%, respectively, p=0.054, unadjusted; p=0.036, adjusted). These results were not noted in the Baseline examination (p=0.982, unadjusted). A review of the relative risk patterns, whether or not statistically significant, for all of the other cardiovascular variables showed general equality, with about half of the risks below unity and half above. This rough equivalence suggests that, although the Ranch Hands have slightly more reported heart disease, the finding is not mirrored by substantial and consistent clinical cardiovascular defects at this time. This observation should not be lightly dismissed, and is cause for continued close surveillance. The most notable cardiovascular finding at the followup examination was the lack of significant peripheral pulse abnormalities, which were unexpectedly found at the 1982 Baseline examination (p=0.05).. The primary contributory cause of the change in pulse significance from Baseline to followup was probably the rigid 4-hour tobacco abstinence required prior to Doppler testing (due to the known vasoconstriction effects of nicotine). Tobacco abstinence, however, was not a requirement for the Baseline manual pulse readings. Although tobacco abstinence was not a requirement prior to manual readings at the followup examination, there was general compliance to the smoking prohibition, particularly if a participant's general physical examination preceded the Doppler testing. Therefore it might be expected that the manual readings would show more pulse abnormalities than Doppler testing; in fact, this was the case (see section on Peripheral Pulses). Whatever the true cause(s), the prevailing fact is that there are no longer significant group differences in pulse abnormalities, as noted by both manual and Doppler techniques, regardless of the poor agreement between the two methods. The close approximation of the estimated relative risks to unity for practically all of the cardiovascular variables is clearly indicative of equivalent cardiovascular health between the two groups. Furthermore, the general similarity of the unadjusted and adjusted results was suggestive of near equivalence of the important cardiovascular risk factors in the Ranch Hands and Comparisons (see Table 15-13), as well as a balance for unanalyzed or hidden covariates of importance. These health assessments of the two groups are considerably strengthened by the almost consistent, classical effects of the covariates in this chapter. In particular, the age effect was uniformly profound, affecting 15-50 �almost all of the dependent variables in the functional categories of reported-verified heart diseases, and central and peripheral vascular function. The covariates of race, percent body fat, and cholesterol (particularly the cholesterol-HDL ratio), and smoking were also generally strong and consistent in their effects. Statistically significant, positive associations were seen between the current level of smoking and posterior tibial, popliteal, and femoral pulses, as well as borderline significant associations between current smoking and other ECG diagnoses, carotid bruits, and reported myocardial infarctions. However, significant negative associations were observed between current smoking and reported and verified essential hypertension. Pack-years of smoking was significantly positively associated with several ECG variables and pulse assessments, although not always in a consistently increasing manner. There was a statistically significant and consistently increasing effect of pack-years of smoking on reported and verified myocardial infarctions, but there was a negative association between pack-years of smoking and verified essential hypertension, with the greatest number of abnormalities in the zero pack-year category. Alcohol was infrequently interactive with the dependent variables, but covariate tests of association generally revealed the classical pattern of more cardiovascular abnormalities in the nondrinking category than in the low drinking category. Personality score, however, usually failed to demonstrate the "expected" aggregation of cardiovascular abnormalities in the Type A direction. In fact, most associations were in the Type B direction. Generally, only cardiovascular studies ascertaining personality type by the Structured Interview technique have shown an association of Type A personality (Type A-l, in particular) to heart disease endpoints, and conversely, studies using questionnaire techniques to measure personality type have not demonstrated the association. Lastly, the strong association between historical-verified cardiovascular events and the specific dependent variables provides assurance that the overall cardiovascular measurements have been accurate and valid. SUMMARY AND CONCLUSIONS The cardiovascular health of both cohorts was assessed by collection of reported and record-verified heart disease events; measurement of central cardiac function by systolic blood pressure, abnormal heart sounds, and electrocardiograph (ECG) findings; and evaluation of peripheral vascular function by diastolic blood pressure, funduscopic examination, presence of carotid bruits, and detailed manual and Doppler measurements of five peripheral pulses. Table 15-20 presents the overall summary of the unadjusted and adjusted results. Where possible, the analyses used the covariates of age, race, occupation, percent body fat, cholesterol, high density lipoprotein (HDL) cholesterol, cholesterol-HDL ratio, smoking history (pack-years and current smoking level), alcohol history (drink-years and current drinking level), personality score, and differential cortisol. The cardiovascular variables did not reveal significant group differences, with the exception of verified heart disease, for which the proportions of recorded cardiac events were 24 and 20 percent in the Ranch Hand and Comparison groups, respectively, (p=0.054 unadjusted, p=0.036 adjusted). This finding was not reinforced by results of individual questionnaire or examination variables showing impairment in the Ranch Hands. There was a remarkable balance in relative risks above and below unity between the groups. 15-51 �TABLE 15-20. Overall Summary Results of Unadjusted and Adjusted Analyses Cardiovascular Variables Variable Statistical/ Clinical Analysis Unadjusted Adjusted Historical and Verified Heart Disease Reported Verified Reported Verified Reported Verified Hypertension Hypertension Heart Disease8 Heart Disease* Heart Attack Heart Attack NS NS NS NS* NS NS " NS NS NS S NS NS Central Cardiac Function Systolic Blood Pressure Discrete Continuous Heart Sounds Electrocardiogram (Overall) ECG: RBBB ECG: LBBB ECG: Nonspecific T-Wave Changes ECG: Bradycardia ECG: Tachycardia ECG: Arrhythmia ECG: Other Diagnoses NS NS NS NS NS NS NS NS NS 15-52 NS **** NS **** NS N/A NS NS N/A **** NS �TABLE 15-20. (continued) Overall Summary Results of Unadjusted and Adjusted Analyses Cardiovascular Variables Statistical/ Clinical Analysis Unadjusted Adjusted Funduscopic Examination NS NS NS NS NS NS Carotid Bruits NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS **** Variable Peripheral Vascular Function Diastolic Blood Pressure Radial Pulses Femoral Pulses Popliteal Pulses Dorsalis Pedis Pulses Posterior Tibial Pulses Leg Pulses Peripheral Pulses All Pulses Discrete Continuous Manual Doppler Manual Doppler Manual Doppler Manual Doppler Manual Doppler Manual Doppler Manual Doppler Manual Doppler NS NS:Not significant (p>0.10). NS*:Borderline significant (0.05<p<0.10). ****Group-by-covariate interaction. a Excluding hypertension. b RH>C (Adj. RR: 1.25;'95% C.I.: [1.02, 1.54], p=0.036). 15-53 NS **** NS **** NS **** NS **** NS **** NS �Other related analyses showed an absence of significant group differences in reported or verified hypertension, reported or verified heart attacks, and reported heart disease. There was good correlation between the verified cardiovascular history and the central and peripheral cardiovascular abnormalities detected at the physical examination, supporting accuracy and validity of the cardiovascular measurements. The adjusted analyses of central cardiac function disclosed a significant group-by-age interaction involving systolic blood pressure in the Black cohort, with a mean systolic blood pressure greater in the Ranch Hands than the Comparisons at younger age levels, but a lower mean pressure at the older ages; the group-by-age interaction was not significant in the nonblack cohort. Additionally, there was a significant group-by-pack-years of smoking interaction for the overall EGG findings, and significant group-by-pack-years of smoking and group-by-percent body fat interactions for arrhythmia, but they all generally pointed to lower adjusted relative risks in the Ranch Hands. In the analysis of peripheral vascular function, no significant group differences were observed for abnormalities involving radial, femoral, popliteal, posterior tibial, dorsalis pedis, or three anatomic aggregates of these pulses, either by manual palpation or Doppler techniques. This overall finding was in distinct contrast to the 1982 Baseline examination, which by the manual palpation method, showed significant peripheral pulse deficits in the Ranch Hands. This favorable pulse reversal over the two examinations is primarily attributed to the rigid 4-hour tobacco abstinence applied prior to Doppler testing, although other factors may be related. The lack of group differences for pulse abnormalities was noted even though the manual and Doppler techniques differed significantly (p<0.05, p<0.001 for most) in the detection of abnormalities for all but one of the pulses or pulse combinations. For manually-determined pulse abnormalities, there was a significant group-by-race interaction for the popliteal pulses, a significant group-bypercent body fat interaction for the leg pulses, and significant group-byoccupation interactions for the posterior tibial, dosalis pedis, and the three pulse aggregates (leg, peripheral, and all pulses). No interactions were encountered in the adjusted analyses of the Doppler results, and none showed significant group differences. Statistical analyses involving the Original Comparisons also showed no significant differences in the cardiovascular measurements between groups, although slightly different interactions were detected in some of the adjusted analyses. For the exposure analyses, the only statistically significant effects were those pointing to less bradycardia and less reported and verified heart disease in the medium exposure level category, as contrasted to the low exposure category, among the enlisted groundcrew. In many cases there were too few abnormalities within the occupational categories to permit formal statistical tests. Overall, the.exposure analyses were deemed as unsupportive of any meaningful dose-response relationships. The longitudinal analysis of the pulse index confirmed the significant difference in the change in the pattern of results from the Baseline examination to the followup examination, largely due to a relatively greater 15-54 �increase of pulse abnormalities in the Comparison group than in the Ranch Hand group. There was no significant change in pattern between the two groups in overall ECG findings between examinations. There was a similar distribution of the covariates between groups, except for a slightly higher level of current Ranch Hand smoking (also observed at Baseline), and a corresponding slightly lower mean percent body fat. The general covariate effects were strong and showed expected, classical associations with the cardiovascular measurements. However, unexpected effects were consistently noted for personality score, with higher proportions of various cardiovascular abnormalities associated with scores in the Type B direction, a finding possibly attributable to the method of personality determination. Nonetheless,^ the repeated demonstration of classical covariate associations with cardiovascular pathology lends considerable credence to the quality of the data. Although smoking was positively associated with many of the cardiovascular measurements, negative associations were seen between current smoking and reported and verified essential hypertension and between pack-years of smoking and verified hypertension. In conclusion, of 27 cardiovascular variables, only one, verified heart disease, showed a significant excess in the Ranch Hands, but this finding was largely unsupported by other cardiac measurements. Both manual palpation and Doppler recordings of five peripheral pulses were similar in both groups, in marked contrast to the 1982 Baseline examination which found significant pulse deficits in the Ranch Hand group. This change at the followup examination was most likely due to required tobacco abstinence prior to the pulse measurements. Exposure index analyses did not support a consistent doseresponse relationship for any variable. Overall, there was remarkable similarity in the cardiovascular health between the Ranch Hand and Comparison groups. 15-55 �CHAPTER 15 REFERENCES 1. Palmer, J.S., and R.D. Radeleff. 1964. The toxicologic effects of certain fungicides and herbicides on sheep and cattle. Ann. N.Y. Acad. Sci. 111:729-736. 2. McConnell, E.E., J.A. Moore, and D.W. Dalgard. 1978. Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Rhesus monkeys (Macaca mulatta) following a single oral dose. Toxicol. Appl. Pharmacol. 43(1):175-187. 3. Kimbrough, R.D., C.D. Carter, J.A. Liddle, R.E. Cline, and P.E. Phillips. 1977. Epidemiology and pathology of a tetrachlorodibenzodioxin poisoning episode. Arch. Environ. Health 32(2):77-86. 4. McConnell, E.E., J.A. Moore, J.K. Baseman, and M.W. Harris. 1978. The comparative toxicity of chlorinated dibenzo-p-dioxins in mice and guinea pigs. Toxicol. Appl. Pharmacol. 44(2):335-356. 5. Schreiweis, D.O., and G.J. Murray. 1976. Cardiovascular malformations in Oryzias latipes embryos treated with 2,4,5-trichlorophenoxyacetic acid (2,4,5-T). Teratology 14(3):287-290. 6. Rifkind, A.B., Y. Hattori, R. Levi, M.J. Hughes, C. Quilley, and D.R. Alonso. The chick embryo as a model for PCB and dioxin toxicity: Evidence of cardiotoxicity and increased prostaglandin synthesis. In Banbury report 18; Biological mechanisms of dioxin action ed. A. Poland and R.D. Kimbrough, pp. 255-266.Cold Spring Harbor, New York: Cold Spring Harbor Laboratory. 7. Dudley, A.W., and N.T. Thapar. 1972. Fatal human ingestion of 2,4-D, a common herbicide. Arch. Path. 94:270-275. 8. Paggiaro, P.L., E. Martino, and S. Mariotti. 1974. A case of 2,4-dichlorophenoxyacetic acid (2,4-D) poisoning. Med. Lavoro 65(3-4):128-135. 9. Berwick, P. 1970. 2,4-Dichlorophenoxyacetic acid poisoning in man. JAMA 214(6):1114-1117. 10. Oliver, R.M. 1975. Toxic effects of 2,3,7,8-tetrachloro-dibenzo-l, 4-dioxin in laboratory workers. Br. J. Ind. Med. 32:46-53. 11. Baader, E.W., and A.J. Bauer. 1951. Industrial intoxication due to pentachlorophenol. Ind. Med. Surg. 20:289-290. 12. Jirasek, L., J. Kalensky, K. Kubec, et al. 1974. Acne chlorina, porphyria cutanea tarda and other manifestations of general intoxication during the manufacture of herbicides, part 2. Czech. Dermatol. 49(3):145-157. 15-56 �13. Pazderova-Vejlupkova, J., M. Nemcova, J. Pickova, L. Jirasek, and E. Lukas. 1981. The development and prognosis of chronic intoxication by tetrachlorodibenzo-p-dioxin in men. Arch. Environ. Health 36:5-11. 14. Poland, A.P., D. Smith, G. Metter, and P. Possick. 1971. A health survey of workers in a 2,4-D and 2,4,5-T plant, with special attention to chloracne, porphyria cutanea tarda, and psychologic parameters. Arch. Environ. Health 22(3):316-327. 15. Moses, M., R. Lilis, K.D. Crow, J. Thornton, A. Fischbein, H.A. Anderson, and I.J. Selikoff. 1984. Health status of workers with past exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in the manufacture of 2,4,5-trichloro-phenoxyacetic acid: Comparison of findings with and without chloracne. Am. J. Ind. Med. 5:161-182. 16. Suskind, R.R., and V.S. Hertzberg. 1984. Human health effects of 2,4,5-T and its toxic contaminants. JAMA 251:2372-2380. 17. Hoffman, R.E., P.A. Stehr-Green, K.B. Webb, G. Evans, A.P. Knutsen, W.F. Schramm, J.L. Staake, B.B. Gibson, and K.K. Steinberg. 1986. Health effects of long-term exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. JAMA 255:2031-2038. 18. Stehr, P.A., G. Stein, H. Falk, et al. 1986. A pilot epidemiologic study of possible health effects associated with 2,3,7,8-tetrachlorodibenzo-p-dioxin contamination in Missouri. Arch. Environ. Health 41:16-22. 19. Troxler, R.G., and H.A. Schwertner. 1985. Cholesterol, stress, lifestyle, and coronary heart disease. Aviat. Space Environ. Med. 56:660-665. 20. American Heart Association Steering Committee for Medical and Community Program. 1980. Risk factors and coronary disease: A statement for physicians. Circulation 62:449A-455A. 21. Castelli, W.P. 1984. Epidemiology of coronary heart disease: Framingham study. Am. J. Med. 76:4-12. The 22. Multiple Risk Factor Intervention Trial Research Group (Neaton, J.D., L.H. Kuller, D. Wentworth, et al.). 1984. Total and cardiovascular mortality in relation to cigarette smoking, serum cholesterol concentration, and diastolic blood pressure among black and white males followed for five years. Am. Heart J. 108:759-769. 23. Morton, W.E., E.D. Crawford, R.A. Maricle, D.D. Douglas, and V.H. Freed. 1975. Hypertension in Oregon pesticide-formulating workers. J. Occ. Med. 17(3):182-185. 24. Martin, J.V. 1984. Lipid abnormalities in workers exposed to dioxin. Br. J. Ind. Med. 41:254-256. 25. Ashe, W.F., and R.R. Suskind. 1949, 1950. Reports on chloracne cases, Monsanto Chemical Company, Nitro, West Virginia. In Report of the Kettering Laboratory, December 1949 and April 1950. 15-57 �26. Lipid Research Clinic Program: The Lipid Research Clinic's Coronary Prevention Trial Results: II. The relationship of reduction in incidence of coronary heart disease to cholesterol lowering. 1984. JAMA 251:365-374. 27. Stamler, J., D. Wentvorth, and J.D. Neaton. 1986. Is relationship between serum cholesterol and risk of premature death from coronary heart disease continuous and graded? JAMA 256:2823-2828. 15-58 �CHAPTER 16 HEMATOLOGICAL EVALUATION INTRODUCTION Although direct impairment of the hematopoietic system may result from exposure to chlorophenols or dioxin, marked abnormalities in many of the circulating hematological elements may also be due to the severe and often endstage toxicity observed in other organs or organ systems. Animal experiments have confirmed both direct and indirect hematopoietic effects of TCDD. In a chronic low-dose feeding study of TCDD in eight monkeys, decreased hemoglobin and hematocrit values were noted at the 6-month mark in all animals. Four of these monkeys expired in 7 to 11 months and all had anemia, leukopenia, and thrombocytopenia. Necropsy of three sacrificed animals at 1 year showed multi-organ pathology including bone marrow degeneration, atrophy of lymphopoietic tissue, and numerous hemorrhages in a variety of organs. In another monkey experiment, using single low and high doses of TCDD, early hematological effects included increased neutrophil counts in the low-dose group and lymphopenia and thrombocytopenia in the high-dose group.2 At the end of the experiment, half the sternal bone-marrow samples revealed a decrease in overall cellular!ty and an increase in the myeloid-erythroid cell ratio. Rat experiments with TCDD demonstrated relatively consistent results. One study revealed elevated erythrocyte, reticulocyte, and neutrophil counts with depressed values for the mean corpuscular volume, mean corpuscular hemoglobin, platelet counts, and clot retraction times. The authors attributed most of these effects to terminal dehydration and nonspecific toxicity. Another rat study using gavage doses of TCDD varying from 0.001 to 1.0 wg/kg demonstrated depressed red blood cell counts and packed cell volumes in the high-dose group.4 In a mixed-dose regimen using rats, mice, and guinea pigs, dose-related decreases in lymphocyte and leukocyte numbers were observed in mice and guinea pigs within 1 week following TCDD administration. Thrombocytopenia and hemoconcentration were found in rats. Because of the lymphopenia in mice and guinea pigs, TCDD was judged to be immunosuppressive. In general, human observational studies showed fewer and less consistent hematological findings than the structured animal experiments. A case report of 2,4-D intoxication with marked neurological findings described transient bone marrow depression with peripheral leukopenia and granulocytopenia. In two industrial accidents involving significant contamination with TCDD and resulting cases of chloracne, only temporary depression of peripheral leukocyte and lymphocyte formation was observed. ' Two contemporary indepth morbidity studies9'10 of the Nitro, Vest Virginia, accident included routine clinical complete blood counts and differential counts, and hemoglobin and hematocrit determinations. Though these studies shared overlapping study cohorts, they did not report any of the 16-1 �hematological results in their publications; presumably, there were no significant differences in any of the parameters between the exposed and the unexposed cohorts. The two pilot studies of TCDD-contaminated residential areas in Missouri also included routine hematological assays of peripheral blood. ' One study paradoxically noted a significantly increased mean platelet count in the high-risk group, although the data were not adjusted for smoking.11 The Quail Run study, predominantly emphasizing cell-mediated immunity, found significant group differences in the mean leukocyte count, mean absolute granulpcyte count, and the mean percentage of monocytes in the differential count. Unfortunately, the authors neglected to identify the group (exposed or unexposed) that had the abnormal hematological findings. However, the finding of a significantly higher proportion of individuals with white blood cell counts exceeding 10,000/mm was in the exposed group. Baseline Summary Results A number of statistically significant group differences and interactions emerged in the analysis of the 1982 Baseline examination. The Ranch Hand group had a significantly higher adjusted mean red blood cell corpuscular volume and corpuscular hemoglobin value than the Comparison group (p»0.05, p=0.04, respectively), although the magnitude of the difference was small in each case. The Ranch Hand adjusted mean values for six other parameters, i.e., red blood cell count, white blood cell count, hemoglobin, hematocrit, mean corpuscular hemoglobin concentration, and platelet count, were nearly identical to the adjusted means of the Comparison group, and all were well within normal range. Similarly, the percent of abnormal values for these eight variables, as established by the upper and lower limits of normal, did not vary by group. Linear models demonstrated the profound effect of smoking, as measured in pack-years. With increased smoking, white blood cell, hemoglobin, hematocrit, mean corpuscular volume, mean corpuscular hemoglobin, and platelet values increased, whereas the mean corpuscular hemoglobin concentration showed a significant negative association with smoking. The red blood cell count revealed a borderline significant negative relationship to smoking. No statistically significant group-by-smoking interactions were detected. The exposure index analyses conducted within the Ranch Hand group disclosed two statistically significant exposure-level effects as well as seven significant or borderline-significant exposure-level-by-smoking interactions. In the officer cohort, the percentage of mean corpuscular hemoglobin abnormalities increased with increasing exposure level. The high-exposure group also had the highest percentage of mean corpuscular hemoglobin concentration abnormalities. No significant associations were found, however, in the enlisted flyer or enlisted groundcrew cohort. Five interactions involved a decreasing association (gradient of slopes) between the hematological measure and pack-years of smoking with increasing exposure level, one showed an increasing association with increasing exposure level, and one was uninterpretable. The report concluded that the overall statistical findings were somewhat consistent among themselves, and that medical morbidity was not significant. 16-2 �Parameters of the 1985 Hematological Evaluation The 1985 hematological assessment was identical to the 1982 Baseline evaluation. The eight hematological variables were red blood cell count (RBC), white blood cell count (WBC), hemoglobin (HGB), hematocrit (HCT), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), and platelet count (PLT); these variables were determined by routine hematological procedures. The normal ranges of the SCRF-determined values differed somewhat from those employed in 1982 by the Kelsey-Seybold Clinic. As before, the analysis of the hematological data included the covariates of age, race, occupation, and smoking. Updated and more comprehensive smoking data, in terms of pack-years and current smoking (including cigar- and pipe-smoking), were used in most analyses. Excluded were three individuals with fever at the time of examination (two Ranch Hands and one Comparison). Hematological variables in the continuous form were analyzed by general linear models adjusting for age, race, occupation, and smoking. The hematological data, trichotomized as abnormally low, normal, or abnormally high, were subjected to log-linear (logit) analysis, adjusted for the same covariates. Minor differences in the table totals within this chapter reflect rare missing data for either the dependent variables or the covariates. Parallel analyses using Original Comparisons can be found in Tables N-4 through N-9 of Appendix N. RESULTS AND DISCUSSION General Eight hematological assays were performed on peripheral blood specimens obtained from all participants on the first day of the physical examination. Table 16-1 lists the assays, the abbreviations used in this chapter, the SCRF laboratory normal range for each assay, and the required laboratory coefficient of variation for each assay. The SCRF laboratory norms varied to some extent from the values used at the Baseline examination (see pages XVI-3-1, Baseline Report). The SCRF laboratory coefficients of variation met or exceeded contract requirements and were uniformly achieved due to the precision of the Coulter 5-Plus automated instrument, in conjunction with rigorous FIR CUSUM quality control techniques (see Chapter 6). The overall precision in the laboratory aspects of the hematological assays, is reflected in the analytic ability to discern minute mean shifts . between groups. Representative statistical power statements are as follows. Sample sizes were sufficiently large to detect a 0.87 percent mean shift in RBC and a 2.5 percent mean shift in PLT values using an a -level of 0.05 (twosided) and a power of 0.80. Further, the sample sizes were sufficient to detect a 1.66-fold increase in the frequency of abnormal values for RBC, and a 1.96-fold increase in the frequency of abnormal values for PLT, with 80 percent certainty. 16-3 �TABLE 16-1. Laboratory Parameters for Hematological Test Variables Hematological Test Abbreviation SCRF Laboratory Normal Range Contract Required Coefficient of Variation (in percent) Red Blood Cell Count RBC 4.3-5.9 million/cubic mm 2.0 White Blood Cell Count WBC A.5-11.0 thousand/cubic mm 2.5 Hemoglobin HGB 13.9-16.3 grams/100 ml 1.1 Hematocrit HCT 39.0-55.0 ml/100 ml 3.0 Mean Corpuscular Volume MCV 80.0-97.0 cubic micra 2.0 Mean Corpuscular Hemoglobin MCH 26.0-34.0 micromicrogram 2.0 Mean Corpuscular Hemoglobin Concentration MCHC 31.0-37.0 percent 2.0 Platelet Count PLT 130-400 thousand/cubic mm 3.5 The statistical analyses in this chapter are presented in the following order: unadjusted tests, covariate tests of association, adjusted analyses, exposure analyses, and longitudinal contrasts. A variable-by-variable discussion summarizes all of the analyses, and representative exposure analyses are also presented. Group-by-covariate interactions are narratively presented, and illustrated by calculating Ranch Hand-Comparison differences at selected covariate levels. The interaction data tables are found in Tables N-2 and N-3 of Appendix N. Unadjusted Categorical Analyses Data from the eight hematological variables were categorized as abnormally low, normal, or abnormally high according to the SCRF laboratory norms cited in Table 16-1. The frequency distribution of these discretized data is presented by group in Table 16-2. As shown, there were no statistically significant, or even marginally significant, differences between the groups. Only one abnormal MCHC value was found among all study participants. 16-4 �TABLE 16-2. Unadjusted Categorical Analyses for Hematological Variables by Group Abnormally Low Variable Group Normal Abnormally High Number Percent Number Percent Number Percent Total p-Value* RBC Ranch Hand Comparison 30 42 3.0 3.2 976 1,239 96.2 95.9 8 11 0.8 0.8 1,014 1,292 0.910 WBC Ranch Hand Comparison 45 63 4.4 4.9 906 1,149 89.4 88.9 62 80 6.1 6.2 1,013 1,292 0.883 HGB Ranch Hand Comparison 39 44 3.8 3.4 752 960 74.2 74.3 223 288 22.0 22.3 1,014 1,292 0.848 HCT Ranch Hand Comparison 11 15 1.1 1.2 1,001 1,274 98.7 98.6 2 3 0.2 0.2 1,014 1,292 0.999 MCV Ranch Hand Comparison 10 13 1.0 1.0 857 1,094 84.5 84.7 147 185 14.5 14.3 1,014 1,292 0.992 MCH Ranch Hand Comparison 7 7 0.7 0.5 943 1,211 93.0 93.7 64 74 6.3 5.7 1,014 1,292 0.755 MCHC Ranch Hand Comparison 1 0 0.1 0.0 1,013 1,292 99.9 100.0 0 0 0.0 0.0 1,014 1,292 — Ranch Hand Comparison 5 3 0.5 0.2 987 1,264 97.4 97.9 21 24 2.1 1.9 1,013 1,291 0.828 PLT *Chi-square test, 2 d.f., except for HCT and PLT which were obtained from continuity adjusted chi-square tests on 1 d.f. (Abnormally high category pooled with normal, and abnormally low category pooled with normal for HCT and PLT, respectively.) —Only one abnormal MCHC value; p-value not given. 16-5 �Unadjusted Analyses of Continuous Data The unadjusted tests of group means from the continuous data for the eight variables are displayed in Table 16-3. The variables WBC and PLT were analyzed in logarithmic units because of their right-skewed original distributions. Antilog values of the means are given for ease of interpretation but their standard error or variance terms are consequently omitted since the relevance of these terms pertains only to the logarithmic scale. The sample sizes were 1,014 for the Ranch Hand group and 1,292 for the Comparisons, except for WBC (Ranch Hands, 1,013; Comparisons, 1,292) and PLT (Ranch Hands, 1,013; Comparisons, 1,291). As shown in Table 16-3, there were no statistically significant group differences between the unadjusted means of each variable. TABLE 16-3. Unadjusted Continuous Analyses for Heoatological Variables (Contrast of Group Means) Group Mean±SE Variable Ranch Hand Difference ±SE Comparison RBC 4.964±0.012 4.982±0.010 WBC8 7.003 6,891 -0.019±0.016 — t-Statistic p-Value -1.19 0.233 1.34 0.182 HGB 15.624±0.033 15.626±0.029 -0.002±0.044 -0.05 0.958 HCT 45.904±0.097 45.952±0.083 -0.048±0.127 -0.38 0.703 MCV 92.596±0.150 92.346±0.132 0.250±0.200 1.25 0.210 MCH 31.544±0.055 31.431±0.049 0.113±0.074 1.53 0.125 MCHC 34.040±0.021 34.009±0.017 0.031±0.027 1.17 0.243 0.96 0.337 PLT* 265.2 263.0 — 'Means transformed from log scale. —Difference and standard errors (SE) not presented, since variables were analyzed on logarithmic scale.. 16-6 �Dependent Variable and Covariate Relationships The data from the Ranch Hand and Comparison groups were pooled for each of the eight hematological variables and analyzed independently with the covariates of age (born in or after 1942, born before 1942), race (Black, nonblack), occupation (officer, enlisted flyer, enlisted groundcrew), and smoking history (0 pack-years; greater than 0 to 10 pack-years; and greater than 10 pack-years). These analyses are summarized in terms of statistical significance (p-values) in Table 16-4. As noted, each of the dependent variables was substantially affected by one or more of the covariates. The exact nature of the covariate influence, e.g., directionality, significance, consistency across related variables, is presented in the variable-byvariable discussion section. Covariate effects were also analyzed in continuous form with the use of linear regression models (see Table 16-6 and discussion following). In addition, covariate distributions were examined between groups (see Table N-l of Appendix N). TABLE 16-4. Association Betveen Hematological Variables and Age, Race, Occupation, and Smoking History in the Combined Ranch Hand and Comparison Groups Variable Age Race Occupation Smoking History RBC 0.010 <0.001 NS NS WBC NS* <0.001 0.001 <0.001 HGB NS 0.002 <0.001 0.003 HCT NS <0.001 NS NS MCV <0.001 <0.001 0.004 <0.001 MCH <0.001 <0.001 0.003 <0.001 — NS — MCHC PLT — 'NS NS NS: Not significant (p>0.10). NS*: Borderline significant (0.05<p<0.10). —Not analyzed due to sparse data. 16-7 — 0.004 �Adjusted Categorical Analyses Log-linear (logit) models for each of the hematological variables were fit to adjust for age, race, occupation, and smoking history. In addition, all significant group-by-covariate interactions were examined. The covariate of current level of smoking (used in the adjusted continuous analyses described below) was not included in the categorical analyses to avoid problems with sparse cells. Adjusted relative risks for Ranch Hand-Comparison contrasts were calculated for the categories of abnormally low values versus normal values and for abnormally high values versus normal values. Adjusted relative risks were not computed for the abnormally high versus normal categories for HCT, or for the abnormally low versus normal categories for PLT, due to sparse data. The results of these analyses are given in Table 16-5 and were quite similar to the unadjusted results, with no statistically significant or borderline significant associations found. TABLE 16-5. Adjusted Categorical Analyses for Hematological Variables (Abnormal Versus Normal), Adjusted for Age, Race, Occupation, and Smoking Abnormally Low vs. Normal Variable Abnormally High vs. Normal Adj. Relative Risk (95% C.I.) Adj. Relative Risk (95% C.I.) p-Value p-Value RBC 0.93 (0.59,1.47). 0.762 1.04 (0.48,2.28) 0.920 WBC 0.96 (0.66,1.42) 0.854 0.97 (0.69,1.36) 0.852 HGB 1.12 (0.74,1.80) 0.522 0.98 (0.80,1.19) 0.824 HCT 1.02 (0.51,2.06)' 0.954" MCV 1.08 (0.52,2.26) 0.787 0.99 (0.78,1.26) 0.960 MCH 1.33 (0.56,3.17) 0.525 1.10 (0.79,1.54) 0.574 1.14 (0.66,1.98)" 0.638b PLT "Abnormally low versus normal/abnormally high. —Not analyzed due to sparse data. b Abnormally high versus normal/abnormally low. 16-8 �Adjusted Analyses of Continuous Data General linear regression models were performed, adjusting for age (at the Baseline examination), race, occupation (OCC), smoking history (packyears (PACKYRJ), and current level of smoking (cigarettes per day [CSMOK]). The linear models were fit to examine the main effects of group (GRP) membership, the covariates, and two- and three-factor interactions among these variables (only three-factor interactions involving group were considered). The hierarchical modeling approach as described in Chapter 7, Statistical Methods, was performed to arrive at a "best model" containing the group effect and all statistically significant covariate main effects and interactions. The results of the adjusted analyses for the hematological variables, along with the significance of the adjusting covariates and covariate interactions are summarized in Table 16-6. These results indicated a lack of significant group differences for RBC, HGB, HCT, MCV, MCH, and MCHC after adjustment for five covariates. Two analyses, WBC and PLT, showed significant group-by-covariate interactions; the statistics of these interactions (along with borderline interactions for RBC) are given in Table N-2 of Appendix N, and the narrative descriptions of these interactions are included in the following variable-by-variable summary presentations. Discussion The following variable-by-variable discussion presents the findings for the unadjusted and adjusted results, main covariate effects, group-associated interactions, and when appropriate, Ranch Hand versus Original Comparison contrasts, and comparisons to Baseline results. The results of the covariate effects and covariate interactions (not involving group) for the adjusted analyses are found in Table 16-6; group-by-covariate interactions are given in Table N-2 of Appendix N. Red Blood Cell Count (RBC) Both the categorical and continuous unadjusted analyses found no statistically significant differences in RBC values between groups. The covariate associations for both groups combined showed a significant effect of age (RBC abnormally low in 4.0% of the older cohort versus 1.9% of the younger; p=0.010) and race (Blacks having 6.3% and 4.2% in the abnormally low and high categories versus 2.9% and 0.6% in nonblacks, respectively; p<0.001). Continuous regression analyses also detected significant effects of current smoking (p=0.004) and an age-by-occupation interaction (p=0.013). The adjusted categorical analysis showed no significant group difference, but the adjusted continuous analysis revealed a borderline significant (p=0.086) three-factor interaction of group-by-occupation-by-smoking history. Estimated Ranch Hand-Comparison contrasts revealed a significant difference (p=0.010) for enlisted groundcrew, 30 pack-years with Ranch Hands exhibiting a slightly lower RBC count than the Comparisons (see Table N-2 of Appendix N). 16-9 �TABLE 16-6. Adjusted Continuous Analyses for Hematological Variables, (Ranch Hand-Comparison Group Differences) Variable RBC WBC Ranch Hand-Comparison Group Difference ± SE p-Value 0.172 **** AGE*OCC (p=0.013) CSMOK (p=0.004) **** -0.021±0.015a Covariate Remarks* GRP*RACE*AGE (p=0.005) GRP*AGE*PACKYR (p=0.004) GRP*RACE*OCC (p=0.004) HGB -0.034±0.042 0.410 AGE*OCC (p»0.002) RACE*OCC (p=0.013) CSMOK (p<0.001) HCT -0.151±0.121 0.210 AGE*OCC (p=0.004) RACE*OCC (p=0.003) OCC*PACKYR (p»0.035) CSMOK (p<0.001) MCV 0.108±0.188 0.565 RACE*AGE (p<0.001) RACE*OCC (p-0.015) RACE*CSMOK (p=0.025) MCH 0.062±0.070 0.378 RACE*AGE (p=0.015) CSMOK (p<0.001) OCC (p<0.001) MCHC 0.032+0.026 0.226 RACE (p=0.001) CSMOK (p»0.042) PLT **** **** GRP*RACE*PACKYR (p<0.001) GRP*RACE*CSMOK (p-0.024) OCC (p=0.039) AGE (p-0.006) *Abbreviations OCC; Occupation CSMOK: Current level of smoking (cigarettes per day) GRP: Group PACKYR: Smoking history (pack-years) "Also, borderline significant three-factor interaction (see text). ****Group-by-covariate interaction; group difference, standard error (SE) and p-value not presented. 16-10 �A similar, but slightly weaker interaction was observed in the analysis of the Original Comparisons versus the Ranch Hands. The general finding of insignificant group differences supported the Baseline observations (despite the use of different statistical procedures), but the followup results differed by the mild three-factor interaction. White Blood Cell Count (WBC) The categorical and unadjusted continuous analyses did not disclose any significant differences in WBC levels between the Ranch Hand and Comparison groups. Covariate tests showed a borderline effect of age (with the older cohort having a slightly lower proportion of abnormally low WBC levels—4.2% versus 5.4% in the younger cohort), and the highly significant effects of race (p<0.001), occupation (p=0.001), and smoking history (p<0.001). Blacks had a much higher proportion of abnormally low WBC counts (15.42) versus nonblacks (4.0%); higher proportions of enlisted flyers and enlisted groundcrew personnel (9.1% and 7.2%, respectively) had abnormally high WBC counts versus officers (3.6%). Increasing frequencies of leukocytosis were associated with increasing levels of smoking. The adjusted categorical analysis was nonrevealing with respect to group differences, but the adjusted continuous analysis disclosed three significant three-factor interactions involving group membership: group-by-race-by-age (p=0.005), group-by-age-by-smoking history (pack-years; p=0.004), and group-by-race-by-occupation (p=0.004). Further analyses were conducted stratifying by race (see Table N-2 of Appendix N). Among Blacks, the best model revealed significant group-byoccupation and group-by-age interactions (p-0.045, p=0.024, respectively). Group differences for covariate levels corresponding to young officers and young enlisted flyers were statistically significant, with the adjusted mean WBC value considerably lower in the Ranch Hand group than in the Comparison group. Conversely, the adjusted difference for the older enlisted groundcrew was in the opposite direction. The results for nonblacks were more precise: The group-by-age-by-smoking history interaction was highly significant (p=0.002), with young heavy smokers having a WBC level approximately 12 percent greater in the Ranch Hands than the Comparisons. Other differences were small in magnitude and not statistically significant. Ranch Hand and Original Comparison contrasts were similar for nonblacks, but for Blacks, the group-by-occupation and group-by-age interactions did not reach statistical significance (p=0.077, p=0.134, respectively). The nonsignificance of the unadjusted and categorical adjusted analyses was equivalent to the findings at the Baseline examination. However, possibly due to different model selections, no interactions were noted at Baseline. Race and occupation were not used as covariates at Baseline. Hemoglobin (HGB) None of the four analyses, unadjusted and adjusted categorical tests and unadjusted and adjusted tests of mean differences, detected a significant difference between groups. 16-11 �Covariate tests of association revealed the profound effects of race (8.4% abnormally low in Blacks versus 3.3% in nonblacks; p=0.002), occupation (25.1% and 25.6% abnormally high in enlisted flyers and groundcrew, respectively, versus 16.7% in officers; p<0.001), and smoking history (with proportions of abnormally high HGB levels associated with increases in pack-years of smoking; p=0.003). Continuous analyses detected significant effects of current smoking (p<0.001), occupation-by-age (p=0.002), and occupationby-race (p=0.013) interactions. No significant group-by-covariate interactions were noted. Analysis of the Ranch Hands and Original Comparisons, however, found significant three-factor interactions of group-by-race-by-age (p=0.030) and group-by-race-by-occupation (p=0.020) (see Tables N-7 and N-8 of Appendix N). For equivalent analyses, the followup results were quite analogous to the Baseline study results. Hematocrit (HCT) All of the unadjusted and adjusted categorical tests and analyses of mean differences failed to detect any group differences. Since there were only five abnormally high values, this category was combined with the normal category in the categorical analyses. The association of race to HCT was highly significant, with 4.9 percent abnormally low values noted in Blacks versus 0.9 percent in nonblacks (p<0.001). Regression analyses also detected significant effects of current smoking (p<0.001) as well as age-by-occupation (p-0.004), race-by-occupation (p-0.003), and occupation-by-smoking history (p»0.035) interactions. In both categorical and continuous adjusted analyses, no significant group-bycovariate interactions were detected. Analyses of data from the Ranch Hands and Original Comparisons, however, detected significant three-factor interactions of group-by-race-by-age (p=0.026) and group-by-race-by-occupation (p=0.011) (see Tables N-7 and N-8 of Appendix N). Mean Corpuscular Volume (HCV) No significant group differences were detected for MCV abnormalities or mean values by any of the unadjusted or adjusted analyses. Main covariate effects were profound for age (p<0.001), race (p<0.001), occupation (p«0.004), and 'smoking history (p<0.001). The older cohort had a greater frequency of abnormally high MCV values than did the younger age group (18.0% vs. 9.4%, respectively), and Blacks had a far greater frequency of abnormally low MCV values than nonblacks (7.7% vs. 0.6%, respectively). Enlisted groundcrew personnel had a lower percentage of abnormally high values than officers or enlisted flyers (12.5%, 15.5%, and 17.0%, respectively), and increases in pack-years of smoking were associated with increasing percentages of abnormally high levels (0 pack-years: 4.7%; greater than 0 to 10 pack-years: 13.1%; and greater than 10 pack-years: 21.0%). Continuous analyses detected significant interactions of race-by-age (p<0.001), race-by-occupation (p=0.015), and race-by-current.smoking (p=0.025). The analysis of the Ranch Hand and Original Comparisons revealed a significant group-by-race interaction (p=0.031) for the categorical analyses and significant group-by-age-by-smoking history (p=0.041) and group-by-age-by-current smoking (p=0.012) interactions in the continuous 16-12 �analyses. Various contrasts are given in Table N-8 of Appendix N. No explanations are apparent for these interactions except chance. The followup examination results of MCV (i.e., significant interactions) differed from the Baseline results, which showed a significantly larger adjusted mean MCV value in the Ranch Hands. Mean Corpuscular Hemoglobin (MCH) MCH abnormalities and mean values did not differ significantly by group in any of the unadjusted or adjusted analyses. Main effects were very significant for all of the covariates. The older cohort had a greater frequency of abnormally high MCH values than the younger group (7.9% vs. 3.2%, respectively; p<0.001), while Blacks had a greater frequency of low abnormalities than nonblacks (4.9% vs. 0.3%, respectively; p<0.001). Enlisted groundcrew had a higher proportion of abnormalities in the lower range than enlisted flyers and officers (1.0%, 0.3%, 0.2%, respectively), but they had a lower proportion of high-range abnormalities compared to the other occupations (4.3%, 7.8%, and 7.3%, respectively). The overall p-value was 0.003. Increasing pack-years of smoking was associated with increasing frequencies of high abnormal MCH results (0 pack-years: 2.1%; greater than 0 to 10 pack-years: 6.0%; and greater than 10 pack-years: 8.3%; p<0.001). Continuous analyses detected a significant race-by-age interaction (p=0.015), as well as significant effects of current smoking (p<0.001) and occupation (p<0.001). The followup findings did not support the Baseline observation of significantly increased MCH in the Ranch Hands, although the mean was still higher (both unadjusted and adjusted) in the Ranch Hand group. In the analysis of the Ranch Hands and the Original Comparisons, a significant three-factor interaction of group-by-age-by-current smoking emerged (p=0.026). Table N-8 of Appendix N presents Ranch Hand-Comparison differences for selected covariate levels corresponding to 35- and 53-yearold nonsmokers, one-pack-per-day current smokers, and two-packs-per-day current smokers. The differences were positive for all contrasts except the 53-year-old smokers, when the differences became increasingly more negative with increasing levels of smoking. Mean Corpuscular Hemoglobin Concentration (MCHC) In both groups, only one abnormal MCHC count was recorded for either the abnormally low or abnormally high categories, precluding unadjusted or adjusted categorical tests, and exploration of main covariate effects. No significant group differences were detected by the unadjusted or adjusted tests of MCHC means, although race (p=0.001) and current smoking (p=0.042) were significantly associated with MCHC (higher MCHC in nonblacks and decreasing MCHC associated with increasing current levels of smoking). Similar findings were noted in the analysis of Ranch Hand and Original Comparisons, and overall, the followup findings were comparable to the 1982 Baseline MCHC results. 16-13 �Platelet Count (PLT) Neither the unadjusted nor the adjusted categorical analysis showed statistically significant group differences. Analysis of continuous data disclosed significant effects due to occupation (p=0.039), age (p=0.006), group-by-race-by-smoking history (p<0.001), and group-by-race-by-current smoking (p=0.024) interactions, with higher PLT values in the heavily smoking Ranch Hands but similar values for nonsmokers (see Table N-2 of Appendix N). The significant interactions of group-by-race-by-smoking history (p=0.011) and group-by-age (p=0.040) were also noted for the analyses involving the Original Comparisons (see Table N-8 of Appendix N). The percentages of abnormally high PLT counts increased with increasing packyears of smoking (0 pack-years: 0.8%; greater than 0 to 10 pack-years: 2.0%; and greater than 10 pack-years: 2.6%). Other than the interactions encountered in the adjusted analyses, the overall findings at the followup were comparable to the Baseline PLT results. EXPOSURE INDEX ANALYSES Exposure index analyses were conducted within each occupational cohort of the Ranch Hand group to search for dose-response relationships (see Chapter 8 for details on the exposure index). Log-linear models were fit to the categorical data to examine the effects of exposure and pack-years of . smoking, as well as the interaction between these variables. The normal and abnormally high categories were pooled for the RBC count, and the abnormally low and normal response categories were pooled for MCV, MCH, and PLT due to empty cells in some strata. Because of the small numbers of abnormal values,' analyses were not conducted for HCT or MCHC. The results of the unadjusted categorical analyses are presented in Table 16-7, and the counterpart adjusted analyses are given in Table 16-8. The unadjusted analyses showed only a statistically significant result for the WBC count in the enlisted flyer category, due primarily to an excess of abnormally low values in the high exposure category. The very sparse data support a trend from low to high exposure, and the finding of abnormally low WBC counts associated with exposure is in the direction expected for an herbicide effect. However, the exposure association with abnormally low WBC counts converted to borderline significance (p=0.082) in the adjusted analysis. There were no statistically significant exposure level-by-smoking history interactions. Similar analyses in the other occupational strata (with much larger sample sizes) did not produce this pattern. The unadjusted analysis of means for all eight hematological variables was carried out by a one-way analysis of variance. The results are arrayed in Table 16-9. These analyses revealed only one statistically significant result (p=0.038), the RBC count in the enlisted groundcrew stratum where individuals in the medium exposure category had a higher mean RBC level than those in the low or high exposure categories. Thus, these significant RBC findings did not demonstrate a dose-response relationship. The results for HCT in the enlisted groundcrew stratum were of borderline significance (p=0.052) with the highest mean HCT level in the medium exposure category. In contrast to the categorical analyses, mean WBC levels in the enlisted flyers were not significantly different among the three exposure levels. 16-14 �IMBUE 16-7. Unadjusted Categorical Exposure Index Analyses for Hanatological Variables by Occupation Abnormally Low Normal Abnormally High Vari- Occu- Exposure able pation Index Number Percent Number Percent Number Percent Total p-Value 123 125 114 96.8 96.2 93.4 1 1 2 0.8 0.8 1.6 127 130 122 0.522* 54 64 54 98.2 98.5 94.7 0 0 0 0.0 0.0 0.0 55 65 57 0.401a 3.9 1.2 2.8 148 158 136 96.1 97.5 95.8 0 2 2 0.0 1.2 1.4 154 162 142 0.329* 7 7 5 5.5 5.4 4.1 115 118 110 90.6 90.8 90.2 5 5 7 3.9 3.8 5.7 127 130 122 0.919 Low Medium High 0 1 6 00 . 1.6 10.5 51 59 47 92.7 92.2 82.5 4 4 4 7.3 6.2 7.0 55 64 57 0.045 Enlisted Low Groundcrev Medium High 4 8 7 2.6 4.9 4.9 139 142 125 90.3 87.6 88.0 11 12 10 7.1 7.4 7.0 154 162 142 0.839 Officer Low Medium High 7 2 6 5.5 1.5 4.9 100 106 92 78.7 81.5 75.4 20 22 24 15.8 16.9 19.7 127 130 122 0.425 Enlisted Flyer Low Medium High 3 3 5 5.4 4.6 8.8 36 51 36 65.4 78.5 63.2 16 11 16 29.1 16.9 28.1 55 65 57 0.350 Enlisted Low Groundcrew Medium High 5 4 4 3.2 2.5 2.8 119 110 102 77.3 67.9 71.8 30 48 36 19.5 29.6 25.4 154 162 142 0.352 3 4 6 2.4 3.1 4.9 Low Medium High 1 1 3 1.8 1.5 . 5.3 6 2 4 Officer Low Medium High Enlisted Flyer HGB Lov Median High Enlisted Low Groundcrev Medium High WBC Officer Enlisted Flyer FBC 16-15 �TMttE 16-7. (continued) Unadjusted Categorical Exposure Index Analyses for Banatological Variables by Occupation Abnormally Low Normal Abnormally High Vari- Occu- Exposure able pation Index Number Percent Number Percent Number Percent MCV Officer Low Medium High 1 1 0 Low Medium High 0 0 Enlisted Low Groundcrew Medium High 2 Enlisted Flyer 0 p-Value 87.4 85.4 83.6 15 18 20 11.8 13.8 16.4 127 130 122 050 .8" 43 54 47 78.2 83.1 12 11 10 21.8 16.9 17.5 55 0.764b 65 57 139 8.4 16.0 15.5 154 162 142 0.091b 127 0.916b 0.8 0.8 0.0 111 111 00 . 0.0 0.0 Total 102 82.5 133 90.3 82.1 3 2.1 117 82.4 13 26 22 1 0 0 08 . 00 . 00 . 117 121 112 92.1 93.1 91.8 9 9 10 7.1 6.9 8.2 130 122 00 . 00 . 00 . 51 60 54 92.7 92.3 94.7 4 5 3 7.3 7.7 5.3 55 65 57 0.855b Median High 0 0 0 Enlisted Low Groundcrew Medium 1 2 3 0.6 1.2 2.1 147 151 130 95.4 93.2 91.6 6 154 162 142 0.626b 9 9 3.9 5.6 6.3 2 1 0 1.6 08 . 00 . 120 126 119 94.5 97.7 97.5 5 2 3 3.9 1.6 2.5 127 129 122 047 .8" High Low Medium High 1 0 0 1.8 00 . 00 . 51 64 57 92.7 5.4 1.5 00 . 55 65 57 0.135b 98.5 100 0. 3 1 0 Enlisted Low Groundcrew Medium High MCH 3 1.3 1.8 0 1 0 0.0 152 158 140 98.7 97.5 9. 86 2 3 2 1.3 1.8 1.4 154 162 142 0.914b Officer Low Medium High Enlisted Flyer Low High PLT Officer Enlisted Flyer Low Medium 0.6 00 . 'Normal pooled with.abnormally high. "Abnormally low pooled with normal. 16-16 �TABLE 16-8. Adjusted Categorical Exposure Index Analyses (Log-Linear Models) for Hematological Variables by Occupation (p-Values) Variable Occupation Exposure Index Effect* Smoking History Effect** Exposure Index-bySmoking History RBC Officer Enlisted Flyer Enlisted Groundcrev 0.593 0.552 0.310 0.246 0.364 0.515 0.472 0.981 0.717 WBC Officer Enlisted Flyer Enlisted Groundcrew 0.928 0.082 0.761 0.001 0.121 0.009 0.616 0.971 0.104 HGB Officer Enlisted Flyer Enlisted Groundcrew 0.444. 0.413 0.299 0.393 0.647 0.104 0.424 0.980 0.143 MCV Officer Enlisted Flyer Enlisted Groundcrew 0.718 0.619 0.101 <0.001 0.020 0.028 0.334 0.490 0.574 MCH Officer Enlisted Flyer Enlisted Groundcrew 0.852 0.800 0.681 0.002 0.168 0.288 0.777 0.514 0.530 PLT Officer Enlisted Flyer Enlisted Groundcrew 0.410 0.178 0.910 0.099 0.816 0.363 0.708 0.976 0.996 *Adjusted for smoking history (no interaction), **Adjusted for exposure index (no interaction), 16-17 �TABLE 16-9. Unadjusted Continuous Exposure Index Analyses for Hematological Variables by Occupation (Analysis of Variance) Exposure Index Mean ± SE Occupation Variable Low (n-127) Officer RBC WBC* HGB HCT MCV MCH MCHC PLT* 4.904±0.030 6.488 15.468±0.084 45.379+0.243 92.648±0.430 31.606±0.161 34.090±0.060 253.66 (n-55) Enlisted Flyer RBC WBC* HGB HCT MCV MCH MCHC PLTa 4.972±0.048 7.531 15.785±0.149 46.345±0.425 93.269±0.618 31.782±0.222 34.065+0.075 272.87 (n=154) Enlisted RBC Groundcrew WBC* HGB HCT MCV MCH MCHC PLT* 4.990+0.032 7.185 15.566±0.099 45.740±0.284 91.737±0.399 31.251±0.154 34.032±0.059 270.97 Medium (n»130) 4.861±0.029 6.553 15.463+0.087 45.313+0.255 93.260+0.365 31.851+0.134 34.123+0.060 255.70* (n-65) 4.942±0.037 7.236 15.629±0.110 45.908±0.315 92.923±0.501 31.675±0.187 34.058±0.072 275.34° (n=162) 5.094±0.031 7.236 15.807±0.075 46.580±0.210 91.672±0.404 31.138±0.151 33.941±0.051 273.42 High (n=122) 4.899+0.034 6.753 15.593+0.094 45.791±0.284 93.548±0.367 31.884±0.123 34.067±0.059 256.72 n=129. c n=64. 16-18 0.560 0.512 0.507 0.380 0.252 0.314 0.801 0.799 (n-57) 4.957±0.053 6.966 15.721±0.180 46.300±0.535 93.400±0.572 31.735+0.208 33.956±0.072 261.13 0.894 0.378 0.744 0.717 0.817 0.933 0.508 0.382 (n»142) 4.999±0.033 7.389 15.685±0.090 46.086±0.252 92.376±0.458 31.468+0.166 34.031±0.057 268.27 'Standard errors (SE) not presented, since variables were analyzed on logarithmic scale. b p-Value 0.038 0.686 0.147 0.052 0.436 0.325 0.409 0.748 �SUMMARY AND CONCLUSIONS The functional integrity of the hematopoietic system was assessed by the measurement of eight peripheral blood variables: red blood cell count (RBC), white blood cell count (WBC), hemoglobin (HGB), hematocrit (HCT), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), and platelet count (PLT). These variables were analyzed in the discrete form to detect differences in the percentages of values outside the designated laboratory range, as well as in the continuous form to detect shifts in mean values between the two groups. A summary of all of these analyses, unadjusted and adjusted for the covariates of age, race, occupation, and smoking, is presented in Table 16-12. The unadjusted discrete analysis of the percent abnormal values, both low and high, showed no statistically significant differences between the Ranch Hand and Comparison groups for any of the hematological variables. Similarly, the adjusted categorical analysis disclosed that none of the adjusted relative risks was significant for either group, and that no significant group-by-covariate interactions were present. The unadjusted continuous analysis did not detect any significant differences in group means for any of the eight variables. The adjusted continuous analysis found no significant group differences for HGB, HCT, MCV, MCH, and MCHC, but encountered significant three-factor interactions for WBC (group-by-race-by-age, group-by-age-by-smoking history, and group-byrace-by-occupation), for PLT (group-by-race-by-smoking history and group-byrace-by-current level of smoking), and a borderline interaction for RBC (group-by-occupation-by-smoking history). Ranch Hand versus Original Comparison analyses revealed further significant interactions for HGB, HCT, MCV, and MCH. As no group strata demonstrated consistent patterns of hematologic impairment, biologic relevance was not assigned to the interactions. The covariate effects of age, race, occupation, and smoking history were highly significant for many of the hematological variables. The effect of race was particularly profound for all variables except PLT. There was fair consistency in the covariate effects upon the RBCrelated variables. Generally, decreasing hematologic values were associated with increasing age and the Black race, and increasing hematologic values were associated with increasing smoking. The detection of these classical covariate effects lends credence to the overall finding of nonsignificant group differences for all of the hematological variables. Significant group differences found for MCV and MCH at the Baseline examination were not significant at the first followup. Other differences (e.g., covariate effects, interactions) between the Baseline and followup examinations may be due to small numeric shifts in the cohorts under study (see Chapter 2) and the selection of alternate statistical models, or due to chance. Unadjusted continuous exposure analyses in the Ranch Hand group revealed only one significant effect (RBC in enlisted groundcrew) and one borderline effect (HCT in enlisted groundcrew), but neither was consistent with a plausible dose-response relationship. The adjusted continuous exposure analyses found only one significant contrast (HCT, medium exposure versus low exposure, enlisted groundcrew). However, seven exposure level-by-covariate interactions were noted for four of the hematological variables. Discrete outcome analyses of the exposure level index revealed a significant result only for WBC in the enlisted flyers. 16-21 �TABLE 16-12. Overall Summary Results of Unadjusted and Adjusted Analyses of Hematological Variables Unadjusted Adjusted Mean Categorical Mean Categorical RBC NS NS NS* NS WBC NS NS **** NS HGB NS NS NS NS HCT NS NS NS NS MCV NS NS NS NS MCH NS NS NS NS MCHC NS PLT NS NS — NS **** — NS NS: Not sgnificant (p>0.10). NS*: Borderline -significant group-by-covariate interaction (0.05<p<0.10). —Analysis not performed due to sparse data. ****Group-by-covariate interaction. Note: Significant group-by-covariate interaction, Ranch Hands versus Original Comparisons only, for HGB, HCT, MCV, and MCH. 16-22 �The longitudinal analyses of MCV, MCH, and PLT found significant differences only for PLT values between the Baseline and followup examinations, with the Baseline group difference in mean values closing to near equivalence at the followup examination. In conclusion, none of the eight hematological variables were found to differ significantly between the Ranch Hand and Comparison groups. In fact, group equivalence was more apparent at the followup examination than at the Baseline examination. The classical effects of age, race, and smoking were demonstrated with most of the hematological variables. The longitudinal analyses also suggested that neither group manifested an impairment of the hematopoietic system. Exposure index analyses did not support a plausible dose-response relationship for any of the hematological variables. 16-23 �CHAPTER 16 REFERENCES 1. Allen, J.R., D.A. Barsotti, J.P. Van Miller, L.J. Abrahamson, and J.J. Lalich. 1977. Morphological changes in monkeys consuming a diet containing low levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Fd. Cosmet. Toxicol. 15:401-410. 2. McConnell, E.E., J.A. Moore, and D.W. Dalgard. 1978. Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Rhesus monkeys (Macaca mulatta) following a single oral dose. Toxicol. Appl. Pharmacol. 43(1):175-187. 3. Weissberg, J.B., and J.G. Zinkl. 1973. Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin upon hemostasis and hematologic function in the rat. Environ. Health Perspect. 5:119-123. 4. Kociba, R.J., P.A. Keeler, C.N. Park, and P.J. Gehring. 1976. 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD): results of a 13-week oral toxicity study in rats. Toxicol. Appl. Pharmacol. 35:553-574. 5. Zinkl, J.G., J.G. Vos, J.A. Moore, and B.N. Gupta. 1973. Hematologic and clinical chemistry effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin in laboratory animals. Environ. Health Perspect. 5:111-118. 6. Todd, R.L. 1962. A case of 2,4-D intoxication. J. Iowa Med. Soc. 52:663-664. 7. May, G. 1973. Chloracne from the accidental production of tetrachlorodibenzodioxin. Br. J. Ind. Med. 30:276-283. 8. Pocchiari, F., V. Silano, and A. Zampieri. 1979. Human health effects from accidental release of tetrachlorodibenzo-p-dioxin (TCDD) at Seveso, Italy. Ann. N.Y. Acad. Sci. 320:311-320. 9. Suskind, R.R., and V.S. Hertzberg. 1984. Human health effects of 2,4,5-T and its toxic contaminants. JAMA 251:2372-2380. 10. Moses, M., R. Lilis, K.D. Crow, J. Thornton, A. Fischbein, H.A. Anderson, and I.J. Selikoff. 1984. Health status of workers with past exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in the manufacture of 2,4,5trichloro-phenoxyacetic acid: comparison of findings with and without chloracne. Am. J. Ind. Med. 5:161-182. 11. Stehr, P.A., G. Stein, H. Falk, et al. 1986. A pilot epidemiologic study of possible health effects associated with 2,3,7,8-tetrachlorodibenzo-p-dioxin contamination in Missouri. Arch. Environ. Health 41:16-22. 16-24 �12. Hoffman, R.E., P.A. Stehr-Green, K.B. Webb, G. Evans, A.P. Knutsen, W.F. Schramra, J.L. Staake, B.B. Gibson, and K.K. Steinberg. 1986. Health effects of long-terra exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. JAMA 255:2031-2038. 16-25 �CHAPTER 17 RENAL ASSESSMENT INTRODUCTION Renal dysfunction and overt renal disease are not considered to be important clinical sequelae of exposure to phenoxy acids, chlorophenols, or TCDD. In man and animals, 2,4-D, 2,4,5-T, and TCDD are excreted by the kidney, largely in the unmetabolized state via a first-order kinetic process. Excretion of these compounds appears to be a function of the proximal convoluted tubules. " In experimental animals, renal damage is generally noted only when very high or lethal doses of TCDD have been administered, an observation that reflects the severe systemic toxicity of TCDD as contrasted to a doubtful role of primary nephrotoxicity. A variety of experimental pharmacokinetjc studies have been conducted in man using both ingested 2,4-D and 2,4,5-T. (1 ' Most of these studies suggested an unconjugated excretion of these compounds by first-order kinetics. No acute deleterious effects, as detected by urinalysis or blood chemistries, were either noted or recorded for the volunteer subjects. In contrast, following significant exposure to a horse arena filled with TCDD-contaminated waste products, a 6-year-old girl developed hemorrhagic cystitis, pyelonephritis, and proteinuria. Horses exposed to this arena and other contaminated arenas also frequently manifested hematuria. A thorough 5-year followup examination of the young girl was essentially normal and did not reveal any renal sequalae. Most dioxin morbidity studies have only briefly mentioned renal disease and function, and then in the context of routine data collected at physical examination rather than as a specific clinical focus. Some studies of significant occupational exposure have been almost devoid of commentary on renal dysfunction.1 ~1 A contemporary study of a residentially exposed cohort showed negative renal findings. The Times Beach, Missouri, pilot study demonstrated historical "trends" of increased urinary tract disease by questionnaire, along with a compatible pattern of leukocyturia and hematuria manifest at physical examination, but none of the observations was statistically significant. The Monsanto industrial morbidity studies reported essentially negative urinalysis findings, although data were not presented. Baseline Summary Results The 1982 Baseline examination assessed renal disease and function by questionnaire and basic urinalysis testing. 17-1 �Based on questionnaire information, the Ranch Hand group reported significantly more kidney disease than the Comparisons (p=0.039), but this finding was not substantiated by laboratory test results, even when all abnormalities were summed over the five tests of BUN, creatinine clearance, presence of occult blood, five or more urine WBC's per high-power field (HPF), and the presence of urine protein. The Comparison group manifested a twofold increase in proteinuria (p-0.055). The distributions of creatinine clearance levels were similar in both groups, as were the means of the BUN, urine specific gravity, and WBC's/HPF. Difficulty in assessing the degree and significance of hidden noncompliance to the full 24-hour urine collection made the interpretation of the creatinine clearance test results somewhat problematic. Of some interest, known noncompliance to urine collection was observed much more frequently (p<0.001) in the elderly participants. Of 18 herbicide exposure analyses, only 1 (enlisted flyer category) was statistically significant vis-a-vis a history of kidney disease, and it did not demonstrate a linear increase from low to high exposure. The validity of the renal assessment was reinforced by the demonstrated effects of the covariates of age (born in or after 1942, born before 1942) and 2-hour status after postprandial glucose levels (less than 120 mg/dl, greater than or equal to 120 mg/dl). Blood urea nitrogen increased with age and specific gravity decreased (p<0.001 for both), while an abnormally high postprandial glucose level indicative of diabetes was associated only with an increasing urine specific gravity, as expected. Overall, the Baseline renal assessment suggested an excess of historical kidney disease in the Ranch Hand group that was not corroborated by laboratory urinalysis testing. Parameters of the 1985 Renal Assessment Because of the essentially negative Baseline results, the fact that kidney disease is not a prime clinical endpoint, and the manifest compliance problems with a 24-hour urine collection, the 1985 examination process did not emphasize further inquiry into renal disease and function. The onsite NORC questionnaire did not specifically probe for a 1982-1985 interval history of kidney disease, although severe cases may be captured by the generic question, "any other major condition?" or by a detailed extraction of review-of-systerns data obtained at the physical examination. Laboratory testing parameters included all the Baseline dependent variables except the creatinine clearance level (omitted because the plasma creatinine assay was deleted from the test battery). Also, the analysis of composite renal abnormalities was deleted. In addition, the 24-hour urine collection was reduced to a 12-hour collection (5:30 a.m. to 5:30 p.m.) to ease participant burden while still maintaining validity for the porphyrin analyses (see Chapter 13). The accuracy of the 12-hour urine collection was not assessed during the 1985 examination. Renal data analyses paralleled the Baseline analysis except for deleting one of the dependent variables and a composite analysis, adding the covariate of race, and defining the covariate of diabetic class as diabetic, impaired, or normal. No clinical exclusion categories applied to the renal analysis. Minor numerical differences in the tables are due to rare missing dependent 17-2 �variable or covariate data. Adjusted statistical analyses using the above covariates were based on 1,016 Ranch Hands and 1,293 Comparisons and used logistic regression and analysis of covariance methods. When age was used as a covariate in the logistic regression models, the continuous form was used mathematically, but for summary table purposes, age is displayed as a dichotomy. Parallel analyses using the Original Comparisons can be found in Appendix 0 (see Tables 0-3 through 0-5). Tests of association between dependent variables and covariates emphasized Fisher's exact test and Pearson's chi-square test for discrete dependent variables and t-tests and analysis of variance techniques for continuous dependent variables. RESULTS AND DISCUSSION Questionnaire Data History of renal disease was assessed by a self-administered review-ofsystems question list at the physical examination. Specific structured questions on renal disease were not incorporated in the NORC questionnaire. The review-of-systems questions, i.e., "kidney trouble?" "kidney stones?" were open-ended with respect to time, and reflected conditions that arose at any time in the past. These questionnaire data did not show a significant difference between the Ranch Hand and Comparison groups, as reflected by the analysis in Table 17-1. Tests of association between the historical presence of kidney disease in both groups and the covariates of race, occupation, diabetes, and age are given in Table 17-2. TABLE 17-1. Unadjusted Analysis of History of Kidney Disease/Kidney Stones by Group History of Kidney Disease/Stones Yes Group Number No Percent Number Percent Est. Relative Total Risk (95% C.I.) p-Value Ranch Hand 94 9.3 920 90.7 1,014 Comparison 128 9.9 1,163 90.1 1,291 0.93 (0.70,1.23) 0.619 17-3 �TABLE 17-2. Association Between Kidney Disease/Kidney Stones and Age, Race, Occupation, and Diabetic Class in the Combined Ranch Hand and Comparison Groups History of Kidney Disease/Stones Yes Covariate Covariate Category Age Born XL942 Born <1942 66 156 Race Nonblack Black Number Percent Occupation Officer Enlisted Flyer Enlisted Groundcrew Diabetic* Class No Diabetic Impaired Normal Number Percent Total p-Value 6.9 11 .6 894 1,189 93.1 88.4 960 1,345 <0.001a 214 8 9.9 5.6 1,949 134 90.1 94.4 2,163 142 0.106 83 9.6 781 90.4 864 36 9.3 350 90.7 386 103 9.8 952 90.2 1,055 14 41 166 8.0 14.5 9.0 161 242 1,677 92.0 85.5 91.0 175 283 1,843 0.969* 0.011 Fisher's exact test. b Pearson's chi-square test. *Unable to classify four participants, due to missing 2-hour postprandial glucose level and no historical evidence of diabetes. These results showed that there was no significant effect due to race or occupation. In contrast, there was a significant effect due to diabetic class (p=0.011), with participants in the impaired diabetic class having a significantly higher proportion of past kidney disease than those in the normal or diabetic classes-. Older participants also had a significantly higher history of past renal events than younger participants (p<0.001). A logistic regression analysis of the history of kidney disease and kidney stones using the above four covariates gave a result very similar to the unadjusted analysis (Adj. RR: 0.95, 95% C.I.: [0.71,1.25], p=0.693). Race and occupation were not significant covariates. However, diabetic class and age were significant covariates (p=0.041 and p<0.001, respectively). These analyses showed that there was no difference in the history of renal disease between the Ranch Hand and Comparison groups, and that the 17-4 �proportions of past kidney disease and kidney stones were significantly influenced by age and diabetic class. While these findings are consistent with traditional expectations in renal disease, they were in direct contrast to the findings of the 1982 Baseline examination, which revealed a significant excess of historical kidney disease in the Ranch Hand group, and group data that were not influenced by age or glucose levels. It is concluded that there were no significant group differences in past renal disease. Physical Examination Data No physical examination procedures were used to evaluate the renal system as most procedures are invasive and beyond the scope of this voluntary examination. Accordingly, the renal system was evaluated primarily by laboratory data. Laboratory Data Five renal variables were quantitated by general laboratory procedures to assess nonspecific renal system function. The presence or absence of urine protein was determined by standard reagent strip testing. Hematuria and leukocyturia were measured by high-power microscopic examination after centrifugation for 5 minutes. Urine specific gravities were measured by Ames' Multisticks; those urines exceeding normal limits were remeasured by standardized refractometers. BUN levels were assayed by a DuPont Automated Chemical Analyzer, model 500. The SCRF laboratory normal values from these variables are given in Table 17-3. TABLE 17-3. Laboratory Norms for Five Renal Variables Renal Variable Normal Abnormal Urine Protein Occult Blood WBC/HPF BUN (mg/dl) Specific Gravity Absent Absent <2 .7-22 1.005-1.03 Present >1 RBC/HPF >2 >23 <1.004 In this section, urinary protein, hematuria, and leukocyturia were analyzed as discrete variables, whereas BUN and urine specific gravity were analyzed as continuous variables. The number and percent of subjects with abnormal values for the discrete variables are displayed in the summary Table 17-4, along with the number of participants, the unadjusted means, and standard errors of the continuous variables. 17-5 �TABLE 17-4. Summary of Renal Laboratory Variables by Group Group Comparison Ranch Hand Renal Variable Number Abnormal Urine Protein Occult Blood WBC/HPF 37 182 102 Percent Abnormal 3.6 17.9 10.0 Number Abnormal 40 208 107 Percent Abnormal 3.1 16.1 8.3 Unadjusted pr-Value 0.485 0.239 0.145 Renal Variable Unadjusted Mean Standard (Sample Size) Error Unadjusted Mean Standard (Sample Size) Error Unadjusted p-Value BUN (mg/dl) Specific Gravity 14.21* (1,016) 1.0157 (1,016) 0.0002 14.30* (1,293) 1.0152 (1,292) 0.0002 0.554 0.082 *Arithmetic mean calculated on square root scale and transformed to original units. —Standard error not given, since analysis performed on square root scale. The following statistical power statements apply to several variables displayed in Table 17-4. At a standard a -level of 0.05 and a power of 0.80, the sample sizes were sufficient to detect a 1.28-fold increase in the frequency of percent abnormal values for urinary occult blood, and a 1.43-fold increase in the percentage of leukocyturia, both over that observed in the Comparison group. Further, the sample sizes were adequate to reveal a 2.9 percent mean shift in the BUN value relative to the mean observed in the Comparison group. Urinary Protein As displayed in Table 17-4, the Ranch Hand group had a prevalence rate of urinary protein of 3.6 percent versus 3.1 percent in the Comparison group (Est. RR: 1.18, 95* C.I.s [0.75,1.86], p=0.485). This difference was not significant. Tests of association were conducted with pooled participant data using the covariates of race, occupation, diabetic class, and age. These tests are presented in Table 17-5. 17-6 �TABLE 17-5. Association Betveen Urinary Protein and Age, Race, Occupation, and Diabetic Class in the Combined Ranch Hand and Comparison Groups Presence of Urinary Protein Yes No Covariate Covariate Category Number Percent Number Percent Age Born XL942 Born <1942 34 43 3.5 3.2 927 1,304 96.5 96.8 961 1,347 0.641* Race Nonblack Black 65 12 3.0 8.4 2,100 131 97.0 91.6 2,165 143 0.002" Occupation Officer 18 Enlisted Flyer 11 Enlisted Groundcrew 48 2.1 845 97.9 863 2.8 376 97.2 387 4.5 1,010 95.5 1,058 20 18 39 11.4 6.4 2.1 155 264 1,808 88.6 93.6 97.9 175 282 1,847 Diabetic* Class Diabetic Impaired Normal Total p-Value 0.010b <0.001b Fisher's exact test. b Pearson's chi-square test. *Unable to classify four participants, due to missing 2-hour postprandial glucose level and no historical evidence of diabetes. These results suggested no age effect, but significant associations for the covariates of race (p*0.002), occupation (p=0.010), and diabetic class (p<0.001) were noted. The significant covariate effects were attributable to higher percentages of urinary protein abnormalities in Blacks versus nonblacks, enlisted groundcrew versus officers or enlisted flyers, and diabetes (past history [unverified] or greater than or equal to 200 mg/dl glucose) versus impaired glucose tolerance (at least 140 but less than 200 mg/dl glucose) versus normal glucose tolerance (less than 140 mg/dl glucose). The prevalence rates of urinary protein abnormalities were adjusted by logistic regression models using the above four covariates. Race and occupation demonstrated significant effects (p=0.023 and p=0.023, respectively), while age did not (p=0.294). Because of a significant interaction between group and diabetic class (p=0.047), stratified analyses were conducted to provide further clarification. The results are shown in Table 17-6. 17-7 �The adjusted relative risk, 95 percent confidence interval, and group p-value for each diabetic class are shown in Table 17-7. TABLE 17-6. Frequency of Urinary Protein by Diabetic Class and Group Presence of Urinary Protein No Yes Diabetic Class Group Number Percent Number Percent Total Diabetic Ranch Hand Comparison 7 13 9.0 13.4 71 84 91.0 86.6 78 97 Impaired Ranch Hand Comparison 5 13 4.7 7.4 101 163 95.3 92.6 106 176 Normal Ranch Hand Comparison 25 14 3.0 1.4 807 1,001 97.0 98.6 832 1,015 TABLE 17-7. Adjusted Relative Risks for Urinary Protein by Diabetic Class Diabetic Class Diabetic Impaired Normal Adjusted Relative Risk 0.66 0.66 2.23 95% C.I. (0.25, 1.77) (0.23, 1.93) (1.15, 4.32) p-Value 0.414 0.453 0.018 This analysis showed that the estimated prevalence of urinary protein is lower in the Ranch Hand group than in the Comparison group for the diabetic and glucose-impaired strata. Conversely, for the normal diabetic class, the Ranch Hand group manifested a significant increased prevalence of positive urinary protein as contrasted with the Comparison group. These followup examination results were different from the 1982 Baseline examination, which showed significantly more proteinuria in the Comparison group. The prevalence of proteinuria in the followup examination was about 75 percent higher than the prevalence observed in the Baseline study. The interaction of group and diabetic class suggested Ranch Hand increases in proteinuria for normal glucose tolerance participants. 17-8 �Urinary Occult Blood Hematuria was determined by microscopic examination. For both groups combined, the frequency distribution of RBC count data was: 0 RBC/HPF, 82.15 percent; 1-2 RBC/HPF, 15.13 percent; 3-5 RBC/HPF, 2.03 percent; and greater than 5 RBC/HPF, 0.69 percent. As noted in Table 17-4, the prevalence of urinary occult blood in the Ranch Hand group (17.9%) was slightly higher than the rate observed for the Comparison group (16.1%). The unadjusted analysis showed no significant group differences for occult blood (Est. RR: 1.14, 95% C.I.: [0.91,1.42], p=0.239). Tests of association with the covariates of race, occupation, diabetic class, and age were conducted using combined group data for urinary occult blood, and these results are given in Table 17-8. TABLE 17-8. Association Between Urinary Occult Blood and Age, Race, Occupation, and Diabetic Class in the Combined Ranch Hand and Comparison Groups Presence of Urinary Occult Blood Yes Covariate Covariate Category Number No Percent Number Percent Total p-Value Age Born XL942 Born <1942 148 242 15.4 18.0 812 1,105 84.6 82.0 960 1 ,347 0.115a Race Nonblack Black 355 35 16.4 24.5 1,809 108 83.6 75.5 2 ,164 143 0.016* Occupation Officer Enlisted Flyer Enlisted Groundcrew 118 13.7 745 86.3 863 .76 19.6 311 80.4 387 196 18.5 861 81.5 1 ,057 Diabetic Class* 33 52 305 18.9 18.5 16.5 142 229 1,542 81.1 81.5 83.5 175 281 1 ,847 Diabetic Impaired Normal 0.005b 0.296b "Fisher's exact test. Pearson's chi-square test. *Unable to classify four participants, due to missing 2-hour postprandial glucose level and no historical evidence of diabetes. 17-9 �As reflected in Table 17-8, there was no significant effect due to diabetic class or age. However, Blacks had a significantly higher prevalence of urinary occult blood than nonblacks (p=0.016), and significant effects were also due to occupation (p=0.005), with officers having a lower proportion of positive occult blood determinations than enlisted personnel. An adjusted analysis of urinary occult blood proportions was conducted by logistic regression techniques. Multiple significant three-factor interactions were noted, e.g., group-by-occupation-by-race (p=0.008), groupby-age-by-diabetic class (p=0.045), and group-by-occupation-by-diabetic class (p=0.017). Consequently, a series of analyses stratified by race were performed to determine adjusted relative risks for nonblacks and Blacks separately. The adjusted results for nonblack participants are given in Table 17-9. TABLE 17-9. Adjusted Analysis for Urinary Occult Blood for Nonblacks by Group Presence of Urinary Occult Blood Yes Group Ranch Hand Comparison Number Percent 166 189 17.4 15.6 No Number Percent Total 789 1,020 82.6 84.4 955 1,209 Summary Statistics Adj. RR: 1 .13 95% C.I.: (0.91,1.42) p- Value: 0.291 The covariates of occupation and age contributed significant effects (p<0.001 and p=0.002, respectively) to this analysis. Diabetic class was not significant (p=0.863), and was consequently not included in the final model. No significant group differences were found (p=0.291). Table 17-10 shows the frequencies for Black participants. The adjusted analysis of the data on Blacks showed a significant interaction of group and occupation (p=0.003). Table 17-11 presents frequencies and percents for the presence of urinary occult blood for each group, stratified by occupation. This table demonstrates that the group-by-occupation interaction for Blacks was due to the Ranch Hand officers having a lesser prevalence of occult blood abnormalities than Comparison officers, while conversely, Ranch Hand enlisted personnel showed a.higher prevalence of abnormalities than enlisted Comparisons. Because of the absence of hematuria in Black Ranch Hand officers, no relative risk was calculated. Consequently, the Black enlisted occupational categories were combined and investigated further through logistic regression techniques. This analysis did not show a difference of urinary occult blood percentages in the Ranch Hand Black 17-10 �TABLE 17-10. Frequency of Urinary Occult Blood for Blacks by Group Presence of Urinary Occult Blood Yes No Number Percent Group Ranch Hand Comparison 16 19 26.7 22.9 Number Percent 44 64 73.3 77.1 Total 60 83 TABLE 17-11. Frequency of Urinary Occult Blood for Blacks by Occupation and Group Presence of Urinary Occult Blood Yes Occupation Officer Enlisted Flyer Enlisted Groundcrew Group Number No Percent Number Percent Total Ranch Hand Comparison 0 3 0.0 42.9 7 4 100.0 57.1 7 7 Ranch Hand Comparison 3 1 30.0 5.9 7 16 70.0 94.1 10 17 Ranch Hand Comparison 13 15 30.2 25.4 30 44 69.8 74.6 43 59 enlisted and the Comparison Black enlisted strata (Est. RR: 1.62, 95% C.I.: [0.73,3.63], (p«0.239). The effects of age (p=0.817), occupation (p-0.171), and diabetic class (p=0.145) were not statistically significant, and were not included in the final adjusted analysis. In conclusion, both unadjusted and adjusted stratified analyses (by race) did not reveal a consistent and plausible excess of hematuria in the Ranch Hand group. The tenfold or greater increase in the cross-sectional prevalence of hematuria compared to the Baseline examination (1.3% of both groups) to this followup examination may be due to a different sensitivity of the laboratory techniques of reagent-strip testing versus microscopic observation. Nonetheless, an approximate prevalence of 17 percent hematuria merits reevaluation at the next followup examination. 17-11 �Urinary White Blood Cell Count Leukocyturia was assessed by microscopic examination. As noted in Table 17-3, more than two white blood cells per high-power field (WBC/HPF) were considered abnormal by the SCRF laboratory. This is in distinct contrast to the cutpoint of five WBC/HPF used at the Baseline examination. Table 17-4 shows the group frequencies of abnormal urine WBC's. The unadjusted analysis revealed a nonsignificant group effect (Est. RR: 1.24, 95% C.I.i [0.93,1.64], p=0.145). Tests of association were conducted between the frequency of abnormal WBC counts in both groups and the covariates of race, occupation, diabetic class, and age. The results revealed a significantly higher prevalence of abnormal counts for Blacks than nonblacks (p<0.001), an effect due to occupation (p=0.023), with a lower prevalence of abnormalities for officers than enlisted personnel and an effect due to diabetic class (p=0.046), with a lower prevalence of abnormal WBC counts in the normal diabetic class than in either the impaired or diabetic classifications. Age was noncontributory (p«0.508). Adjusted analyses of leukocyturia by group were performed by logistic regression techniques. A significant three-way interaction for group, age, and race was detected (p=0.004), requiring further stratified analyses. A summary of the frequencies £or nonblacks is presented in Table 17-12. TABLE 17-12. Frequency of Urinary WBC/HPF for Nonblacks by Group Urinary WBC/HPF Abnormal Group Ranch Hand Comparison Number 92 88 Normal Percent Number 9.6 7.3 864 1,121 Percent 90.4 92.7 Total 956 1,209 The logistic regression adjustment of the data for nonblacks showed significant covariate effects for occupation (p»0.046) and diabetic class (p=0.031), and a significant interaction between group and age (p=0.018). Consequently, additional analyses were conducted stratifying by age (born in or after 1942, born before 1942), and are shown in Table 17-13. 17-12 �TABLE 17-13. Adjusted Analyses for Urinary WBC/HPF for Nonblacks by Age Category and Group Urinary WBC/HPF Abnormal Age Group Normal Number Percent Number Percent Total Summary Statistics Born XL942 Ranch Hand Comparison 41 24 10.8 4.8 339 478 89.2 95.2 380 502 Adj. RR: 2.42 95* C.I.: (1.43,4.09) p-Value: 0.001 Born <1942 Ranch Hand Comparison 51 64 8.9 9.1 525 643 91.1 90.9 576 707 Adj. RR: 0.99 95* C.I.: (0.67,1-46) p-Value: 0.956 As depicted by the above table, the adjusted rate of nonblack young Ranch Hands with abnormal urinary white blood cell counts was significantly greater than that for nonblack Comparisons (p=0.001 adjusted for occupation and diabetic class). Demonstrating the interaction involving age and group, the adjusted rate of nonblack older Ranch Hands with abnormal urinary WBC counts was nonsignificant and less than older nonblack Comparisons (p=0.956 adjusted for occupation and diabetic class). Similar analyses were conducted for Black participants. Rates of abnormal urinary white blood cell count levels were 16.7 percent and 22.9 percent (n=60 and 83) for Black Ranch Hands and Black Comparisons, respectively. Significant interactions involving group and occupation (p=0.002) and group and age (p^O.OOl) were found. Additional analyses stratified by occupation were performed. Frequencies stratified by occupation are shown in Table 17-14. This table clearly shows how the proportions of WBC abnormalities vary by group within the various occupational categories. However, because of the lack of abnormalities in the Black Ranch Hand officer stratum, an adjusted relative risk was not calculated for this occupation. Thus, Black enlisted categories were combined and subjected to further logistic regression techniques. The analysis showed yet another interaction, between group and age (p=0.026), requiring an additional stratification by age. Results of these analyses are presented in Table 17-15. 17-13 �TABLE 17-14. Frequency of Urinary WBC for Blacks by Occupational Category and Group Urinary WBC/HPF Count Abnormal Occupation Number Percent Group Normal Number Percent Total Officer Ranch Hand Comparison 0 2 0.0 28.6 7 5 100.0 71.4 7 7 Enlisted Flyer Ranch Hand Comparison 3 3 30.0 17.6 7 14 70.0 82.4 10 17 Enlisted Groundcrew Ranch Hand Comparison 7 14 16.3 23.7 36 45 83.7 76.3 43 59 TABLE 17-15. Adjusted Analyses for Urinary WBC/HPF for Black Enlisted Flyers and Groundcrew by Age and Group Urinary WBC/HPF Count Abnormal Age Group Normal Number Percent Number Percent Total Summary Statistics Born XL942 Ranch Hand Comparison 4 13 13.8 28.3 25 33 86.2 71.7 29 Adj. RR: 0.41 46 95% C.I.: (0.12,1.40) p-Value: 0.153 Born <1942 Ranch Hand Comparison 6 4 25 .0 13 ,3 18 26 75 .0 86 .7 24 Adj. RR: 2.17 79) 30 95% C.I. : (0.53,8. p-Value: 0.279 17-14 �In the presence of relatively small sample sizes, these results demonstrated that the prevalence of abnormal urinary white cell counts in Black enlisted personnel did not vary significantly by group for either age category, although the reversal of group proportions for different ages was prominent and fully reflective of the group-by-age interaction. It is noted that the Black group-by-age interaction is opposite the nonblack group-by-age interaction (see Table 17-13), explaining the significant three-way interaction involving group, age, and race. In summary, the unadjusted analysis of urinary WBC/HPF abnormalities showed no group differences, but the adjusted analyses showed significant effects for diabetic class and occupation for nonblack enlisted participants, and a group-by-age interaction for both Black and nonblack enlisted participants. Only for younger nonblack participants was a significant group effect seen (Ranch Hands>Comparisons). The observations from this examination were consistent with the negative Baseline findings. Blood Urea Nitrogen (BUN) BUN was analyzed as a continuous variable using two sample t-tests, analysis of variance, and analysis of covariance techniques. The data were transformed to the square root scale for analysis. Adjusted analyses used the covariates of race, occupation, diabetic class, and age, as in analysis of discrete dependent variables. As noted in Table 17-4, unadjusted group summary statistics revealed no significant differences in mean BUN levels (p=0.554). The groups were combined and contrasted to the covariates, and results are presented below. These tests of covariate association showed a significant racial effect (p=0.007), with a higher mean BUN level for nonblacks than Blacks; a significant effect for occupation (p<0.001), with officers having a higher mean level than both enlisted categories; a significant age effect (p<0.001), with a higher mean BUN level for older than for younger participants; and a marginally significant (p=0.059) difference due to diabetic class, with participants in the impaired category having the highest mean BUN level. An analysis of covariance using the above four covariates demonstrated the significant effects of age (p<0.001), occupation (p=0.015), and significant group-by-race (p=0.022) and race-by-diabetic class (p=0.024) interactions. Table 17-16 presents mean BUN values, adjusted by the covariates and covariate interactions, stratified by race. Test results for the equality of adjusted means between groups are given in the p-value column. As noted from this table, Black Comparisons had a significantly higher adjusted mean BUN level than Black Ranch Hands (p=*0.017), and there was no group difference for nonblacks. These results were analogous to the findings at the Baseline examination (although race was not used as a covariate), i.e., no detriment to the Ranch Hand group and a significant covariate effect of age. 17-15 �TABLE 17-16. Adjusted Analysis of BUN by Race and Group Race Group Total Adjusted Mean* p-Value Nonblack Ranch Hand Comparison 956 1,206 14.15 14.17 0.907 Black Ranch Hand Comparison 60 83 12.40 13.75 0.017 *Converted from square root scale. Urinary Specific Gravity The unadjusted means of the urine specific gravity disclosed a marginally significant difference between the Ranch Hand and Comparison groups (p=0.082). The summary statistics of the unadjusted analysis are given in Table 17-4. By t-tests and analysis of variance, tests of association were performed on the combined groups using the covariates of race, occupation, diabetic class, and age. These tests showed a significant effect of occupation (p<0.001), with officers having the lowest mean urine specific gravity and the enlisted groundcrew category having the highest, and a significant effect (p-0.018) due to diabetic class, with the diabetic category having the highest specific gravity and the normal (nondiabetic) class having the lowest mean value. The effects of age and race were not statistically significant (p=0.382 and p=0.065, respectively). An analysis of covariance with these four covariates showed significant effects due to diabetic class (p=0.019), and significant group-by-race (p=0.017) and group-by-occupation (p=0.034) interactions/ Adjusted group mean specific gravities were stratified by race and by occupation. The results are presented in the summary Table 17-17. These stratified group data showed a difference for nonblack enlisted groundcrew, but Comparisons had a lower adjusted mean urine specific gravity level than Ranch Hands (low specific gravity representing renal dysfunction). Noteworthy is the contrast of results between this followup examination and the Baseline examination in 1982. The urine specific gravities of the followup examination appeared to be very substantially lower than those of the Baseline. A probable explanation was the difference in methods of assessing specific gravity. At the Baseline, the Ames' Clinilab automated procedure (falling drop) was used, as contrasted to the Ames' Multistick procedure at the followup. Both examinations used specimens obtained early on the second examination day, and did not use aliquots of 12- or 24-hour urine collections that were used for the porphyrin analyses. Although the 17-16 �TABLE 17-17. Adjusted Analysis of Urine Specific Gravity by Race, Occupation, and Group Group Total Adjusted Mean p-Value Ranch Hand Comparison 373 474 1.0153 1.0151 0.734 Ranch Hand Comparison 167 193 1.0158 1.0161 0.631 Ranch Hand Enlisted Groundcrew Comparison Race 416 538 1.0174 1.0157 <0.001 Occupation Nonblack Officer Enlisted Flyer Officer Ranch Hand Comparison 7 7 1.0158 1.0186 0.462 Enlisted Flyer Ranch Hand Comparison 10 17 1.0144 1.0158 0.624 Enlisted Groundcrew Black Ranch Hand Comparison 43 59 1.0162 1.0183 0.157 covariate effect of age upon specific gravity was not observed at the followup as it had been at the Baseline, both examinations demonstrated the marked effect of diabetes upon specific gravity, i.e., a higher specific gravity was detected in diabetics than in nondiabetics. EXPOSURE INDEX ANALYSES Exposure index analyses were conducted within each occupational cohort of the Ranch Hand group to search for dose-response relationships (see Chapter 8 for details on the exposure index). The variables of kidney disease, urinary protein, urinary occult blood, and urinary white blood cell count were investigated (unadjusted for any covariates) using Pearson's chi-square test and Fisher's exact test. Adjusted analyses were performed by logistic regression for these variables, using age, race, diabetic class, and any significant pairwise interactions between the exposure index and these covariates. Overall significance in the proportion of abnormalities among the exposure index levels of low, medium, and high was determined, as well as contrasts of the proportion of abnormalities between medium and low exposure levels, and between the high and low exposure levels. Age was used as a continuous variable in the adjusted analyses, and dichotomized (born in or after 1942, born before 1942) when age was involved in an interaction with the exposure index. 17-17 �Analyses of mean blood urea nitrogen and urine specific gravity (continuous variables) were performed, unadjusted for any covariates or interactions, using analysis of variance techniques and t-tests. Analysis of covariance models were used in adjusted analyses. Contrasts of medium versus low exposure and high versus low exposure were also studied. A square root transformation was applied to the 'blood urea nitrogen data. Results of the adjusted analyses for these six variables are presented in Tables 17-18 and 17-19, and counterpart results for unadjusted analyses are presented in Table 0-1 of Appendix 0. Results from further investigation of exposure index-by-covariate interactions are given in Table 0-2 of Appendix 0. Unadjusted analyses revealed no significant differences among exposure index levels for any occupation. Further investigation of these variables, for which the medium versus low and the high versus low contrasts were also examined, revealed only two variables having borderline significance: kidney disease in enlisted flyers, high versus low (Est. RR: 0.25, 95% C.I.: [0.05,1.26], psO.091), and urinary occult blood in enlisted groundcrew, high versus low (Est. RR: 1.77, 95% C.I.: [1.00,3.13], p=0.061). The results for urinary occult blood in enlisted groundcrew supported an increase in the proportion of abnormalities from low to high exposure, whereas the kidney disease data showed the opposite effect. The frequency of abnormalities (or mean levels closer to the abnormal range for continuous variables) for the different exposure index levels exhibited no graduated pattern across exposure levels. The number of combinations for which the medium exposure level had the smallest proportion of abnormalities (or more abnormal mean level) was greater than the other exposure levels. Adjusted analyses revealed no significant differences among exposure index levels for any occupational stratum. Interactions were present for four of the six variables, however, and were observed in all occupations. A summary of these interactions is presented in Table 17-20. No interaction patterns in either the covariates or occupations were observed. The only contrast observed approaching significance for an adverse effect at higher exposure levels was observed for urinary protein (officers in normal diabetic class, high versus low, p=0.097), but this contrast was highly affected by sparse cell sizes (see Table 0-2 of Appendix 0). In summary, six renal variables showed no evidence of an increasing dose-response relationship at the followup examination. No patterns in the relationship of prevalence rates among the exposure index levels were seen within occupational strata. The exposure index level patterns observed at the Baseline examination for kidney disease in the enlisted flyer stratum were not seen at the first followup examination. Overall, both the Baseline and followup examinations showed very little evidence of a dose-response relationship. 17-18 �TABLE 17-18. Adjusted Categorical Exposure Index Analyses for Renal Variables by Occupation Exposure Index Variable Occupation Officer Low Total 127 Kidney Disease Enlisted Flyer • Enlisted Groundcrew 153 Officer 127 Urinary Protein Enlisted Flyer Enlisted Groundcrew 55 55 154 Medium Total 130 65 163 130 65 163 High Total 123 57 141 123 57 142 Contrast Adj. Relative Risk ( 5 C.I.) 9% p-Value Overall M vs. L H vs. L 0.93 (0.37,2.34) 1.67 (0.72,3.88) 0.314 0.878 0.236 Overall M vs. L H vs. L 1.05 (0.34,3.22) 0.26 ( . 5 1 3 ) 00,.1 0.124 0.935 0.102 Overall M vs. L H vs. L 0.57 (0.26,1.26) 0.58 (0.25,1.31) 0.269 0.163 0.189 Overall M vs. L H vs. L **() **! **() **! **() **! **() **! **() **! Overall M vs. L H vs. L 0.34 (0.03,4.61) 0.41 (0.03,4.99) 0.657 0.420 0.486 Overall M vs. L H vs. L ****(2) ****(2) ****(2) ****(2) ****(2) �TABLE 17-18. (continued) Adjusted Categorical Exposure Index Analyses for Renal Variables by Occupation Exposure Index Variable Occupation Officer Urinary Occult Blood Enlisted Flyer Low Total 127 55 Medium Total 130 65 High Total 123 57 Contrast Adj. Relative Risk (95% C.I.) p-Value Overall H vs. L B vs. L 0.80 (0.38,1.71) 1.40 ( . 0 2 8 ) 07,.0 0.299 0.566 0.345 Overall M vs. L H vs. L 0.97 ( . 9 2 4 ) 03,.3 0.43 (0.15,1.24) 0.187 0.950 0.118 Enlisted Groundcrew Urinary Vhite Blood Cell 154 163 141 Overall M vs. L H vs. L ****(3) ****(3) ****(3) ****(3) ****(3) Officer i o NJ 127 130 123 Overall H vs. L H vs. L 0.55 (0.20,1.51) 0.85 (0.34,2.10) 0.488 0.247 0.718 Overall M vs. L H vs. L * * ( , 3) **! * * ( , 3) **! * * ( , 3) **! * * ( , 3) **! * * ( , 3) **! Overall M vs. L B vs. L 0.68 (0.33,1.38) 1.05 (0.53,2.08) 0.424 0.284 0.886 Enlisted Flyer Count Enlisted Groundcrew 55 154 65 163 57 142 * * ( ) exposure index-by-diabetic class interaction — relative risk, confidence interval, and p-value not **!: presented. ****(2): exposure index-by-race interaction — relative risk, confidence interval, and p-value not presented. ****(3): exposure index-by-age interaction — relative risk, confidence interval, and p-value not presented. ****(!,3): exposure index-by-diabetic class and exposure index-by-age interaction — relative risk, confidence interval, and p-value not presented. �TABLE 17-19. Adjusted Continuous Exposure Index Analyses for Renal Variables Exposure Index Blood Urea Nitrogen i to Occupation Statistic Low Medium High Contrast p-Value Officer Variable n Adj . Mean 95% C.I. 127 ****(2) ****(2) 130 ****(2) ****(2) 123 ****(2) ****(2) Overall M vs. L H vs. L ****(2) ****(2) ****(2) Enlisted Flyer n Adj , Mean 95% C.I. 55 13.59 (12.02, 15.26) 65 13.76 (12.32, 15.27) 57 13.77 (12.30, 15.32) Overall M vs. L H vs. L 0.961 0.808 0.804 Enlisted Groundcrev n Adj . Mean 95% C.I. 154 13.31 (12.50, 14.15) 163 13.18 (12.41, 13.98) 142 13.08 (12.30, 13.88) Overall M vs. L H vs. L 0.829 0.722 0.544 �TABLE 17-19. (continued) Adjusted Continuous Exposure Index Analyses for Renal Variables Exposure Index l-» 1 Statistic Low Medium High Contrast p-Value n Ad j . Mean 95% C.I. 127 1.0161 (1.0131, 1.0191) 130 1.0167 (1.0138, 1.0197) 123 1.0165 (1.0136, 1.0194) Overall M vs. L H vs. L 0.755 0.457 0.647 Enlisted Flyer n Adj. Mean 95% C.I. 55 1.0159 (1.0128, 1.0191) 65 1.0157 (1.0129, 1.0185) 57 1.0142 (1.0113, 1.0171) Overall M vs. L H vs. L 0.378 0.861 0.205 Enlisted Groundcrev Urine Specific Gravity Occupation Officer Variable n Adj . Mean 95% C.I. 154 1.0166 (1.0148, 1.0184) 163 1.0166 (1.0149, 1.0183) 142 1.0164 (1.0147, 1.0182) Overall M vs. L H vs. L 0.974 0.976 0.854 NJ NJ * * ( ) exposure index-by-race interaction — adjusted mean, confidence interval, and p-value not presented. **2: �TABLE 17-20. Summary of Exposure Index-by-Covariate Interactions for Renal Variables Variable Occupation Covariate p-Value Urinary Protein Urinary Protein Urinary Occult Blood Urinary White Blood Cell Count . Urinary White Blood Cell Count Blood Urea Nitrogen Officer Enlisted Groundcrew • Enlisted Groundcrew Diabetic Class Race Age 0.004 0.023 0.032 Enlisted Flyer Age 0.015 Enlisted Flyer Officer Diabetic Class Race 0.029 0.009 LONGITUDINAL ANALYSES One variable, the BUN level, was used to assess longitudinal differences between the 1982 Baseline examination and the 1985 followup examination. This variable was selected from the five renal assays because it was judged that serial BUN levels would be more indicative of long-term renal health than the others; further, both examination measurements were made by the same high-precision automated analyzer, permitting a more valid comparison. Other commentary, contrasting general results of the other four renal variables to the Baseline, has been made for each variable above. BUN was analyzed as a continuous variable by repeated measurements analysis of variance (see Chapter 7, Statistical Methods). A square root transformation was used. The data were not adjusted by covariates. The sample base for this analysis was the number of participants who attended both examinations; the results are given in Table 17-21. These data indicated a slight and relatively symmetrical increase in the BUN level in both groups. Based upon longitudinal analyses of BUN, there was no evidence to assert a detriment in the renal health of the Ranch Hand group. SUMMARY AND CONCLUSIONS A summary of all renal variables, including unadjusted and adjusted analyses, is displayed in Table 17-22. 17-23 �TABLE 17-21. Longitudinal Analysis of BUN: A Contrast of Baseline and First Follovup Examination Laboratory Means BUN Means Group 1982 Baseline 1985 Followup Total Ranch Hand 13.72 14.21 971 Comparison 13.93 14.30 1,139 p-Value (Equality of Difference) 0.48 TABLE 17-22. Overall Summary Results of Unadjusted and Adjusted Analyses for Renal Variables Variable Unadjusted Adjusted Reported Kidney Disease NS NS Urinary Protein NS **** Urinary Occult Blood NS **** Urinary Leukocytosis NS **** BUN NS **** Urine Specific Gravity NS* **** NS: Not significant (p>0.10). NS*: Borderline significant (0.05<p<0.10). ****Group-by-covariate interaction. 17-24 �A historical assessment of kidney disease/kidney stones by a review-ofsysteras questionnaire showed no significant differences between the Ranch Hand and Comparison groups. An adjusted analysis did not alter this conclusion as an adjusted relative risk of 0.95 (95* C.I.: [0.71,1.25], p=0.693) was demonstrated. These statistics appeared to be in marked contrast to the Baseline historical findings. Differences vis-a-vis the Baseline were most likely due to a difference in questionnaire techniques. Current renal function was evaluated by five laboratory variables: urine protein, occult blood, urine, white blood cell counts (WBC's), blood urea nitrogen (BUN), and urine specific gravity. Invasive procedures were not used. The unadjusted analysis of proteinuria showed no group differences (Est. RR: 1.18, 95% C.I.: [0.75,1.86], p=0.485), but the adjusted analysis showed an interaction of group and diabetic class; appropriate stratified analyses revealed that the prevalence of proteinuria was lower in the Ranch Hands than in the Comparisons in the diabetic and impaired strata, but higher in the normal strata for the Ranch Hands. These results were in contrast to the Baseline findings, which showed a marginally significant proteinuria in the Comparison group (p=0.055), and overall, lower prevalence rates of proteinuria. The unadjusted prevalence rates for hematuria were similar for both groups (Est. RR: 1.14, 95% C.I.: [0.91,1.42], p=0.239). Three significant interactions involving group membership and covariates precluded a direct adjusted comparison of the estimated prevalence rates. Covariate analyses indicated increased hematuria in Blacks and among enlisted personnel. Ultimately via a series of stratified analyses, statistical equivalence was determined for the Black enlisted strata of both groups. Of particular note was the approximate tenfold increase in hematuria in both groups over that observed at Baseline, a finding most likely due to different laboratory techniques (reagent-strip testing versus microscopic observation). Similar results were found for leukocyturia, i.e., a nonsignificant unadjusted analysis (Est. RR: 1.24, 95% C.I.: [0.93,1.64], p=0.145), and a significant three-way interaction (group, age, race) in the adjusted analysis. Significant covariate effects were noted for diabetic class and occupation for nonblack participants, whereas age was a significant adjusting variable for Blacks. A significant group difference was found only for the younger, nonblack Ranch Hands. The overall results were consistent with the Baseline findings. BUN levels did not vary significantly by group (p=0.554, unadjusted). Adjusted analyses showed significant covariate effects for age and occupation and interactions for group and race and for race and diabetic class. An analysis stratified by race revealed no significant group differences for nonblacks, but a significantly higher adjusted mean BUN level in Black Comparisons than in Black Ranch Hands. Overall, the BUN results were similar to those observed at the Baseline examination. Urine specific gravity levels manifested marginally significant group differences (p=0.082, unadjusted). The adjusted analysis disclosed significant covariate effects of diabetic class and the interactions of group and race and group and occupation. < Analyses by race showed no strata with significantly lower mean levels for Ranch Hands. In contrast to the Baseline 17-25 �values, the followup urine specific gravities were lower, a finding most likely attributable to differences in laboratory methodology (falling drop method versus multistick procedure). Exposure index relationship at the prevalence rates or within occupational analyses showed very little evidence of a dose-response followup examination. No patterns in the relationship of mean levels among the exposure index levels were seen strata. The longitudinal analysis was based solely upon a contrast of BUN levels between the two examinations. The unadjusted mean BUN value increased slightly from the Baseline to the followup examination, but the increases were symmetrical in the two groups and nonsignificant (p=0.48). In conclusion, none of the six renal assessment variables showed a significant difference between the Ranch Hand and Comparison groups by unadjusted tests. However, in the adjusted analyses, all renal measurements except reported kidney disease revealed group-by-covariate interactions. These interactions were often complex, making it impossible to reach a firm conclusion as to the presence of an herbicide effect. 17-26 �CHAPTER 17 REFERENCES 1. St. John, L.E., D.G. Wagner, and D.J. Lisk. 1964. kuron, silvex, and 2,4,5-T in the dairy cow. 47:1267-1270. 2. Erne, K. 1966. Fate of atrazine, J. Dairy Sci. Studies on the animal metabolism of phenoxyacetic herbicides. Acta Vet. Scand. 7:264-271. 3. Matsumura, A. 1970. man. The fate of 2,4,5-trichlorophenoxyacetic acid in Jap. J. Environ. Health 12:20-25. 4. Gehring, P.J., C.G. Kramer, B.A. Schwetz, J.Q. Rose, and V.K. Rove. 1973. The fate of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) following oral administration to man. Toxicol. Appl. Pharmacol. 26:352-361. 5. Kohli, J.D., R.N. Khanna, B.N. Gupta, M.M. Dhar, J.S. Tandon, and K.P. Sircar. 1974. Absorption and excretion of 2,4,5-trichlorophenoxyacetic acid in man. Arch. Int. Pharmacodyn. 210:250-255. 6. Bjorklund, N.E., and K. Erne. 1971. Phenoxy-acid-induced renal changes in the chicken: I. Ultrastructure. Acta Vet. Scand. 12:243-256. 7. Fowler, B.A., G.E.R Hook, and G.tf. Lucier. 1977. Tetrachloro-dibenzo-pdioxin induction of renal microsomal enzyme systems: Ultrastructural effects on pars recta (S,) proximal tubule cells of the rat kidney. J. Pharmacol. Exp. Ther. 203(3):712-721. 8. Koschier, F.J., and M. Acara. 1979. Transport of 2,4,5-trichlorophenoxyacetate in the isolated, perfused rat kidney. J. Pharmacol. Exp. Ther. 208:287-293. 9. Gupta, B.N., J.G. Vos, J.A. Moore, J.G. Zinkl, and B.C. Bullock. 1973. Pathological effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin in laboratory animals. Environ. Health Persp. 5:125-140. 10. Pegg, D.G., W.R. Hewitt, K.M. McCormack, and J.B. Hook. 1976. Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on renal function in the rat. J^ Toxicol. Environ. Health 2:55-65. 11. Courtney, K.D., J.P. Putnam, and J.E. Andres. 1978. Metabolic studies with TCDD (dioxin) treated rats. Arch. Environ. Contain. Toxicol. 7(4):385-396. 12. Hook, J.B., K.M. McCormack, and W.M Kluwe. 1978. Renal effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin. In Pentachlorophenol; Chemistry, pharmacology and environmental toxicology^ed. K.R. Rao, pp. 381-388.New York:Plenum Press. 17-27 �13. Kohli, J.D., R.N. Khanna, B.N. Gupta, M.M. Dhar, J.S. Tandon, and K.P. Sircar. 1974. Absorption and excretion of 2,4-dichlorophenoxyacetic acid in man. Xenobiotica 4(2):97-100. 14. Sauerhoff, M.W., W.H. Braun, G.E. Blau, and P.J. Gehring. 1977. The fate of 2,4-dichlorophenoxyacetic acid (2,4-D) following oral administration to man. Toxicology 8:3-11. 15. Carter, C.D., R.D. Kimbrough, J.A. Liddle, R.E. Cline, M.M. Zack, W.F. Barthel, R.E. Koehler, and P.E. Phillips. 1975. Tetrachlorodibenzodioxin: An accidental poisoning episode in horse arenas. Science 188(4189):738-740. 16. Beale, M.G., W.T. Shearer, M.M. Karl, and A.M. Robson. effects of dioxin exposure. Lancet 1(8014):748. 1977. Long-term 17. Poland, A.P., D. Smith, G. Metter, and P. Fossick. 1971. A health survey of workers in a 2,4-D and 2,4,5-T plant, with special attention to chloracne, porphyria cutanea tarda, and psychologic parameters. Arch. Environ. Health 22(3):316-327. 18. Oliver, R.M, 1975. Toxic effects of 2,3,7,8-tetrachloro-dibenzo-l, 4-dioxin in laboratory workers. Br. J. Ind. Med. 32:46-53. 19. Pazderova-Vejlupkova, J., M. Nemcova, J. Pickova, L. Jirasek, and E. Lukas. 1981. The development and prognosis of chronic intoxication by tetrachlorodibenzo-p-dioxin in men. Arch. Environ. Health 36:5-11. 20. Hoffman, R.E., P.A. Stehr-Green, K.B. Webb, G. Evans, A.P. Knutsen, W.F. Schramm, J.L. Staake, B.B. Gibson, and K.K. Steinberg. 1986. Health effects of long-term exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. JAMA 255:2031-2038. 21. Stehr, P.A., G. Stein, H. Falk, et al. 1986. A pilot epidemiologic study of possible health effects associated with 2,3,7,8-tetrachlorodibenzo-p-dioxin contamination in Missouri. Arch. Environ. Health 41:16-22. 22. Moses, M., R. Lilis, K.D. Crow, J. Thornton, A. Fischbein, H.A. Anderson, and I.J. Selikoff. 1984. Health status of workers with past exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in the manufacture of 2,4,5-trichloro-phenoxyacetic acid: Comparison of findings with and without chloracne. Am. J. Ind. Med. 5:161-182. 23. Suskind, R.R., and V.S. Hertzberg. 1984. Human health effects of 2,4,5-T and its toxic contaminants. JAMA 251:2372-2380. 17-28 �CHAPTER 18 ENDOCRINE ASSESSMENT INTRODUCTION The human endocrine system is generally not thought to be influenced by chlorophenol or TCDD exposure. This is not so in animals, however. A wide range of endocrine abnormalities in many animal species has been induced experimentally by TCDD, and includes hypoglycemia, hypothyroxinemia, ' reduced progesterone levels,3 and increased testosterone levels, the latter presumably reflecting decreased liver catabolism due to parenchymal liver damage or an inhibition of the cytochrome P-450 system.4 Further, thymic atrophy, one of the most sensitive indicators of TCDD toxicity, has been shown not to be mediated by the pituitary-adrenal axis. Comparable animal data for the isolated effects of 2,4-D and 2,4,5-T have been noticeably meager. Other animal studies have emphasized the endocrine system, and thyroid function in particular, as important in causing or ameliorating TCDD toxicity, and not simply as an endpoint response. ' Mounting experimental evidence suggests that both natural and radiation-induced hypothyroidism protect against TCDD lethality and that this favorable process can be quickly reversed by treatments with T4. ' If the protective reaction of hypothyroidism in animals can be extrapolated to humans, it suggests that cases of hypothyroidism or altered patterns of thyroid hormones may aggregate in groups of highly exposed workers (particularly in those with chloracne) and/or, alternatively, that severe sequelae of TCDD exposure may be associated with hyperthyroidism. In fact, such thyroid findings have not been commonly reported in dioxin morbidity studies. Occasional cases of hypothyroidism and thyromegaly have been linked to exposures to polybrominated biphenyls and hexachlorobenzene, but the data were too sparse and oblique to support a causal relationship for hypothyroidism and TCDD exposure. °' An assessment of the Times Beach, Missouri, residents, whose community was contaminated with TCDD, did not reveal TSH or T4 differences between the high- and low-risk groups. Temporary glycosuria and impaired glucose tolerance tests were noted in two studies of industrial workers exposed to TCDD. ' However, neither abnormal glucose metabolism nor frank diabetes was specifically noted in other comparable studies. -1 Overall, dioxin morbidity studies have not rigorously assessed the clinical or biochemical parameters of the endocrine system. A detailed description of endocrine function following TCDD exposure was the 1984 AFHS Baseline Morbidity Report, summarized below. 18-1 �Baseline Summary Results The 1982 Baseline examination did not explore historical endocrinological disorders by questionnaire sufficiently to merit analysis. Hence, a comprehensive biochemical assessment of the endocrine system was used for analysis. Five measures of endocrine status were assessed, T3 % Uptake, T4, free thyroxine index (FTI), testosterone, and 2-hour postprandial glucose. Three hormones, follicle stimulating hormone, leutinizing hormone, and cortisol, and correlations of all hormones to various fertility measurements remain for future analysis. Results showed significant group differences for T, % Uptake, predominantly in Ranch Hands 40 years old or less, and abnormally low T % Uptake values, highest for those with high percent body fat. No group difference was noted for elevated 2-hour postprandial glucose values, and as classically expected, the prevalence of abnormal values was associated with older ages and higher percent body fat. Similarly, low testosterone levels were identical in both groups and were associated with increasing age and increasing percent body fat. Higher mean testosterone values (although still within "normal range") were significantly more prevalent in the Ranch Hand group. Significant mean shifts were not noted for the T3 % Uptake, T4, or FTI variable, although the T3 % Uptake was associated with a group-by-age interaction. The exposure index analyses were essentially negative for the T3 % Uptake and T4 variables. FTI, postprandial glucose, and testosterone analyses were marked by a series of covariate interactions in varying occupational categories. Of some note were the significant percent body fat-by-exposure interactions in two occupational strata in the glucose determination. In summary, the endocrine system, as measured by five biochemical assays, did not reveal clinically apparent abnormalities that could be attributed to Herbicide Orange exposure. However, significant mean shifts in several values (although still in normal range) presented trends that were both consistent and conflicting vis-a-vis an herbicide etiology. These data, coupled with the emerging animal literature on the profound influence of the endocrine system on lethality and body fat metabolism following TCDD exposure, clearly underscore the importance of evaluating the endocrine system more comprehensively, as was done in the third-year followup study in 1985. Parameters of the 1985 Endocrine Assessment The 1985 AFHS endocrine test battery was slightly altered from Baseline and included T3 % Uptake, TSH, testosterone, 2-hour postprandial glucose, and timed paired cortisols. The 100 gram glucose load was standardized by a Glucola* challenge (as contrasted to an estimated 100 gram carbohydrate breakfast at Baseline) in preparation for a more definitive assessment of diabetes. Specific questionnaire data on past diabetes and thyroid disease were collected for assessment. 18-2 �Thus, the analyses of endocrine function were comparable to those conducted on Baseline data. Additional refinements included adding diabetes (past and current) as a dependent variable, and the covariates race and personality type, when appropriate. Continuous dependent variables were dichotomized into normal/abnormal categories when necessary using the SCRF values of normal range. Numerous exclusion criteria, e.g., thyroidectomy, orchiectomy, supplemental steroid medication, and diabetes, were used for specific dependent variables. Variations in the numbers of observations in the tables, therefore, reflect these exclusions in addition to rare missing data from the dependent or adjusting variables. Comparable analyses using the Original Comparisons are found in Tables P-4 to P-6 of Appendix P. Log-linear models (BMDP®-4F), general linear models (SAS«-GLM), and logistic regression models (BMDP*-LR) formed the core of the statistical approach. RESULTS AND DISCUSSION Questionnaire Data General screening questions on thyroid function and disease were posed to each participant. Two instruments were used: a self-administered review-of-systerns form containing five questions (e.g., goiter or thyroid trouble, use of thyroid medication?) and the interval health questionnaire with the single question, "thyroid problems?" administered by a trained interviewer. These data are summarized in Table 18-1. Table 18-1 shows that past and current thyroid problems vary according to the interview technique; the group difference in the self-administered questionnaire response was not significant, but the group difference in the interviewer-obtained response was borderline significant. The higher proportion of thyroid disease with the review-of-systerns questionnaire was most likely due to the broader range of prompting questions or interpretation of the questions by the study participant. Since the interviewer-administered questionnaire contained medical provider information for each positive response, verification by medical record review was possible. These data are summarized in Table 18-2 and demonstrated equivalent verification findings in the Ranch Hand and Comparison groups. Thus, the relative absence of reported thyroid disease in the Ranch Hand group appears valid. Physical Examination Data Physical examination of the endocrine system was necessarily limited to manual palpation of the thyroid gland and the testes. Thyroid abnormalities consisted of an enlarged gland with or without nodules or tenderness, while abnormal testes were noted for atrophied glands. The overall palpation results are summarized in Table 18-3. The physical examination data for thyroid abnormalities were clearly supportive of the findings of the questionnaire/review of systems analysis. The proportion of testicular abnormalities (only atrophy represented in the above analysis) was essentially equivalent in both groups. 18-3 �TABLE 18-1. Unadjusted Analysis for Reporting of Thyroid Symptoms/Disease by Questionnaire Method by Group Group Ranch Hand Questionnaire Method Statistic Self-Administered n Diseased3 Not Diseased n Diseased Not Diseased Interviewer Administered Comparison Percent Number Percent Est. Relative Risk (95% C.I.) p-Value* 1,016 48 968 4.7 95.3 1,293 57 1,236 4.4 95 .6 1.08 (0.73,1.59) 0.763 1,016 7 1,009 0.7 99.3 1,293 21 1,272 1.6 98 .4 0.42 (0.18,0.99) 0.054 Number 00 I *Fisher's exact test. a Participants answered positively to having thyroid or goiter trouble, high thyroid level, low thyroid level, lump in throat, or taking thyroid medication. 'Participants answered positively to having thyroid problems since last interviewed. �TABLE 18-2. Medical Record Verification Results of Reported Thyroid Disease by Group Group Ranch Hand Verification Status Comparison Number with Reported Thyroid Conditions 7 21 Medical Records Reviewed 7 21 Medical Records Pending 0 0 100 100 Percent Thyroid Conditions Verified TABLE 18-3. Unadjusted Analysis for Thyroid and Testicular Conditions by Group Group Ranch Hand Comparison Variable Statistic Number Percent Number Percent Thyroid* Est. Relative Risk (95% C.I.) p-Value n 1,015 Abnormal 342 Normal 673 33.7 66.3 1,293 431 862 33.3 1.02 (0.85,1.21) 66.7 0.860 Testicular" n 1,002 Abnormal 26 Normal 976 2.6 97.4 1,289 41 1,248 3.2 0.81 (0.49,1.34) 96.8 0.454 *Thyroidectomies omitted; thyroid abnormal if palpably tender or enlarged, or if nodules present. Orchiectomies omitted; testes abnormal if atrophied (compared to normal). 18-5 �Laboratory Test Data General The collection of relatively scant endocrinological data by questionnaire and physical examination techniques was due to competing priorities of the examination and to the primary reliance upon laboratory testing as established by the 1982 Baseline examination. With research-grade laboratory quality control and reasonably large sample sizes, it was judged that even small mean shifts could be discerned in the test variables. In the presence of corroborating data, these shifts may be ascribed to an herbicide effect if, in fact, one exists. The endocrinological assessment centered upon analysis of laboratory data for T3 % Uptake, TSH, testosterone, timed paired cortisol specimens (the latter three assays conducted by radioimmunoassay [RIA]), 2-hour postprandial glucose, and a composite indicator of past and current diabetes. Normal values of these measurements, as determined by the SCRF Laboratory, are categorized in Table 18-4. It is noted that some of these variables have associated "cutpoints" that differ considerably from those used by the 1982 examining laboratory. Based upon the SCRP laboratory norms, the endocrinological variables distributed into normal and abnormal proportions as displayed in Table 18-5. Unadjusted Ranch Hand and Comparison group means are also provided for quick contrast. TABLE 18-4. Laboratory Endocrinological Variables: SCRF Normal and Abnormal Ranges Variable T3 % Uptake <24% TSH Testosterone Abnormally High 24-32% <7.5 uU/ml <7 ug/dl >7.5 uU/ml 270-1,100 mg/dl >1,100 mg/dl <140 mg/dl <270 mg/dl 2-Hour Postprandial Glucose Cortisol Normal Abnormally Low >140-<200 mg/dl (impaired) >200 mg/dl (diabetic) 7-25 ug/dl >25 ug/dl 18-6 �TABLE 18-5. Unadjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group Group Variable Statistic T3 % Uptake n Mean 95% C.I. Number/% Low Normal High 1,003 27.79 (27.67,27.91) 1,270 27.73 (27.62,27.84) 7 969 27 0.7% 96.6% 2.7% 18 1.4% 1,221 96.1% 31 2.4% n Mean 95% C.I. Number/% Normal High 1,003 1.158 (1.13,1.19) 99.3% 0.7% 1,264 99.5% 6 0.5% n Mean 95% C.I. Number/% Lov Normal High 1,000 597.3 (584.0,610.8) oo i Testosterone 996 7 38 949 13 3.8% 94.9% 1.3% Comparison Est. Relative Risk (95% C.I.) 1,270 1.107 (1.08,1.13) TSH Ranch Hand Contrast — Overall Low vs. Normal High vs. Normal 0.457a 0.49 ( .20,1.18) 0 1.10 (0.65,1.85) 0.248b 0.110C 0.789C 0.019a 1.48 (0.50,4.42) 1,288 578.3 (566.9,589.9) 49 3.8% 1,225 95.1% 14 1.1% p-Value 0.579° 0.035a Overall Lov vs. Normal High vs. Normal 1.00 (0.65,1.54) 1.20 (0.56,2.56) b 0.896C 0.999 0.698C �TABLE 18-5. (continued) Unadjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group Group Ranch Hand Comparison Variable Statistic Initial Cortisol n Mean 95% C.I. Number/% Low Normal High 1,009 11.62 (11.39,11.85) 5.2% 94.2% 0.7% 64 5.0% 1,207 94.0% 13 1.0% n Mean 95% C.I. Number/% Low Normal High 1,009 9.30 (9.10 ,9.51) 0 1,281 3 n Mean 95% C.I. 1,009 2.30 (2.05 ,2.55) 1,284 2.46 (2.24 ,2.69) Est. Relative Risk (95% C.I.) p-Value 1,284 9.27 ( . 0,9.44) 91 0 0.0% 1,005 99.6% 4 0.4% Contrast 2-Hour Cortisol 03 00 Differential Cortisol 52 950 7 1,284 11.68 (11.48,11.89) 0.0% 99.8% 0.2% 0.668a Overall Low vs. Normal High vs. Normal 1.03 (0.71,1.50) 0.68 (0.27,1.72) 0.708b 0.924C 0.501C 0.793a 1.70 (0.38,7.61) 0.706C 0.349a �TABLE 18-5. (continued) Unadjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group Group Variable 2-Hour Postprandial Glucose £ i VO Statistic n Mean 95% C.I. Number/% Normal Impaired Diabetic Diabetes (Composite Indicator) n Number/% Yes No Ranch Hand 976 107.9 (105.9 ,110.0) 836 106 . 34 85.7% 10.9% 3.5% 1,016 74 942 Comparison Contrast Est. Relative Risk (95% C.I.) p-Value 1,235 109.0 (107.3 ,110.7) 0.435a 1,026 83.1% Overall 176 14.3% Impaired vs. Normal 33 2.7% Diabetic vs. Normal 0.74 (0.57,0.96) 1.26 ( . 8 2 0 ) 07,.6 0.038b 0.024° 0.382C 1.09 (0.79,1.50) 0.622° 1,293 7.3% 92.7% 87 6.7% 1,206 93.3% —Relative risk not given for continuous analyses of variables. a t-test. b Chi-square test. c Fisher's exact test. �The following representative statistical power statements (for power 0.8, 2-sided a =0.05) may be applied to parameters of several variables listed in Table 18-5. The sample sizes were sufficient to detect a 1.9-fold increase in the frequency of percent abnormal high values for T3 % Uptake and a 2.5-fold increase in percent abnormal high values for testosterone, relative to that observed in the Comparison group. In addition, the sample sizes were sufficient to detect a 2.7 percent mean shift in TSH and a 1.5 percent mean shift in the first cortisol specimen, over those means observed in the Comparison group. Table 18-5 shows remarkably comparable unadjusted group means and distributional parameters for Ranch Hands and Comparisons in T3 % Uptake, initial cortisol, and 2-hour cortisol. For TSH, testosterone, and 2-hour postprandial glucose, however, there was disparity between the statistical results of the means test and the distributional chi-square test, suggesting that significant differences may exist between the Ranch Hand and Comparison groups. Since all endocrinological variables were known to depend upon classical covariates such as age and race, each variable was reanalyzed by general linear models (using transformations when necessary), logistic regression analyses, or log-linear models adjusted for these covariates. The results of these adjusted analyses are presented in a series of functional endocrine groups below. Table 18-6 presents complete details on the adjusted analyses for all the endocrinological variables. Thyroid Function: T3 % Uptake and Thyroid Stimulating Hormone (TSH) Assessment of both thyroid assays excluded all participants on thyroid medication (as determined by both the self-administered questionnaire and the structured NORC questionnaire) as well as participants with partial or total thyroidectomies. Thus, 13 Ranch Hands and 20 Comparisons were omitted from the following analyses. T3 % Uptake The T3 % Uptake categorical data, as summarized in Table 18-5, were reanalyzed controlling for the covariate effects of occupation, race, age, and personality type. Group data were pooled to reveal the marginal effects of the four covariates. These data are summarized in Table 18-7. The analysis of these data showed a significant effect of occupation (p=0.024) on the percentage of participants with abnormal T3 % Uptake results. Specifically, this was mostly attributable to a relatively high percentage of officers with high T3 % Uptake levels (31 observed versus 21.5 expected, see Table 18-7) and a low percentage of enlisted flyers with high T3 % Uptake results (5 observed versus 9.8 expected). Table 18-7 also shows a marginal effect of personality type on T3 % Uptake results (however, this effect was significant [p=0.035] when analysis was restricted to Ranch Hands and Original Comparisons). Most of the personality-type effect was due to larger numbers than expected of Type A 18-10 �T6HEE 18-6. Adjusted Continuous and Categorical Analyses for Group Variable Statistic T3 %Uj>take n TSH n Adj. Mean 95% C.I. 998 1.158 (1.13,1.19) 1,267 1.109 (1.08,1.14) n 1,003 1,270 n Adj. Mean 95% C.I. 1,000 AArArA' A A AA 1,270 Contrast p-Value Covariate Remarks* 0.250 0.117 0.809 OCC (p=0.025) 0.025 AGE*PERSTYPE(p=O.037) icJcJck 1,003 Comparison Adj. Relative Risk (95% C.I.) 1,287 A1 A"A A Testosterone Ranch Hand Overall Low vs. Normal 0.50 (0.21,1.19) High vs. Normal 1.10 (0.50,2.44) — High vs. Normal 1.48 (0.50,4.42) GRP*BFAT (p=0.024) __ Overall Low vs. Normal 1.00 (0.64,1.55) High vs. Normal 1.13 (0.48,2.64) Initial Cortisol n Adj. Mean 95% C.I. 1,004 11.42 (10.59,12.31) 1,280 11.49 (10.66,12.38) 0.579 — -l-l-l-lr XXXVT A3E*BFAT (p=0.024) RACE (p=0.004) 0.949 0.986 0'.774 ACS (pO.OOl) %BEAT (pO.OOl) 0.659 PGE (p<0.001) %BFAT (p<0.001) PERSTYPE (p=0.002) RACE*OCC (p=0.009) �TABLE 18-6. (continued) Adjusted Continuous and Categorical Analyses for Laboratory Endocrinological Variables by Group Group^ Variable Statistic Differential n Adj. Mean Ranch Hand Comparison 1,004 ycMcfc Adj. Relative Risk (95% C.I.) p-Value 1,280 "kXTCK Contrast ™ KTCfCX Covariate Remarks* GRP*AGE*RACE Cortisol (p=0.032) PERSTYPE (p=0.005) %BFAT (p<0.001) Diabetes (Composite Indicator) n Adj. Mean 95%C.I. n 0.487 %BFAT (p<0.001) OCC (p<D.001) AGE*RAOE (p=0.002) Overall Impaired vs. Normal 0.73 (0.56,0.%) Diabetic vs. Normal 1.26 (0.72,2.22) 2-Hour Postprandial Glucose 0.034 0:022 0.421 AGE (pO.OOl) RACE (p=<0.016) JffiFAT (p<0.001) Diseased vs. Normal 1.12 (0.80,1.56) 0.500 %BFAT (p<0.001) AGE*RACE (p=0.005) 976 1,234 115.3 114.4 (107.3,122.0) (108.2,123.0) 1,016 — 1 292 *Abbreviations: GKP: Group OCC: Occupation PEBSTYPE: Personality type (A or B) • %BFAT: Percent body fat —No relative risk or confidence interval given for continuous analyses. ****Group-by-covariate interaction-^Adjusted mean/relative risk, confidence interval, and p-value are not presented. �TABLE 18-7. Association Betveen T % Uptake and Age, Race, Occupation, ana Personality Type in the Combined Ranch Hand and Comparison Groups Percent Abnormal Covariate Covariate Category Total Low High p-Value Age Born XL942 Born <1942 953 1,320 1.05 1.14 2.52 2.58 0.977 Race Black Nonblack 143 2,130 1.40 1.08 1.40 2.63 0.628 Occupation Officer Enlisted Flyer Enlisted Groundcrew 842 383 0.59 1.04 3.68 1.31 0.024 1,048 1.53 2.10 997 1,268 1.60 0.71 2.91 2.21 Personality Type A Direction B Direction 0.071 participants with lower T % Uptake levels. The covariates age and race were not correlated with T, % uptake abnormalities. Log-linear models were then used to assess possible group differences in T3 % Uptake abnormalities, adjusting for occupation (OCC), race, age, and personality type (PERSTYPE). The covariates age, race, and personality type did not contribute significantly to the fit of the adjusted model and were deleted to yield the simplest model, which included occupation. This analysis was summarized in terms of adjusted relative risks and is displayed in Table 18-8. There were no significant differences in percent abnormalities of T3 % Uptake between the Ranch Hand and the Comparison groups. Occupation demonstrated a significant effect (p=0.025). Personality type, although marginally significant (p-0.068), did not affect the assessment of group differences. Thyroid Stimulating Hormone (TSH) TSH laboratory values were analyzed in both discrete and continuous forms. As noted in Table 18-5, an unadjusted t-test of group means showed a statistically significant elevation of TSH in the Ranch Hand group, whereas the categorical analysis did not reveal a statistically significant group difference in the percentage of abnormalities. Exclusion categories and the number of participants were identical to the T3 % Uptake analyses. 18-13 �TABLE 18-8. Adjusted Categorical Analysis for T % Uptake Analysis Contrast Adjusted Relative Risk. 95% C.I. Overall p-Value 0.250 Abnormally Low vs. Normal 0.50 (0.21,1.19) 1.10 (0.50,2.44) Occupation(p=0.025) 0.117 Abnormally High vs. Normal Covariate Remarks 0.809 *Chi-square test (2 d.f.) for group difference. Unadjusted covariate analyses of discrete TSH data from the combined Ranch Hand and Comparison groups showed a borderline significant difference (p=0.071) among occupational groups, with a higher proportion of enlisted flyers with abnormally high TSH levels than observed in the officer or enlisted groundcrew population. The covariates age (born in or after 1942, born before 1942), race, and personality type were nonsignificant. A stepwise logistic regression analysis was performed. The final model was identical to the unadjusted analysis as none of the covariates were significantly associated with TSH. The adjusted percent TSH abnormalities by group were expressed as relative risks. For completeness this summary analysis is shown again in Table 18-9. TSH was subsequently analyzed as a continuous variable. The unadjusted group contrast (determined by a t-test following transformation of TSH values to an inverse square root scale) showed a statistically significant (p=0.019) increase in the mean TSH of the Ranch Hand group, as depicted in Table 18-5. After suitable model fitting, group mean data were adjusted for age (continuous), personality type, and an age-by-personality type interaction. Adjusted results are shown in Table 18-10. As shown, the Ranch Hand TSH mean was significantly elevated over the Comparison group mean after covariate adjustment. However, the group mean values were well below the observed cutoff value of 7.5 uU/ml. The herbicide literature suggests a possibility of primary or secondary hypothyroidism as an endpoint following TCDD exposure. Hypothyroidism, as manifest by the test parameters in this study, should produce a 1 tendency toward depressed T3 % Uptake levels and increased levels of TSH. In the Ranch Hand group, the T3 % Uptake did not indicate hypothyroidism, whereas the TSH mean value showed an increase consistent with hypothyroidism. Questionnaire, physical examination, and laboratory data on thyroid function 18-14 �TABLE 18-9. Adjusted Categorical Analysis for TSH Adjusted Relative Risk 95% C.I. p-Value 1.48 (0.50,4.42) 0.579 TABLE 18-10. Adjusted Continuous Analysis for TSH by Group Group Ranch Hand Total* Adjusted Mean 95% C.I. 998 1.158 (1.13,1.19) p-Value 0.025 Comparison 1,267 1.109 Covariate Remarks Age-by-Personality Type (p=0.037) (1.08,1.14) *Eight participants excluded because of missing data on personality type; 35 participants excluded because of thyroid medication. 18-15 �and disease led to the conclusion that there were no essential differences indicating thyroid disease between the Ranch Hand and the Comparison groups. Testosterone Serum testosterone levels were measured by RIA on all participants. Normal range values from the SCRF Laboratory were used to categorize all data into abnormally low, normal, abnormally high determinations (see Table 18-4). All analyses omitted participants with unilateral or bilateral orchiectomies, and those participants on supplemental testosterone medication. The unadjusted categorical analysis (see Table 18-5) showed no significant differences (p=0.896) in the proportions of abnormalities between the Ranch Hand group and the total Comparison group. The groups were combined and the relationships between categorized testosterone levels and the covariates occupation, race, age, percent body fat (%BFAT), and personality type were examined. Significant statistical differences were noted for occupation (p-0.012), increasing age (p<0.001), and increasing percent body fat (p<0.001). No effect was found due to race or personality type. An adjusted analysis was done to determine the simplest model using the significant covariates, and relative risks were calculated. This analysis is depicted in Table 18-11. These results showed that neither percent low testosterone abnormalities nor percent high testosterone abnormalities were excessive in the Ranch Hand group, as the confidence interval of the adjusted relative risks included the value 1.00. TABLE 18-11. Adjusted Categorical Analysis for Testosterone Analysis Contrast Adjusted Relative Risk 95% C.I. Overall* p-Value Covariate Remarks 0.949 Abnormally Low vs. Normal 1.00 (0.64,1.55) 0.986 Abnormally High vs. Normal 1.13 (0.48,2.64) 0.774 *Chi-square test (2 d.f.) for group difference. 18-16 Age(p<0.001) Percent Body Fat (p<0.001) �In contrast to the negative categorical analyses, the unadjusted test of testosterone means showed a significant elevation in the Ranch Hand group (see Table 18-5). Using similar covariates as in the adjusted categorical analyses, the group means were contrasted by an analysis of covariance. A significant group-by-percent body fat interaction was found (p=0.024). This was due to Ranch Hands having a significantly lower mean than Comparisons (654.4 mg/dl versus 1042.8 mg/dl, p=0.012), for the less than 10 percent body fat category, but a significantly higher mean for the 10 to 25 percent body fat category (603.3 mg/dl versus 582.4 mg/dl), and a nonsignificantly higher mean for the greater than 25 percent body fat category (463.0 mg/dl versus 456.7 mg/dl). However, the number of participants in the less than 10 percent body fat category was very small: six Ranch Hands and four Comparisons, and without these, the overall Ranch Hand mean testosterone level was higher than that for Comparisons. An age-by-percent body fat interaction (p=0.024) and race (p=0.004) were significant covariates. The group interaction is summarized in Table P-l of Appendix P. The adjusted analysis showed a significantly elevated mean testosterone level in the Ranch Hand group for the 10 to 25 percent body fat category, which comprised 80 percent of the Ranch Hand and 79 percent of the Comparison participants, whereas the categorical analyses did not reveal any group differences. These findings might be viewed as supportive of an herbicide effect. Cortisol: Initial, 2-Hour, and Differential Cortisol measurements were obtained in the AFHS for two reasons: as a general indicator of the integrity of the endocrine system (and specifically as a functional measure of the pituitary-adrenal circuit), and as an important secondary risk factor in coronary artery disease (CAD). As cholesterol is a metabolic precursor to cortisol, there has been longstanding scientific interest on cause-and-effect relationships between these substances. Clearly, steroid and ACTH treatments have been implicated in induced hypercholesterolemia and possibly resulting CAD. " Cholesterol elevations have been consistently .noted following exposure to TCDD (see Chapter 15) and, therefore, are of prime interest in this study. Consequently, exploration of the cholesterol-TCDD or cholesterol-CAD relationship must also account for cortisol differences, if any. Timed serum specimens were obtained from all participants at a 2-hour interval early on the second day of the examination. The difference between the timed paired specimens was termed the "differential cortisol." The value of the first specimen was generally higher than the value of the second specimen (due to liver catabolism). The mean values of the two cortisol determinations (initial and 2-hour) for the Ranch Hand and the Comparison groups (as reflected in Table 18-5) did not differ by unadjusted t-tests (p=0.668, p=0.793, respectively), Further, the unadjusted categorical analyses for both specimens based on the normal values of the SCRF Laboratory also did not demonstrate significant group differences (p=0.708, p=0.706, respectively). 18-17 �By an analysis of covariance, using the covariates age, occupation, race, percent body fat, and personality type, the mean value of the initial cortisol specimen was adjusted and contrasted by group. These results are given in Table 18-12, and as indicated, there was no statistically significant group difference. Tests of association between the differential cortisol and the covariates (Table 18-13) disclosed significant effects by percent body fat and personality type (p=0.002, p=0.006, respectively). Age was only slightly suggestive of an effect. An adjusted analysis was performed using the above covariates. A groupby-age-by-race interaction was found (p=0.032). Personality type (p=0.005) and percent body fat (p<0.001) were significant covariates. The interaction found a significantly lower mean differential cortisol level for Black Ranch Hands (p»0.003) born in or after 1942 (unadjusted mean 0.17 ug/dl, adjusted mean -0.46 ug/dl) versus corresponding Comparisons (unadjusted mean 2.78 ug/dl, adjusted mean 2.33 ug/dl); no significant differences were found for older Blacks or nonblacks. The interaction is summarized in Table P-l of Appendix P. The analyses discussed above showed that the Ranch Hand and Comparison groups did not differ with regard to both paired cortisol specimens, and the differential cortisol of those specimens for all nonblacks and Blacks born before 1942. For Blacks born in or after 1942 (32 Ranch Hands, 47 Comparisons) the mean differential cortisol level was lower for Ranch Hands than Comparisons. The mean cortisol levels for each personality type and percent body fat category were plotted over time. Figure 18-1 shows the rate of decrease in cortisol for Type A and Type B personalities, adjusted for percent body fat and age. Similarly, Figure 18-2 shows the rate of decrease in cortisol in three categories of percent body fat, adjusted by personality type and age. The effect of personality type and percent body fat upon the levels of cortisol and the rate of change of cortisol over the 2-hour period are noteworthy. Age was also a significant covariate. Type A personalities began with slightly lower cortisol levels but had a lower rate of decrease of cortisol over the next 2 hours as contrasted to Type B personalities. This analysis demonstrated the ability of the Jenkins Activity Scale to differentiate personality type in this cohort, as measured by differential cortisol levels. The strong effect of percent body fat upon cortisol was not expected. Glucose Metabolism: 2-Hour Postprandial Glucose and Composite Diabetes Indicator The 1985 examination at SCRF presented two major changes in the assessment of glucose metabolism as contrasted to the 1982 Baseline examination: (1) the accepted laboratory criteria by which to diagnose diabetes shifted from the standard of 120 mg/dl or more at 2 hours to a designation of "impaired" glucose tolerance (at least 140 but less than 200 mg/dl) and "diabetic" glucose tolerance (at least 200 mg/dl), and (2) participants were given a standardized 100 gram Glucola* challenge rather than an estimated 100 gram carbohydrate breakfast. Further, most known diabetics were encouraged not to take the Glucola* challenge. 18-18 �TABLE 18-12. Adjusted Continuous Analysis for Initial Cortisol by Group Group Total* Ranch Hand Adjusted Mean 1,004 11.42 0.659 Comparison 1,280 Covariate Remarks p-Value 11.49 Age (p<0.001) Personality Type (p=0.002) Percent Body Fat (p<0.001) Occupation-by-Race (p=0.009) * Nine participants omitted due to missing data on personality type and body fat. TABLE 18-13. Association Betveen Differential Cortisol and Age, Race, Occupation, Percent Body Fat, and Personality Score in the Combined Ranch Hand and Comparison Groups Covariate Covariate Category Total Mean Differential Cortisol Level p-Value Age Born > 1942 Born < 1942 955 1,338 2.24 2.51 0.122s Race Black Nonblack 143 2,150 2.16 2.41 0.575* Occupation Officer Enlisted Flyer Enlisted Groundcrev 852 385 1,056 2.55 2.48 2.23 0.203b <10% 10-25% >25% 10 1,846 436 1.80 2.54 1.79 0.002b 1,002 1,283 2.12 2.60 0.006* Percent Body Fat Personality A Direction Type B Direction "By t-test. b 18-19 L �12- 11.511- 10.5Cortisol Units/ml 10- 9.5- 98.5- 87:30 a.m. 9:30 a.m. Time of Serum Collection D Type A * Type B Figure 18-1. Mean Cortisol Levels by Personality Type, Adjusted for Age and Percent Body Fat, by Time of Specimen Collection 18-20 �12- 11.511- 10.5Cortisol Units/ml 10- 9.598.5- 8- I I 7:30 a.m. 9:30 a.m. Time of Serum Collection D <10% Body Fat x 10-15% Body Fat 0 >25% Body Fat Figure 18-2. Mean Cortisol Levels by Percent Body Fat, Adjusted for Age and Personality Type, by Time of Specimen Collection 18-21 �All participants were provided high carbohydrate menus preceding the examination, and were encouraged to consume high calorie meals for 3 days immediately before their examination to improve the diagnostic efficiency of the glucose tolerance test. At the examination site, compliance or noncompliance to the carbohydrate diet was recorded but reported compliance was not analyzed. These data, however, were not used to exclude participants from the analyses, as the 1984 Baseline Report showed that compliance to the diet was inconsequential to the analyses. All known diabetics, as determined by the Baseline history and the 1982-1985 interval questionnaire, were excluded from the glucose tolerance analyses. However, the 43 Ranch Hands and the 59 Comparisons comprising the exclusion group were included in the composite diabetes analysis. 2-Hour Postprandial Glucose As noted in Table 18-5, a trichotomized contrast of the 2-hour postprandial glucose showed a statistically significant difference (p=0.038) between the Ranch Hand and the Comparison groups. This was due to a slightly higher percentage of Ranch Hands in the diabetic category, and a lower percentage of them in the impaired category relative to the Comparison group. Both study groups were pooled to assess the covariate main effects of age, race, occupation, and personality type. The results showed a significant effect for occupation (p=0.030), largely due to a higher proportion of enlisted flyers having impaired glucose levels. Race, age, and percent body fat were significant covariates (p=0.037, p<0.001, p<0.001, respectively), with Blacks, older ages, and high body fat categories having many more observed abnormalities than nonblack, younger age, and normal body fat categories. Personality type showed no effect (p=0.562). Using the three covariates age, race, and percent body fat, the percent impaired and percent high glucose categories were adjusted and relative risks were calculated. These data are summarized in Table 18-14 and revealed that ' significantly fewer Ranch Hands had impaired glucose levels (at least 140 but less than 200 mg/dl) than did Comparison members, as demonstrated by the fact that the relative risk was bracketed by a confidence interval with upper limit less than 1.00. Conversely, more Ranch Hands had diabetic levels of glucose (at least 200 mg/dl) on the 2-hour postprandial test than did the Comparisons, but this excess was not statistically significant. The 2-hour postprandial glucose level was also analyzed as a continuous variable. Group data were transformed to a logarithmic scale and were adjusted by a general linear model using the covariates age, race, occupation, and percent body fat. This analysis is reflected in Table 18-15. These results showed no group difference for the 2-hour postprandial glucose variable. Significant covariate effects are noted for percent body fat (p<0.001), occupation (p<0.001), and the age-by-race interaction (p*0.002). 18-22 �TABLE 18-14. Adjusted Categorical Analysis for 2-Hour Postprandial Glucose Analysis Contrast Adjusted Relative Risk. 95% C.I. Covariate Remarks p-Value Overall* 0.034 Impaired vs. Normal 0.73 (0.56,0.96) 0.022 Diabetic vs. % Normal 1.26 (0.72,2.22) Age (p<0.001) Race (p=0.016) Percent Body Fat (p<0.001) 0.421 "Chi-square test (2 d.f.) for group difference. TABLE 18-15. Adjusted Continuous Analysis for 2-Hour Postprandial Glucose by Group Group Total Ranch Hand 976 Comparison 1,234 Adjusted Mean 95% C.I. 114.4 (107.3,122.0) p-Value Covariate Remarks Age-by-Race(p*0.002) 0.487 Occupation(p<0.001) 115.3 (108.2,123.0) Percent Body Fat(p<0.001) 18-23 �Composite Diabetes Indicator This variable was constructed by selecting participants with a known history of diabetes via the Baseline or interval (1982-1985) questionnaire, and adding them to the group whose 2-hour postprandial glucose level was at least 200 mg/dl at the 1985 examination. Thus, this pool represents all "true diabetics," past and present. These data were contrasted to the "nondiabetics," recognizing the mild degree of misclassification introduced by considering glucose-impaired individuals as normal. The unadjusted frequencies (Table 18-5) were 7.3 percent diabetics in the Ranch Hand group and 6.7 percent diabetics in the Comparison group (p=0.622). A series of analyses were conducted to determine the best adjusting model for these data, using stepdown procedures from a model containing all main effects and two- or three-way interactions. The final adjustment used the significant covariates of percent body fat and an age-by-race interaction to adjust the proportions of diabetes in each group. These results, formulated as a relative risk, are presented in Table 18-16. The adjusted results indicated no significant difference in the frequency of past and current diabetes in the Ranch Hand and Comparison groups. The analyses above provide a firm platform to conclude that both study groups were essentially equal with respect to glucose metabolism, and past and current diabetes. Although the herbicide literature suggests a possible endpoint of diabetes, this followup study provides no support for that notion. The slight discrepancies between the categorical tests of glucose abnormalities and the assessment of mean values are probably explained on distributional grounds. EXPOSURE INDEX ANALYSES Within each occupational category, exposure index analyses were carried out to assess possible dose-response relationships (see details in Chapter 8). The variables T3 % Uptake, TSH, testosterone, initial and 2-hour cortisol, differential cortisol, and 2-hour postprandial glucose were analyzed as continuous variables by t-tests and analysis of variance (unadjusted by any of the covariates). Adjusted analyses were performed using general linear models; adjusting covariates were age, race, occupation, and as appropriate, percent body fat and personality type. Group-by-covariate interactions were explored for each analysis, and tests were made of differences in means among the three exposure levels as well as contrasts of means between the medium and low exposure levels, and between the high and low exposure levels. The dependent variables were transformed prior to analysis as described earlier in this chapter. TABLE 18-16. Adjusted Analysis for Diabetes (Composite Indicator) Adjusted Relative Risk 95% C.I. p-Value 1.12 (0.80,1.56) 0.500 18-24 Covariate Remarks Age-by-Race ( p»0 . 005 ) Percent Body Fat (p<0.001) �Results of the adjusted analyses are presented in Table 18-17 and parallel results for unadjusted analyses are given in Table P-2 of Appendix P. Results of investigation of any exposure index by covariate interactions are given in Table P-3 of Appendix P. Unadjusted analyses showed significant differences either among exposure levels or in the high versus lov or medium versus low exposure level contrasts for testosterone for officers, and initial cortisol, differential cortisol, and 2-hour postprandial glucose for enlisted flyers. For officers, a significantly lower mean testosterone level was seen for the medium exposure level as contrasted to the low exposure level (547.4 mg/dl versus 599.4 mg/dl, p=0.041). Enlisted flyers had significantly lower mean initial cortisol in the medium as contrasted with low exposure level (11.08 ug/dl versus 11.97 ug/dl, p=0.001); participants in the high exposure level also had a much lower mean, 11.13 ug/dl, as contrasted with the low exposure level but the difference was not significant. Enlisted flyers had a significant difference in'differential cortisol among exposure index levels (p=0.003). The mean differential cortisol levels were 3.43 ug/dl, 1.20 yg/dl, 2.30 ug/dl for the low, medium, and high exposure levels, respectively; the medium versus low contrast was very significant (p<0.001), and the high versus low contrast was marginally significant (p=0.092). Mean 2-hour postprandial glucose for enlisted flyers in the medium exposure category was much higher than in the low category: 118.0 mg/dl versus 100.9 mg/dl (p-0.015). However, the mean glucose level for the high exposure category was not as high as that for the medium level, 110.9 mg/dl. The difference among all the exposure levels was close to significance (p=0.051). Adjusted analyses (Table 18-17) showed patterns very similar to unadjusted analyses. A summary of exposure index by covariate interactions found are listed in Table 18-18. The adjusted mean TSH level for enlisted flyers was significantly higher in the high exposure level as contrasted with the low exposure level (p=0.045); moreover, there was a steady trend upwards with low, medium, and high exposure levels. Enlisted flyers in the medium exposure level had a higher adjusted mean 2-hour cortisol level than the low exposure level (p=0.034), but no trend was apparent. There was a significant difference in differential cortisol among the exposure levels of enlisted flyers (p=0.008) and the medium exposure level had a much lower adjusted mean than the low exposure level (p=0.002). No clear trend with increasing exposure was apparent. Further, enlisted flyers in the medium exposure level had a higher mean postprandial glucose than the lower level (p=0.012), and the overall test for differences among the three levels was significant (p=0.042). In summary, the emergent pattern was that the enlisted flyers in the medium exposure level were significantly different from those in the low exposure level for 2-hour cortisol, differential cortisol, 2-hour postprandial glucose and marginally significantly different (p=0.098) for testosterone. However, the corresponding high versus low contrasts were not statistically significant. LONGITUDINAL ANALYSES Three endocrine variables were chosen for longitudinal analysis: testosterone, T3 % Uptake, and TSH. Only participants attending both examinations were eligible. The three variables were measured by relatively comparable laboratory techniques at the Kelsey-Seybold Laboratory in 1982 and 18-25 �TABLE 18-17. Adjusted Exposure Index Analyses for Endocrinological Variables' by Occupation Exposure Index Variable Medium High Low Officer n Adj . mean 95% C.I. 124 120 Overall 126 28.15 28.17 28.58 M vs. L (27.36,29.01) ( 7 7 , 9 4 ) (27.35,28.98) H vs. L 2.82.1 0.180 0.120 0.928 n Adj . mean 95% C.I. 55 55 65 Overall 27.45 27.62 27.95 M vs. L (26.69,28.24) (26.90,28.36) (27.24,28.68) H vs. L 0.388 0.639 0.178 Enlisted Groundcrew »-» 00 Statistic Enlisted Flyer T3 % Uptake Occupation n Adj . mean 95% C.I. 140 Overall 153 160 27.87 28.00 27.96 M vs. L (27.60,28.41) (27.56,27.96) (27.45,28.30) H vs. L 0.853 0.857 0.579 Officer n Adj . mean 95% C.I. Overall 124 120 126 1.212 1.343 M vs. L 1.263 (1.045,1.555) (1.011,1.479) (1.107, 1.664) H vs. L 0.262 0.513 0.332 Enlisted Flyer n Adj . mean 95% C.I. 55 55 65 Overall 1.058 M vs. L 0.899 1.005 (0.768,1.067) (0.860,1.191) (0.904,1.254) H vs. L 0.120 0.155 0.045 Enlisted Groundcrew n Adj . mean 95% C.I. 140 Overall 153 160 M vs. L 1.135 1.174 1.151 (1.041,1.243) (1.054,1.263) (1.070,1.294) H vs. L 0.807 0.775 0.513 1 N} TSH Contrast p-Value �TABLE 18-17. (continued) Adjusted Exposure Index Analyses for Endocrinological Variables by Occupation Exposure Index 00 Initial Cortisol Statistic Low Medium Officer n Adj. mean 95% C.I. 125 482.6 (414.7,555.5) 128 116 Overall 464.5 461.0 M vs. L (395.8,531.2) (397.9,536.2) H vs. L 0.560 0.312 0.405 n Adj. mean 95% C.I. 55 57 63 Overall M vs. L 507.7 571.3 536.1 (427.6,594.7) (492.7,655.7) (462.6,614.9) H vs. L 0.251 0.098 0.454 Enlisted Groundcrev Testosterone Occupation Enlisted Flyer Variable n Adj. mean 95% C.I. 153 **** **** 161 **** **** 137 **** **** Overall M vs. L H vs. L ****a **** **** Officer n Adj. mean 95% C.I. 124 **** **** 130 **** **** 119 **** **** Overall H vs. L H vs. L ***** **** Enlisted Flyer n Adj. mean 95% C.I. 55 65 11.69 11.11 (10.38,13.17) (9.96,12.39) 57 11.08 (9.97,12.32) Overall M vs. L H vs. L 0.533 0.335 0.320 Enlisted Groundcrew n Adj. mean 95% C.I. 154 11.11 (9.96,12.40) 140 11.01 (9.88,12.27) Overall M vs. L H vs. L 0.948 0.757 0.809 160 10.98 (9.87,12.23) High Contrast p-Value �Adjusted Exposure Index Analyses for Endocrinological Variables by Occupation Exposure Index Variable 2-Hour Cortisol Occupation Officer Enlisted Flyer Enlisted Groundcrev Differential Cortisol Officer Enlisted Flyer Enlisted Groundcrew Statistic Low Medium High Contrast p-Value �TABLE 18-17. (continued) Adjusted Exposure Index Analyses for Endocrinological Variables by Occupation Exposure Index Medium 00 I Statistic Low Officer n Ad j . mean 95% C.I. 121 124 111.1 106.8 (100.8,122.4) (97.1,117.5) n Ad j . mean 95% C.I. Enlisted Groundcrew 2-Hour Postprandial Glucose Occupation Enlisted Flyer Variable n Ad j . mean 95% C.I. High Contrast p-Value 111 108.1 (98.1,119.3) Overall M vs. L H vs. L 0.411 0.191 0.383 54 113.7 (97.9,132.0) 62 56 Overall 134.4 121.9 M vs. L (117.2,154.2) (106.5,139.6) H vs. L 0.042 0.012 0.286 150 107.1 (96.2,119.3) 155 110.4 (99.2,122.7) 138 109.2 (98.2,121.6) Overall M vs. L H vs. L ****Group-by-covariate interaction—adjusted mean, confidence interval, and p-value not given. "Exposure index-by-percent body fat interaction. b Exposure index-by-race interaction. °Exposure index-by-race and exposure index-by-personality type interactions. 0.706 0.413 0.597 �TABLE 18-18. Summary of Exposure Index-by-Covariate Interactions Encountered in Analyses of Endocrinological Variables Variable Occupation Covariate p-Value Testosterone Enlisted Groundcrew Percent Body Fat 0.001 Initial Cortisol Officer Percent Body Fat 0.037 2-Hour Cortisol Officer Percent Body Fat 0.011 2-Hour Cortisol Enlisted Groundcrew Differential Cortisol Enlisted Groundcrew Race Race Personality Type 0.006 0.007 0.021 at the SCRF Laboratory in 1985. As described in Chapter 7, "Statistical Methods," each variable was analyzed continuously by a repeated measurements analysis of variance. Testosterone data were subjected to a square root transformation, and TSH values received a logarithmic transformation. Results of the analysis are shown in Table 18-19. As shown in Table 18-19, all three variables declined from their Baseline values, but the reductions over time were relatively proportional for each group by variable. It is concluded that significant differences between groups did not exist for the change in levels between the Baseline examination and the first followup examination. The symmetrical changes in the testosterone and T3 % Uptake variables are speculatively attributed to a 3-year aging effect, but the change in TSH values is suggestive of a change in laboratory methods. There is no suggestion of an adverse rate change in either the Ranch Hand or Comparison group. SUMMARY AND CONCLUSIONS The physical examination and laboratory testing results of all endocrinological variables are summarized in Table 18-20. Questionnaire and review-of-systerns data for past thyroid disease were essentially equivalent in both the Ranch Hand and Comparison groups. These historical data were confirmed by medical record reviews. Physical examination findings were necessarily limited to data from palpation of thyroid glands and testicles; the unadjusted results showed no significant group differences. 18-30 �Longitudinal Analysis for Testosterone, T3 % Uptake, and TSH: A Contrast of Baseline and First Followup Examination Test Means Means Variable Group Testosterone p-Value (Equality of Difference) Ranch Hand Comparison TSH 1985 Followup Ranch Hand Comparison T3 % Uptake Total 1982 Baseline Ranch Hand Comparison Evaluation of the endocrine system was conducted primarily by laboratory testing of hormone levels. The thyroid test battery consisted of T3 % Uptake and TSH assays. The T3 % Uptake data showed no group differences for either mean values or frequency of abnormally low or high values. Occupation was a significant covariate. TSH results revealed a significantly higher mean level in the Ranch Hand group, but this difference was not found by categorical testing of proportions of abnormally high TSH results. Mean levels of testosterone were significantly elevated' among Ranch Hands as contrasted with Comparisons in the 10 to 25 percent body fat category, but this was not reflected by the categorical tests. For the few participants with less than 10 percent body fat (six Ranch Hands, four Comparisons), mean testosterone levels were lower for Ranch Hands than for Comparisons. Age, occupation, and percent body fat were significant adjusting variables. Two timed cortisol specimens showed no significant group differences in mean values and percent abnormalities. The difference between the timed cortisol results, termed the differential cortisol, showed no significant group differences for nonblacks or Blacks born before 1942, but Black Ranch Hands born in or after 1942 had a lower mean differential cortisol level than Comparisons. Age, percent body fat, and personality type were significant covariates in these analyses. Group means of 2-hour postprandial glucose levels were not statistically different, but categorical testing revealed that there was a significantly higher frequency of glucose-impaired (at least 140 but less than 200 mg/dl) Comparisons than Ranch Hands. A constructed variable comprised of known diabetics and individuals classified as diabetic by the glucose tolerance test, showed no difference between the Ranch Hand and Comparison groups. As expected, past and current diabetes were highly influenced by the covariates age, race, and percent body fat. �Overall Summary Results of Unadjusted and Adjusted Continuous and Categorical Analyses of Endocrinological Variables Unadjusted Test Mean Categorical Questionnaire and Physical Examination Past Thyroid Disease (SelfAdministered) Past Thyroid Disease (Interviewer Administered) Thyroid Abnormalities Testicular Abnormalities Laboratory Testing T, % Uptake TSH Testosterone Initial Cortisol 2-Hour Cortisol Differential Cortisol 2-Hour Postprandial Glucose Diabetes (Composite Indicator) ~aAnalysis not feasible. NS^ Not significant (p>0.10). — Analysis not performed. NS*j Borderline significant (0.05<p<0.10). ****Group-by-covariate interaction. Ad j us t ed Mean Categorical �Exposure index analyses did not reveal any pattern consistent with a dose-response relationship. Enlisted flyers in the medium exposure level were significantly different from those in the low exposure level for 2-hour cortisol, differential cortisol, and 2-hour postprandial glucose. However, the corresponding high versus low contrasts were not statistically significant. Longitudinal analyses of T3 % Uptake, TSH, and testosterone levels on all individuals attending both the Baseline and followup examinations revealed only symmetrical and nonsignificant changes in the Ranch Hand and Comparison groups in the interval between examinations. In conclusion, both limited historical and physical examination data, seven endocrinological laboratory variables, and a composite indicator of diabetes did not demonstrate consistent patterns indicating an herbicide effect. However, there was a significant interaction between group and percent body fat for testosterone that could be interpreted as an herbicide effect. TSH and testosterone means tests were statistically significant, and in the expected direction of an herbicide effect, but these results were not confirmed by categorical testing. Also significant was the impaired category of the glucose tolerance test, which showed an excess in the Comparison group. The consistent demonstration of the classical effects of the covariates age, race, occupation, and percent body fat on appropriate endocrine variables provided support for these conclusions. Overall, the endocrine health status of both groups was reasonably comparable. 18-33 �Chapter 18 REFERENCES 1. Potter, C.L., I.G. Sipes, D.H. Russell. 1983. Hypothyroxinemia and . ... , hyothermia in rats in response to 2,3,7,8-tetrachlorodibenzo-p-dioxin ' ' ' ion. Tnxirnl. Annl. Pharmarnl. 69:89-95. administration. Toxicol. Appl. Pharmacol. fiQiflQ-QS. 2. Bastomsky, C.H. 1977. Enhanced thyroxine metabolism and high uptake goiters in rats after a single dose of 2,3,7,8-tetrachlorodibenzop-dioxin. Endocrinology 101:292-296. 3. Barsotti, D.A., L.J. Abrahamson, and J.R. Allen. 1979. Hormonal alterations in female Rhesus monkeys fed a diet containing 2,3,7,8-tetrachlorodibenzo-p-dioxin. Bull. Environ. Contam. Toxicol. 21(4-5):463-469. " 4. Nienstedt, ¥., M. Parkki, P. Uotila, and A. Aitio. 1979. Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on the hepatic metabolism of testosterone in the rat. Toxicology 13:233-236. 5. Van Logten, M.J., B.N. Gupta, E.E. McConnell, and J.A. Moore. 1980. Role of the endocrine system in the action of 2,3,7,8-TCDD on the thymus. Toxicology 15(2):135-144. 6. Neal, R.A., P.W. Beatty, and T.A. Gasiewicz. 1979. Studies on the mechanism of toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). In Health effects of halogenated aromatic hydrocarbons, ed. W.J. Nicholson and J.A. Moore, 320:204.New York:The New York Academy of Sciences. 7. Rozman, K., T. Rozman, and H. Greim. 1984. Effect of thyroidectomy and thyroxine on 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) induced toxicity. Toxicol. Appl. Pharmacol. 72:372. 8. Rozman, K., E. Scheufler, T. Rozman, T. Pazdernick, and H. Greim. 1984. Effect of thyroxine (T4) and triiodothyronine (T3) on TCDD toxicity in thyroidectomized rats. Toxicologist 4:189. 9. Rozman, K.K. 1984. Role of thyroid hormones and brown adipose tissue in the toxicity of TCDD. In Banbury Report 18; Biological mechanisms of dioxin action, ed. A. Poland and R.O. Kimbrough, ppT 345-354.Cold Springs Harbor, New York: Cold Spring Harbor Laboratory. 10. Bahn, A.K., J.L. Mills, P.J. Snyder, P.H. Gann, L. Houten, 0. Bialik, L. Hollmann, and R.D. Utiger. 1980. Hypothyroidism in workers exposed to polybrominated biphenyls. N. Engl. J. Med. 302:31-33. 18-34 �11. Peters, H.A., A. Gocmen, D.J. Cripps, G.T. Bryan, and I. Dogramaci. 1982. Epidemiology of hexachlorobenzene-induced porphyria in Turkey. Arch. Neurol. 39:744-749. 12. Stehr, P.A., G. Stein, H. Falk, et al. 1986. A pilot epidemiologic study of possible health effects associated with 2,3,7,8tetrachlorodibenzo-p-dioxin contamination in Missiouri. Arch. Environ. Health 41:16-22. 13. May, G. 1973. Chloracne from the accidental production of tetrachlorodibenzodioxin. Br. J. Inds. Med. 30:276-283. 14. Pazderova-Vejlupkova, J., M. Nemcova; J. Pickova, L. Jirasek, and E. Lukas. 1981. The development and prognosis of chronic intoxication by tetrachlorodibenzo-p-dioxin in men. Arch. Environ. Health 36:5-11. 15. Poland, A.P., D. Smith, G. Metter, and P. Possick. 1971. A health survey of workers in a 2,4-D and 2,4,5-T plant, with special attention to chloracne, porphyria cutanea tarda, and psychologic parameters. Arch. Environ. Health 22(3):316-327. 16. Suskind, R.R., and V.S. Hertzberg. 1984. Human health effects of 2,4,5-T and its toxic contaminants. JAMA 251:2372-2380. 17. Hoffman, R.E., P.A. Stehr-Green, K.B. Webb, G. Evans, A.P. Knutsen, V.F. Schramm, J.L. Staake, B.B. Gibson, and K.K. Steinberg. 1986. Health effects of long-term exposure to 2,3,7,8-tetrachlorodibenzo-pdioxin. JAMA 255:2031-2038. 18. Moses, M., R. Lilis, K.D. Crow, J. Thornton, A. Fischbein, H.A. Anderson, and I.J. Selikoff. 1984. Health status of workers with past exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in the manufacture of 2,4,5trichloro-phenoxyacetic acid: Comparison of findings with and without chloracne. Am. J. Ind. Med. 5:161-182. 19. Gruhn, J.G., C.P. Barsano, and Y. Kumar. 1987. The development of tests of thyroid function. Arch. Pathol. Lab. Med. 111:84-100. 20. Schwertner, H.A., R.G. Troxler, G.S. Uhl, and W.G. Jackson. 1984. Relationship between cortisol and cholesterol in men with coronary artery disease and type A behaviour. Arteriosclerosis 4:59-64. 21. Troxler, R.G. and H.A. Schwertner. 1985. , lifestyle, and coronary heart disease. 56:660-665. 22. Aldersberg D., L. Schaefer, S.R. Drachman. 1950. Development of hypercholesterolemia during cortisone and ACTH therapy. JAMA 144:909-914. 23. Stern, M.P., O.G. Kolterman, J.F. Fries, H.O. McDevitt, G.M. Reaven. 1973. Adrenocortical steroid treatment of rheumatic diseases: Effects on lipid metabolism. Arch. Intern. Med. 132:97-100. 18-35 Cholesterol, stress, Aviat. Space Environ. Med. �24. Friedman, M., S.O. Byers, and R.H. Rosenman. 1972. Plasma ACTH and cortisol concentration of coronary-prone subjects. Proc. Soc. Exp. Bio. Med. 140:681-684. 25. Rubenstein, E., and D.D. Federman, eds. 1986. Metabolism: Diabetes Mellitus. Chap. 9. in Scientific American Medicine. New York: Scientific American, Inc. 18-36 �CHAPTER 19 IMMUNOLOGICAL EVALUATION INTRODUCTION Overt damage to organs of the immune system and depressed immunologic function have been noted in a variety of animals exposed to TCDD. As the fields of immunology and immunotoxicology have grown within the past 10 years, a significant spectrum of subtle immunotoxic effects has also been described in animals, but for many possible reasons, comparable adverse effects have not been consistently recorded in exposed human individuals or cohorts. Thus, an intensive search is underway to ascertain the effects of TCDD on the human immune system, particularly with respect to the development of cancer. Every major ongoing dioxin morbidity study in the United States has now incorporated comprehensive laboratory assessments of the immune system. Numerous animal studies have demonstrated significant immunotoxicity following the administration of TCDD. The relatively consistent observations of decreased thymus weight (with cortical atrophy and a depletion of lymphocytes), atrophy of other lymphoid tissue, depressed cellular bone marrow, and decreased humoral and cell-mediated immunity and increased susceptibility to infection have been noted in a variety of animals, including monkeys, rabbits, guinea pigs, rats, and mice. ~ The immune-response effects varied by species, species strain, age, integrity of the endocrine system, dose, and route of administration. Generally, the immunologic parameters returned to normal or approximate normal values over time, even following moderate to high doses of TCDD. While experiments have been conducted to assess the immunotoxicity of TCDD, little has been published on the immunotoxicity of 2,4-D or 2,4,5-T. The. immune system is so sensitive to TCDD that immune function has frequently been used as a marker of toxicity in the absence of other biologic effects. The mechanism of TCDD immunotoxicity is under intensive investigation by molecular biologists, pathologists, and geneticists. In general, TCDD toxicity is probably linked to the Ah receptor, and specifically to the Ah allele which governs microsomal enzyme induction as reflected by aryl hydrocarbon hydroxylase and cytochrome P-448/450 levels.1 This premise underscores the questions of the degree to which the human response to TCDD is dependent upon the Ah locus or other genetic receptors, and how this response is manifested in the immune system. Animal studies and several observational studies in humans have shown variable results. Data from the Times Beach, Missouri, episode disclosed no group differences for various T lymphocyte populations, proliferative responses to PHA, concanavalin A, pokeweed or tetanus toxoid stimulation, and in skin testing with seven antigens.1 A report of the assessment of the immune system of men exposed to TCDD in an industrial accident in Britain 19-1 �did not discuss the results of the measurement of the immunoglobulin profile, lymphocytes, T and B cells, response to PHA, and three hematologic variables. A prior publication on the same cohort cited unpublished findings of Ward, suggesting a reduction in the capacity of the "primary" immune system. A longitudinal evaluation of 48 highly exposed children (one-half with chloracne) from the Seveso incident showed significantly elevated complement hemolytic activity over six measurements during the period 1976 to 1979 (although the biologic significance of this is unknown) and an increased Pf9liferative response to PHA and pokeweed during the first three screenings. This study (as others) was characterized by shifting a study population over the observation period and by excessive laboratory variation that may have masked other true group differences. Nonetheless, the Seveso data may be interpreted as indicative of a stimulated immune system, particularly cellmediated immunity, differing substantially from the bulk of animal studies, which showed decreased activity. A recent study of residents of the TCDD-contaminated Quail Run Mobile Home Park in Missouri also revealed data that conflicted with the Seveso experience. A statistically significant amount of anergy and relative anergy was detected in the TCDD-exposed group, as determined by the multitest applicator (seven-antigen test system). Inter-reader variation presented major interpretive difficulties. Nevertheless, findings suggestive of decreased cell-mediated immunity were provided by decreased T , T4, and T X1 cell percentages. Also noted was an increased lymphoproliferative response to pokeweed mitogen (PWM). The overall depression of immunologic response was not correlated with an increase in clinical disease. Baseline Summary Results Immunologic function and phenotypic marker studies were performed on 592 participants (297 Ranch Hands, 295 Comparisons) randomly selected by the terminal digit of their case number. Because of laboratory problems, e.g., fluctuating quality control and lack of simultaneous differential counts on the peripheral mononuclear cells, a special Immunology Review Committee was convened to determine which data were relevant for analysis. Such decisions were made on a case-by-case basis without knowledge of Ranch Hand or Comparison group membership. The Committee concluded that the data could be analyzed on a group basis, but interpretation of data on an individual basis was inappropriate. Analyses of the cell surface markers (T11, T3, T4, T8, B, the T4/Tg ratio, and the total lymphocyte count [TLC]) showed no significant group differences. However, the increased smoking was significantly associated with increases in cell counts but not with the T4/T8 ratio and B cells, whereas increasing age was significantly associated with decreasing TLC and T. cells. o Functional studies of T and B cells via reaction to antigenic (tetanus toxoid) or mitogen (phytohemagglutinin, concanavalin A, and pokeweed) stimulation showed no group differences. Similarly, unadjusted and adjusted mean values of the four assays were not significantly different between groups, but one unstimulated control value (reflecting Baseline thymidine uptake by T cells) was significantly decreased in the Ranch Hands. The biologic relevance of this finding was unclear. 19-2 �Further, in the covariate analysis of the functional studies, group-bysmoking and group-by-alcohol interactions were noted. Of greater importance, however, was the finding that lymphocytic response increased with increased smoking, but was depressed in association with increasing age. In summary, both immunologic function and cell marker studies did not show significant impairment in the Ranch Hand group, or patterns supportive of an herbicide effect. Smoking, for the first time to the knowledge of the authors, was associated with a significant increase in the marker cells T l l f T3, T4, and T8, and in the total lymphocyte count, with a concomitant increase in lymphocytic response to PWM. Parameters of the 1985 Immunologic Profile The format for the 1985 AFHS physical examination placed more emphasis on the immunologic assessment than did the 1982 Baseline profile. The random sampling scheme was expanded to produce an approximate 50-percent sample of the cohorts, and included the same terminal digits of the case number used at Baseline in order to include all individuals evaluated in 1982 and thus establish a longitudinal data base. All immunologic tests were performed at the Scripps Immunology Reference Laboratory (SIRL). The battery of phenotypic marker assays and functional tests was slightly modified from the Baseline profile. The assay for HLA-DR cells was added to the battery of marker studies, and a functional test for natural killer cells (with and without interferon) was substituted for the concanavalin A stimulation assay. A comprehensive set of skin tests for the antigens Candida, mumps, Trichophyton, and staph-phage-lysate was added to evaluate the integrity of the delayed hypersensitivity response. The dependent variables of the analyses in this chapter arise from three distinct measurement systems: phenotypic marker studies, functional studies, and skin testing. There were more covariates than in the Baseline study, namely age, race, occupational category, exposure index, and new smoking and alcohol data from the 1985 questionnaire. Participants deleted from the immunological analyses included those with recent radiation or chemotherapy, and those individuals on immunosuppressive or systemic steroid medication. Marginal totals in the tables below vary somewhat due to missing covariate data. Thus, numbers in the table also vary according to which immunologic data sources were used in the analysis. In general, most analyses are based upon data from 465 Ranch Hands and 585 Comparisons. Analytic tests included t-tests, general linear models (SAS®-GLM), logistic regression (BMDP®-LR), and Fisher's exact test. Parallel analyses using Original Comparisons are in Tables Q-5 through Q-10 of Appendix Q. Rationale of the Immunologic Measurements Because of rapid changes in,our knowledge of the immune system, Table 19-1 is provided as an aid in interpreting the medical significance of the immunological data. 19-3 �TABLE 19-1. Medical Significance of the Immunological Data Immunologic Measure Rationale of the Measurement Disease/Syndrome/Condition Endpoint MARKER STUDIES Measures total T cells coincident with sheep rosette receptor on cell surface (most are T4 and Tg cells). Decreased in immune deficiency/ increased with lymphoproliferative disorders. Measures peripheral blood B cells, no reaction with T cells, granulocytes, or monocytes. Decreased in immunodeficiency/ increased in lymphoproliferative disorders. Measures T cells which exhibit helper/ inducer phenotype. Decreased in AIDS/increased in autoimmune diseases. OKT Measures T cells which exhibit suppressor/ cytotoxic functions. Variable in autoimmune diseases. Increased in some viral illnesses and immunodeficiencies. 4 Leu M3 Measures mature monocytes in peripheral blood. Increases with inflammation. HLA-DR Measures cells expressing HLA-DR antigen; includes B cells and monocytes. B cell deficiency/ Agamraaglobulinemia. Mixed Leukocyte Culture (MLC) Measures reactivity of T cells to foreign histocompatiblity antigens on unrelated lymphocytes. HLA sensitization/transplantation. PHA Measures functional capability of T cells to become activated by mitogen and undergo proliferation. T cell deficiency. NKC (with interferon) NKC (without interferon) Measures natural killer cell lytic activity with and without interferon treatment of the natural killer cells. Decreased natural defenses. Leu 12 vO FUNCTIONAL STUDIES �TABLE 19-1. (continued) Medical Significance of the Immunological Data Immunologic Measure Rationale of the Measurement Disease/Syndrome/Condition Endpoint FUNCTIONAL STUDIES (continued) PUM Measures functional capability of T cells to become activated by mitogen and undergo proliferation. T cell deficiencies. Each measures skin reactivity induced by specific antigen injected intradermally and correlates with recall T cell sensitivity to the antigen. Antigen reactivity or sensitivity/ Anergy. SKIN TESTS Candida Mumps Tricophyton Staph-phage-lysate vO I �Immunology Methodologies The isolation of peripheral blood mononuclear (PBM) cells was the first step to prepare for testing immune competence and enumeration of phenotypic markers. Heparinized whole blood was obtained from each patient. PBM's were isolated by Ficoll-Hypaque density gradient centrifugation. The PBM's were then washed and resuspended in HB101 media (HANA Biologies, Inc.) supplemented with 10M units/ml penicillin, 10,000 meg/ml streptomycin, 1 percent sodium pyruvate (100 mM), and 1 percent L-glutamine (200 mM). To determine percent monocyte and granulocyte contamination of the PBM cell preparations, an aliquot of the cells was stained with a nonspecific esterase stain. PBM concentration was adjusted for each individual assay, Cell Surface Marker Analysis Mouse monoclonal antibodies directed against specific surface markers were used to identify and quantitate different cell populations in the peripheral blood of the participants. Mononuclear cell concentrations adjusted to 1.0 x 10 cells/ml were incubated with the following fluorescein isothiocyanate conjugated monoclonal antibodies: CD2(OKT11*), CD4(OKT4*), CD8(OKT8*), CD19(Leul2**), CD14(LeuM3**), and HLA-DR(OKDR*). These cell surface antibodies measure total numbers of T and B lymphocytes, monocytes, helper T lymphocytes, suppressor T lymphocytes, and those cells carrying the HLA-DR antigen. A flow cytometer (Spectrum III, Ortho Diagnostic Systems, Raritan, New Jersey) was used to measure percent positive for each specific surface marker and absolute numbers were calculated. Phytohemagglutinin (PHA) and Pokeveed Mitogen Stimulation Assays Mitogens were used to stimulate the proliferation of lymphocytes in vitro. During the proliferative response, the lymphocytes undergo blast transformation and incorporate radioactive thymidine into their DNA. Participant lymphocyte concentrations were adjusted to 2.0 x 10 cells/ml in supplemented HB101 media. Samples were cultured in quadruplicate. Individual cultures consisted of 0.1 ml of cell suspension and 0.1 ml of mitogen solution in microtiter plates. The cultures were incubated in an atmosphere of 5 percent CO, at 37 degrees Centigrade. Participant cells were cultured with PHA (12 ug/ml, Sigma Chemical Co., St. Louis, Missouri) for a total of 4 days and pokeweed mitogen (0.05 ug/ml, Sigma Chemical Co., St. Louis, Missouri) for a total of 5 days. The cultures were pulsed with tritiated thymidine (1.0 uCi/ microtiter well) for 4 hours and then harvested on a multiple automated harvester. Cellular proliferation was assessed by tritiated thymidine uptake measured by liquid scintillation counting. Mixed Lymphocyte Reaction Histocompatibility antigens (HLA) can also stimulate lymphocytes causing blast transformation. Donor lymphocytes were used to stimulate the proliferation of participants' lymphocytes in vitro. A pool of donor lymphocytes *0rtho Diagnostic Systems, Raritan, New Jersey **Becton Dickinson Monoclonal Center, Inc., Mountain View, California 19-6 �was frozen and used as a stimulator pool throughout the course of the study. An aliquot of this pool was thawed daily. Viability of this pool was assessed by trypan blue exclusion. A pool of freshly isolated lymphocytes was prepared daily and also used as stimulator cells. Both pools of stimulator cells were inactivated by irradiation ( , 0 rad). 300 Stimulator pools and participant lymphocyte concentrations were adjusted to 1.0 x 10 cells/ml in supplemented HB101 media. Samples were cultured in quadruplicate. Individual cultures consisted of 0.1 ml of participant cell suspension and 0.1 ml of stimulator cell suspension, in microtiter plates. The cultures were incubated in an atmosphere of 5 percent CO, at 37 degrees Centigrade for 6 days. The cultures were pulsed with tritiated thymidine (1.0 uCi/ microtiter well) for 16 hours and then harvested on a multiple automated cell harvester. Cellular proliferation was assessed by tritiated thymidine uptake measured by liquid scintillation counting. Natural Killer Cell Assays Mononuclear cells from the participant were evaluated to assess the ability of 'certain peripheral blood cells to kill target cells from a K-562 leukemia cell line. The K-562 target cells were preincubated with radioactive chromium ( Cr) at 37 degrees Centigrade in 5 percent C02 for 1 hour, washed, and the cell concentration adjusted to 1.6 x 10 cells/ml. A 50 ul aliquot of radioactive K-562 cells was added to each microtiter well. Participant lymphocytes were adjusted to three different concentrations: 0.53, 1.6, and 2.7 x 10 cells/ml. One ml of each of these concentrations was incubated with 20 units of recombinant y-interferon (Genentech, Inc., San Francisco, California) for 1 hour at about 37 degrees Centigrade. Quadruplicate 150 ul aliquots of each concentration, with and without interferon preincubation, were dispensed in a microtiter plate. Four wells contained media alone to determine the spontaneous release of radioactivity from the K-562 cells. Four wells contained 1 percent Triton X-100 to determine the maximal release of radioactivity. The final effector to target ratios were 50;1, 30:1, and 10:1. The microtiter plates were centrifuged briefly at low speed and incubated at 37 degrees Centigrade in 5 percent C02 for 3 hours. A 100 pi aliquot of the supernatant was removed from each well and counted on a gamma counter. Percent chromium release from the K-562 target cells was determined for each effector:target cell ratio. Interpretive Considerations The values of the results of assays of immunologic status are more variable than those found in routine single reactant clinical chemistry assays. Often there are numerous biochemical factors/metabolites that affect the immunologic assay results so that interpretations of normalcy must be in the context of those 'obtained concurrently in a normal control cohort group. Such controls allow for proper adjustments of the raw assay data in order to minimize the broad range of technical and reagent effects in the various immunologic assays. These adjustments in the raw assay data results will correct for such variability and allow for the detection of any significant biologic abnormalities. Because of the need for these control adjustments, the immunologic assay results cannot be meaningfully compared to existing normal ranges determined on different groups of individuals at other times. 19-7 �RESULTS AND DISCUSSION Cell Surface Marker (Phenotypic) Studies Immunological tests were carried out on 47 percent (1,085) of the participants because of the complexity of the assay and the expense of these tests. The participants were randomly selected so that approximately 50 percent of each group of participants arriving for the physical examination had blood drawn for the immunological tests. Logistical delay during the initial weeks of the examination reduced the number to less than 50 percent. Within each group, blood was drawn for the immunological tests from about one-half of the selected participants on the first day of the physical examination, and from the remainder on the following day. Skin tests, which were scheduled for the first day, were therefore carried out after the blood draw on the first day for the first half of the immuno-tested participants. Skin tests were not done for those participants selected for immunological testing on day two in order to avoid any effect the skin test antigens might have on the cell counts and functions. Thus, 553 participants received both the iinmunological tests and the skin tests, 532 received the immunological tests but not the skin tests, 1,206 received the skin tests but not the immunological tests, and 18 received neither. Table 19-2 gives the frequencies of the participants in each exposure group who had the tests. Participants who were taking anti-inflammatory or immunosuppression medication or who had recently received x-ray treatment or chemotherapy for cancer were excluded from all the analyses. Participants taking aspirin, however, were not excluded. Frequencies of Participants Who Took the Immunological Tests and the Skin Tests, by Group Skin Tests Group Ranch Hand Immunology Tests No Yes Total Comparison No Yes Total No Yes Total �For those participants who were given the immunological tests, following dependent variables were examined: total T cells, helper suppressor T cells, B cells, monocytes, HLA-DR cells, and the T4/Tg suppressor cell) ratio. These variables were treated as continuous analysis. the T cells, (helper/ in the The covariates considered in the analysis were the matching variables (age, race, occupation), smoking history (current cigarettes/day and total pack-years of smoking), and alcohol consumption (average number of drinks per day during the 2 weeks prior to the physical examination and total drinkyears). The covariates age and the smoking history and alcohol consumption variables were used as continuous variables in the analyses since the relationships between the dependent variables and the covariates were generally monotonic. Considerable day-to-day variation exists in the results of immunological tests due to a number of extraneous factors, including temperature, humidity, and sensitivity of the instrumentation. Significant batch-to-batch variation (among examination groups) was apparent for total T cells, suppressor T cells, B cells, and the T4/T8 ratio, and significant blood-draw day variation was apparent for helper T cells, monocytes, and HLA-DR cells. Adjustments in the analyses were made for these sources of variation by using batch or blooddraw day indicators. Throughout this section, appropriate adjustment was carried out in the assessment of group differences of the dependent variables; this analysis was unadjusted for the covariates listed above and is referred to as the "unadjusted" analysis. Adjustment was also made for batch-to-batch or blood-draw day variation in the analyses of the associations of the dependent variables with the covariates. Further, this adjustment was also used in the fitting of general linear models to assess the group differences, adjusted for the covariates. Prior to analysis, group data were pooled for each continuous variable and were examined to determine whether transformation would enhance normality or distributional symmetry. The following transformations were used in the analyses: Variable Transformation Total T cells Square root Helper T cells Square root Suppressor T cells Logarithm B cells Square root Monocytes Logarithm HLA-DR cells Square root T4/T8 ratio , Logarithm The results of the analyses in this section are summarized in Tables 19-3 through 19-5. Table 19-3 presents the unadjusted analyses for the cell surface markers, Table 19-4 displays the covariate associations, and Table 19-5 gives the adjusted results. These tables are accompanied by 19-9 �Unadjusted Analyses for Cell Surface Markers by Group Group Variable Total T Cells Helper T Cells (T4) Suppressor T Cells (T8) B Cells Monocytes HLA-DR Cells T4/T8 Ratio Statistic Ranch Hand Comparison p-Value �Association Between Cell Surface Marker Variables and the Covariates in the (Combined Ranch Hand and Comparison Groups (Directionality Shown) Variable Race Occupation Age Current Alcohol (Drinks/Day) DrinkYears Current Smoking (Cigarettes/Day) Lifetime Smoking (Pack-years) Total T Cells Helper T Cells Suppressor T " Cells B Cells Monocytes HLA-DR Cells T./T. Ratio 4 o "Monotone decreasing. NS: Not significant (p>0.10) b N: Nonblack °Monotone increasing. B: Black Generally decreasing trend. 0: Officer e E: Enlisted personnel (flyer and lncreases, drop-off at highest category. lncreases from 0 category, then decreases, but not back to same level. Increases from 0 category, then steady decrease with increasing levels. 9 groundcrew) Generally increasing trend. F: Enlisted flyer Flat for 0 and first few categories, then increases. G: Enlisted groundcrew �TABLE 19-5. Adjusted Analyses for Cell Surface Markers by Group Group Variable Statistic Ranch Hand Comparison p-Value Total TCells n Adj. Mean 95% C.I. 442 567 iLJ.iti .!• TvTOwC TCnrCn Helper TCells n Adj. Mean 95% C I .. 439 566 894 6. 885 7. 062 .6 (3., 9 3 2 (4., 984 861 0 . ) 892 0.) BATCH (p=0.021) nAY(BATCH) (p=0.014) AGE (p<0.001) ALOOCC (p=0.008) CSMCK*OCC (p=0.023) ALC*CSMCK (p=0.006) (p=o.oi2) Suppressor TCells n Adj. Mean 95% C I .. 463 50 8 530.8 537.9 0.640 (506.8, 556.0) (516.1, 560.5) BATCH (p<0.001) OCC (p=0.014) AGE (p=0.004) ALC (p=0.020) CSMCK (p<0.001) BCells n Adj. Mean 95% C I .. 435 561 BATCH (p<0.001) ALC (p=0.006) AGE*CSMCK (p=0.025) E8KHl*RACE (p=0.026) GRP*PACm (p=0.018) GRP*RACE*OCC (p=0.046) Monocytes n Adj. Mean 95% C.I. 440 568 BATCH (pO.OOl) DAY(BATCH) (p<0.001) PACE (p=0.032) EBKXR (p=0.013) CSMOK (p<0.001) PACKYR (p=0.006) GRP*OCC (p=0.044) GRP*ALC (p=0.010) 19-12 Covariate Remarks* BATCH (p=0.029) AGE (p=0.009) ALC (p=0.001) CSMCK (p<0.001) GRP*RACE (p=0.033) (p=0.015) �TABLE 1 - . (continued) 95 Adjusted Analyses for Gall Surface Markers by Group Group Variable Statistic Ranch Hand Comparison p-Value HLA-CR Cells n Adj. Mean 95% C I .. 49 5 50 8 T4/T 8 Ratio n Adj. Mean 95% C I .. 461 577 150 .7 152 .5 068 .7 (1.501, 1.643) (1.491, 1.616) Covariate Remarks* MKH (p<0.001) DAY(MDCH) (p=0.004) OCC (p=0.035) CSMCK (p<0.001) GRP*ALC (p=0.045) AG£*PACKYR (p=0.005) B^TCH (p<0.001) OCC (p=0.020) CSMOK (p<0.001) *Abbreyiations BATCH: batch-to-batch variation among examination groups DAY(MICH): blood-draw day variation ALC: current alcohol use CSMOK: current smoking OCC: occupation GRP: group ERKXR: lifetime alcohol use (drink-years) PACKYR: lifetime smoking (pack-years) ****Significant group-by-covariate interaction—adjusted mean, confidence interval, and p-value not presented. 19-13 �discussion of each variable. The results of adjusted analyses with group-bycovariate interactions are found in Table Q-l of Appendix Q. Total T Cells ( ^ T ) No significant difference was found between groups in the mean values of total T cells (p=0.736). These data were analyzed without adjustment for any covariates except batch-to-batch variation. The data were pooled for the two groups, and the relationship with the covariates was examined. Significant associations were found with occupation (p=0.005), age (p=0.002), current smoking (p<0.001), and pack-years (p<0.001). A marginal association (p=0.069) was found with current alcohol use due to a steady decrease in mean counts with higher drinking levels. Officers had a lower mean count (1,539 cells/mm ) | a enlisted flyers (1,668 cells/mm ), or hn enlisted groundcrew (1,647 cells/mm ) . The mean count decreased with age: 1,663 cells/mm , 1,582 cells/mm , and 1,404 cells/mm for those born in or after 1942, born between 1923 and 1941, and born in or before 1922, respectively. The mean count increased with increasing current smoking and increasing lifetime smoking history (pack-years). A general linear model was fitted to assess the group difference in mean count of total T cells with adjustment for each covariate and any interactions that made significant contributions to the model. Batch-to-batch variation was a significant covariate (p=0.029). A significant group-by-race interaction was found (p=0.033); Black Ranch Hands had a significantly lower adjusted mean count than Black Comparisons (1,566 cells/mm versus 1,888 cells/mm ; p=0.039), but the group difference for nonblacks was not significant (p=0.619) (see Table Q-l of Appendix Q). The following covariates were significant: age (p=0.009), current alcohol use (p=0.001), current smoking (p<0.001), and a drink-year-by-pack-year interaction (p=0.015). Analyses using only Original Comparisons showed the same results as when using the total Comparison group (see Tables Q-6 and Q-7 of Appendix Q), with a group-by-race interaction present (p=0.028). Helper T Cells ( 4 T) No significant difference was found between groups in the mean values of helper T cells (p=0.610). This contrast was analyzed without adjustment for any covariates except blood-draw day variation. The data were pooled for the two groups, and the relationship with the covariates was examined. Significant associations were found with occupation (p=0.024), age (p<0.001), current smoking (p<^).001), and pack-years (p<0.001). Officers had gt lower mean count (831 cells/mm ) than enlisted flyers ( 8 cells/mm ) or enlisted groundcrew ( 9 cells/mm ). There was a decrease 85 84 in the mean count with increasing age: 907 cells/mm , 850 cells/mm , and 713 cells/mm for those born in or after 1942, born between 1923 and 1941, and born in or before 1922, respectively. The mean count increased with increasing levels of current smoking and with increasing pack-years of lifetime smoking. 19-14 �Adjusted analyses assessed the group difference in mean count of helper T cells with adjustment for each covariate and any significant interactions. Adjustment for the blood-draw day variation was included. Age made a significant contribution to the model (p<0.001)1 The following interactions between covariates were significant: current alcohol use-by-occupation (p=0.008), current smoking-by-occupation (p=0.023), current alcohol use-bycurrent smoking (p=0.006), and drink-years-by-pack-years (p=0.012). The adjusted group difference in mean count was not significant (p=0.662): 869 cells/mm for the Ranch Hand group versus 879 cells/mm for the Comparison group. Adjusted analyses using Original Comparisons (Table Q-6 of Appendix Q) also revealed a nonsignificant group difference (p=0.835). Suppressor T Cells (TB) No significant difference was found between groups in the mean values of suppressor T cells (p=0.671). This contrast was analyzed without adjustment for any covariates except batch-to-batch variation. The data were pooled for the two groups, and the relationship with the covariates was examined. Significant associations were found with occupation (p<0.001), age (p<0.001), and current smoking (p<0.001). The mean count for officers was less than the mean count for enlisted groundcrew, which was in turn less than the mean count for enlisted flyers; the means were 492 cells/ mm , 540 cells/mm , and.575 cells/mm , respectively. The mean counts decreased with increasing age: 557 cells/mm , 512 cells/mm , and 439 cells/ mm for participants born in or after 1942, born between 1923 and 1941, and born in or before 1922, respectively. The mean counts increased with increasing levels of current smoking. Marginally significant associations were found with current alcohol use (p=0.058, mean T8 counts decreased with increasing current levels of drinking) and pack-years (p=0.076, mean counts increased with increasing pack-years). The adjusted analysis of group differences in mean count of suppressor T cells was made with adjustment for each covariate and any interactions that made significant contributions, including significant batch-to-batch variation (p<0.001). Significant adjusting covariates were occupation (p=0.014), age (p=0.004), current alcohol use (p=0.020), and current smoking (p<0.001). The adjusted group difference was not significant (p=0.640). A marginal (p=0.063) group-by-race interaction was not retained in the final model, but was explored. Black Ranch Hands had a lower adjusted mean count than Black Comparisons (512 cells/mm versus 649 cells/mm , p=0.056), whereas the difference between nonblack groups was negligible (531 cells/mm3 for Ranch Hands and 532 cells/mm for Comparisons, p=0.974). Analyses involving the Original Comparisons showed a significant interaction between group and race (p=0.010) (see Tables Q-6 and Q-7 of Appendix Q), with the same pattern seen for the group-by-race interaction for the total Comparison group. B Cells No significant difference was found between groups in the mean values of B cells (p=0.594). This contrast was analyzed without adjustment for any covariates except batch-to-batch variation. 19-15 �Significant associations using pooled data were found between B cells and occupation (p<0.001), age (p<0.001), current alcohol use (p=0.001), drink-years (p=0.047), current smoking (p<0.001), and pack-years (p=0.005). Officers had a lower mean count than enlisted flyers and groundcrew (166 cells/mm , 205 cells/mm , and 201 cells/mm , respectively). The mean count decreased with increasing age: 206 cells/mm , 177 cells/mm , and 146 cells/mm for those born in or after 1942, born between 1923 and 1941, and born in or before 1922, respectively. The mean counts decreased with an increasing number of drinks per day, and also with higher levels of total lifetime drinking, except for the "never-drinkers," whose level was lower than the greater than 30 to 100 drink-year group; the means for the drinkyear categories weres 0, 180 cells/mm ; greater than 0 to 5, 199 cells/mm ; greater than 5 to 30, 189 cells/mm ; greater than 30 to 100, 183 cells/mm3; and greater than 100, 150 cells/mm . The nonsmokers had a lower mean count than current smokers, whereas among the smokers the mean counts decreased with higher current smoking levels. The means for the different currentsmoking (cigarettes/day) categories were: 0, 166 cells/mm ; greater than 0 to 20, 237 cells/mm 5 greater than 20 to 40, 222 cells/mm ; and greater than 40, 202 cells/mm3. Lifetime smokers had a higher mean count than "neversmokers"; otherwise the pattern was not clear. Adjusted analyses, including adjustment for the significant (p<0.001) batch-to-batch variation, were used to investigate the mean count of B cells. Adjustment was made for each covariate and any interactions that made significant contributions. A significant group-by-race-by-occupation interaction was found (p=0.046), along with a group-by-pack-year interaction (p=0.018). Significant contributions were made by current alcohol use (p=0.006), an age-by-current smoking interaction (p=0.025), and a drink-years-by-race interaction (p=0.026). The analysis consequently was performed separately for nonblacks and Blacks. For nonblacks, the group-by-pack-year interaction persisted (p=0.021) (see Table Q-l of Appendix Q). Ranch Hands who had never smoked had a much lower adjusted mean count than the corresponding Comparisons, 154 cells/mm versus 190 cells/mm (p=0.004). Among smokers, the adjusted mean count for the greater than 0 to 20 pack-year category was less for Ranch Hands than for Comparisons. For both the greater than 20 to 40 and the greater than 40 pack-year categories, the adjusted mean count was higher for Ranch Hands than for Comparisons. The p-values for these three contrasts were greater than 0.10. For Blacks, the unadjusted group difference was not significant (p=0.808; Ranch Hands, 186 cells/mm , versus Comparisons, 194 cells/mm ). Adjusted means were not calculated because no covariates made any significant contribution to an adjusted model, and moreover, adjustment for batch-to-batch variation was not possible because of the small number of Black participants. Other significant covariates and interactions in the adjustment for nonblacks included occupation (p=0.047), drink-years (p<0.001), and an age-by-current smoking interaction (p=0.039). Monocytes No significant difference was found between groups in the mean value of monocytes (p=0.427). This contrast was analyzed without adjustment for any covariates except blood-draw day variation. 19-16 �The data were pooled for the two groups, and the relationship with the cbvariates was examined. Significant associations were found with race (p«0.027), occupation (p=0.019), current drinking (p=0.031), drink-years (p<0.001), current smoking (p<0.001), and pack-years (p<0.001). Blacks had a lower mean count than nonblacks (37.1 cells/mm3 versus 45.7 cells/mm , respectively). Officers had a lower mean count (42.3 cells/mm ) than enlisted flyers ( 4 4 cells/mm ), who had a lower mean count than enlisted 4. groundcrew ( 8 2 cells/mm3). Higher mean counts were associated with higher 4. current drinking levels. There were increases in mean counts with higher drink-years and with increasing amounts of both current and lifetime smoking. Assessment of the group difference in mean count of monocytes was done with adjustment for each covariate and any interactions that made significant contributions, including blood-draw day variation. A significant group-by-occupation interaction (p=0.044) and a significant group-by-current alcohol use interaction (p=0.010) were found. For interpretation, these were explored in a model including the group-byoccupation-by-current alcohol use interaction, with the alcohol variable discretized (see Table Q-l of Appendix Q). Except for those men consuming more than two to four drinks per day, Ranch Hand officers had a higher adjusted mean count than Comparison officers, the difference being large ( 4 2 cells/mm versus 32.3 cells/mm ) for nondrinkers, (p=0.060). For 4. enlisted flyers, except those in the greater than four drinks per day category, Ranch Hands had a lower adjusted mean count than corresponding Comparisons. For the greater than two to four drinks per day category, a large difference between adjusted means (32.7 cells/mm for Ranch Hands, 56.2 cells/mm for Comparisons) was observed (p=0.097). Further, it was found that for enlisted groundcrew not currently drinking, Ranch Hands had a lower adjusted mean count than the corresponding Comparisons, whereas the Ranch Hand current drinkers had higher adjusted mean counts than the corresponding Comparisons. The difference was large ( 8 9 cells/mm versus 35.3 6. cells/mm ) for the greater than four drinks per day category (p=0.003). Significant effects on the monocyte counts were also seen for race (p=0.032), drink-years (p=0.013), current smoking (p<0.001), and pack-years (p=0.006). Analyses using Original Comparisons revealed a significant (p=0.040) group-by-age interaction (see Tables Q-6 and Q-7 of Appendix Q). This was due to a lower count for Ranch Hands than Comparisons for those born in or after 1942 ( 1 4 cells/mm versus 48.0 cells/mm , p=0.048), a higher 4. count for Ranch Hands than Comparisons for those born between 1923 and 1941 ( 8 2 cells/mm versus 42.8 cells/mm , p=0.058), and very little difference 4. for those born in or before 1922 (p=0.924). HLA-DR Cells No significant difference was found between groups in the mean values of HLA-DR cells (p=0.842). This contrast was analyzed without adjustment for any covariates except blood-draw day variation. Significant associations were found using pooled data with occupation (p<0.001), age (p=0.010), current smoking (p<0.001), and pack-years (p<0.001). Officers had a lower mean count than enlisted participants (526 cells/mm versus 597 cells/mm for flyers and 598 cells/mm for groundcrew). The average mean count was higher for younger participants than for 19-17 �older participants: 588 cells/mm3, 557 cells/mm3, and 555 cells/mm3 for those born in or after 1942, born between 1923 and 1941, and born in or before 1922, respectively. There was a significant increase in average mean counts with increasing levels of both current and lifetime smoking. There was a marginally significant increase in mean cell counts with drink-years (p=0,083). Analyses, with adjustment for blood-draw day, each covariate, and any interactions, were carried out to assess the group difference in mean count of HLA-DR cells. A significant group-by-current alcohol use interaction was found (p=0.045); for Ranch Hands drinking more than four drinks per day, the adjusted mean count was greater, 564 cells/mm versus 473 cells/mm , than for Comparisons (p=0.052), whereas no appreciable group differences were apparent for the participants drinking four or fewer drinks per day (see Table Q-l of Appendix Q). Significant effects were seen with occupation (p=0.035), current smoking (p<0.001), and an age-by-pack-year interaction (p=0.005). Analyses using Original Comparisons (Table Q-6 of Appendix Q) did not show a significant group-by-current alcohol use interaction (p=0.152), and no significant difference between groups was observed (psO.887). T4/T8 Ratio No significant difference was found between groups in the mean value of the T4/T8 ratio (p=0.499). This contrast was analyzed without adjustment for any covariates except batch-to-batch variation. The data were pooled for the two groups, and the relationship with the covariates was examined. Significant associations were found with drinkyears (p=0.049), current smoking (p<0.001), and pack-years (p=0.001). The mean T4/T8 ratio generally increased with increasing drink-years, and increased with increasing amounts of current smoking and total pack-years. There was a marginally significant association with occupation (p=0.063). Enlisted flyers had a lower average ratio than officers and enlisted groundcrew ( . 8 versus 1.62 and 1.60, respectively). 14 The adjusted group difference in the T4/T8 ratio was not significant (p=0.678; Ranch Hands 1.57 versus Comparisons 1.55). Significant effects were seen for occupation (p=0.020), current smoking (p<0.001), and batch-tobatch variation (p<0.001). In the analysis of the Original Comparisons, a significant group-bycurrent smoking interaction was found (p=0.016). Further analysis snowed a significant difference between groups (Ranch Hand mean ratio of 1.84 versus Original Comparison mean ratio of 1.51, p=0.004), for the greater than 20 to 40 cigarettes per day category (see Tables Q-6 and Q-7 of Appendix Q). Functional Stimulation Studies Statistical analyses were performed on cell function responses to PHA, PWM, and MLC. For each stimulated cell population, autologous controls were also studied. The measurements resulting from each test were the average counts over four samples for the stimulated cell population and for the autologous controls. The net average count, defined as the difference 19-18 �between the average counts per minute (CPM) for the stimulated and the control cells, was also calculated. Cell response data were obtained from the 1,085 immunologically tested participants. The exclusion conditions were the same as in the previous section, namely, participants who were taking anti-inflammatory or immunosuppressant medication, or who had recently experienced radiation therapy or chemotherapy for cancer. Review of the immunological data base by the Air Force and SIRL resulted in certain test exclusions due to technical error, and equipment malfunction, those identified by quality control procedures (Grubbs' test2 ), and unexplained outliers. For the mean cell counts per minute (CPM) analyzed for this section, a total of 17 data points were excluded as unexplained outliers from the data base, involving eight participants (three Ranch Hands and five Comparisons): 1 point was invalid due to technical error in the assay for MLC stimulated cells (Ranch Hand) and the remainder were outliers in the PWM controls or stimulated cells (two Ranch Hand controls, 13 Comparison controls, and one Ranch Hand stimulated cells). This meant that, for one participant, the PWM control mean was omitted from the analysis and, for the other seven participants, the means were calculated from fewer than four points. No unexplained data points were found for the PHA-stimulated cells or corresponding controls. All analyses were adjusted for significant blood-draw day variation, and the same covariates were used as in the adjusted analyses of the cell surface markers. The covariates age, current smoking, pack-years, current alcohol use, and drink-years were discretized because marginal examination showed generally nonlinear responses of the cell function variables with these covariates. Thus, the p-values given in this section for the marginal association of the variables with each covariate indicate the significance of the differences among the categories defined by the levels of the covariate. Prior to analysis, the data were transformed to enhance normality or at least distributional symmetry. The following transformations were used: Variable Transformation Unstimulated Response (PHA) logarithm PHA Net Response none Pokeweed Net Response square root MLC Net Response square root The summarized results of this section are given in Tables 19-6 through 19-8 (see Table Q-l of Appendix Q for results involving group-by-covariate interactions). Only results for the unstimulated controls for the PHA assay are presented as an assessment of the function of the immune system in the unchallenged state. However, separate controls were run for each assay since incubation periods vary for each1test procedure. In the analysis of data on the net response for each assay, the appropriate control was used. Analysis of each control assay was performed, and no significant group differences were noted. 19-19 �Unadjusted Analyses for Functional Stimulation Tests by Group Variable Statistic* Group Ranch H a n d C o m p a r i s o n p - V a l u e Unstimulated Response (PHA) PHA Net Response Pokeweed Net Response MLC Net Response Unstimulated Response (PHA) No significant difference was found between groups in the mean values of PHA Unstimulated responses (p=0.979). These control values were derived from Unstimulated cells and reflect baseline cell function. This contrast was analyzed without adjustment for any covariates except blood-draw day variation. The data were pooled for the two groups, and the relationship with the covariates was examined. Significant associations"were found with race (p<0.001), age (p<0.001), and drink-years (p=0.048). The average mean count for nonblacks was lower than for Blacks: 1,629 CPM and 2,210 CPM, respectively. There was a strong decrease in mean count with increasing age. For those born in or after 1942, the mean was 1,770 CPM; for those born between 1923 and 1941, the mean was 1,606 CPM; and for those born in or before 1922, the mean count' was 1,238 CPM. The mean count generally decreased with increasing drink-years, with a maximum mean count of 1,726 CPM for nondrinkers and a minimum mean count of 1,414 CPM for participants with greater than 100 drink-years. A marginally significant association was found with occupation (p=0.086). The average mean count for officers was lower than that for enlisted flyers, which was in turn lower than that for enlisted groundcrew: the means were, respectively, 1,592 CPM, 1,662 CPM, and 1,713 CPM. This relationship with occupation was not seen with the PWM or MLC Unstimulated responses. Since these values were derived from the same blood specimens and were Unstimulated, this observation may represent a chance occurrence. There was a marginally significant association with pack-years (p=0.051); the response generally increased with increasing pack-years. �Association Between Functional Stimulation Test Variables and the Covariates in the Combined Ranch Hand and Comparison Groups (Directionality Shown) Variable Race Occupation Age Current Alcohol (Drinks/Day) DrinkYears Unstimulated Response (PHA) PHA Net Response Pokeweed Net Response MLC Net Response a Monotone decreasing. Generally decreasing trend. c Flat for 0 and first few categories, then increases. d lncreases,- drop-off at highest category. e Generally increasing trend. f Monotone increasing. 9 Increases from 0 category, then decreases, but not back to same level. NS: N: B: 0: F: G: Not significant ( > . 0 . p01) Nonblack Black Officer Enlisted flyer Enlisted groundcrew Current Smoking (Cigarettes/Day) Lifetime Smoking Pack-years �TABLE 19-8. Adjusted Analyses for fractional Stimulation Tests by Group Group Variable Statistic* Ranch Hand Unstimulated n Response Mean (PHA) 95% C I .. 44 6 1,741 (,9, 191 155 ,0) Covariate Comparison p-Value 54 8 1,731 (,9, 182 153 ,8) 085 .5 Remarks BATCH ( < . 0 ) p001 DAY (BATCH) ( < . 0 ) p001 RACE ( < . 0 ) p001 AGE ( < . 0 ) p001 BATCH (p<0.001) DAY (BATCH) (p<0.001) RACE (p=0.011) AGE*CSMOK (p=0.007) AUXSMOK (p=0.008) PHANet Response n Adj. Mean 95% C I .. 461 581 023 .3 1892 8,5 1320 9,8 (7,3, 2059 (7,8, 2604 1602 1,2) 1189 0,1) Net Pokeweed Response n Mean 95% C I .. 463 9,6 157 (219 1 1 4 1 8,8 0 , 5 ) 52 8 9,9 007 (108 9 , 6 ) 8,0, 9 6 9 059 .7 BATCH (p<0.001) DAY (BATCH) (p<0.001) RACE*OCC (p=0.024) ALC*OCC (p=O.036) ALC*CSMCK (p=0.009) Net MLC Response n Adj. Mean 95% C I .. 40 3 50 5 ** ** BATCH (p<0.001) DAY (BATCH) (p=0.001) EMOR (p<0.001) ALC (p=0.001) GRP*PACKSR (p=0.046) RACE*CSMOK (p=0.043) *Group means and confidence intervals expressed as counts per minute (CPM). ****Group-by-covariate interaction—adjusted mean, confidence interval, and p-value not presented. 19-22 �(The means were 1,657 CPM, 1,696 CPM, 1,519 CPM, and 1,712 CPM for 0, greater than 0 to 20, greater than 20 to 40, and greater than 40 pack-years, respectively.) Adjusted analyses to assess the group difference in mean counts of PHA controls were performed with adjustment for each covariate and any interactions that made significant contributions, including significant blood-draw day variation. The group difference in adjusted mean count was not significant (p=0.855; Ranch Hand group mean of 1,741 CPM versus Comparison group mean of 1,731 CPM). Race and age were significant covariates (p<0.001 for both). Adjusted and unadjusted analyses using the Original Comparisons (see Tables Q-8 and Q-9 of Appendix Q) showed similar results; i.e., no significant group difference (p=0.608, unadjusted; p=0.613, adjusted). PHA Net Response No significant difference was found between groups in the mean values of net response to PHA (p=0.339). This contrast was analyzed without adjustment for any covariates except blood-draw day variation. The data were pooled for the two groups and the relationship with the covariates was examined. Significant associations were found for race (p=0.002), age (p<0.001), current alcohol use (p<0.001), and drink-years (p=0.002). Nonblacks had a lower net count than Blacks (208,953 CPM, 233,622 CPM, respectively). There was a steady decrease in net count with increasing age: the means were 217,003 CPM, 206,901 CPM, and 184,419 CPM for those born in or after 1942, born between 1923 and 1941, and born in or before 1922, respectively. Those currently drinking more than four drinks per day had a lower mean net count than those drinking less. The participants in the greater than 100 drink-year category had a lower mean net count than those with fewer drink-years. Using a general linear model with adjustment for each covariate and any significant interactions including blood-draw day variation, the adjusted group difference was found to be not significant (p=0.233; Ranch Hand count of 193,280 CPM versus Comparison count of 188,952 CPM). Significant contributions were made b'y race (p=0.011), an age-by-current smoking interaction (p=0.007), and a current alcohol use-by-current smoking interaction (p=0.008). A marginally significant (p=0.057) group-by-occupation interaction was excluded from the final model. However, this interaction was explored, and was found to be due to a group difference among enlisted flyers (p=0.014); the adjusted mean Ranch Hand net stimulated count was greater than that of the Comparisons (207,050 CPM and 185,344 CPM, respectively). Analyses using the Original Comparisons (see Tables Q-9 and Q-10 of Appendix Q) revealed a significant group-by-occupation interaction (p=0.017), with results similar to the Ranch Hand versus total Comparison contrast of net counts; namely, enlisted flyer Ranch Hands had an adjusted mean count greater than enlisted flyer Original Comparisons (p=0.003). 19-23 �Pokeweed Net Response No significant difference was found between groups in the mean values of net response to pokeweed (p=0.317). These data were analyzed without adjustment for any covariates except blood-draw day variation. Significant associations were found using the pooled group data with drink-years (p=0.038), current smoking (p<0.001), and pack-years (p=0.001). The mean count was higher for those with greater than 100 drink-years and lower for never-drinkers, but with no pattern for the in between categories. For both current and lifetime smoking (pack-years), there was a steady upward trend in mean counts with increasing levels of smoking. The difference in adjusted group means was not significant: Ranch Hands, 91,567 CPM, and Comparisons, 90,097 CPM (p=0.579). The following interactions were significant: race-by-occupation (p=0.024), current alcohol use-by-occupation (p=0.036), and current alcohol use-by-current smoking (p=0.009). Net Response to MLC Stimulation No significant difference was found between groups in the mean response to MLC stimulation (p=0.185). These data were analyzed without adjustment for any covariates except blood-draw day variation. The data were pooled for the two groups, and the relationship with the covariates was examined. Significant associations were found for age (p=0.035), drink-years (p=0.008), current smoking (p<0.001), and pack-years (p=0.015). The net mean count generally decreased with increasing age: 84,543 CPM, 72,408 CPM, and 79,081 CPM for those born in or after 1942, between 1923 and 1941, and in or before 1922, respectively. The net mean count was lowest for never-drinkers, with no clear pattern among the drinkers: 66,933 CPM, 78,555 CPM, 80,713 CPM, 84,236 CPM, and 80,416 CPM for the 0, greater than 0 to 5, greater than 5 to 30, greater than 30 to 100, and greater than 100 drink-year categories, respectively. There was a monotonically increasing trend in net average count with current smoking, the nonsmokers having a much lower value than the smokers. An equivalent pattern was found for lifetime smoking (pack-years). Adjusted analyses were carried out to assess the group difference in mean counts of MLC net response, including adjustment for the significant blood-draw day variation. A significant group by pack-year interaction was found (p=0.046). Never-smoking Ranch Hands had a lower adjusted mean count (68,921 CPM) than the corresponding Comparisons (77,232 CPM) (p=0.053). Ranch Hands in the greater than 0 to 20 pack-year category had a lower adjusted mean count (67,976 CPM) than the corresponding Comparisons (74,333 CPM) (p=0,057). The adjusted means for the Comparisons decreased with increasing pack-years, whereas those of the Ranch Hands generally increased (see Table Q-l of Appendix Q). Significant contributions were made to the model by drink-years (p<0,001), current alcohol use (p=0.001), and a race-by-current smoking interaction (p=0.043). 19-24 �Discussion The performance of the phenotypic and cell stimulation studies was monitored daily by highly structured quality assurance techniques (see Chapter 6). This resulted in a remarkably error-free data set, in contrast to the immunologic tests at the Baseline study that required the assistance of a review group to determine which data were appropriate for analysis. The finding of significant blood-draw day and batch-to-batch variation at the followup examination was judged to be totally normal and inherent within the test procedures; only a few data points within specific variables were omitted because of outlying values. The unique use of a "batch" variable for adjustment of all the phenotypic and stimulation studies permitted unadjusted and covariate-adjusted group contrasts while controlling for inherent laboratory variation. All unadjusted and adjusted analyses (without group interactions) showed no significant group differences. Analysis of MLC revealed a group-by-packyear of smoking interaction with lower counts in the Ranch Hand group than in the Comparison group for 0 and greater than 0 to 20 pack-year categories. Despite differences in the quality of Baseline and followup results, slight changes in cohort numbers, and different mathematical models, there was remarkable concordance in the immunologic results of both examinations, both for the dependent variables and for the effects of the covariates. No judgment of adverse immunologic competence was made for any variable, or sets of variables, or in substrata examined because of group-by-covariate interactions for the cell surface marker and cell stimulation studies. EXPOSURE INDEX ANALYSES Within each occupational category, exposure index analyses were conducted to assess possible dose-response relationships (see details in Chapter 8). Analyses were performed for the cell surface marker variables (total T cells, helper T cells, suppressor T cells, B cells, monocytes, HLA-DR cells, and the T4/T8 ratio) and for the functional stimulation tests (the control counts per minute for the PHA test, and the net PHA, PWM, and MLC counts per minute). Analyses were not done for the skin test responses. Unadjusted and adjusted analyses were performed using general linear models. Exposure index-by-covariate interactions were explored in the adjusted analyses. Covariates were age, race, current and lifetime alcohol use (drink-years), and current and lifetime cigarette smoking (pack-years). For each analysis, an overall test was made of the differences among the means corresponding to the low, medium, and high exposure index levels. Medium versus low and high versus low contrasts of means were also made. Results of the adjusted analyses are .presented in Table 19-9 for cell surface markers and 19-10 for functional stimulation tests. Parallel results of unadjusted analyses are given in Tables Q-2 and Q-3, Appendix Q. Results of exposure index-by-covariate interactions are also given in Table Q-4 of Appendix Q. 19-25 �Adjusted Exposure Index Analyses for Cell Surface Markers by Occupation Exposure Index Variable Total T Cells Occupation Officer Enlisted Flyer Enlisted Groundcrew Helper T Cells Officer Enlisted Flyer Enlisted Groundcrew Statistic Low Medium High Contrast p-Value �Adjusted Exposure Index Analyses for Cell Surface Markers by Occupation Exposure Index Variable Suppressor T Cells Occupation Officer Enlisted Flyer Enlisted Groundcrew B Cells Officer Enlisted Flyer Enlisted Groundcrew Statistic Low Medium High Contrast p-Value �Adjusted Exposure Index Analyses for Cell Surface Markers by Occupation Exposure Index Variable Monocytes Occupation Officer Enlisted Flyer Enlisted Groundcrew HLA-DR Cells Officer Enlisted Flyer Enlisted Groundcrew Statistic Low Medium High Contrast p-Value �Adjusted Exposure Index Analyses for Cell Surface Markers by Occupation Exposure Index Variable T4/T8 Ratio Occupation Officer Enlisted Flyer Enlisted Groundcrew Statistic Low Medium High Contrast p-Value �Adjusted Exposure Index Analyses for Functional Stimulation Tests by Occupation Exposure Index Variable Unstimulated Response (PHA) Occupation Officer Enlisted Flyer Enlisted Groundcrew PHA Net Response Officer Enlisted Flyer Enlisted Groundcrew Statistic Low Medium High Contrast p-Value �Adjusted Exposure Index Analyses for Functional Stimulation Tests by Occupation Exposure Index Variable Pokeweed Net Response Occupation Statistic Low Medium High Contrast Officer Enlisted Flyer Enlisted Groundcrew MLC Net Response Officer Enlisted Flyer Enlisted Groundcrew *Group-by-covariate interaction—adjusted mean, confidence interval, and p-value not presented. p-Value �Cell Surface Markers Unadjusted analyses revealed very few significant results. Among enlisted groundcrew, the medium exposure level had a significantly lower mean total T cell count than the low exposure level (1,555 cells/mm versus 1,759 cells/mm , p=0.032), and the high exposure level mean was marginally significantly (p=0.091) lower than the low exposure level mean (1,586 cells/ mm versus 1,759 cells/mm ) . Suppressor T cells, for enlisted groundcrew in the low exposure level, were marginally significantly higher than in the medium or high exposure levels (575.5, 502.3, 505.5 cells/mm , respectively? medium versus low, p=0.063, high versus low, p=0.097). For enlisted flyers, the trends with exposure level were steadily downwards for total T cells, helper T cells, B cells, and the T4/T8 ratio, and upwards for suppressor T cells and monocytes, but no contrasts were significant. Adjusted analyses revealed marginally significant differences among exposure levels of enlisted groundcrew for total T cells (p=0.068) and suppressor T cells (p=0.088). For both total and suppressor T cells, the means for the medium and high exposure levels were much lower than those of the low exposure level. For total T cells, the adjusted means were: low, 1,737 cells/mm ; medium, 1,533 cells/mm 5 and high, 1,558 cells/mm (medium versus low: p=0.029, high versus low: p-0.085). For suppressor T ce].ls, the adjusted means were: low, 558.6 cells/mm ; medium, 4 0 8 cells/mm ; and 8. high, 483.7 cells/mm (medium versus low p=0.044, high versus low p=0.081). A similar but less marked pattern was seen for helper T cells. In summary, there was no consistent evidence of any significant doseresponse pattern in an occupational category. For the enlisted flyer cohort, six of the seven variables revealed nonsignificant dose-response trends in the unadjusted analyses, but only two trends persisted after adjustment by the covariates. Functional Stimulation Tests Exposure index analyses were performed on PHA unstimulated responses, and net PHA, PWM, and MLC counts. For officers, the unadjusted mean PHA unstimulated response counts varied significantly among exposure levels (p=0.047). The means were 1,705 CPM, 1,428 CPM, and 1,809 CPM, respectively, for low, medium, and high exposure levels (medium versus low: p«0.071, high versus low: p=0.557). The PWM net count for enlisted flyers was significantly lower for the high versus the low exposure levels (55,480 CPM versus 92,847 CPM, p=0.011). The PHA net count had a downward trend with increasing exposure level for enlisted groundcrew. The PWM net count for officers had an increasing trend but there was no statistically significant difference among exposure levels. In the adjusted analyses, officers had a significant exposure index-bydrink-year interaction (p=0.011) for PHA controls, and an exposure index-byage interaction for the PHA net count (p=0.003) (see Table 19-11 for a summary of these interactions). .Although the numbers were small, the high 19-32 �Summary of Exposure Index by Covariate Interactions for Functional Stimulation Tests Variable Occupation Covariate Unstimulated Response (PHA) Officer Drink-years PHA Net Response Officer p-Value Age exposure-level nondrinking officers had a lower mean PHA control count than the low exposure level (1,557 CPM versus 3,273 CPM, p=0.031), and the high exposure level officers with more than 100 drink-years had a higher PHA control count than the corresponding low exposure group ( , 0 CPM versus 670 1,983 CPM, p=0.049). Officers born in or after 1942 had a lower PHA net count in the medium exposure level as contrasted with the low exposure level (153,534 CPM versus 261,397 CPM, p=0.002). Other adjusted analyses revealed that enlisted flyers had a lower PWM net count in the high exposure level as compared to the low exposure level (111,772 CPM versus 173,897 CPM, p=0.014), as in the unadjusted analysis. Enlisted groundcrew in the medium exposure level also had a marginally significantly higher MLC net count as compared to the low exposure level (78,259 CPM versus 61,403 CPM, p=0.097). In summary, there was no evidence for a strong dose-response relationship, but there was a trend for declining PWM and MLC net counts for enlisted flyers with increasing exposure level. SKIN TESTING RESULTS General Four skin test antigens, mumps, Candida albicans, Trichophyton, and staph-phage-lysate, were intradermally administered to 76 percent (1,759) of the participants on the first day of the examination. Skin tests were not given to the remaining 24 percent of the population because they had been selected to give blood for the immunological tests on the second day of their examination. Candida albicans and Trichophyton tests were administered ( . ml) at a 1:1000 weight/volume dilution because of clinical concern that 01 a 1:100 or higher concentration might induce significant skin reactions and cause morbidity in the active pilot population. Mumps was given at a dose of 2 complement-fixing units, and staph-phage-lysate was administered at a dose of 6-9 x 10 colony-forming units of Staph. aureus and 0.5 - 5 x 10 bacteriaphage plaque-forming units. �Three experienced technicians from the SCRF Allergy Division measured the size of both induration and skin erythema by the "pen method" at 24 and 48 hours after administration. Each reader was required to measure the skin reactions by a millimeter ruler and record length and breadth measurements at each of the four sites, refer exaggerated reactions to an allergist, collect medication use data, and sign the data form. The skin test data were interpreted by defined criteria, as given in Table 19-12. Other categories included: impairment noted, clinical correlation required; normal (versus abnormal) results; with medications noted; and refusal. Of the 1,759 participants with skin tests, 269 were excluded from the analyses for the following reasons; 205 due to missing reader signature or failure of the participant to report for the 48-hour reading; 58 because of immunosuppressive medication, cancer chemotherapy, or x-ray therapy; 3 for impaired hypersensitivity requiring more tests; and 3 due to refusal. Readings at 24 hours were not analyzed since these readings occurred prior to peak response to the antigens. Statistical Analyses and Interpretations The initial analytical intent was to test Ranch Hand-Comparison group differences in skin test response by standard models, using both discrete and continuous data. In the preanalysis of the continuously distributed data (length by width measurement of the skin reactions), there was a suggestion of profound reader variation. This observation generated a series of contrasts between the readers prior to group testing. Figures 19-1 through 19-6 show contrasts of the three skin test readers from the tests of mumps and Trichophyton. Each graph shows the individual plots of the 48-hour induration square area measurement versus the 48-hour TABLE 19-12. Clinical Interpretation Categories of Skin Test Results by Specific Measurement Criteria at SCRF Clinical Interpretation Category Measurement Criteria Normal (Delayed Cutaneous Hypersensitivity Intact) Length (L) induration @ 48 Hrs >10 mm Width (W) induration @ 48 Hrs >~10 mm on any one of four skin tests Probably Normal (Probably Intact Delayed Cutaneous Hypersensitivity) L induration @ 48 Hrs > 5-<10 mm W induration @ 48 Hrs > 5-<10 mm on any one of four skin tests Possibly Anergic L, W induration o_r erythema @ 48 Hrs >0-<5 mm on any one or more of four skin tests Anergic L and W, induration at 48 hrs = 0 on all skin tests. 19-34 �10 - v£> I U> Ul 4 5 Log (Area of 48-Hour ERYTH Meas + 1) Figure 19-1. Relationship of Induration Measurements to Erythema Measurements for the Mumps Skin Test Reader 1 Results �10 - 4 5 6 Log (Area of 48-Hour ERYTH Meas + 1) Figure 19-2. Relationship of Induration Measurements to Erythema Measurements for the Trichophyton Skin Test. Reader 1 Results i 9 1 10 �104 Log (Area of 48-Hour ERYTH Meas + 1) Figure 19-3. Relationship of Induration Measurements to Erythema Measurements for the Mumps Skin Test Reader 2 Results �0 10 �Log (Area of 48-Hour INDUR Meas + 1) i 3 4 5 6 Log (Area of 48-Hour ERYTH Meas + 1) Figure 19-5. Relationship of Induration Measurements to Erythema Measurements for the Mumps Skin Test Reader 3 Results I 10 �' o (Area of 48-Hour INDUR Meas + 1) I 0 I 2. I I I I 3i 4 5 6 Log (Area of 48-Hour ERYTH Meas + 1) Figure 19-6. Relationship of Induration Measurements to Erythema Measurements for the Trichophyton Skin Test Reader 3 Results i 8 10 �square area erythema measurement by specific skin test and reader. These measurements are presented in log units to centralize the outlying values. These analyses were done because the size of induration rarely exceeds the size of the erythema reaction. Thus, each of the depicted graphs shows a line of values with the sizes of erythema equal to the size of induration; this line, and all values on, or to the lower right of the line, are labeled "clinically acceptable" values. All values above and to the left of the line, deemed "clinically unacceptable," are probably due to hurried measurements by inspection (rather than the pen method) or recording errors. These figures demonstrated a marked difference in the occurrence of clinically unacceptable results between readers for comparable tests. Specifically, Reader 2's measurements revealed a higher proportion of clinically unacceptable results than those observed with Readers 1 and 3. Further, the graphs supported some variation in the clinically acceptable measurement values between Reader 1 and Reader 3. Because of these discordances, further analyses of the continuously distributed data were abandoned in favor of discretized analyses. Categorical analyses were conducted on two parameters of the skin testing results, the area measurement relationship of induration to erythema, and the clinical interpretation of the skin test readings. Each of the three readers was compared for 48-hour measurements on the same skin test, categorizing the induration-erythema relationship as (1) Induration (I) equals Erythema (E) (both values equal to zero), (2) E greater than I, (3) I equals E, and (4) I greater than E. As previously noted, only the category of I greater than E was judged clinically unacceptable. An analysis of these four categories, by reader, for each of the four skin tests, showed a profound statistical difference (p<0.001) between the readers for all four skin tests. An average of the percentages for each category by reader is shown in Table 19-13, exemplifying the marked differences (a p-value is inappropriate due to the averaging). Induration Erythema Relationships in Average Percentage Over Four Skin Tests, by Reader In Percent Reader *I:Induration E:Erythema I=E=Zero* E>I I=E I>E �These data show marked reader differences for the category I greater than E. The magnitude of clinically unacceptable results ( 0 0 on the 3.% average for four skin tests) for Reader 2 (visually shown in Figures 19-3 and 19-4) strongly suggested that this entire data set was invalid. Further, the data pattern from Reader 2 was shown to be uniform over time, confirming the existence of a consistent bias. In this light, the existence and magnitude of a reverse error for Reader 2, i.e., misreadings of I equals E equals 0, E greater than I, and I equals E, seem plausible, but unestimable. Of these three categories, I equals E equals 0 is the most clinically important (suggesting anergy), and Table 19-13 provides clear evidence of a negative bias, with Readers 1 and 3 showing over three times more average detection of anergy than Reader 2. Also of interest in Table 19-13 are the substantial differences in the categories E greater than I and I equals E for Readers 1 and 3. Analyses of the four skin tests by erythema-induration relationships showed statistically significant differences beween Readers 1 and 3 for all four tests (p<0.001 for mumps, Candida albicans, and Trichophyton, and p=0.036 for staph-phage-lysate). The decision to remove Reader 2 data from subsequent analysis was agreed to by all the Principal Investigators, recognizing the minimal role of erythema as a contemporary indicator of anergy. This decision was based on the concern that an error in erythema measurement likely indicated an error in measurement of induration (the predominant indicator of anergy). In preparation for the analysis of group data remaining from Readers 1 and 3, it was noted that the clinical interpretations (see Table 19-12) from these valid readings were inconsistent over time of the study. Specifically, 80 percent of relative anergy and anergy occurred in the first 10 of 81 groups of participants (or 2 1/2 months of the 9-month examination period). Further, the proportion of diagnoses of anergy between the allergists was disproportionate. The value of these analyses was therefore reduced. SUMMARY AND CONCLUSIONS Immunologic competence was measured by cell surface marker (phenotypic) studies and cell stimulation studies on 47 percent of the study population, and by a four antigen series of skin tests in 76 percent of participants to assess the delayed hypersensitivity response. Table 19-14 summarizes the results of all unadjusted and adjusted analyses on 11 primary variables spanning the first two of these three functional areas. Cell surface marker studies were conducted for total T cells (T-1), helper T cells (T4), suppressor T cells (Tg), B cells, monocytes, and HLA-DR cells; the ratio of T /T8 cells was included in the analysis. Because of inherent significant day-to-day and batch-to-batch variation, all results (including functional stimulation studies) were adjusted for blood-draw day variation. Statistical testing of the seven phenotypic cell markers did not reveal any significant group differences (interactions excepted), either unadjusted or adjusted for the covariates of age, race, occupation, current smoking, lifetime smoking history (pack-years), current alcohol use, or lifetime alcohol use (drink-years). Similarly, none of the unadjusted or adjusted analyses of the functional stimulation studies (for phytohemagglutinin, pokeweed mitogen, or mixed lymphocyte culture) showed any 19-42 �Overall Summary Results of Unadjusted and Adjusted Analyses of Immunological Variables Variable Unadjusted Adjusted Total T Cells (T ) Helper T Cells (T4) Suppressor T Cells (T ) B Cells Monocytes HLA-DR Cells T4/T8 Ratio Unstiraulated Response (PHA) PHA Net Response Pokeweed Net Response MLC Net Response NS:Not significant ( > . 0 . p01) ****Significant group-by-covariate interaction. statistically significant group differences. However, the adjusted analyses for total T cells, B cells, raonocytes, HLA-DR cells, pokeweed mitogen, and net mixed lymphocyte culture stimulation showed some significant 'group-bycovariate interactions, precluding direct adjusted group contrasts. Overall, no discernible pattern was identified to suggest a detriment in any subgroup of either the Ranch Hands or Comparisons. Results were similar between the analyses of the total Comparison group and the analyses of the Original Comparisons. The covariate effects of age, race, smoking, and alcohol use were generally profound on most variables in the phenotypic and stimulation studies. Consistently decreasing values of all cell markers and stimulated cells were associated with increasing age, whereas increased levels of smoking were usually associated with increases in the values of those variables. Blacks had consistently higher stimulated cell counts than nonblacks, but this effect was not observed for counts of T cells, B cells, or HLA-DR cells. Enlisted personnel generally had higher cell surface marker counts than officers. Exposure index analyses of cell surface markers revealed no pattern consistent with a dose-response relationship. For enlisted groundcrew, the mean total T cell and suppressor T cell counts for the medium exposure level were significantly lower than those of the low exposure level, but were slightly lower than those of the.high exposure level. The exposure index analyses of the functional stimulation tests revealed no consistent significant dose-response patterns for net PHA counts or net MLC counts. For net pokeweed counts, enlisted flyers in the high exposure level had a significantly lower adjusted count than enlisted flyers in the low exposure level, and a decreasing trend was apparent. �The delayed hypersensitivity response was assessed by the skin test antigens of mumps, Candida albicans, Trichophyton, and staph-phage-lysate. The 48-hour measurements of skin induration and erythema for the four tests showed marked inter-reader variation. Analyses showed that one of the three skin test readers too often measured induration larger than erythema (a clinically unacceptable finding), in an average of 30 percent of the readings, and did not yield measurements that detected a case of possible or overt anergy, whereas the other two readers found this condition in 5.6 percent of the participants. Remaining data from Readers 1 and 3, however, were found to vary significantly in clinical interpretation over duration of the examination. Consequently, all skin test data were declared invalid, and were not used in the assessment of group differences. The skin test reading problems led to the use of additional clinical quality control procedures for the AFHS followup examination begun in May 1987. In conclusion, no significant group differences were judged present for the comprehensive cell surface marker or functional stimulation studies. The profound effects of age, smoking, and alcohol use were observed in these immunologic tests. The assessment of delayed hypersensitivity skin responses was precluded by poor data quality and excluded from further analysis. Overall, there was no indication of impaired immunologic competence in either group. 19-44 �CHAPTER 19 REFERENCES 1. Vos, J.G., J.A. Moore, and J.G. Zinkl. 1973. Effect of 2,3,7,8tetrachlorodibenzo-p-dioxin on the immune system of laboratory animals. Environ. Health Perspec. 5:149-162. 2. Zinkl, J.G., J.G. Vos, J.A. Moore, and B.N. Gupta. 1973. Hematologic and clinical chemistry effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin in laboratory animals. Environ. Health Perspec. 5;111-118. 3. Vos, J.G., and J.A. Moore. 1974. Supression of cellular immunity in rats and mice by maternal treatment with 2,3,7,8-tetrachlorodibenzop-dioxin. Int. Arch. Allerg. Appl. Immunol. 47:777-794. 4. Thigpen, J.E., R.E. Faith, K.E. McConnell, and J.A. Moore. 1975. Increased susceptibility to bacterial infection as a sequela of exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Infect, and Immun. 12(6):1319-1324. 5. Faith, R.E., and J.A. Moore. 1977. Impairment of thymus-dependent immune functions by exposure of the developing immune system to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). J. Toxicol. Environ. Health 3:451-465. 6. McNulty, W.P. 1977. Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin for Rhesus monkeys: Brief report. Bull. Environ. Contam. Toxicol. 1()1819 81:0-0. 7. Faith, R.E., M.I. Luster, and J.A. Moore. 1978. Chemical separation of helper cell functions and delayed hypersensitivity responses. Cellular Immunol. 40:275-284. 8. Vos, J.G., J.G. Kreeftenberg, H.V.B. Engel, A. Minderhoud, and L.M. Van Noorle Jansen. 1978. Studies on 2,3,7,8-tetrachlorodibenzo-p-dioxin induced immune suppression and decreased resistance to infection: Endotoxin hypersensitivity, serum zinc concentrations and effect of thymosin treatment. Toxicology 9:75-86. 9. McConnell, E.E., J.A. Moore, and D.W. Dalgard. 1978. Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Rhesus monkeys (Macaca mulatta) following a single oral dose. Toxicol. Appl. Pharmacol. 43(10):175-187. 10. Sharma, R.P., and P.J. Gehring. 1979. Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on splenic lymphocyte transformation in mice after single and repeated exposures. Ann. N.Y. Acad. Sci. 320:487-497. ~~ 11. Faith, R.E., and M.I. Luster. 1979. Investigations on the effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on parameters of various immune functions. Ann. N.Y. Acad. Sci. 320:564-571. 19-45 �CHAPTER 19 REFERENCES (Continued) 12. Dean, J.H., M.I. Luster, G.A. Boorman, K. Chae, L.D. Lauer, R.W. Luebke, L.D. Lawson, and R.E. Wilson. 1981. Assessment of immunotoxicity induced by the environmental chemicals 2,3,7,8-tetrachlorodibenzop-dioxin, diethylstilbestrol and benzo(a)pyrene. In Advances in Immunopharinacology, ed. J. Hadden, L. Chedid, P. Mullen, and F. Spreafico, pp. 37-50. New York: Pergamon Press. 13. Clark, D.A., J. Gauldie, M.R. Szevczuk, and G. Sweeney. 1981. Enhanced suppressor cell activity as a mechanism of immunosuppression by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Proc. Soc. Exp. Biol. Med. 168:290-299. 14. Poland, A. 1984. Reflections on the mechanism of action of halogenated aromatic hydrocarbons. In Banbury report 18: Biological mechanisms of dioxin action, ed. A. Poland and R.D. Kimbrough, pp. 109-117. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory. 15. Knutsen, A.P. 1984. Immunologic effects of TCDD exposure in humans. Bull. Environ. Contam. Toxicol. 33:673-681. 16. May, G. 1982. Tetrachlorodibenzodioxin: A survey of subjects ten years after exposure. Br. J. Ind. Med. 39:128-135. 17. Hay, A. 18. Sirchia, G.G. 1982. Exposure to TCDD: Immunologic effects. In Plans for clinical and epidemiologic follovup after area-wide chemical contamination; proceedings of an international workshop, Washington, D.C., March 1980. Washington,. D.C.: National Academy Press. 19. Hoffman, R.E., P.A. Stehr-Green, K.B. Webb, G. Evans, A.P. Knutsen, W.F. Schramm, J.L. Staake, B.B. Gibson, and K.K. Steinberg. 1986. Health effects of long-term exposure to 2,3,7,8-tetrachlorodibenzop-dioxin. JAMA 255:2031-2038. 20. Grubbs, F.E. 1969. Procedures for detecting outlying observations in samples. Technometrics XI:1-21. 1981. Dioxin hazards: Secrecy at Coalite. 19-46 Nature 290:729. �CHAPTER 20 PULMONARY DISEASE INTRODUCTION Pulmonary dysfunction and overt pulmonary disease are not recognized clinical entities resulting from exposure to chlorophenols or TCDD. Acute exposure to chlorophenols, phenoxy herbicides, and TCDD, have caused the traditional acute symptoms of cough, nasal/lung irritation, shortness of breath, and, occasionally, bronchitis. These symptoms have been noted almost exclusively in industrial workers and not in individuals experiencing casual contact. Long-term sequelae arising from the acute symptom stage in ill individuals have not been generally known because of minimal followup and surveillance of the pulmonary symptoms. Only one contemporary morbidity study has attributed pulmonary dysfunction to phenoxy herbicide and TCDD exposure. The percent abnormal pulmonary parameters of forced expiratory volume (FEV), forced vital capacity (FVC), forced expiratory volume in one second (FEV^/FVC ratio, and forced midexpiratory flow rate (FEF25_75) were significantly higher in exposed workers who currently smoke, than in nonexposed workers who smoke. In considerable contrast, these test parameters were essentially equal in nonsmokers and former smokers of both the exposed and nonexposed groups. The effect of current smoking persisted after a logistic regression analysis adjusting for pack-years of cigarette smoking. Adjusted means of the test parameters FEV, FVC, and FEV1/FVC also showed significant differences for current smokers but not for nonsmokers or former smokers. As with other nonclassical clinical endpoints, prior investigators perhaps undervalued the incorporation of pulmonary disease and function into their study protocols. Further, due to the profound effect of smoking on pulmonary function, great emphasis must be placed in the collection of highly accurate, detailed, and validated smoking data as an adjustment variable, a process that is not straightforward in today's environment of antismoking. The only recent data comparable to this study are found in the 1984 AFHS Baseline Morbidity Report, which is reviewed below. Baseline Summary Results The 1982 Baseline examination explored historical pulmonary disease by questionnaire and active pulmonary function by standardized spirometric technique at the physical examination. These areas were of significant interest because of routine operational inhalation of Herbicide Orange by all Ranch Hand flying crewmen as well as ground maintenance personnel (Baseline Report Chapter 1, Buckingham). 20-1 �The questionnaire revealed no group differences for historical diagnoses of tuberculosis and fungal infections, pneumonia, cancer, or chronic sinusitis and upper respiratory disease. At the physical examination the unadjusted means for FEV1 (percent predicted), FVC, and the FEV /FVC ratio were almost identical between the Ranch Hands and Comparisons. Adjusted mean values were not calculated due to significant interactions (age, group, and pulmonary function for FEV1 and FVC; smoking with FEV1/FVC). Detailed exposure analyses showed two significant associations in the enlisted flyer and enlisted groundcrew strata, but neither was indicative of linear dose response. Attempts to adjust the means of the pulmonary function values for age and smoking revealed several interactions, but essentially negative results. Overall, there were no pulmonary disease or pulmonary function data or associations of concern. Parameters of the 1985 Pulmonary Examination Because of the essentially negative pulmonary analyses from the Baseline examination, pulmonary function (spirometric) studies were not performed during the first followup examination. Collection of pulmonary data was limited to a questionnaire history of respiratory disease, physical examination of the thorax and lungs, and pulmonary abnormalities detected on a routine chest x ray. Thus, t,he data analyses consist of group assessments of respiratory disease incidence, physical examination abnormalities, and the current prevalence of x-ray abnormalities. Covariate adjustments are made for age and smoking (yes, no, former, and pack-years). Minor numeric differences in the tables are due to rare missing dependent variable or covariable data. The analyses are based on 1,016 Ranch Hands and 1,293 Comparisons. No exclusions based on clinical conditions were made. Mortality due to respiratory disease, as of 31 December 1985, in the Ranch Hand and the 1:5 matched Comparison cohort is summarized. Morbidity data are analyzed using linear and loglinear models. RESULTS AND DISCUSSION Mortality Experience The mortality of the Ranch Hand and Comparison groups through 31 December 1985 was evaluated. There were seven deaths from respiratory system conditions in the Comparison group and none in the Ranch Hand group. This analysis was based on the 1:5 Ranch Hand to Comparison mortality study cohorts. Two of these deaths were Comparison flying officers, three were enlisted flyers, and the remaining two were enlisted groundcrew. Unadjusted Morbidity Analyses Analyses were performed on the history of respiratory illnesses as provided by the participants during the physical examination. The results of the 20-2 �radiological and clinical examination of the lungs and chest were also analyzed. These unadjusted analyses are summarized in Tables 20-1 and 20-2. As shown, no significant group differences were observed for history of asthma, bronchitis, pleurisy, pneumonia, or tuberculosis. Similar nonsignificant results were found in the evaluation of the clinical variables. Parallel analyses were conducted using data from the Original Comparisons, with comparable results (Appendix R, Table R-l). Adjusted Morbidity Analyses Statistical adjustment for the effects of age and lifetime smoking did not alter the findings of group similarity seen in the unadjusted analyses. Lifetime smoking was categorized as nonsmoking (0 pack-years), moderate (greater than 0 to 10 pack-years) and heavy (greater than 10 pack-years). These results are shown in Table 20-3. Lifetime smoking consistently exerts significant effects on nearly all historical illness and clinical examination variables, and age was an important factor for the history of pneumonia and the clinical assessment of thorax and lungs (representing an overall clinical assessment of normality/ abnormality in the respiratory system), chest asymmetry, the presence of hyperresonance, rales, and the presence of x-ray abnormality. There were significant or borderline significant group-by-pack-year interactions in analyses of a history of pleurisy and tuberculosis, for the presence of rales on examination, and for x-ray abnormality. There was also an interaction for asthma of borderline significance (p=0.068). A significant group-by-age interaction was seen for the presence of rales. The results of analyses stratified to clarify these interactions are shown in Table 20-4. Nonsmoking Ranch Hands had significantly more asthma (p=0.050) than their nonsmoking Comparisons, while the history of asthma was not significantly different in either category of smokers. Pleurisy was significantly more frequent in moderately smoking Ranch Hands (p=0.0001), but bordered on being significantly-increased in heavily smoking Comparisons (p=0.060). Analyses of a history of tuberculosis and the presence of rales was hampered by small numbers of cases in both groups (a total of 13 cases). The presence of several cells containing zeros makes interpretation of these analyses extremely difficult. Except in those strata with zero cells, no statistical significance was noted. In the analysis of x-ray abnormalities, the nonsmoking Ranch Hands had significantly less abnormality (p=0.030) than the nonsmoking Comparisons. Analyses of other strata did not reveal any significant group differences. These adjusted analyses were performed on data from the Original Comparisons, with similar results (see Tables 20-2 and 20-3). EXPOSURE ANALYSES The pulmonary data from the Ranch Hands were analyzed using the exposure index as a covariate (categorized as high, medium, or low within each occupational stratum). The percent abnormality at each level of exposure for each clinical or historical variable is presented in Tables 20-5, 20-6, and 20-7. 20-3 �Unadjusted Analyses of Reported History of Respiratory Illness by Group Group Ranch Hand Variable Asthma Bronchitis Pleurisy Pneumonia Tuberculosis Statistic Comparison Number Percent Number Percent Est. Relative Risk ( 5 C.I.) 9% p-Value �Unadjusted Analyses of Radiological and Clinical Respiratory System Findings by Group Group Ranch Hand Variable Thorax and Lungs Asymmetrical Expiration Hyperresonance Dullness Wheezes Rales X Ray Statistic Comparison Number Percent Number Percent Est. Relative Risk (95% C.I.) p-Value �Adjusted Analyses of Respiratory Variables by Group* Group Variable Ranch Hand Total Comparison Total Adj. Relative Risk. (95% C.I.) p-Value Covariate Remarks** As thma PACKYR (p=0.023) GRP*PACKYR (Borderline: p=0.068) Bronchitis None Pleurisy GRP*PACKYR (p^O.0026) Pneumonia AGE (p»0.0001) Tuberculosis GRP*PACKYR (p=0.034) Thorax and Lungs AGE (p<0.0001) PACKYR (p<0.001) Asymmetrical Expiration AGE*PACKYR (p=0.036) Hyperresonance AGE (p<0.0001) PACKYR (p<0.0001) Dullness None Wheezes PACKYR (p<0.0001) Rales GRP*AGE (p=0.046) GRP*PACKYR (Borderline: p=0.070) AGE*PACKYR (Borderline: p=0.090) AGE (p<0.0001) PACKYR (p»0.0019) GRP*PACKYR (Borderline: p=0.060) *Group-by-covariate interactions are described in Table 20-4. **Abbreviations PACKYR: Lifetime smoking history (pack-years) GRP: Group ****Group-by-covariate interaction—relative risk, confidence interval, and p-value not presented. �Suaaary of Group-by-Covariate Interactions for Respiratory Variables Group Ranch Hand Variable Interaction Asthma Group-byPack-Year Pleurisy Group-byPack-Year Stratification Statistic Number Percent Comparison Number Percent Adj. Relative Risk (95% C.I.) p-Value �Summary of Group-by-Covariate Interactions for Respiratory Variables Group Ranch Hand Variable Interaction Tuberculosis Group-byPack-Year Rales Group-byAge Stratification Statistic Number Percent Comparison Number Percent Adj. Relative Risk (95% C.I.) p-Value �Siuuary of Group-by-Covariate Interactions for Respiratory Variables Group Ranch Hand Variable Interaction Rales Group-byPack-Year X Ray Group-byPack-Year Stratification Statistic Number Percent Comparison Number Percent Adj. Relative Risk (95% C.I.) p-Value �TABLE 20-5. Exposure Index Analysis Results for Officers p-Values of Dependent Variable-by-Covariate Association*'b Variable D*EXP D*AGE D*PACKffi D*EXP *AGE D*EXP D*AGE D*EXP* *PAOOR *PACm AGE*PAOQR Abnormal Total Percent Asthma Bronchitis Pleurisy Pneumonia Tuberculosis Thorax and Lungs Asynrnetrical Exp. Hyperresonance Dullness Wheezes Pales XRay 'Dependent variable indicated by 0 in coluon headings. b Abbreviations: EXP: Exposure index. PACKYR: Pack-years. TABLE 20-6. Exposure Index Analysis Results for Enlisted Flyers p-Values of Dependent Variable-by-Covariate Association* Variable Asthma Bronchitis Pleurisy Pneumonia Tuberculosis Thorax and Lungs Asymmetrical Exp. Hyperresonance Dullness Wheezes Bales XRay D*EXP D*AGE D*PAOOR D*EXP *AGE D*EXP D*AGE D*EXP* *PAOOR *PAOCZR AGE*PAOOR Abnornal Total Percent �TABLE 20-7. Exposure Index Analysis Results for Bilisted Gcoundcrew: p-Values of Dependent Variable by Covariate Association* Overall Variable D*EXP D*AGE D*PACm D*EXP* *AG£ D*EXP D*AGE D*EXP* *PACKXR *PAOOR AG£*PACm Abnormal Total Percent Asthma Bronchitis Pleurisy Pneumonia Tuberculosis Thorax and Lungs Asynmetrical Exp. Hyperresonance Dullness Vheezes Rales XRay Two sets of analyses were performed on enlisted groundcrew data. In the first set of analyses, all three year-of-birth categories (born after 1942, born between 1922 and 1942, born before 1922) were used. In the second set of analyses, only those born between 1922 and 1942 and after 1942 were used, since only one enlisted groundcrew Ranch Hand was born before 1922. All testing results in the two sets of analyses were the same, except for the asthma-by-age interaction shown in Table 20-6. Each of the dependent variable-by-exposure category interactions are noted by occupation category in Appendix R, Tables R-4 through R-18. These data are considered too sparse for meaningful interpretation. SUMMARY AND CONCLUSIONS A summary of the results on the analyses of reported history of respiratory illness and of radiological and clinical findings is given in Table 20-8. Based on the 31 December 1986 mortality data, there were seven deaths from respiratory conditions in the Comparison group and none in the Ranch Hand group. �TABLE 2 - . 08 Overall Summary Results of Unadjusted and Adjusted Analyses of Pulmonary Disease Pulmonary Disease Unadjusted Adjusted Reported History of Respiratory Illness Asthma Bronchitis Pleurisy Pneumonia Tuberculosis Radiological and Clinical Findings Thorax and Lungs Asymmetrical Expiration Hyperresonance Dullness Wheezes Rales X Ray NS: Not significant (p>0.10) ****Group-by-covariate interaction. There were no group differences found for reported history of asthma, bronchitis, pleurisy, or tuberculosis based on the unadjusted analyses. Adjustments for age and lifetime smoking did not alter the findings of group similarity, although there was a significant group-by-pack-year interaction for pleurisy and for tuberculosis. Similarly, there were no significant group differences in the unadjusted analyses for the radiological and clinical respiratory findings of thorax and lungs, asymmetrical expiration, hyperresonance, dullness, wheezes, rales, and x-ray interpretations. These findings were supported by the adjusted analyses, although there was a group-by-age interaction for rales. The exposure index analyses revealed no consistent dose-response pattern. Analyses of past history of respiratory illness and the clinical and radiological examination of the chest and lungs did not reveal any statistically significant differences between the Ranch Hand and Comparison groups suggestive of herbicide related disease. Several group-by-covariate interactions did exhibit statistical significance, but these findings did not indicate any consistent patterns suggesting different disease experience in the two groups. �REFERENCES CHAPTER 20 1. Suskind, R.R., and V.H. Hertzberg. 1984. Human health effects of 2,4,5-T and its toxic contaminants. JAMA 251:2372-2380. 2. Lathrop, G.D., P.M. Moynahan, R.A. Albanese, and W.H. Wolfe. Epidemiologic Investigation of Health Effects in Air Force Following Exposure to Herbicides—Baseline Mortality Study Epidemiology Division, Data Sciences Division, USAF School Aerospace Mecicine, Brooks Air Force Base, Texas. 20-13 1983. An Personnel Results. of �CHAPTER 21 INTERPRETIVE CONSIDERATIONS This chapter reviews several scientific issues that should be considered when attempting to reach conclusions on a study of this size and complexity. These issues are critical to the interpretation of the data analyses in this report. Data patterns observed in many clinical chapters of this report are also summarized so that hypothesis testing of group differences may be placed in better perspective. DIOXIN ENDPOINTS Based upon data in this report, final conclusions on herbicide causality must consider results of the various clinical areas, reflected in the separate chapters. Each chapter introduction has attempted to highlight the major organ systems that are known or suspected to be significantly affected by the ingredients of Agent Orange with particular emphasis on the effects of dioxin. Categories of clinical endpoints and their generally accepted degree of association with dioxin are presented in Table 21-1. These associations are based on the scientific literature. TABLE 21-1. Summary Associations of Adverse Health Effects to TCDD Exposure Reported in the Literature Degree of Association by Clinical Chapter Confirmed Highly Suspected Moderately Suspected Dermatology Neurology Hepatic Malignancy General Health Immunology Negative or Weakly Suspected Psychology Cardiovascular Hematology Endocrine Renal Pulmonary It is recognized that alternative conclusions based on these patterns of association are possible within the framework of current knowledge, particularly for the highly and moderately suspected areas (malignancy, general health, immunology). However, for illustrative purposes, two extremes are presented: multiple adverse findings in the Ranch Hand group for the areas 21-1 �of dermatology, neurology, hepatic (discussed in Chapter 13), and cancer would suggest a case for TCDD causality, whereas multiple adverse findings in the weakly suspected areas, and not in any of the confirmed areas, would be difficult to ascribe to an overall TCDD causation. The aspects of biological plausibility and specificity require balanced interpretation across clinical chapters, with careful attention placed on nonsignificant findings as well as significant findings. The chapters in this report should be viewed as artificial boundaries for convenience of presentation, and should not discourage consideration of their relatedness, or of the individual variables within them. EXPOSURE Approximately 600 exposure index analyses have been conducted in this study, underscoring attempts to associate increasing proportions of various abnormalities to estimates of increasing exposure. To determine whether the results of the exposure analyses varied by chance, several perspectives were taken. Of the 255 adjusted exposure analyses (excluding 39 with interactions), 13 were statistically significant, a figure which is the expected number (based on a =0.05). It is recognized that this contrast is a crude yardstick, considering the relatedness of the dependent variables, statistical power, disproportionate representation of chapter variables, and the presence of interactions. The six possible patterns of exposure response (increasing, decreasing, V-shaped with fewer abnormalities at the low exposure level than the high exposure level, V-shaped with more abnormalities at the low exposure level than at the high exposure level, inverted V-shaped with fewer abnormalities at the low exposure level than the high exposure level, and inverted V-shaped with more abnormalities at the low exposure level than at the high exposure level) were tabulated (regardless of statistical significance) for the clinical chapters of dermatology, neurology, psychology, and renal. As noted in Table 21-1, two of these chapters contain clinical variables that have had confirmed associations to TCDD exposure, and two chapters have had negative or weakly suspected associations to TCDD. Of the 126 exposure analyses in these four chapters, 21 (or one-sixth) showed the primary pattern of interest, an increase—exactly the number expected. Taken together, these analyses suggest that statistically significant exposure analyses may have occurred due to chance among the data set, and that the pattern of dose-response may also have been random. These inferences, or that the exposure index was unrelated to actual exposure, together with the acknowledged limitations of the exposure index, indicate that estimated exposure may only be weakly relied upon to assert a causal relationship. Based upon the current exposure index calculations, either of the above inferential alternatives is possible. The use of serum dioxin levels (see Chapter 23, Future Directions) in the next report will clarify the exposure calculations of this report and the Baseline Report. Thus, from an interpretive context, final conclusions on dose-response, and the implications to herbicide causation are based on current knowledge available for this report. These conclusions could change with future analyses using a factual exposure concept. 21-2 �TYPES OF MEASUREMENTS This report includes all types of measures traditionally used in morbidity followup epidemiologic studies, e.g., self-reports, structured interview responses, medical record data, physician findings, scalar measurements, biopsy results, laboratory determinations, morbidity indices, and mortality results. At many points in this report, various terms have been used to qualitatively describe the data and analyses arising from the measurement processes. In particular, the terms "subjective," "objective," "continuous," and "categorical," and "constructed indices" have been used to connote differences in data or data sets that are important in making statements of inference. From the perspective of the Study Protocol, significant group differences for subjective historical variables, not mirrored by significant group differences in medical record findings or physician/laboratory testing, may be viewed as preliminary evidence of over-reporting by a group. The opposite finding of significant group differences for physical examination variables in the absence of reported symptoms may support the primary conclusion of significant subclinical group differences. Either of these alternatives may greatly affect an overall inference of herbicide causality. Hence, the descriptive phrases "subjective data" and "objective data" have not been used as value judgments of the worth of the data, but simply as inferential qualifiers. This report contains numerous comments on the differences in results between analyses of continuous versus categorical data from the same variable (exclusively laboratory data). Because the statistical power is stronger for detecting mean shifts than categorical differences, it was anticipated that very small mean shifts might be more easily discerned.than differences in proportions of abnormalities between the two groups. Both methods of examining the data reveal important aspects of the distribution. Inferentially, when both types of analyses were done, greater weight has been given to significant group differences when analyses of both data forms agree. Lesser weight was given to significant differences seen in only one analysis, and least weight to significant shifts in means if both group means were within normal range, and the mean difference was not supported by other statistical findings in related variables (e.g., hepatic test battery). Consistent patterns of findings within an organ system, or between related organ systems, is required to strongly suggest an inference of causality. Several summary indices were constructed in this report, e.g., dermatology index, cranial nerve function index, and anatomic categories of abnormal peripheral pulses, and are similar to some indices in the 1984 Baseline Report. They were formed by summing or grouping related abnormalities for the purposes of assessing increased numbers and/or showing group directionality of overall results. They should not be strongly considered in final inferences because they are artificially derived. BASELINE-FOLLOVUP EXAMINATION DIFFERENCES A common difficulty of followup studies is the inherent variation in measurement systems from one observation period to the next. To the maximum extent possible, the USAF has restricted clinical variation by requiring the use of identical laboratory equipment for most clinical chemistries, by the 21-3 �use of 50 samples from the Baseline serum bank to evaluate interexamination laboratory differences, and by the use of carefully prescribed written clinical procedures that allow little room for variation. Nonetheless, some interexamination variability must be expected, but in the presence of blindness to group membership, there is no reason to expect biases in the results with respect to either the Ranch Hand or Comparison groups. This report has cited classical longitudinal analyses to assess changes in variables between the examinations by group. Of 21 variables examined, 5 showed statistically significant group differences in the changes between examinations. Four of these significant results were attributed to actual changes over time, while the other (e.g., sedimentation rate) was believed due to a change in laboratory methodology. Other less refined longitudinal contrasts consisting of narrative discussions of Baseline results versus followup results have been presented in all chapters. Interpretive caution is required in assessing examination similarities or differences because of the slight changes in cohort composition between the examinations (see Chapter 2, Population), the use of slightly different statistical models and modeling strategy (see Chapter 7, Statistical Methods), and sometimes the use of the Original Comparison group. The relative contribution of these changes was not explored mathematically, but is believed to have played a minimal role in accounting for any large group shifts between examinations. In the context of comparing results between examinations, there has been a subtle but consistent observation that group differences have substantially narrowed over the 3-year period, either by decreased findings in the Ranch Hands, increased findings in the Comparisons, or a combination of both mechanisms. In general, several broad interpretations are possible: any bona fide herbicide effect decreases over time, that the convergence is largely attributable to unquantifiable factors, that both examinations have produced chance results, or that these observations have been affected by the slight shifts in cohort composition and modeling strategy. Several segments of this report have noted marked differences in the prevalence rates of abnormalities found at the Baseline and followup specialty examinations, e.g., the dermatology and neurology clinical assessments. The followup dermatological examination detected substantially more abnormalities than the Baseline examination, whereas far greater numbers of neurological abnormalities were noted at the Baseline examination than at the followup for some variables. These examination variances were affected by differences in "clinical sensitivity" between the examining teams, although clearly other factors (such as a true change in disease-abnormality status or slight cohort differences) contributed. The phrase "clinical sensitivity" refers to the inherent differences in clinical styles and interpretations of possible abnormalities that often prevail. Because of examiner blindness to exposure status, and because of the judgment that the interexamination variation was within the artful bounds of accepted medical practice, no bias was thought to have resulted from this inherent variation. STUDY BIASES Each reviewer of this report must reach a conclusion on whether the results of this study have been seriously flawed by the design, the operation 21-4 �of significant biases, or both. The Protocol authors believe that the comprehensive multifaceted design is the chief strength of this study, although it is recognized that each and every published phase of the study must invite renewed inspection of fundamental scientific aspects of the study design. It is believed that, with the exception of skin test readings, all data in this study were collected accurately and validly, and that blindness to group membership was well maintained throughout the collection process. This opinion is important from an inferential perspective in that both misclassification of data (tending to dilute true group differences) and bias in data (creating a false group effect) most likely did not occur appreciably in this study. Thus, it is believed that both the magnitude and direction of the group results found in this study reflect truth to the maximum degree possible, within the inherent boundaries of statistical models to account for all important adjusting variables. GROUP INTERACTIONS: PATTERN RECOGNITION Many of the adjusted analyses in this report have demonstrated significant group-by-covariate interactions, requiring stratified analyses to determine the nature of significant group differences. All significant twoand three-factor interactions have been included in the main text or in appendices. The analysis of followup data has found substantially more interactions than the analysis of Baseline data, due primarily to the larger number of covariates used in the followup analyses. Several related viewpoints have aided in the overall interpretation of group-by-covariate interaction in the report. In the presence of a significant interaction, a direct conclusion on main group effects cannot be made, and the focal point of interpretation resides with the covariate stratum containing the significant group effect (or a reversal in nonsignificant group effects across strata). Past this point, however, there appears to be little consensus in how to best place the interaction into inferential context. Further interpretations appear to be largely individualistic. No consistent pattern has emerged to support a finding of impairment in the Ranch Hands for any specific stratum of one or more covariates. In fact, of all the two- and three-factor interactions encountered, only one was thought to have possible biologic relevance. Other interactions may have such relevance, but the reason was not apparent. As with tests of group differences, significant interactions may occur by chance, but the method to calculate an expected number of group-by-covariate interactions, unfortunately, remains an open research question. Because of the possible diverse interpretations of" interactions, all significant two- and three-factor interactions involving group with statistically significant strata are presented in Table 21-2 for detailed inspection. No particular covariate or group pattern is noted, although the variables in psychology and gastrointestinal showed Ranch Hands at a relative detriment, while the interactions in the cardiovascular .chapter indicated detrimental findings in the Comparisons. Most variables without interactions in this report have shown remarkable concordance between unadjusted and adjusted results, both in terms of absolute value of relative risk and of statistical significance. 21-5 �TABLE 21-2. Summary of Significant Covariate Strata (or Covariate Level Difference) Found Within Significant Two- and Three-Factor Group-by-Covariate Interactions by Clinical Chapter and Dependent Variable (Group Direction and p-Value) Clinical Chapter Dependent Variable Covariate Stratum RH>C ORH p-Value General Health Self-Perception of Health Enlisted Groundcrew * 0.003 Malignancy Basal Cell Carcinoma (Verified Interval) Systemic Cancer (Verified plus Suspected, Interval) Basal Cell Carcinoma (Verified plus Suspected, Lifetime) Systemic Cancers (Verified, Lifetime) Systemic Cancer (Verified plus Suspected, Lifetime) Enlisted Flyer * 0.019 Enlisted Flyer * Neurology Pin Prick Impaired (Diabetic Class) Psychology Paranoia Schizophrenia Social Introversion Validity Total CMI Born Before 1942 High School Combat Index—Low Black High School * * * 0.027 0.033 002 .0 0.038 <0.001 Gastrointestinal SCOT Alkaline Phosphatase Direct Bilirubin Triglycerides (cont.) Triglycerides (disc.) Uroporphyrins 1-4 Drinks Exposed to Exposed to Born In or Officer BUN<14 * * * * * 0.010 <0.001 0.035 0.039 0.035 <0.001 to (-» I 0.038 Intermediate Skin Reaction to Sun Enlisted Flyer * Enlisted Flyer * per Day Ind. Chems. Ind. Chems. Before 1922 0.042 0.019 0.004 0.021 �Summary of Significant Covariate Strata (or Covariate Level Difference) Found ¥ithin Significant Two- and Three-Factor Group-by-Covariate Interactions by Clinical Chapter and Dependent Variable (Group Direction and p-Value) Clinical Chapter Dependent Variable Covariate Stratum RH>C* Dermatology Dermatology Index Pre-SEA Acne: Cardiovascular Systolic Blood Pressure ECG (Overall) ECG (Arrhythmia) Posterior Pulses (Manual) Leg Pulses (Manual) Peripheral Pulses (Manual) Black/53 Yrs Old 0 Pack-years 7 Pack-years/10% Body Fat Enlisted Flyer Officer/21% Body Fat Officer Hematology WBC Nonblack/30 Pack-years/ 35 Yrs Old Black/Officer/35 Yrs Old Black/EFL/35 Yrs Old Nonblack/30 Pack-years and 1 pack/day Black/30 Pack-years and 1 pack/day WBC WBC PLT PLT Renal BUN Urine Specific Gravity Endocrinology Urinary Protein Urinary WBC 1 vs. 0 Normal (Diabetic Class) Nonblack/Born In or After 1942 Black Nonblack/Enlisted Groundcrew Testosterone Testosterone Differential Cortisol <10% Body Fat 10-25% Body Fat Black/Born In or After 1942 ORE p-Value �TABLE 21-2. .(continued) Summary of Significant Covariate Strata (or Covariate Level Difference) Found Within Significant Two- and Three-Factor Group-by-Covariate Interactions by Clinical Chapter and Dependent Variable (Group Direction and p-Value) Clinical Chapter Dependent Variable Covariate Stratum Immunology Total T Cells B Cells Monocytes Black Nonblack/0 Pack-years Enlisted Groundcrew/ 4 Drinks/Day Pulmonary Pleurisy Tuberculosis X-ray RH>C* 1-10 Pack-years 1-10 Pack-years 0 Pack-years Total Interactions: 43 *Relative risk greater than one, or Ranch Hand mean greater than Comparison mean. ORH p-Value �CLASSICAL COVARIATES Many of the dependent variables in this report are known to be significantly affected by risk factors also measured in this study. The use of these covariates in the adjusted analyses has served to clarify Ranch HandComparison group differences in the presence of significant covariate group differences. Such adjustments, whether by a single covariate, multiple covariates, or covariate interactions, have given results on group differences generally quite similar to the unadjusted analyses both in terms of relative risk and statistical signficance. In fact, in only one instance in this report has an unadjusted result of pX).10 changed to a value of p<0.05 in the adjusted analysis. The covariates used in this study were not effect modifiers (which may be synergistic with exposure and also be equally distributed between groups). Consistent effects were observed for almost all of the classical covariates of age, race, occupation, education, alcohol, smoking, percent body fat, and glucose tolerance. In only a few instances were unexpected effects noted, e.g., personality type, wine consumption, and a few smoking and alcohol "inversions." The overall covariate effects observed in this study indeed reflect the mainstream of results found in well-conducted epidemiologic studies, and lend credence to the validity of the clinical endpoints and covariate values in this report. MULTIPLE COMPARISONS As noted in Chapter 7, Statistical Methods, the problem of multiple comparisons is complex and not easily adjudicated because of the total number of statistical tests, the number of tests performed on each dependent variable, and the biologic relatedness of many of the variables. A conscious effort has been made to expand inferential interest to borderline group associations (0.05<p<0.10) thereby increasing the probability of the acceptance of a false association. Each chapter summary has carefully flagged all borderline associations to provide expanded summary statements for possible inclusion in deriving final conclusions. Additional confidence in the final acceptance or rejection of an overall herbicide effect would be warranted if the majority of borderline associations were in the same consistent direction as the significant associations. Multiple analyses on the same variable have been conducted in this report. Continuous and categorical data have been subjected to both unadjusted and adjusted analyses, and multiple adjusted analyses were sometimes conducted with different covariates or slightly different covariate sets. The question arises as to which results best reflect the truth when different results are found. In general, the following approach has been followed: the statistical significance of both continuous and categorical analyses is convincing, while significance for only the continuous analysis must be viewed in terms of the biologic relevance of the mean shift detected. Overall,-the multiple comparison issue is due to repeated hypothesis testing for group, exposure, and interaction strata differences. The calculation of expected numbers of significant associations for these tests is difficult (if not impossible) because of the relatedness of the dependent variables, the relatedness of the covariates, and the often difficult analytic decisions that arise in a "step-down, best model" strategy. Thus, 21-9 �the final assessment of whether the frequency of significant associations does not meet, or exceeds expectation, must remain an interpretive judgment of each reader. CAUSALITY The AFHS is an inferential assessment of observed group differences. The inference of herbicide causality will be determined by a balanced judgment of the following factors; biological plausibility, consistency, specificity, coherence, time relationships, and strength of association. Except for aspects of association strength, most of these causality factors have been discussed in the preceding sections of this chapter. Nearly every statistically significant group difference in this report has only been of moderate to weak strength. Highly significant p-values (p<0.001) were not found for main group associations, but were observed for covariate tests. A few strata in the group interactions were highly significant. Most of .the statistically significant estimated relative risks were below the value of 2.0 (a traditional boundary of interest in epidemiology). The few relative risks above 2.0 generally had very wide confidence intervals due to low proportions of detected abnormalities. Weakly significant associations, in particular, are cause to reassess the element of chance and the possible presence of other causality factors before a final conclusion of cause and effect is determined. 21-10 �CHAPTER 22 CONCLUSIONS INTRODUCTION This chapter summarizes the conclusions drawn from the statistical analyses that have been conducted on the Air Force Health Study data base. The followup study, which began in 1985, was the logical extension of the 1982 Baseline study, building upon the strengths of the Baseline study and utilizing the data collected at both the Baseline and the followup. The high level of Government support and outstanding participation of the study subjects that characterized the Baseline study were maintained through this first followup. STUDY PERFORMANCE ASPECTS Of the living Baseline study participants, 99.2 percent were located and asked to participate in the followup. Participation in the followup physical examination and questionnaire was very high. Of the fully compliant Baseline participants, 971 of the 1,045 Ranch Hands (92.9%) and 1,139 of the 1,224 Comparisons (93,1%) participated in the followup. Thus, there was no group difference in compliance of the Baseline participants at the followup. Overall, the 2,309 participants in the followup (1,016 Ranch Hands and 1,293 Comparisons) represented a loss of 159 individuals and a gain of 199 since Baseline. One percent of the fully compliant Baseline population died between 1982 and the 1985 followup examination. The bias/compliance analyses suggested that there had been no change between Baseline and the followup in the way replacements volunteered for entry into the study, and that no additional bias had been introduced at the followup due to scheduling differences. Although replacements were not health-matched at Baseline as they were at the followup, they were similar to refusals with respect to reported health, medication use, and income level. The results supported the use of the total Comparison group in the main analyses presented in this report. POPULATION CHARACTERISTICS Overall, the Ranch Hands and Comparisons reported similar social and behavioral characteristics. No significant differences were found in age, educational background, religious preference, current military status, and income level. Significantly more Ranch Hands smoked cigarettes at the time of the followup examination than did Comparisons, but there was no significant difference between groups on past cigarette, cigar, and pipe use and on recent and past use of marijuana. A much higher percentage of participants 22-1 �reported past marijuana use at the followup than at Baseline. This difference was most likely due to a greater level of confidentiality afforded by the questionnaire technique. Risk taking behavior, assessed by questions on potentially dangerous recreational activities, revealed borderline significance. Slightly more Comparisons were scuba divers and more Ranch Hands raced motor vehicles. The difference in scuba diving was also significant at Baseline. Patterns of Results Both the chapter conclusions and the final conclusions of this report have been predicated upon concepts of consistency, specificity, coherence, strength, and plausibility as they apply to the interpretation of group differences. In particular, careful consideration has been given to a variety of data and patterns of results that have emerged from the clinical evaluations. Specifically, there were few differences in the proportions of abnormalities between groups; the positive associations have not aggregated in the clinical areas of prime dioxin concern, nor have they been of serious clinical importance; the unadjusted results have been remarkably concordant with the adjusted results, both in terms of relative risk and p value; the analyses using the Original Comparison set have largely mirrored the results found with the total Comparison group; many of the group differences noted at Baseline have disappeared at the followup examination, and only a few new associations have emerged; almost all of the covariates have acted as expected in the adjusted analyses; and the exposure index analyses and the group-by-covariate interactions have not demonstrated biological patterns of concern and appeared to be more likely due to chance than not. Due to the acknowledged limitations of the exposure index used in this report (and considering the potential use of dioxin body burden levels at the next followup), dose-response relationships have not been emphasized in reaching final conclusions. The overall pattern of these findings indicates that this followup study cannot be viewed as alarming from the traditional perspectives of clinical medicine or epidemiology. This study, in fact, demonstrates similarity in current health status between the Ranch Hand and Comparison groups. CLINICAL ASPECTS General Health The nonspecific assessment of general health showed relatively close similarity between the two groups. ' Ranch Hands rated their health as fair or poor more frequently, but this difference was found only in the enlisted groundcrew and not in the officers nor enlisted flyers. The perception of health in both groups had improved since Baseline. Physician-rated appearance of relative age was not found to be significantly different at the followup in contrast to the Baseline finding that a higher percent of Ranch Hands than Comparisons looked younger than their stated age. The categorical analysis of sedimentation rate showed that the Ranch Hands had more abnormalities than the Comparisons. These results were not supported by the continuous analysis of mean sedimentation rates and were opposite to the 22-2 �Baseline results, which showed that younger Comparisons had elevated sedimentation rates. The categorical analysis of percent body fat showed no significant differences between the two groups, which was consistent with Baseline. However, the continuous analysis found that the Ranch Hands had a significantly lower mean percent body fat using age, race, and occupation as covariates. The detailed exposure analyses revealed no consistent exposure effects, and this result was consistent with the Baseline analysis. No longitudinal difference was found on perception of health. A significant group difference was found over time for the longitudinal analysis of sedimentation rate due to the change in the findings between the two examinations, possibly related to a change in laboratory methodology. Malignancy Skin and systemic cancers, both suspected and verified by medical records, showed no significant group differences for the Baseline-followup interval (1982-1985). However, for all neoplasms combined (malignant, benign, and uncertain), a borderline significant excess in the Ranch Hand group was noted in an unadjusted analysis. The analyses of interval cancers revealed group interactions for verified and verified plus suspected basal cell carcinoma and verified plus suspected systemic cancers. Nonsignificant findings were observed for verified and verified plus suspected sun exposurerelated cancers. Verified systemic cancers did not differ significantly between groups. The analyses of lifetime cancer found significant results for verified basal cell carcinoma and verified sun exposure-related skin cancers. Group interactions were noted for systemic cancer categories and for verified plus suspected basal cell carcinoma. The higher rate of basal cell carcinoma in the Ranch Hands versus the Comparisons found at Baseline was nonsignificant for the followup interval, but due to the effect of the larger number of Baseline cases and the significant confounding of average residential latitude, the adjusted analysis of lifetime basal cell carcinoma emerged as statistically significant. There were several disparities in the distribution of testicular, colon, and smoking-related tumors in the groups. Further, one case of soft tissue sarcoma and one possible lymphoma (both in Ranch Hands) were diagnosed in the interval, balancing the two similar cases found in the Comparison group at Baseline. Considering that the systemic cancer curves are in their early stages for both groups, with perhaps insufficient latency, the cancer results of the followup examination should not be viewed as disturbing, but as cause for continued monitoring. Neurological Assessment None of the 27 neurological variables demonstrated a significant group difference, although several variables had relative risks which were greater than one. There was no group difference in reported neurological illnesses for the interval or for a lifetime history. Of the cranial nerve variables, speech and tongue position were marginally significant, with the Ranch Hands at a slight detriment. The analyses of peripheral nerve function showed no significant differences between the Ranch Hands and the Comparisons. In the analysis of central nervous system function, hand tremor was found to be of borderline significance, with the Ranch Hands faring slightly worse than the Comparisons. A borderline significant group interaction (Ranch Hand hand tremor by insecticide exposure) may have had biological and operational 22-3 �significance. Overall, substantially fewer neurological abnormalities were detected at the followup examination than at the Baseline examination. The exposure analyses showed only occasional statistically significant results, although no consistent pattern with increasing exposure was evident. In the longitudinal analysis of the Babinski reflex, a significant change over time was observed. This was due to a nonsignificant finding in the Ranch Hands at the followup, which differed from the significant adverse finding at Baseline. The covariates of age, alcohol history, and diabetes showed classical effects with many of the neurological measurements. Overall, the followup examination results were quite similar to the Baseline findings. Psychological Assessment The reported and verified data on lifetime psychological illnesses showed no significant differences between groups. Distributional tests of the 14 Minnesota Multiphasic Personality Inventory (MMPI) scales, stratified by occupation, revealed that only 2 of the 42 results approached significance. For the total Cornell Medical Index (CMI), separate distributional tests were conducted with stratification by age, race, occupation, education, and current drinking status; a significant difference was found for one statum of each of the covariates. In all cases, the mean of the Ranch Hand distribution was greater than the mean of the Comparisons. The analysis of the 14 MMPI scales showed that there was a significant difference between the two groups for denial and masculinity/femininity, with more abnormalities in the Comparisons than the Ranch Hands. The results of the analyses for hysteria were of borderline significance, with more abnormalities in the Ranch Hands. There were more abnormalities in the Ranch Hands than the Comparisons for social introversion, which was of borderline significance. Differences in the total CMI and A-H area subscore were found to be significant, with more abnormalities in the Ranch Hands. There was no significant difference between the two groups on the Halstead-Reitan Battery impairment index, a measure of the functional integrity of the CNS. The exposure index analyses did not reveal any pattern consistent with a dose-response relationship. As expected, the effects of age, educational level, and alcoholic history showed profound effects on many of the psychological measurements. Gastrointestinal Assessment Although the followup gastrointestinal assessment disclosed more statistically significant findings than the Baseline examination, the abnormalities were distributed equally between the two groups, and there was no clinical, statistical, or exposure pattern consistent with an herbicide-related effect on health. No historical or biochemical evidence was found to suggest an increased likelihood of porphyria cutanea tarda (PCT) in the Ranch Hand group. Only sparse and nonsignificant liver disorders were reported for the interval between Baseline and followup. Also, for the lifetime history of liver disorders, there were no significant differences between groups. Further, there were no significant group differences in reported lifetime peptic ulcer disease. A review of digestive system mortality showed a relative excess in the Ranch Hands but a relative lack of malignant neoplasms. The results of the physical examination showed a borderline increase of hepatomegaly in the Ranch Hand group. There was a significantly lower mean serum glutamic-pyruvic transminase (SGPT) level, a greater mean alkaline phosphatase level, and a lower mean uroporphyrin level in the Ranch 22-4 �Hand group. The analysis of coproporphyrin was of borderline significance, with the mean of the Ranch Hands in excess of the mean of the Comparisons. No group differences were found for serum glutamic-oxaloacetic transminase (SCOT), gamma-glutamyl transpeptidase (GGTP), total and direct bilirubin, lactic dehydrogenase (LDH), cholesterol, or triglycerides. The numerous group-by-covariate interactions did not disclose any consistent subgroup patterns detrimental to the Ranch Hands. These findings were generally consistent with the results of the 1982 assessment. The longitudinal analyses for SCOT, SGPT, and GGTP showed no significant differences between results by group over time. Dermatological Evaluation No significant group differences were identified in the dermatological evaluation. None of the questionnaire data showed an increased likelihood of past chloracne, as determined by anatomic patterns of acne, and no cases were diagnosed in the physical examination. Analyses were conducted on six dermatologic disorders (comedones, acneiform lesions, acneiform scars, inclusion cysts, depigmentation, and hyperpigmentation) and on a composite variable of 16 other minor conditions (the latter not generally associated with chloracne). Exposure index analyses did not reveal consistent patterns suggestive of a dose-response relationship. The longitudinal analysis, based on a composite dermatology index, showed no significant differences between the results over time. Substantially more dermatologic abnormalities were detected at the followup examination than at the Baseline examination. In general, however, the followup results were consistent with the findings at Baseline. Cardiovascular Evaluation Overall there was general similarity in the cardiovascular health of the Ranch Hands and the Comparisons. Of the 27 cardiovascular variables, there was a. significant difference for only one, verified heart disease, with an excess in the Ranch Hand group. This finding was largely unsupported by other cardiac measurements. The cardiovascular assessment was based on reported and verified heart disease; the measurement of central cardiac function by systolic blood pressure, abnormal heart sounds, and ECG findings; and the evaluation of peripheral vascula'r function by diastolic blood pressure, funduscopic examination, presence of carotid bruits, and detailed manual and Doppler measurements of five peripheral pulses. Doppler recordings of five peripheral pulses were similar in both groups, a finding which was in marked contrast to the Baseline examination that found significant pulse deficits in the Ranch Hand group. This change was most likely due to a required 4-hour abstinence from tobacco prior to the pulse measurements. Overall, the exposure analyses were unsupportive of any meaningful doseresponse relationship. The longitudinal analyses confirmed the change in pulse abnormalities in the Ranch Hand group over time, but showed no significant group change in overall ECG findings between the examinations. Hematological Evaluation The hematological evaluation found that neither group manifested an impairment of the hematopoietic system, consistent with similar findings at 22-5 �the Baseline. The evaluation was based on eight peripheral blood variables: red blood cells (RBC), white blood cells (WBC), hemoglobin (HGB), hematocrit concentration (HCT), corpuscular volume (MCV), corpuscular hemoglobin (MCH), corpuscular hemoglobin concentration (MCHC), and platelet count (PLT). Both the discrete and categorical analyses revealed no significant group differences. The covariate effects of age, race, occupation, and smoking history were highly significant for many of the variables. Two group-by-covariate interactions in the analyses of mean differences did not appear to have a meaningful interpretation. The exposure index analyses did not support any plausible dose-response relationship. The longitudinal analyses of MCV, MCH, and PLT found significant differences only for PLT between the Baseline and the followup, with the Ranch Hands exhibiting a slight decline in mean level from Baseline and the Comparisons showing an opposite change. Renal Assessment None of the six renal variables of reported kidney disease, urine protein, occult blood, urine white blood cell count, blood urea nitrogen, and urine specific gravity showed a significant difference between the two groups based on the unadjusted analyses. In the adjusted analyses of the laboratory variables, however, there were significant group-by-covariate interactions that did not yield a consistent pattern to suggest a renal detriment to either group. The finding of group equivalence for past kidney disease was in contrast to the Baseline examination, which found significantly more reported disease in the Ranch Hand group. The difference in findings is more likely due to a change in questionnaire wording than to a true change in renal health. Like the Baseline findings, the exposure index analyses showed very little evidence of a dose-response relationship. In the longitudinal analyses of blood urea nitrogen, there was no significant group difference in the change between the examinations. Endocrine Assessment In general, the endocrine health status of the Ranch Hands and the Comparisons was reasonably comparable. The examination found no significant differences between the two groups for past thyroid disease, or thyroid and testicular abnormalities determined by palpation. In the analyses of the seven laboratory values (T3 % Uptake; thyroid stimulating hormone [TSH]; testosterone; initial, second, and differential cortisol; and postprandial glucose), significant differences were found for TSH and testosterone, with higher mean levels in the Ranch Hands. These analyses were not supported by the categorical analyses. The thyroid test results were conflicting with respect to an assertion of hypothyroidism in the Ranch Hands (a possible dioxin effect). Mean levels of testosterone were significantly elevated in the Ranch Hand group as contrasted with the Comparisons in the 10-25 percent body fat category. The effects of personality score and percent body fat on the differential cortisol levels were not fully expected. Although tests of 2-hour postprandial mean values showed no significant group differences, comparable categorical tests revealed that significantly fewer Ranch Hands had impaired glucose levels, but conversely, had more (nonsignificant) diabetic levels of glucose. Analyses of the composite diabetes indicator (history plus 2-hour postprandial results) did not disclose significant group differences. The exposure index analyses suggested that the enlisted flyers in the medium exposure level were significantly different from those in the 22-6 �low exposure level for differential cortisol, postprandial glucose, and testosterone. The corresponding high to low contrasts were not significant. The longitudinal analyses were based on T^ % Uptake, TSH, and testosterone, and revealed only symmetrical and nonsignificant changes in the Ranch Hand and Comparison groups over the time interval. Immunological Evaluation Overall, there were no significant group differences or any indication of impaired immunological competence in either group based on comprehensive cell surface marker and functional stimulation studies. Six cell surface markers (total T cells, helper T cells, suppressor T cells, B cells, monocytes, HLA-DR cells, and a constructed helper/suppressor ratio variable) and three functional stimulation studies (PHA, pokeweed, and mixed lymphocyte culture) were conducted on 47 percent of the study population. No significant differences were revealed for five of these variables. In the analyses of the other five variables, there were significant group-by-covariate interactions, but no discernible pattern was identified to suggest a detriment in any subgroup of either group. Skin test assessments of delayed hypersensitivity were characterized by inter-reader variation and shifting diagnostic criteria for anergy. The skin test data were judged invalid and were not subjected to statistical testing for group differences. No consistent pattern of immunological deficits could be associated with increasing levels of herbicide exposure in the Ranch Hand group. Pulmonary Disease The pulmonary assessment did not reveal any statistically significant differences between the Ranch Hand and Comparison groups that were suggestive of an herbicide-related disease. The analyses consisted of group assessments of respiratory disease incidence, physical examination abnormalities, and the current prevalence of x-ray abnormalities. There were no significant differences between the Ranch Hands and Comparisons for history of asthma, bronchitis, pneumonia, or for six of seven clinical variables (excluding rales) determined by x-ray or auscultation. Analyses of history of pleurisy, history of tuberculosis, and rales showed significant but inconsistent groupby-covariate interactions. These findings did not indicate any patterns suggesting a different disease experience in the two groups. The exposure index analyses did not reveal any consistent pattern suggestive of an increasing dose response. CONCLUSION The results of the first followup study in 1985 have shown a, subtle but consistent narrowing of medical differences between the Ranch Hands and Comparisons since the Baseline Study in 1982. The 1985 examination results provide reassuring evidence that the current state of health of the Ranch Hand participants is unrelated to herbicide exposure in Vietnam. Continued close medical surveillance of these military populations is strongly indicated. This followup report concludes that there is not sufficient plausible or consistent scientific evidence at this time to implicate a causal relationship between herbicide exposure and adverse health in the Ranch Hand group. 22-7 �CHAPTER 23 FUTURE DIRECTIONS The scope and complexity of the AFHS has required gradual refinement and correction to meet the challenges of changing technology and scientific direction, and to ensure continued participation of all enrolled members. This chapter outlines some of the changes incorporated in the fifth-year followup examination and identifies several areas of future work expected to significantly augment the study. FIFTH-YEAR FOLLOWUP EXAMINATION Since the fifth-year followup examination was initiated prior to the full analysis of the data from the third-year examination, most modifications were founded upon quality control issues and the desire to make the clinical content of the examination more responsive to the medical needs of the participants. Clinical quality control enhancements were made to improve measurement techniques. The digit preference noted in systolic and diastolic blood pressure readings led to the use of automated blood pressure recording; all other parameters of the blood pressure readings (e.g., sitting position, three recordings, nondominant arm at heart level) were not changed. The problem in skin test reading was met by a rigorous quality control plan that included the following elements: refresher training for readers; a required reading of the four skin tests of all participants by both readers, each blind to the results of the other; a required reread of 10 percent of all tests by each of the readers, each blind to the previous reading; and a required weekly report citing numbers and proportions of participants with possible anergy, reversal of induration-erythema measurements, and untoward skin reactions or other reading problems (e.g., participant refusal). In addition, new skin test forms were developed to facilitate accurate recording and transcription; specific clinical criteria were formulated to require consultation by an allergist; and the skin test measurement criterion for possible anergy, consistent with current World Health Organization guidelines, was adopted for the clinical interpretation of all skin test readings. It is anticipated that this clinical quality control program will standardize both readings and interpretations, and will produce a uniformly superior data set. EXPOSURE INDEX REFINEMENTS Since the development of the Study Protocol and the analysis of the 1982 Baseline data, there has been concern among some scientists and the principal 23-1 �investigators over the accuracy and validity of the exposure estimates. It is unclear whether statistically significant differences in some variables between the Ranch Hand and Comparison groups, unsupported by dose-response estimates, have been due to chance, or whether true differences are obscured by an inadequate exposure index or group misclassification. In mid-1986, strong correlations between dioxin levels in fat tissue and serum were demonstrated by the CDC and other institutions. Because of these results, the Air Force is currently engaged in a collaborative study with CDC to determine whether serum dioxin levels vary significantly in the Ranch Hand population. Approximately 200 AFHS volunteers have supplied a pint of blood to be analyzed for dioxin at the CDC laboratories. If clear and meaningful exposure findings are evident from this study, several additional studies are feasible: testing can be expanded to the entire study population and a meaningful exposure index based on total current TCDD body burden may be developed; and by means of archived AFHS serum samples from the Baseline study, it may be possible to calculate a reasonably precise half-life of TCDD in humans. These expanded studies will allow the estimation of body burdens of TCDD at the time of departure from SEA (assuming the absence of intervening vocational and recreational exposures). If, in fact, these potential studies become reality within the next 2 years, the fifth-year followup study data will be statistically analyzed using a more appropriate exposure index. In anticipation of this advance, the AFHS is currently collecting 280 to 350 ml of blood from all volunteers attending the fifth-year followup study. ADDITIONAL ANALYSES AND STUDIES As in the 1984 Baseline Report, not all of the measured dependent variables were subjected to statistical analysis (e.g., prothrombin, leutinizing hormone, follicle stimulating hormone), largely because they were not within the bounds of the Air Force-prescribed analyses. Exploration of many of the unanalyzed variables is contemplated as time and resources permit. Similarly, many analytic opportunities to define possible symptom-clinical sign clusters or syndromes by multivariate analysis of variance techniques were passed over due to time and charter. Particularly challenging as an area of future work may be the changing relationships of some immunological variables over time and the biological impact of these changes on the induction of diseases such as cancer. Likewise, future efforts to define shifting cardiovascular disease patterns are a logical extension of the rich longitudinal data base of the AFHS. Such efforts await future analysis and publication. The assessment of possible selection and participation bias has been addressed in a comprehensive manner in this report (see Chapter 5). The analyses and discussion suggest that statistical use of the total Comparison group (versus the Original Comparison group) is justified in this report, and that the impact of selection and participation biases have been minimal. As the followup studies continue, it is anticipated that a wealth of data on 23-2 �compliance-participation factors will be available for continued comprehensive bias analyses. In particular, it is hoped that more complete data will exist to examine the true differences in current health status between refusals and their replacements. As the data set grows over time, the bias analyses will become more complex and will have to deal with changing motivations of the participants to continue in this study. Such bias analyses and assessments will always be of great importance to this study as they ultimately set the bounds for an inference on herbicide causality. 23-3 � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 059 Folder The folder containing the original item. 1585 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. William H. Wolfe Joel E. Michalek Judson C. Miner Michael R. Peterson Stella G. Machado Theodore G. Karrison William D. Grubbs Wanda F. Thomas Date A point or period of time associated with an event in the lifecycle of the resource 1987-10-01 Title A name given to the resource Air Force Health Study: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Volume I, First Followup Examination Results, January 1985-September 1987 Subject The topic of the resource Air Force Health Study dioxin morbidity trends ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/1f9b2eb8741a985801b0699f1d446087.pdf f8d1a09cc62f75b69d8f463ffd33d77a PDF Text Text Item ID Number °1584 AlltllOr Lathrop, George D. Corporate Author RBpOrt/ArtldO TltlB Air F°rce Health Study: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Summary, First Followup Examination Results, January 1985-September 1987 Journal/BookTItlQ Year 1987 Month/Day October n Color Number of Images 50 Dnsnrlnton NotBS Contract no. F41689-85-D-0010 and SAIC Project no. 2816-XX-195/254-XX. Item also includes 4 page report that was found inserted into the summary. Report is titled "Air Force Completes Second Ranch Hand Health Study" and is a short summary of the summary. V Wednesday, May 23, 2001 Page 1585 of 1608 �AIR FORCE OCMPLETES SBOOND RANCH HAND HEALTH STODY The Air Force today released the second health, study of its Herbicide Orange study. The study is an investigation of the possible health, effects from herbicide exposure of Air Force members who conducted aerial spraying missions in Southeast Asia (Operation Ranch Hand). Like the initial 1982 study, this follow-up concludes that there is insufficient evidence to support a cause and effect relationship between herbicide exposure and health affects in the Ranch Hand group at this time. A number of minor medical findings have been reported that require continued surveillance. The Air Force is currently engaged in a collaborative study with the Centers for Disease Control to determine whether blood dioxin levels vary significantly in the Ranch Hand population. Initial serum analysis indicates clear and meaningful dioxin exposure did occur in the Ranch Hand group. The follow-up study examined 2,309 participants (1,016 Ranch Hands and 1,293 comparisons). It was conducted under contract to the Air Force by Science Applications International Corporation, in conjunction with the Soripps Clinic and Research Foundation and the National Opinion Research Center. This follow up study marks nearly 9 years of intensive Air Force research into the herbicide health question. Nearly 100 government, academic and industry scientists have guided and contributed to the study since its inception. The current advisory committee, (The Advisory Committee on Special Studies Related to the Possible Long-term Health Effects on Ehenoxy Herbicides and Contaminants) comprised of renowned epidemiologists and chaired by Dr. Robert W. Miller of the National Cancer Institute, has reviewed and approved this report. A detailed report on the study follows. �No significant differences between the Ranch Hand and comparison groups were seen in the.1982-1985 interval for skin or systemic cancers. However, when overall lifetime basal cell carcinoma rates were adjusted for risk factors involved in the cause of such cancers (e.g. sun exposure, skin color, skin reaction to sun), Ranch Hands had a significantly higher proportion of basal cell carcinoma (the most common neoplasm in the white population in the United States) than comparisons. Yet the degree of difference between the two groups is decreasing with the passage of time. No group differences were observed for systemic cancer, although one soft tissue sarcoma and one suspected lymphoma were noted in the Ranch Hands. Two similar cases were previously found in the comparison group bringing the lifetime total to two of these cancers in each group. Overall, the cancer findings were not viewed as disturbing but as a reason for continued medical surveillance. The neurological assessment of cranial nerve function, peripheral nerve function, and central nervous system coordination did not reveal any consistently significant group differences. There were fewer neurological abnormalities noted than in the baseline study. Age, alcohol, and diabetes had the medically expected effects on many neurological measures. There were no group differences for current or past psychological or psychiatric illness. Age, educational level, and alcohol history showed strong and expected effects on the psychological measures. In evaluating the gastrointestinal and hepatic functions, both groups had an equivalent history of liver disease and ulcers. The follow-up examination disclosed more statistically significant findings for tests of liver function than the Baseline examination, but they were equally divided between the two groups. There were no demonstrated clinical, statistical, or exposure patterns consistent with an herbicide-related health effect. No evidence was found to suggest an increased likelihood of porphyria cutanea tarda in the Ranch Hand group. In the dermatological assessment, no cases of chloracne were diagnosed on examination, nor was there a history of acne that suggested past chloracne in the Ranch Hand group. Current and past non-malignant skin disease was equivalent in both groups. Overall, there was similarity in the cardiovascular health between the Ranch Hand and comparison groups. The cardiovascular evaluation showed no significant group differences for reported or verified hypertension, reported heart disease, or reported or verified heart attacks. However, verified heart d1sea.se was significantly greater in the Ranch Hands than the Comparisons. This category included a wide variety of cardiac conditions. Comparisons of the renal, and endocrine systems were very similar. Extensive study of the immune system revealed group equivalence in this important body function. �The pulmonary assessment, consisting of past history, physical examination, and x-ray results did not indicate any consistently different disease patterns in the two groups. Age and lifetime smoking history were important risk factors for most pulmonary measures. The exposure index analyses, which were specific to each occupation, revealed sporadic differences between exposure levels; however, there were no consistent dose-response relationships that supported an herbicide effect for any clinical area. These analyses found a subtle but consistent narrowing of medical differences between the Ranch Hands and comparisons since the baseline study in 1982. Continued close medical surveillance of these military populations is strongly indicated. At this time, there is not sufficient plausible or consistent scientific evidence to implicate a causal relationship between herbicide exposure and health effects in the Ranch Band group. Exposure Index Refinements Since the development of the study protocol and the analysis of the baseline data, there has been concern over the accuracy and validity of herbicide exposure estimates. The current herbicide exposure index is based on an estimate of the gallons of herbicide sprayed per month ( 9 2 1 7 ) 16-91; the levels of dioxin contamination and the number of men assigned each month. A different index was established for each occupation group. It was recognized that a better method of determining exposure was needed. In mid 1986, strong correlations between dioxin levels in fat tissue and blood were demonstrated by the Centers for Disease Control (CDC) and other institutions. The Air Force is currently engaged in a collaborative study with the CDC to determine whether serum dioxin levels vary significantly in the Ranch Hand population. Approximately 200 volunteers have supplied a pint of blood for analysis at the CDC laboratories. Initial serum dioxin analysis indicates clear and meaningful exposure did occur in the Ranch Hand group. The range of dioxin exposure in the Ranch Hand group (3-314 parts per trillion) is significantly higher than the control group ( - parts per trillion.) A part per trillion is equivalent to one second 39 in 32 thousand years. It is anticipated that serum dioxin testing will be expanded to all study participants and a more meaningful exposure index based on total current dioxin contamination may be developed for each participant. It may also be possible to calculate a precise half-life of dioxin in humans. These expanded studies will allow the estimation of dioxin in the participants since their departure from Southeast Asia. Study participants are providing blood for this dioxin test during the fifth year follow-up study currently underway. �Within the next two years, another morbidity report, mortality reports, and an expanded birth, defects study will be completed. The Air Force Health Study is a planned twenty year research, effort, concluding in 2002. This follow-up morbidity report has been reviewed and approved by the Advisory Committee on Special Studies Related to the Possible Long-Terra Health Effects on Fhenoxy Herbicides and Contaminants. �Air Force Health Study An Epidenriologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides Air Force Team SAJC Team George D. Lathrop, M.D., M.P.H., Ph.D. Stella G. Machado, Ph.D. Theodore G. Karrison, Ph.D. William D. Grubbs, Ph.D. Wanda F. Thomas, M.S. COL Joel LTC LTC William H. Wolfe, M.D., M.P.H. E. Michalek, Ph.D. Judson C. Miner, D.V.M., M.P.H. Michael R. Peterson, D.V.M., M.P.H., Dr.P.H. Project Manager: W.F. Thomas Program Manager: R.W. Ogershok SCIENCE APPLICATIONS INTERNATIONAL CORPORATION 8400 Westpark Drive McLean, Virginia 22102 EPIDEMIOLOGY DIVISION USAF School of Aerospace Medicine Human Systems Division (AFSC) Brooks Air Force Base, Texas 78235 October 1987 SUMMARY First Followup Examination Results January 1985 to September 1987 Contract Number F41689-85-D-0010 SAIC Project Number 2-8I6-XX-195/254-XX (Distribution Unlimited) �INTRODUCTION FIRST FOLLOWUP MORBIDITY STUDY The Air Force Health Study is an epidemiological study conducted to determine whether adverse health effects exist and can be attributed to occupational exposure to Herbicide Orange. The study consists of mortality and morbidity components, with followup studies. The Baseline study was conducted in 1982, and the first followup physical examinations were performed in 1985. The purpose of this report is to present the results of the first followup study. In the Baseline study, each living Ranch Hand was matched to the first living and compliant member of a randomly selected Comparison group based on age, race, and military occupation. The Comparisons had served in numerous flying organizations that transported cargo to, from, and within Vietnam but were not involved in the aerial spray operations of Herbicide Orange. Recruitment for the first followup was in accordance with the Study Protocol: All previous participants and refusals, newly located study members, and replacements (matched to noncompliant Comparisons on self-perception of health) were invited. Of the living Baseline study participants, 99.2 percent were contacted to enroll in the followup on a strictly voluntary basis. Participation was very high, with 93 percent of both the Ranch Hands and the Comparisons who were fully compliant at Baseline also participating in the followup. Overall, the 2,309 followup participants (1,016 Ranch Hands and 1,293 Comparisons) represented a loss to the study of 159 individuals but a gain of 199 new participants since Baseline. Statistical analyses of selection and participation bias supported the use of the total Comparison group for the main analyses presented in this report. 1 The followup study was conducted under contract to the Air Force by Science Applications International Corporation, in conjunction with the Scripps Clinic and Research Foundation and the National Opinion Research Center. Most of the data were collected through face-to-face interviews and physical examinations conducted at the Scripps Clinic in La Jolla, California. Other data sources included medical and military records and the 1982 Baseline data base. As a contract requirement, all data collection personnel were blind to exposure status, and all phases of the study were monitored by stringent quality control. The statistical analyses were based on analysis of variance and covariance, chi-square tests, Fisher's exact tests, general linear models, Kolmogorov-Smirnov tests, logistic regression, proportional odds models, t-tests, and log-linear models. The release of this 1987 followup Morbidity Report marks more than 8 1/2 years of intensive Air Force research in the herbicide question. Since the commitment to Congress in October 1978 to conduct an epidemiologic investigation of Air Force.personnel who aerially disseminated herbicides in the Vietnam War (code-named Operation Ranch Hand), the United States Air Force Surgeon General has issued the following publications: a Study Protocol, four annual mortality reports, the Baseline Morbidity Report, and this first followup morbidity report. Within the next 2 years, the second followup morbidity report, other annual mortality reports, and an expanded birth defects study are expected for publication. This level of commitment has used approximately $40 million of contract research funds, excluding significant Air Force in-house expenditures. �Nearly 100 Government, academic, and industry scientists have guided and contributed to the Air Force Health Study (AFHS) since its inception. The Air Force's current advisory committee, chaired by Dr Robert W. Miller of the National Cancer Institute, is responsible for providing assistance on all scientific and medical matters pertaining to the AFHS. The chief strength of the AFHS is its design. The interwoven study elements of multiple mortality assessments, a Baseline morbidity study, and five followup morbidity studies over 20 years provide a comprehensive approach to the detection of adverse health effects. The weakest feature of the design is the mortality assessment which cannot detect group differences for very rare conditions (e.g., soft tissue sarcoma) because of the inherent constraints of the limited size of the population. The strength of the mortality studies should increase with the aging of the study populations and the increase in death with the passage of time. All four mortality assessments have shown that the Ranch Hand population is faring about the same as the Comparison group, with no unusual causes o-f death, increased frequency of death, or evidence suggesting death at younger ages. Because of the healthy veteran .effect, both groups are surviving significantly longer than similarly aged civilians. The morbidity assessment, released in 1984, disclosed only minor differences between the Ranch Hands and the Comparisons, and these differences were not traditional indicators of dioxin>-related disease. Both the content and the progress of the AFHS has been presented on many occasions to Congress, to the media, and to scientific meetings around the world. On the whole, the AFHS has been very well received in these circles, giving additional strength and credence to this work. The questionnaire and physical examination data were analyzed by major organ system. The primary focus was on the assessment of differences between the Ranch Hand and Comparison groups based on data from the first followup. Additionally, dose-response relationships within the Ranch Hand group were examined, and longitudinal assessments of differences in the changes of the two groups between the examinations were conducted for selected variables. Reported significant findings in subsequent major organ systems are understood to be significant at the 0.05 level or less. GENERAL HEALTH General physical health was evaluated by five measures, three of which were subjective (self^-perception of health, appearance of distress, and appearance of relative age), and two of which were objective (percent body fat and sedimentation rate). Table 1 presents a summary of all the unadjusted and adjusted analyses of these five variables. The mean age of the Ranch Hands was 46.9 years and the comparisons mean age was 46.8 years. The Ranch Hands rated their health as fair or poor more often than the Comparisons (9.1$ versus 7.3%, respectively), but this difference was not significant by categorical testing. However, further analysis revealed a significant group-by-occupation interaction; differences were largely confined to the enlisted groundcrew category. Both the Ranch Hand and Comparison groups noticeably improved their perceptions of health since the 1982 Baseline examination. �Only 10 individuals were reported as appearing acutely ill or distressed at the followup examination, 4 were Ranch Hands and 6 were Comparisons. This difference was not statistically significant and the data were insufficient for adjusted analyses. TABLE 1. Overall Suuury Results of Unadjusted and Adjusted Analyses of General Health Variables Unadjusted Variable Categorical Adjusted Mean Categorical Mean **** Self-Perception of Health NS Appearance of Illness/Disstress NS Appearance of Relative Age NS Sedimentation Rate 0.013 NS 0.011 NS Percent Body Fat NS 0.037 NS 0.035 - • **** —Analysis not performed. ****Group-by-covariate interaction. 'Analysis not possible due to sparse data. Appearance of relative age, as determined by the examining physician, showed 1.6 percent of the Ranch Hands appearing younger than their stated age, 94.3 percent appearing the same, and 4.1 percent appearing older (as contrasted to 0.7$, 95.4$, and 3.9$, respectively, in the Comparison group). There was a significant group-by-occupational interaction, but none of the estimated relative risks for the occupational categories was significant. This observation at the followup examination contrasted with the significant tendency at the Baseline for a higher percentage of Ranch Hands than Comparisons to appear younger than their stated ages. The geometric mean sedimentation rates (5.05 mm/hr Ranch Hand versus 4.93 mm/hr Comparison) did not differ significantly by group, either unadjusted or after adjustment for age, race, occupation, personality score, and an ageby-personality score interaction. However, in the dichotomous form, 5.8 percent of the Ranch Hands had sedimentation rate abnormalities as contrasted to 3.6 percent in the Comparison group. This difference was significant by �both unadjusted and adjusted tests. Also, this finding was opposite to that of the Baseline examination, where it was noted that younger Comparisons had significantly elevated sedimentation rates. The mean percent body fat of the Ranch Hands was significantly lower than the Comparisons (21.10J+0.15, 21.5^+0.14, respectively, p-0.037), and was of nearly the same magnitude after adjustment for age, race, and occupation. However, both unadjusted and adjusted categorical tests did not reveal significant group differences, although the percent obese was lower in the Ranch Hands than in the Comparisons. No group differences in percent body fat were noted at the Baseline examination. Detailed exposure analyses were done on four general health variables (appearance of acute distress was too sparse for testing). Only one analysis demonstrated statistical significance, i.e., a positive association of sedimentation rate abnormalities with increasing exposure in the enlisted flyer cohort. Overall, no consistent pattern of exposure effects was discernible, and the exposure findings at the thirds-year followup were similar to the findings at Baseline. Longitudinal differences between the 1982 Baseline and the 1985 followup examination were assessed by analyses of two discrete variables, selfperception of health and sedimentation rate. Perceived health showed no significant group differences over time, but both the Ranch Hand and Comparison groups paradoxically reported symmetrical improvements in their perceptions over the 3~year period. The sedimentation rate analysis revealed a highly significant group difference (p-0.002), due to a reversal of findings between examinations, i.e., a significant detriment in the younger Comparisons at the Baseline versus a significant detriment in the Ranch Hands at the followup. The cause(s) and biological relevance of this observation are unclear. In conclusion, a nonspecific assessment of general physical health has shown relatively close similarity between the Ranch Hand and Comparison groups, with the Ranch Hands continuing to perceive their health more negatively than the Comparisons, having a slightly more favorable percent body-<fat proportion, but a higher proportion of abnormal sedimentation rates that reflects a marked change since the Baseline examination. These findings must be placed in context with the organ and system-specific evaluations found in the succeeding chapters. MALIGNANCY The cancer analysis focused on cancer occurrences in the Baselinefollowup interval, and also included analyses of the Baseline plus interval cancer history. A listing of systemic malignancies occurring in the study particpants is shown in table 2 and a summary of the cancer findings is given in Table 3No significant unadjusted differences were found between nonblack Ranch Hands and Comparisons in the Interval (Baseline^Followup) incidence rates of basal cell carcinoma, melanoma, squamous cell carcinoma, all malignant skin cancers, sun^exposure related malignant neoplasms (comprising basal cell carcinoma, melanoma, and epithelial neoplasms NOS) or all malignant skin cancers as a group. The unadjusted group contrast of all skin neoplasms �TABLE 2. Summary of Pollowup Participants With Lifetime Incidence of Verified Malignant Systemic Neoplasms by Group Group Ranch Hand Comparison Total 0 1 Oral Cavity and Pharynx 1 3..b 0 3 Larynx 0 1 1 Thyroid Gland 0 2 2 Esophagus 0 lc 1 Bronchus and Lung 2 1 Colon 0 0 5d.. 5 Kidney and Bladder 4 3 7 Prostate 2 2 4 Testicles 3 0 3 Connective and Other Soft Tissue 1 1 2 Hodgkin's Disease 0 1 1 Ill-Defined Sites lf I9 2 Site Eye Total 17 17 34 'includes one Ranch Hand with separate malignancies of tongue and epiglottis and also malignant neoplasm of bone. b Includes one Ranch Hand with separate malignant neoplasms of tongue and oropharynx and secondary malignant neoplasm of other site. c Also has malignant neoplasm of bone. d lncudes one Comparison with secondary malignant neoplasms of liver and bone and bone marrow. 'includes one Comparison with secondary malignant neoplasm of liver. 'Malignant neoplasm of thorax. g Malignant neoplasm of face, head, or neck. �TABLE 3. Overall Siunary Table: Unadjusted and Adjusted Analysis of Interval and Lifetime Skin and Systeaic Cancer Incidence Cancer Type Baseline-Followup Interval UnadjustedAdjusted Lifetime (Baseline & Follovup) UnadjustedAdjusted Malignant Skin Cancer (Nonblacks only) Verified Basal Cell Carcinoma NS **** NS Verified plus Suspected Basal Cell Carcinoma NS **** NS Verified Melanoma NS —* NS Verified plus Suspected Melanoma NS —* NS Verified Squamous Cell Carcinoma NS —* NS Verified plus Suspected Squamous Cell Carcinoma NS —* NS Verified Sun Exposure Skin Cancers NS NS NS* Verified plus Suspected Sun Exposure Skin Cancers NS NS NS —" NS All Verified Malignant Skin Cancers NS Verified plus Suspected Malignant Skin Cancers NS NS Verified Skin Cancers of Any Type NS* S Verified plus Suspected Skin Cancers of Any Type NS NS* **** NS �TABLE 3. Overall Summary Table: Unadjusted and Adjusted Analysis of Interval and Lifetime Skin and Systemic Cancer Incidence (continued) Baseline-Followup Interval UnadjustedAdjusted Cancer Type Lifetime (Baseline & Follovup) Unadjusted Adjusted Malignant Systemic Cancer (Blacks and Nonblacks) Verified Systemic Cancer NS NS NS **** Verified plus Suspected Systemic Cancer NS **** NS **** All Neoplasms (Blacks and Nonblacks) Any Type, Any Location13 Verified NS* NS: Not significant (p>0.10). ****Group-by-covariate Interaction. —"Analysis not done. NS*: Borderline significant (0.05<p<0.10). Comprises malignant, benign, uncertain behavior. S: Significant (p<0.05). �(comprising malignant and benign neoplasms, and neoplasms of uncertain behavior or unspecified nature) was marginally significant, with a higher rate among Ranch Hands. When suspected malignant skin cancers (noted at Followup but not verified at the time of writing) were included in the analyses with the verified conditions, all the unadjusted group contrasts were nonsignificant. The covariates used for the adjusted analyses of basal cell carcinoma and the sun exposure related skin malignancies were age, occupation, skin color, reaction of skin to sun, and average latitude, all of which were highly associated with skin cancer incidence. Other host factors were related to skin cancer incidence, but not as strongly as those included in the analysis. A borderline association with smoking history was noted, and was determined to be partly an age effect. Analysis of the incidence of interval basal cell carcinoma revealed a significant group«-by-occupation interaction, due to a significant group difference for enlisted flyers, but not for officers or enlisted groundcrew. .Inclusion of suspected basal cell carcinoma resulted in a group-bysun reaction index interaction. This was due to Ranch Hands with an intermediate reaction to sun having a higher relative risk than the corresponding Comparisons. The adjusted group contrast of the incidence rates of verified sunexposure related skin cancers was not significant; inclusion of suspected conditions did not alter this lack of significance. There was no significant group difference for Blacks and nonblacks in the unadjusted incidence rates of all interval verified malignant systemic neoplasms combined, nor was there a significant difference in the adjusted group rates. Analysis of the verified plus suspected interval systemic cancers showed a nonsignificant unadjusted group difference, but a group by occupation interaction was found in the adjusted analysis. This was due to a significant group difference of verified plus suspected systemic malignancies among the enlisted flyers with five occurrences among the Ranch Hands, but none among the Comparisons. Age and a race^by-packyear interaction were important adjusting factors. The Baseline and Followup data were combined for the assessment of lifetime incidence of cancer; occurrences of cancer prior to Vietnam were excluded. There were no significant unadjusted group differences in lifetime incidence rates among nonblacks for basal cell carcinoma, melanoma, squamous cell carcinoma, the sun-exposure related skin cancers, or all malignant skin cancers combined. The unadjusted group contrast of all lifetime skin malignan*cies was significant, with a higher rate among Ranch Hands. Inclusion of suspected cancers with the verified cancers reduced the difference between the groups for all these malignant skin contrasts, except for the sun-exposure related skin cancers, for which a marginally significant group difference was found. However, the contrast of all skin malignancies remained close to significance. Adjusted analysis of the incidence rates of lifetime basal cell carcinoma revealed a significantly higher incidence rate among Ranch Hands (Adj. RR: 1.56, p-0.035). Significant effects of an occupation-by-age interaction, a �skin color-by-sun reaction index interaction, and a sun reaction index-byaverage residential latitude interaction were seen. The adjustment resulted in a significant relative risk that, moreover, was higher than the unadjusted relative risk. Average residential latitude, associated with both group and skin cancer, and skin color, which was associated with the disease and marginally associated with group, played a major part in the change from the unadjusted analysis due to confounding. Inclusion of suspected basal cell carcinoma in the adjusted analysis resulted in a group by sun reaction index interaction, as was noted for the interval analysis. The adjusted group contrast in incidence rates of the sun-exposure related skin cancers was also significant (Adj. RR: 1.54, p=0.030), which is not surprising since the majority are basal cell carcinoma. Inclusion of the suspected conditions resulted in a nonsignificant group contrast. The unadjusted group contrasts of the incidence rates of all systemic cancers combined were not significant, both for verified and verified plus suspected conditions. There was one new occurrence of a soft tissue sarcoma (Ranch Haind) and one suspected cancer of the lymphatic system (Ranch Hand), in addition to the one previously reported soft tissue sarcoma and one Hodgkin's disease in the Comparison group. Adjusted analysis of all lifetime malignant systemic neoplasms as a group, however, revealed a group by occupation interaction, due to a significantly higher rate for Ranch Hand enlisted flyers as contrasted to Comparisons. The same result was found for verified plus suspected systemic cancers. In conclusion, there were no adjusted or unadjusted differences between groups in basal cell carcinoma incidence in the Baseline-followup interval. At Baseline, a significantly higher rate of basal cell carcinoma was found for Ranch Hands when contrasted with Original Comparisons. When the Baseline data were combined with the interval data, adjusted analysis, but not the unadjusted analysis, revealed a significantly higher rate of basal cell carcinoma among the Ranch Hands than among all Comparisons. The relative risk of basal cell carcinoma appears to be declining over time. Relative risks of basal cell carcinoma and systemic cancer were found to be consistently larger than 1. Most of the skin cancers were basal cell carcinomas, upon which most of the skin cancer analysis focused, thus relative risks for swvexposure related skin neoplasms and all malignant skin cancers as a group were very similar to those for basal cell carcinoma. The number of occurrences of systemic cancer was small, in part because the cohort is relatively young, and although the relative risks (lifetime and interval) are greater than 1, the difference between groups is not significant. Sufficient time may not have elapsed since Vietnam to enable a group difference in systemic neoplasms, if one exists, to be apparent. NEUROLOGY Interval questionnaire data (1982 through 1985) on neurological illnesses, verified by medical records, revealed no significant group differr ences. These data were added to verified Baseline historical information to �assess possible differences in the lifetime experience of neurological disease. Again, there was no significant difference between the Ranch Hand and Comparison groups. A detailed neurological examination evaluated neurological integrity in three broad areas: cranial nerve function, peripheral nerve function, and central nervous system (CNS) coordination. The summary analytic results for all measurement variables comprising these three functional areas are presented in Table 4. Assessment of the 12 cranial nerves was based on the measurement of 1M variables. Two summary indices were constructed. Both the unadjusted and adjusted analyses did not disclose any statistically significant group differences, although two variables, speech and tongue position, were of borderline significance, with Ranch Hands faring worse than Comparisons. One of the two cranial nerve summary indices was marginally significant, again with the Ranch Hands at a slight detriment. The unadjusted and adjusted analyses of peripheral nerve function, as measured by eight variables (four reflexes, three sensory determinations, and muscle mass), did not reveal significant group differences. CNS coordination was evaluated by four measurements and a constructed summary variable. Hand tremor was found to be of borderline significance, with the Ranch hands faring slightly worse than the Comparisons. The CNS summary index showed a significant detriment to the Ranch Hands. The exposure analyses for neurological variables with reasonable counts of abnormalities showed only occasional statistically significant results. No consistent pattern with increasing exposure was evident for any occupational category of the Ranch Hand group. In a longitudinal analysis of the Romberg sign and the Babinski reflex, only the Babinski reflex revealed a significant difference between the Baseline and followup examination, with the Ranch Hands converting from significant adverse findings at Baseline to favorable nonsignificant findings at the followup examination. Overall, the followup examination findings are quite similar to the Baseline findings. However, several distinct patterns were evident from the analyses: (1) The followup examination detected substantially fewer abnormalities for almost all measurement variables, (2) the decrease in abnormalities was equivalent in both groups, (3) most of the covariate effects were classical, although exceptions were evident, CO the adjusted analyses were uniformly similar to the unadjusted analyses, (5) the constructed summary variables were generally statistically significant, or of borderline significance (however, some indices were created after the data were examined), and (6) although statistical significance at the pre^-assigned '-level of 0.05 was not achieved for any of the measurement variables, abnormalities tended to cluster in the Ranch Hand group. Of the three group-by-covariate interactions in the adjusted analyses, only one, a borderline group-by-insecticide exposure interaction for hand tremor, where Ranch Hands exposed to insecticides had a marginally significant adverse effect, was of probable biologic (and operational) significance. 10 �TABLE 4. Overall Suwury Results of Unadjusted and Adjusted Analyses of Neurological Variables Variable Unadjusted Adjusted Direction of Results** Questionnaire* Physical Examination Neurological Disease (Interval) Neurological Disease (History) NSb NS Cranial Nerve Function Smell Visual Fields Light Reaction Ocular Movements Facial Sensation Corneal Reflex Jaw Clench Smile Palpebral Fissures Balance Gag Reflex Speech Tongue Position Relative to Midline Palate and Uvula Movement Neck Range of Motion Cranial Nerve Function Index* Cranial Nerve Function Index (excluding Neck Range of Motion) NS NS NS NS NS —c NS NS NS NS NS NS* RH>C NS* NS NS NS NS NS NS* NS* RH>C Peripheral Nerve Function Pin Prick Light Touch Muscle Status Vibratory Sensation Patellar Reflex Achilles Reflex Biceps Reflex Babinski Reflex NS NS NS NS NS NS NS NS 11 **** NS NS NS RH>C �TABLE 4. (Continued) Overall Suuury Results of Unadjusted and Adjusted Analyses of Neurological Variables Variable Unadjusted Adjusted Direction of Results** Central Nervous System Coordination Tremor Coordination Romberg Sign Gait CNS Summary Index NS* NS NS NS 0.036 NS* Rfl>C — — NS O.OA2 RH>C **RH>C: More abnormalities in Ranch Hand group than in Comparison group. 'Disease categories include: inflammatory diseases, heriditary and degenerative diseases, peripheral disorders, disorders of the eye, disorders of the ear, and other disorders. NS:Not significant (p>0.10). b No inflammatory diseases noted; borderline significant (p-0.069, RH>C) for other disorders; not significant for remaining categories. —Analysis not performed because of sparse number of abnormalities. c No abnormalities present. NS*Borderline significant (0.05<p<0.10). Constructed variable. ****Group-by-covariate interaction. 12 �In conclusion, none of the 27 neurological variables demonstrated a significant group difference, although several showed an aggregation of abnor*malities in the Ranch Hand group, which merits continued surveillance. Historical reporting of neurologic disease was equal in both groups. The clinical sensitivity in detecting neurological deficits varied substantially between the Baseline and the followup examinations, but the number of statistically significant variables remained about the same. None of the exposure analyses revealed dose-response patterns in the Ranch Hand occupational categories. The longitudinal analyses disclosed a favorable reversal of significant Babinski reflex abnormalities at Baseline to nonsignificant findings at . the followup examination for the Ranch Hands. The similarity in results between unadjusted and adjusted statistical tests is evidence of group equality for the traditionally important neurological covariates of age, alcohol, and diabetes. Of three group-by-covariate interactions in the adjusted analyses, only the Ranch Hand insecticide interaction with hand tremor was biologically plausible. PSYCHOLOGY Questionnaire data (verified by medical record reviews) for the lifetime events of psychotic illness, alcohol dependence, anxiety, or other neuroses disclosed no significant differences between groups for these conditions. (Table 5). TABLE 5 Distribution of Reported Psychological Illness By Group - Baseline and Followup Studies Combined Type of Illness Psychoses Alcohol Dependence Anxiety Other Neuroses Abnormalities Ranch H a n d s C o m p a r i s o n s NumberPercent NumberPercent 14 9 7 72 9 8 13 74 1.4 0.9 0.7 7.1 *Fisher's Exact Test 13 0.7 0.6 1.0 5.7 Total 23 17 20 146 p-Value* 0.138 0.473 0.501 0.197 �Analyses of the followup psychological examination emphasized Hi scales from the Minnesota Multiphasic Personality Inventory (MMPI), 3 parameters of the Cornell Medical Index (CMI), and the Halstead-Reitan Battery (HRB) impairment index. The similarity of the group distribution for the 14 MMPI variables, each stratified by the 3 occupational categories, was examined, and only 2 of the 42 tests approached statistical significance. The group distributions of the total CMI score were similarly contrasted, with separate analyses performed with stratification by the five covariates of age, race, occupation, education, and current drinking status. For one stratum of each of these covariates, a significant difference in the distribution of the Ranch Hand and Comparison scores was found. In all cases for the CMI, the Ranch Hand mean was greater than the Comparison mean. Distributional analyses using Original Comparisons generally reflected the same results as those involving the total Comparison group. Results of unadjusted and adjusted analyses on all of the 18 psychological variables are given in Table 6. The unadjusted analyses showed a significant difference for the MMPI scales of denial (p<0.001) and masculinity/femininity (p-0.017), the total CMI (p<0.001), and the Section A^H area subscore (p-0.003). A borderline significant difference was observed for the MMPI scales of hysteria (p-0.067) and social introversion (p-0.069). Comparisons had a greater percentage of abnopmal scores for the denial and masculinity/femininity scales, whereas Ranch Hands showed adverse findings for the other four variables. The overall MMPI results have been interpreted in light of the significant increased denial in the Comparison group. The covariates age, education, drink-years, current alcohol use, and occupation had pronounced effects on the psychological variables, with a significant association or a borderline significant association with at least twof-thirds of the 18 psychological variables. Many dependent variables in this chapter were affected by age in an expected pattern. Very few variables exhibited this pattern of consistency with drink--years. The intermediate category of greater than 0 to 50 drink-years often had'the smallest proportion of abnormalities. The post?-traumatic stress disorder (PTSD) variable, derived from a subset of'the MMPI, was strongly associated with the CMI measures, but not with the HRB Impairment Index. Race and the Vietnam combat index (used for the MMPI subscales) had significant associations with a lesser amount of the psychological variables (6 of 18 variables and 3 of 14 variables, for race and combat index, respectively). The adjusted analyses were generally quite similar to the unadjusted analyses with respect to group differences, although a direct comparison of these analyses was often clouded by the presence of a substantial number of interactions (six group4bycovariate interactions were significant, and three interactions approached significance [0.05<p£0.10]). The MMPI scales of denial and masculinity/femininity were statistically significant in both the adjusted and unadjusted analyses, where Comparisons showed an adverse effect over Ranch Hands. The A--H area subscore of the CMI (suggesting diffuse medical problems) was also significant, where the Ranch Hands had higher mean scores than the Comparisons, suggesting the Ranch Hands had more illness. Education was often involved in significant group interactions with high 14 �TABLE 6. Overall Summary Results of Adjusted and Unadjusted Analyses of Psychological Variables Variable Unadjusted Questionnaire! Psychological Illness Adjusted Direction of Results* NS Psychological Examination; MMPI Anxiety Consistency Oefensiveness Denial Depression Hypochondria Hysteria Nania/Hypomania Masculini ty/Feminini ty Paranoia Psychopathic/Deviate Schizophrenia Social Introversion Validity CMI Total CMI M-R Subscore A-H Area Subscore HRfi Impairment Index • NS NS NS <0.001 NS NS b NS*b NS 0.017 NS NS NS **** NS <0.001 NS MS b NS* NS 0.020 **** ORH RH>C ORH NS NS **** h NS*b NS **** **** RH>C <0.001 NS 0.003 **** NS 0.040 RH>C RH>C NS NS *RH>C - more abnormalities in Ranch Hands; ORH - more abnormalities in Comparisons. b lllnesses include psychosis, alcohol dependence, anxiety, and other neuroses. —Analysis not performed. NS: Not significant. NS*: Borderline significant (0.05<p<0.10). ****Interaction involving group. 15 �school-educated Ranch Hands demonstrating a higher percentage of abnormal scores than high school-educated Comparisons. No group differences were observed in the college-educated stratum. The M-R subscore of the CMI, a broad indicator of emotional health, was not statistically different between the two groups. The HRB impairment index, a measure of central nervous system (CNS) functional integrity, did not d i f f e r significantly between the Ranch Hand and Comparison groups. Strong covariates in the adjusted analysis were age, race, and education. Because of alternate statistical models and slightly different psychological testing parameters, a direct contrast between the psychological results of the Baseline and followup examinations was not always possible,, However, several broad patterns were observed: ( 1 ) the discordance between distributional tests and results from traditional statistical models of the MMPI variables was noted with data from both examinations; (2) there was a narrowing of group differences at the followup examination for most subjective' variables, either by a decrease in Ranch Hand reporting, or by an increase in Comparison reporting; and (3) as at the Baseline, functional CNS testing, as measured by the HRB impairment index, showed no group differences, and did not support an organic basis for differences in self-reported symptomatology. The longitudinal analysis of two MMPI scales, depression and denial, showed a significant reversal of depression seen at Baseline in the high schooleducated Ranch Hands. The determination of PTSD in both Air Force cohorts by a relatively new MMPI scale showed a prevalence rate of less than 1 percent. This low rate is strongly influenced by characteristics of the study population (e.g., age, education, and officer ratio). Unadjusted exposure index analyses did not reveal any patterns consistent with a dose-response relationship. For the adjusted exposure analyses, approximately one-third presented exposure interactions with the covariates of race, education, and age, but no consistent pattern could be identified. In conclusion, some test measures of psychological health (MMPI and CMI) did not show substantial adverse effects for either group. Significant test results were present in both groups or were noted in specific subgroups of a covariate. Educational level, age, and alcohol use showed strong effects on the psychological scales and scores in this psychological assessment. There was a subtle but consistent trend for more favorable subjective test results at the followup examination for the Ranch Hands relative to the Comparisons. Testing of the CNS by the HRB demonstrated an almost identical prevalence of pathology in both groups. DIGESTIVE SYSTEM The interval questionnaire revealed sparse reporting of liver disorders from 1982 to 1985 that was not significantly different between groups. Historical liver disease was verified by medical records, and these data were added to the verified Baseline history to assess possible lifetime differences. No significant differences were found. The medical record verifica- 16 �tion process showed that the historical data were generally correctly reported and classified between groups, except for the category of enlarged liver which showed a higher verification rate in the Comparison group. Digestive system mortality showed an overall nonsignificant excess in the Ranch Hands, but a relative nonsignificant excess of malignant neoplasms in the Comparisons. No differences were found for past or current peptic ulcer disease for the Ranch Hand and Comparison groups, adjusted for standard covariates as well as blood type. The physical examination disclosed a borderline significant increase of hepatomegaly in the Ranch Hand group. Emphasis was placed on nine laboratory test variables measuring liver function, i.e., serum glutamic-oxaloacetic transaminase (SCOT), serum glutamic^pyruvic transaminase (SGPT), gammaglutamyl transpeptidase (GGTP), alkaline phosphatase, total and direct bilirubin, lactic dehydrogenase (LDH), cholesterol, and triglycerides. In addition, • uroporphyrin and coproporphyrin measurements were obtained to assess liver function and the likelihood of porphyria cutanea tarda (PCT). The nine hepatic variables were subjected to continuous and categorical statistical tests, and were adjusted for the covariates age, race, occupation, current alcohol consumption, and unprotected exposure to both industrial chemicals and degreasing chemicals. Final statistical models used only the significant covariates and two-way interactions for adjustment. The two porphyrin measurements were analyzed only in the continuous form. The overall summary re-' suits of the analyses of these 11 variables are given in Table 7. The results showed a significantly lower mean SGPT level, a greater mean alkaline phosphatase level, a lower mean uroporphyrin level for Ranch Hands as contrasted with Comparisons, and a marginally significant greater mean coproporphyrin level. Only in the instance of alkaline phosphatase did the discrete analysis approach statistical significance. No group differences were noted for SCOT, GGTP, total and direct bilirubin, LDH, cholesterol, or triglycerides. However, an analysis using only the Original Comparisons revealed a significantly greater mean cholesterol level in the Comparison group. A review of the covariate effects in the adjusted statistical models revealed that all covariates behaved as expected with the exception of alcohol consumption for the alkaline phosphatase analysis, which showed an inverse relationship with wine consumption. Exploration of group"by-group covariate interactions for alkaline phosphatase, direct bilirubin, triglycerides, SCOT, and uroporphyrins revealed significant group differences within specific covariate strata. In particular, Ranch Hands exposed to industrial chemicals had a significantly higher adjusted mean level of alkaline phosphatase and a significantly higher abnormal prevalence rate of direct bilirubin than similarly exposed Comparisons. For triglycerides, Ranch Hands born in or before 1922 had a significantly higher adjusted mean level than similar aged Comparisons, while Ranch Hand officers exhibited a significantly higher abnormal prevalence rate than Comparison officers. For SCOT, Ranch Hand moderate current drinkers (more than one to four drinks per day) had a significantly higher mean level than corresponding Comparisons. In the opposite direction, Comparisons with a mean BUN level less than or equal to 1*1 (median participants) were found to have a 17 �TABLE 7. Overall Stunary Results of Unadjusted and Adjusted Analyses of Nine Hepatic Function Variables and Two Porphyrin Metabolite Tests Unadjusted Adjusted* Mean Categorical Mean CC Variable Questionnaire Liver Disease (Lifetime History) Hepatitis Jaundice Cirrhosis Enlarged Liver Miscellaneous Liver Disorders Peptic Ulcer Disease CD NS Physical Examination Hepatomegaly Categorical Direction of Results** NS* NS NS NS NS NS NS' RH>C Laboratory Testing SCOT SGPT GGTP Alkaline Phosphate Total Bilirubin Direct Bilirubin LDH Cholesterol Triglycerides Uroporphyrin Coproporphyrin NS NS* NS 0.009 NS NS NS NS NS 008 .4 NS* NS NS NS NS NS NS NS NS NS NS **** NS 0.048 NS 0.008 NS NS **** NS **** **** 0.029 NS **** NS NS NS NS NS NS NS NS* NS **** NS NS **** NS* Questionnaire-Laboratory Correlation Skin Bruises, Patches,-0.001 and Sensitivity ORH RH>C ORH Rfl>C RH>C *C: Continuous D: Discrete **RH>C: more abnormalities, or higher mean value, in Ranch Hands, ORH: more abnormalities, or higher mean value, in Comparisons. 'Adjusted for blood type. NS: Not significant (p>0.10). NS*: Borderline significant (0.05<p<0.10). --Analysis not performed. ****Group-by-covariate interaction. 18 �significantly higher adjusted mean uroporphyrin level than similar Ranch Hands. These results did not disclose any common pattern detrimental to the Ranch Hand group. These findings were generally consistent with the 1982 Baseline data, which disclosed a significantly increased mean cholesterol level in the Comparisons and nonsignificant Ranch Hand mean elevations for GGTP and LDH. Slight differences in analytic results are probably due to the use of more fully adjusted models used for the followup examination data. Overall, the followup examination laboratory data showed no adverse clinical or exposure patterns in either group. Further, the detection of significant mean shifts (still within normal range) by the continuous statistical tests, not mirrored by the categorical tests, suggests a circumstance of statistical power rather than findings of biological relevance. Of the five significant or marginally significant results that were found in the adjusted exposure index analyses, four exhibited a trend suggest!ve of an increasing dose-response relationship. In the enlisted flyer cohort, the percentages of SGPT abnormalities were significantly different and increased from the low to the high exposure categories. The corresponding mean values were marginally significantly different among exposure levels. Also, the mean levels of total bilirubin were marginally significantly different among exposure levels, increasing with exposure level. For enlisted groundcrew, the percentage of SCOT abnormalities significantly differed among exposure levels. Within the enlisted flyer cohort, all nine laboratory tests of hepatic function had the lowest percentage of abnormalities in the low exposure category; correspondingly, six of the nine mean levels were lowest for the low exposure category. Of the ten group-by-covariate interactions that were found, three (SGOT, SGPT, and GGTP) supported a dose-response relationship in the enlisted groundcrew cohort. Exploration of these interactions revealed a trend that showed an increasing association between current alcohol consumption and the dependent variables for increasing exposure levels. Longitudinal analyses for SGOT, SGPT, and GGTP disclosed no statistically significant group differences in the mean shifts from the Baseline to the followup examination. Interval reporting of PCT-like symptoms of skin patches, bruises, and sensitivity was significantly increased in the Ranch Hands (p-0.001). However, when these historic data were contrasted to both uroporphyrin and coproporphyrin abnormalities, no correlation was apparent, nor were there any significant group differences. Since an elevation in the uroporphyrin level is required for a diagnosis of PCT, the historic data were retabulated with only uroporphyrin abnormalities; again, no group differences were apparent, and, in fact, uroporphyrin abnormalities in both groups were higher in those participants without a history of skin disorders than in those participants with such a history. The likelihood of bona fide PCT among study participants, and particularly among the Ranch Hands, appears to be remote. In conclusion, the followup examination disclosed more statistically significant findings for tests of liver function than the Baseline examination, but they were equally divided between the two groups and did not demon- 19 �strate clinical, statistical, or exposure patterns consistent with an herbicide-related effect on health. No evidence was found to suggest an increased likelihood of PCT among the Ranch Hand group. DERMATOLOGY Interval questionnaire data on the occurrence, time, and location of acne were analyzed to assess the possible historical diagnosis of chloracne. No significant difference was observed between groups for reported occurrence of acne, although the Ranch Hand cohort reported slightly more acne. The occur-' rence of acne relative to 1961 was comparable between groups. A marginally significant difference in the occurrence of post-1961 acne was found, with more Ranch Hands than Comparisons reporting acne strictly post-SEA. The duration of post-31961 acne was not significantly different between the two groups. For participants with post-SEA acne, the spatial eyeglass distribution of acne (suggesting chloracne) was observed to be similar for the Ranch Hand and Comparison groups, both for individual sites and the combination of acne on the eyelids, ears, and temples. This analysis suggested that the occurrence of skin disease compatible with chloracne was not different in the two groups. Analyses of the followup physical examination data, as with the Baseline examination, placed primary emphasis on six dermatologic disorders: comedones, acneiform lesions, acneiform scars, inclusion cysts, depigmentation, and hyperpigmentation. Secondary emphasis was given to 16 other minor conditions (generally not associated with chloracne) recorded at the physical examination. No significant findings occurred in any variable, as reflected in Table 8.• No significant difference was found for any of these variables in the unadjusted analyses. The variable consisting of the 16 secondary conditions, labeled "other abnormalities," had the largest difference in the prevalence of abnormalities for the Ranch Hand cohort over the Comparison group (Est. RR: 1.08, 95$ C.I.: [0.92,1.28], p-0.3^9), but the difference was clearly nonsignificant. The covariate effects of age, race, occupation, and the presence of pre-SEA acne were often profound with respect to the recorded dermatologic conditions. The adjusted analyses closely mirrored the unadjusted analyses, with no significance noted between groups for any variable. Only one group-bycovariate interaction was observed in the adjusted'analysis of the dermatology index, with a group~by-presence of pre-SEA acne interaction noted. However, further analysis of this interaction did not show an adverse effect in the Ranch Hand group. Exposure index analyses did support dose-response relationships for some of the variables in certain occupational strata, but did not reveal a strong pattern of results suggesting a relationship between skin disease and herbicide exposure. Overall the followup examination results paralleled the Baseline findings. Although the followup examination detected more dermatologic abnormalities than those present at Baseline, slightly more abnormalities were found in 20 �TABLE 8. Overall Summary Results of Unadjusted and Adjusted Analyses of Questionnaire and Physical Examination Dermatologies! Variables Variable Unadjusted Adjusted Questionnaire Incidence of Acne Occurrence NS Relative to 1961 NS Relative to SEA (Post -1961 Cases) NS* Duration of Acne NS Location of Acne NS NS Physical Examination Comedones NS NS Acneifora Lesions NS NS Acne i form Scars NS NS Depigmentation NS NS Inclusion Cysts NS NS Hyperpigmentation NS NS Other Abnormalities NS NS Dermatology Index NS **** NS: Not significant (p>0.10). — Analyses not performed. NS*: Borderline significant (0.05<p<0.10). ****Group-by-covariate interaction. �the Comparisons, and moat relative risks approached unity. The longitudinal analysis for the dermatology index showed no statistically significant differences between groups in the change in results from the Baseline to the followup examination. In conclusion, none of the questionnaire results disclosed an increased likelihood of past chloracne in the Ranch Hands. The physical examination did not diagnose a current case of chloracne. The dermatological data were similar between the Ranch Hand and Comparison groups, and the longitudinal analy.sis of the dermatology index suggested equivalence between the Baseline and followup examination results. CARDIOVASCULAR SYSTEM The cardiovascular health of both cohorts was assessed by collection of reported and record-verified heart disease events; measurement of central cardiac function by systolic blood pressure, abnormal heart sounds, and electrocardiograph (ECG) findings; and evaluation of peripheral vascular function by diastolic blood pressure, funduscopic examination, presence of carotid bruits, and detailed manual and Doppler measurements of five peripheral pulses. Table 9 presents the overall summary of the unadjusted and adjusted results. Where possible, the analyses used the covariates of age, race, occupation, percent body fat, cholesterol, high density lipoprotein (HDL) cholesterol, cholesterol-HDL ratio, smoking history (packryears and current smoking level), alcohol history (drink-years and current drinking level), personality score, and differential cortisol. The cardiovascular variables did not reveal significant group differences, with the exception of verified heart disease, for which the proportions of recorded cardiac events were 24 and 20 percent in the Ranch Hand and Comparison groups, respectively, (p-0.054 unadjusted, p-0.036 adjusted). This finding was not reinforced by results of individual questionnaire or examination variables showing impairment in the Ranch Hands. There was a remarkable balance in relative risks above and below unity between the groups. Other related analyses showed an absence of significant group differences in reported or verified hypertension, reported or verified heart attacks, and reported heart disease. There was good correlation between the verified cardiovascular history'and the central and peripheral cardiovascular abnormalities detected at the physical examination, supporting accuracy and valid-' ity of .the cardiovascular measurements. The adjusted analyses of central cardiac function disclosed a significant group^-byrage interaction involving systolic blood pressure in the Black cohort, with a mean systolic blood pressure greater in the Ranch Hands than the Comparisons at younger age levels, but a lower mean pressure at the older ages; the group-by-age interaction was not significant in the nonblack cohort. Additionally, there was a significant group-*by-pack-years of smoking interaction for the overall ECG findings, and significant group-by-pack-years of smoking and group^-by?-percent body fat interactions for arrhythmia, but they all generally pointed to lower adjusted relative risks in the Ranch Hands. 22 �TABLE 9. Overall Summary Results of Unadjusted and Adjusted Analyses Cardiovascular'Variables Variable Statistical/ Clinical Analysis Unadjusted Adjusted NS NS MS' Historical and Verified Heart Disease Reported Verified Reported Verified Reported Verified Hypertension Hypertension Heart Disease* Heart Disease* Heart Attack Heart Attack NS NS NS* NS NS NS NS Central Cardiac Function Systolic Blood Pressure Discrete Continuous Heart Sounds Electrocardiogram (Overall) ECG: RBBB ECG: LBBB ECG: Nonspecific T-Vave Changes ECG: Bradycardia ECG: Tachycardia ECG: Arrhythmia ECG: Other Diagnoses NS NS NS NS NS NS NS NS NS 23 NS **** NS **** NS N/A NS NS N/A **** NS �TABLE 9. (Continued) Overall Sunary Results of Unadjusted and Adjusted Analyses Cardiovascular Variables Statistical/ Clinical Analysis Variable Unadjusted Adjusted Funduscopic Examination NS NS NS NS NS NS Carotid Bruits NS NS ' Peripheral Vascular Function Oiastolic Blood Pressure Radial Pulses Femoral Pulses Popliteal Pulses Oorsalis Pedis Pulses Posterior Tibial Pulses Leg Pulses Peripheral Pulses All Pulses Discrete Continuous Manual Doppler Manual Doppler Manual Doppler Manual Doppler Manual Doppler Manual Doppler Manual Doppler Manual Doppler NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS:Not significant (p>0.10). NS*:Borderline significant (0.05<p<0.10). ****Group-by-covariate interaction. 'Excluding hypertension. b RH>C (Est. RR: 1.25; 95* C.I.:. (1.02, 1.54], p-0.054). NS NS NS NS **** NS **** NS **** NS **** NS **** NS **** NS �In the analysis of peripheral vascular function, no significant group differences were observed for abnormalities involving radial, femoral, popliteal, posterior tibial, dorsalis pedis, or three anatomic aggregates of these pulses, either by manual palpation or Doppler techniques. This overall f i n d i n g was in distinct contrast to the 1982 Baseline examination, which by the manual palpation method, showed significant peripheral pulse deficits in the Ranch Hands. This favorable pulse reversal over the two examinations is primarily attributed to the rigid 4-hour tobacco abstinence applied prior to Doppler testing, although other factors may be related. The lack of group differences for pulse abnormalities was noted even though the manual and Doppler techniques differed significantly (p<0,05, p<0.001 for most) in the detection of abnormalities for all but one of the pulses or pulse combinations. For manually-determined pulse abnormalities, there was a significant group-by-race interaction for the popliteal pulses, a significant groupby-percent body fat interaction for the leg pulses, and significant groupby-occupation interactions for the posterior tibial, dorsalis pedis, and the three pulse aggregates (leg, peripheral, and all pulses). No interactions were encountered in the adjusted analyses of the Doppler results, and none showed significant group differences. Statistical analyses involving the Original Comparisons also showed no significant differences in the cardiovascular measurements between groups, although slightly different interactions were detected in some of the adjusted analyses. For the exposure analyses, the only statistically significant effects were those pointing to less bradycardia and less reported and verified heart disease in the medium exposure level category, as contrasted to the low exposure category, among the enlisted groundcrew. In many cases there were too few abnormalities within the occupational categories to permit formal statistical tests. Overall, the exposure analyses were deemed as unsupportive of any meaningful dose-response relationships. The longitudinal analysis of the pulse index confirmed the significant difference in the change in the pattern of results from the Baseline examination to the followup examination, largely due to a relatively greater increase of pulse abnormalities in the Comparison group than in the Ranch Hand group. There was no significant change in pattern between the two groups in overall ECG findings between examinations. There was a similar distribution of the covariates between groups, except for a slightly higher level of current Ranch Hand smoking (also observed at Baseline), and a corresponding slightly lower mean percent body fat. The general covariate effects were strong and showed expected, classical associations with the cardiovascular measurements. However, unexpected effects were consistently noted for personality score, with higher proportions of various cardiovascular abnormalities associated with scores in the Type B direction, a. finding possibly attributable to the method of personality determination. Although smoking was positively associated with many of the cardiovascular measurements, negative associations were seen between current smoking and reported and verified essential hypertension and between pack-years of smoking and verified hypertension. 25 �In conclusion, of 27 cardiovascular variables, only one, verified heart disease, showed a significant excess in the Ranch Hands, but this finding was largely unsupported by other cardiac measurements. Both manual palpation and Doppler recordings of five peripheral pulses were similar in both groups, in marked contrast to the 1982 Baseline examination which found significant pulse deficits in the Ranch Hand group. This change at the followup examination was most likely due to required tobacco abstinence prior to the pulse measurements. Exposure index analyses did not support a consistent dose-response relationship for any variable. Overall, there was remarkable similarity in the cardiovascular health between the Ranch Hand and Comparison groups. HEMATOLOGY The functional integrity of the hematopoietic system was assessed by the measurement of eight peripheral blood variables: red blood cell count (RBC), white blood cell count (WBC), hemoglobin (HGB), hematocrit (HCT), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), and platelet count (PLT). These variables were analyzed in the discrete form to detect differences in the percentages of values outside the designated laboratory range, as well as in the continuous form to detect shifts in mean values between the two groups. A summary of all of these analyses, unadjusted and adjusted for the covariates of age, race, occupation, and smoking, is presented in Table 10. The unadjusted discrete analysis of the percent abnormal values, both low and high, showed no statistically significant differences between the Ranch Hand and Comparison groups for any of the hematological variables. Similarly, the adjusted categorical analysis disclosed that none of the adjusted relative risks was significant for either group, and that no significant group-by covariate interactions were present. The unadjusted continuous analysis did not detect any significant differences in group means for any of the eight variables. The adjusted continuous analysis found no significant group differences for HGB, HCT, MCV, MCH, and MCHC, but encountered significant three^factor interactions for WBC (group^ by-race-by-age, group-by-age-by-smoking history, and group~by-race-byoccupation), for PLT (group-byrace-by-smoking history and group^by^raceby-current level of smoking), and a borderline interaction for RBC (group?by-occupation-by-smoking history). Ranch Hand versus Original Comparison analyses revealed further significant interactions for HGB, HCT, MCV, and MCH. As no group strata demonstrated consistent patterns of hematologic impairment, biologic relevance was not assigned to the interactions. The covariate effects of age, race, occupation, and smoking history were highly significant for many of the hematological variables. The effect of race was particularly profound for all variables except PLT. There was fair consistency in the covariate effects upon the RBC-related variables. Generally, decreasing hematologic values were associated with increasing age and the Black race, and increasing hematologic values were associated with increasing smoking. The detection of these classical covariate effects lends credence to the overall finding of nonsignificant group differences for all of the hematological variables. Significant group differences found for MCV and MCH at the Baseline examination were not significant at the first followup. Other differences (e.g., covariate effects, interactions) 26 �TABLE 10. Overall SUBMIT? Results of Unadjusted and Adjusted Analyses of Heaatologlcal Variables Unadjusted Adjusted Mean Categorical Mean Categorical RBC NS NS NS* NS VBC NS NS **** NS HGB NS NS NS NS HCT NS NS NS NS HCV NS NS NS NS HCH NS NS NS NS HCHC NS PLT NS NS — NS *_-* • J WwWK — NS NS: Not sgnifleant (p>0.10). NS*: Borderline significant group-by-covariate interaction (0.05<p<0.10). —Analysis not performed due to sparse data. ****Group-by-covariate interaction. Note: Significant group-by-covariate interaction, Ranch Hands versus Original Comparisons only, for HGB, HCT, MCV, and MCH. 27 �between the Baseline and followup examinations may be due to small numeric shifts in the cohorts under study (see Chapter 2) and the selection of alternate statistical models, or due to chance. Unadjusted continuous exposure analyses in the Ranch Hand group revealed only one significant effect (RBC in enlisted groundcrew) and one borderline effect (HCT in enlisted groundcrew) but neither was consistent with a plausible dose-response relationship. The adjusted continuous exposure analyses found only one significant contrast (HCT, medium exposure versus low exposure, enlisted groundcrew). However, seven exposure level-by-covariate interactions were noted for four of the hematological variables. Discrete outcome analyses of the exposure level index revealed a significant result only for WBC in the enlisted flyers. The longitudinal analyses of MCV, MCH, and PLT found significant differences only for PLT values between the Baseline and followup examinations, with the Baseline group difference-in mean values closing to near equivalence at the followup examination. In conclusion, none of the eight hematological variables were found to differ significantly between the Ranch Hand and Comparison groups. In fact, group equivalence was more apparent at the followup examination than at the Baseline examination. The classical effects of age, race, and smoking were demonstrated with most of the hematological variables. The longitudinal analyses also suggested that neither group manifested an impairment of the hematopoietic system. Exposure index analyses did not support a plausible dose-response relationship for any of the hematological variables. RENAL A summary of all renal variables, including unadjusted and adjusted analyses, is displayed in Table 11. A historical assessment of kidney disease/kidney stones by a reviewof-systems questionnaire showed no significant differences between the Ranch Hand and Comparison groups (see Table 12). An adjusted analysis did not alter this conclusion as an adjusted relative risk of 0.95 (95$ C.I.: [0.71,1.25], p-0.693) was demonstrated. These statistics appeared to be'in marked contrast to the Baseline historical findings. Differences vis-a-'vis the Baseline were most likely due to a difference in questionnaire techniques. Current renal function was evaluated by five laboratory variables: urine protein, occult blood, urine, white blood cell counts (WBC's), blood urea nitrogen (BUN), and urine specific gravity. Invasive procedures were not used. The unadjusted analysis of proteinuria showed no group differences (Est. RR: 1.18, 95% C.I.: [0.75,1.86], p-0.485), but the adjusted analysis showed an interaction of group and diabetic class; appropriate stratified analyses revealed that the prevalence of proteinuria was lower in the Ranch Hands than in the Comparisons in the diabetic and impaired strata, but higher in the normal strata for the Ranch Hands. These results were in contrast to the 28 �TABLE 11. Overall Summary Results of Unadjusted and Adjusted Analyses for Renal Variables Variable Unadjusted Adjusted Reported Kidney Disease NS **** Urinary Occult Blood NS **** Urinary Leukocytosis NS **** BUN NS **** Urine Specific Gravity NS*: NS Urinary P r o t e i n NS: NS NS* **** Not s i g n i f i c a n t (p>0.10). Borderline s i g n i f i c a n t (0.05<p<0.10). ****Group-by-covariate interaction. TABLE 12. Unadjusted Analysis of History of Kidney Disease/Kidney Stones by Group History of Kidney Disease/Stones No Yes Group Ranch Hand Number 94 Percent 9.3 Percent Total 920 90.7 1,163 90.1 1,291 Est. Relative Risk (95% C.I.) p-Value 1,014 Number 0.93 (0.70,1.23) 0.619 Comparison 128 9.9 29. �Baseline findings, which showed a marginally significant proteinuria in the Comparison group (p-0.055), and overall, lower prevalence rates of proteinuria. The unadjusted prevalence rates for hematuria were similar for both groups (Est. RR: 1.14, 95$ C.I.: [0.91,1.42], p-0.239). Three significant interactions involving group membership and covariates precluded a direct adjusted comparison of the estimated prevalence rates. Covariate analyses indicated increased hematuria in Blacks and among enlisted personnel. Ultimately via a series of stratified analyses, statistical equivalence was determined for the Black enlisted strata of both groups. Of particular note was the approximate tenfold increase in hematuria in both groups over that observed at Baseline, a finding most likely due to different laboratory techniques (reagent-strip testing versus microscopic observation). Similar results were found for leukocyturia, i.e., a nonsignificant unadjusted analysis (Est. RR: 1.24, 95% C.I.: [0.93,1.64], p-0.145), and a. significant three-way interaction (group, age, race) in the adjusted analysis. Significant covariate effects were noted for diabetic class and occupation for nonblack participants, whereas age was a significant adjusting variable for Blacks. A significant group difference was found only for the younger, nonblack Ranch Hands. The overall results were consistent with the Baseline findings. BUN levels did not vary significantly by group (p-0.554, unadjusted). Adjusted analyses showed significant covariate effects for age and occupation and interactions for group and race and for race and diabetic class. An analysis stratified by race revealed no significant group differences for nonblacks, but a significantly higher adjusted mean BUN level in Black Comparisons than in Black Ranch Hands. Overall, the BUN results were similar to those observed at the Baseline examination. Urine specific gravity levels manifested marginally significant group differences (p-0.082, unadjusted). The adjusted analysis disclosed significant covariate effects of diabetic class and the interactions of group and race and group and occupation. Analyses by race showed no strata with significantly lower mean levels for Ranch hands. In contrast to the Baseline values, the followup urine specific gravities were lower, a finding most likely attributable to differences in laboratory methodology (falling drop method versus multistick procedure). Exposure index relationship at the prevalence rates or within occupational analyses showed very little evidence of a dose-response followup examination. No patterns in the relationship of mean levels among the exposure index levels were seen strata. The longitudinal analysis was based solely upon a contrast of BUN levels between the two examinations. The unadjusted mean BUN value increased slightly from the Baseline to the followup examination, but the increases were symmetrical in the two groups and nonsignificant (p-0.48). In conclusion, none of the six renal assessment variables showed a significant difference between the Ranch Hand and Comparison groups by unadjusted tests. However, in the adjusted analyses, all renal measurements except 30 �reported kidney disease revealed group-by-covariate interactions. These interactions were often complex, making it impossible to reach a firm conclusion as to the presence of an herbicide effect. ENDOCRINOLOGY Questionnaire and review-of-systems data for past thyroid disease were essentially equivalent in both the Ranch Hand and Comparison groups. These historical data were confirmed by medical record reviews. Physical examination findings were necessarily limited to data from palpation of thyroid glands and testicles; the unadjusted results showed no significant group differences. TABLE 13. Medical Record Verification Results of Reported Thyroid Disease by Group Group Verification Status Ranch Hand Comparison Number vith Reported Thyroid Conditions 7 21 Medical Records Reviewed 7 21 Medical Records Pending 0 0 100 100 Percent Thyroid Conditions Verified The physical examination and laboratory testing results of all endocrino-logical variables are summarized in Table HJ. Evaluation of the endocrine system was conducted primarily by laboratory testing of hormone levels. The thyroid test battery consisted of T^ % Uptake and TSH assays. The T3 $ Uptake data showed no group differences for either mean values or frequency of abnormally low or high values. Occupation was a significant covariate. TSH results revealed a significantly higher mean level in the Ranch Hand group, but this difference was not found by categorical testing of proportions of abnormally high TSH results. Mean levels of testosterone were significantly elevated among Ranch Hands as contrasted with Comparisons in the 10 to 25 percent body fat category, but this was not reflected by the categorical tests. For the few participants 31 �TABLE 14. Overall Summary Results of Unadjusted and Adjusted Continuous and Categorical Analyses of Bndocrinological Variables Unadjusted Test Mean Categorical Mean Questionnaire and Physical Examination * Past Thyroid Disease (SelfAdministered) • Past Thyroid Disease (Interviewer Administered) • Thyroid Abnormalities • Testicular Abnormalities Laboratory Testing T, Z Uptake NS TSH Testosterone Initial Cortisol Adjusted Categorical • b • :b NS • b NS • b NS NS* Overall: NS Lov vs. Normal: NS High vs. Normal: NS 0.019 NS 0.03S Overall: NS Lov vs. Normal: NS High vs. Normal: NS NS Overall: NS Lov vs. Normal: NS High vs. Normal: NS NS NS « NS —b Overall: NS Lov vs. Normal: NS High vs. Normal: NS 0.025 NS **** Overall: NS Lov vs. Normal: NS High vs. Normal: NS NS b b 2-Hour Cortisol * * * * —* Differential Cortisol Overall: 0.034 2-Hour Postprandial NS Overall: 0.038 NS Glucose Impaired vs. Normal: 0.024 Impaired vs. Normal: 0.022 Diabetic vs. Normal: NS Diabetic vs. Normal: NS • NS Diabetes (Composite NS Indicator) —'Analysis not feasible. NS: Not significant (p>0.10). — Analysis not performed. NS*: Borderline significant (0.05<p<0.10). ****Group-by-covariate interaction. 32 �with less than 10 percent body fat (six Ranch Hands, four Comparisons), mean testosterone levels were lower for Ranch hands than for Comparisons. Age, occupation, and percent body fat were significant adjusting variables. Two timed cortisol specimens showed no significant group differences in mean values and percent abnormalities. The difference between the timed cortisol results, termed the differential cortisol, showed no significant group differences for nonblacks or Blacks born before 1942, but Black Ranch Hands born in or after 1942 had a lower mean differential cortisol level than Comparisons. Age, percent body fat, and personality type were significant . covariates in these analyses. Group means of 2-hpur postprandial glucose levels were not statistically different, but categorical testing revealed that there was a significantly higher frequency of glucose-impaired (at least 140 but less than 200 mg/dl) Comparisons than Ranch Hands. A constructed variable comprised of known diabetics and individuals classified as diabetic by the glucose tolerance test, showed no difference between the Ranch Hand and Comparison groups. As expected, past and current diabetes were highly influenced by the covariates age, race, and percent body fat. Exposure index analyses did not reveal any pattern consistent with a dose^response relationship. Enlisted flyers in the medium exposure level were significantly different from those in the low exposure level for 2-hour cortisol, differential cortisol, and 2«hour postprandial glucose. However, the corresponding high versus low contrasts were not statistically significant. Longitudinal analyses of T, % Uptake, TSH, and testosterone levels on all individuals attending both the Baseline and followup examinations revealed only symmetrical and nonsignificant changes in the Ranch Hand and Comparison groups in the interval between examinations. In conclusion, both limited historical and physical examination data, seven endocrinological laboratory variables, and a composite indicator of diabetes did not demonstrate consistent patterns indicating an herbicide effect. However, there was a significant interaction between group and percent body fat for testosterone that could be interpreted as an herbicide effect. TSH and testosterone means tests were statistically significant, and in the expected direction of an herbicide effect, but these results were not confirmed by categorical testing. Also significant was the impaired category of the glucose tolerance test, which showed an excess in the Comparison group. The consistent demonstration of the classical effects of the covariates age, ' race, occupation, and percent body fat on appropriate endocrine variables provided support for these conclusions. Overall, the endocrine health status of both groups was reasonably comparable. IMMUNOLOGY Immunologic competence was measured by cell surface marker (phenotypic) studies and cell stimulation studies on 47 percent of the study population, and by a four antigen series of skin tests in 76 percent of participants to assess the delayed hypersensitivity response. Table 15 summarizes the results of all unadjusted and adjusted analyses on 11 primary variables spanning the first two of these three functional areas. 33 �TABLE 15. Overall Suaaary Results of Unadjusted and Adjusted Analyses of Isaunologlcal Variables Variable Unadjusted Adjusted Total T Cells (T ) NS A•^ ^ WwWn Helper T Cells ( , T) Suppressor T Cells ( , T) B Cells Monocytes HLA-DR Cells T4/T, Ratio NS NS NS NS NS NS NS NS **** **** **** NS Unst initiated Response (PHA) PHA Net Response Pokeveed Net Response HLC Net Response NS NS NS NS NS NS NS **** NS:Not significant (p>0.10). ****Significant group-by-covariate interaction. Cell surface marker studies were conducted for total T cells (T^), helper T cells (T^), suppressor T cells (Tg), B cells, monocytes, and HLA-DR cells; the ratio of TVTg cells was included in the analysis. Because of inherent significant day«to»day and batch*to-batch variation, all results (including functional stimulation studies) were adjusted for blood-draw day variation. Statistical testing of the seven phenotypic cell markers did not reveal any significant group differences (interactions excepted), either unadjusted or adjusted for the covariates of age, race, occupation, current smoking, lifetime smoking history (pack^-years), current alcohol use, or lifetime alcohol use (drink-years). Similarly, none of the unadjusted or adjusted analyses of the functional stimulation studies (for phytohemagglutinin, pokeweed mitogen, or mixed lymphocyte culture) showed any statistically significant group differences. However, the adjusted analyses for total T cells, B cells, monocytes, HLA^DR cells, pokeweed mitogen, and net mixed lymphocyte culture stimulation showed some significant group-by-covariate interactions, precluding direct adjusted group contrasts. Overall, no discernible pattern was identified to suggest a detriment in any subgroup of either the Ranch Hands or Comparisons. Results were similar between the analyses of the total Comparison group and'the analyses of the Original Comparisons. The covariate effects of age, race, smoking, and alcohol use were generally profound on most variables in the phenotypic and stimulation studies. Consistently decreasing values of all cell markers and stimulated cells were associated with increasing age, whereas Increased levels of smoking were usually associated with increases in the values of those variables. Blacks �had consistently higher stimulated cell counts than nonblacks, but this effect was not observed for counts of T cells, B cells, or HLA-DR cells. Enlisted personnel generally had higher cell surface marker counts than officers. Exposure index analyses of cell surface markers revealed no pattern consistent with a dose-response relationship. For enlisted groundcrew, the mean total T cell and suppressor T cell counts for the medium exposure level were significantly lower than those of the low exposure level, but were slightly lower than those of the high exposure level. The exposure index analyses of the functional stimulation tests revealed no consistent significant dose;-response patterns for net PHA counts or net MLC counts. For net pokeweed counts, enlisted flyers in the high exposure level had a significantly lower adjusted count than enlisted flyers in the low exposure level, and a decreasing trend was apparent. The delayed hypersensitivity response was assessed by the skin test antigens of mumps, Candida alb leans, Trichophyton, and staph-'phage-lysate. The JtSrhour measurements of skin induration and erythema for the four test's showed marked inter-reader variation. Analyses showed that one of the three skin test readers too often measured induration larger than erythema (a clinically unacceptable finding), in an average of 30 percent of the readings, and did not yield measurements that detected a case of possible or overt anergy, whereas the other two readers found this condition in 5.6 percent of the participants. Remaining data from Readers 1 and 3, however, were found to vary significantly in clinical interpretation over duration of the examination. Consequently, all skin test data were declared invalid, and were not used'in the assessment of group differences. The skin test reading problems led to the use of additional clinical quality control procedures for the AFHS followup examination begun in May 1987. In conclusion, no significant group differences were judged present for the comprehensive cell surface marker or functional stimulation studies. The profound effects of age, smoking, and alcohol use were observed in these immunologic tests. The assessment of delayed hypersensitivity skin responses was precluded by poor data quality and excluded from further analysis. Overall, there was no indication of impaired immunologic competence in either group. RESPIRATORY SYSTEM A summary of the results on the analyses of reported history of respiratory illness and of radiological and clinical findings is given in Table 16. Based on the 31 December 1986 mortality data, there were seven deaths from respiratory conditions in the Comparison group and none in the Ranch Hand group. 35 �TABLE 16. Overall Summary Results of Unadjusted and Adjusted Analyses of Pulmonary Disease Pulmonary Disease Unadjusted Adjusted NS NS NS NS NS NS NS Reported History of Respiratory Illness Asthma Bronchitis Pleurisy Pneumonia Tuberculosis **** NS **** Radiological and Clinical Findings Thorax and Lungs Asymmetrical Expiration Hyperresonance Dullness Wheezes Rales X Ray NS NS NS NS NS NS NS NS NS NS NS NS **** NS NS: Not significant (p>0.10) ****Group-by-covariate interaction. There were no group differences found for reported history of asthma bronchitis, pleurisy, or tuberculosis based on the unadjusted analyses. Adjustments for age and lifetime smoking did not alter the findings of group similarity, although there was a significant group-bypack-year interaction for pleurisy and for tuberculosis. Similarly, there were no significant group differences in the unadjusted analyses for the radiological and clinical respiratory findings of thorax and lungs, asymmetrical expiration, hyperresonance, dullness, wheezes, rales and xrray interpretations. These findings were supported by the adjusted analyses, although there was a group-by-age interaction for rales. The exposure index analyses revealed no consistent dose^response pattern. Analyses of past history of respiratory illness and the clinical and radiological examination of the chest and lungs did not reveal any statistically significant differences between the Ranch Hand and Comparison groups suggestive of herbicide related disease. Several group^byrcovariate interactions did exhibit statistical significance, but these findings did not indicate any consistent patterns suggesting different disease experience in the two groups. 36 �CONCLUSIONS This chapter summarizes the conclusions drawn from the statistical analyses that have been conducted on the Air Force Health Study data base. The followup study, which began in 1985, was the logical extension of the 1982 Baseline, building upon the strengths of the Baseline study and utilizing the data collected at both the Baseline and the followup. The high level of Government support and outstanding participation of the study subjects that characterized the Baseline study were maintained through this first followup. STUDY PERFORMANCE ASPECTS Of the living Baseline study participants, 99.2 percent were located and asked to participate in the followup. Participation in the followup physical examination and questionnaire was very high. Of the fully compliant Baseline participants, 971 of the 1,045 Ranch Hands (92.9$) and 1,139 of the 1,224 . Comparisons (93.1$).participated in the followup. Thus, there was no group difference in compliance of the Baseline participants at the followup. Overall, the 2,309 participants in the followup (1,016 Ranch Hands and 1,293 Comparisons) represented a loss of 159 individuals and .a gain of 199 since Baseline. One percent of the fully compliant Baseline population died between 1982 and the 1985 followup examination. The bias/compliance analyses suggested that there had been no change between Baseline and the followup in the way replacements volunteered for entry into the study, and that no additional bias had been introduced at the followup due to scheduling differences. Although replacements were not health-matched at Baseline as they were at the followup, they were similar to refusals with respect to reported health, medication use, and income level. This result supported the conclusion that there has been little, if any, selection bias due to differential participation in the Comparison group and supported the use of the total Comparison group in the main analyses presented in this report. POPULATION CHARACTERISTICS Overall, the Ranch Hands and Comparisons reported similar social and behavioral characteristics. No significant differences were found in age, educational background, religious preference, current military status, and income level. Significantly more Ranch Hands smoked cigarettes at the time of the followup'examination than did Comparisons, but there was no significant difference between groups on past cigarette, cigar, and pipe use and on recent^ and past use of marijuana. A much higher percentage of participants reported past marijuana use at the followup than at Baseline. This difference was most likely due to a greater level of confidentiality afforded by the questionnaire technique. Risk taking behavior, assessed by questions on potentially dangerous recreational activities, revealed borderline significance. Slightly more Comparisons were scuba divers and more Ranch Hands raced motor vehicles. The difference in scuba diving was also significant at Baseline. 37 �Patterns of Results Both the chapter conclusions and the final conclusions of this report have been predicated upon concepts of consistency, specificity, coherence, strength, and plausibility as they apply to the interpretation of group differences. In particular, careful consideration has been given to a variety of data and patterns of results that have emerged from the clinical evaluations. Specifically, there were few differences in the proportions of abnormalities between groups; the positive associations have not aggregated in the clinical areas of prime dioxin concern, nor have they been of serious clinical importance; the unadjusted results have been remarkably concordant with the adjusted results, both in. terms of relative risk and p value; the analyses using the Original Comparison set have largely mirrored the results found with the total Comparison group; many of the group differences noted at Baseline have disappeared at the followup examination, and only a few new associations have emerged; almost all of the covariates have acted as expected in the adjusted analyses; and the exposure index analyses and the group-by-covariate interactions have not demonstrated biological patterns of concern and appeared to'be more likely due to chance than not. Due to the acknowledged limitations of the exposure index used in this report (and considering the potential use of dioxin body burden levels at the next followup), dose-response relationships have not been emphasized in reaching final conclusions. The overall pattern of these findings indicates that this followup study cannot be viewed as alarming from the traditional perspectives of clinical medicine or epidemiology. This study, in fact, demonstrates similarity in current health status between the Ranch Hand and Comparison groups. CLINICAL ASPECTS General Health The nonspecific assessment of general health showed relatively close similarity between the two groups. Ranch Hands rated their health as fair or poor more frequently, but this difference was found only in the enlisted groundcrew and not in the officers nor enlisted flyers. The perception of health in both groups had improved since Baseline. Physician-rated appearance of relative age was not found to be significantly different at the followup in contrast to the Baseline finding that a higher percent of Ranch Hands than Comparisons looked younger than their stated age. The categorical analysis of sedimentation rate showed that the Ranch Hands had more abnormalities than the Comparisons. These results were not supported by the continuous analysis of mean sedimentation rates and were opposite to the Baseline results, which showed that younger Comparisons had elevated sedimentation rates. The categorical analysis of percent body fat showed no significant differences between the two groups, which was consistent with Baseline. However, the continuous analysis found that the Ranch Hands had a significantly lower mean percent body fat using age, race, and occupation as covariates. The detailed exposure analyses revealed no consistent exposure effects, and this result was consistent with the Baseline analysis. No longitudinal difference was found on perception of health. A significant group difference was found over time for the longitudinal analysis of sedimentation rate due to the change in the findings between the two examinations, possibly related to a change in laboratory methodology. 38 �Malignancy Skin and systemic cancers, both suspected and verified by medical records, showed no significant group differences for the Baselinerfollowup interval (1982-1985). However, for all neoplasms combined (malignant, benign, and uncertain), a borderline significant excess in the Ranch Hand group was noted in an unadjusted analysis. The analyses of interval cancers revealed group interactions for verified and verified plus suspected basal cell carcinoma and verified plus suspected systemic cancers. Nonsignificant findings were observed for verified and verified plus suspected sun exposure-related cancers. Verified systemic cancers did not differ significantly between groups. The analyses of lifetime cancer found significant results for verified basal cell carcinoma and verified sun exposures-related skin cancers. Group interactions were noted for systemic cancer categories and for verified plus suspected basal cell carcinoma. The higher rate of basal cell carcinoma in the -Ranch Hands versus the Comparisons found at Baseline was nonsignificant for the followup interval, but due to the effect of the larger number of Baseline cases and the significant confounding of average residential latitude, the adjusted analysis of lifetime basal cell carcinoma emerged as statistically significant. There were several disparities in the distribution of testicular, colon, and smoking-related tumors in the groups. Further, one case of soft tissue sarcoma and one possible lymphoma (both in Ranch Hands) were diagnosed in the interval, balancing the two similar cases found in the Comparison group at Baseline. Considering that the systemic cancer curves are in their early stages for both groups, with perhaps insufficient latency, the cancer results of the followup examination should not be viewed as disturbing, but as cause for continued monitoring. Neurological Assessment None of the 27 neurological variables demonstrated a significant group difference, although several variables had relative risks which were greater than one. There was no group difference in reported neurological illnesses for the interval or for a lifetime history. Of the cranial nerve variables, speech and tongue position were marginally significant, with the Ranch Hands at a slight detriment. The analyses of peripheral nerve function showed no significant differences between the Ranch Hands and the Comparisons. In the analysis of central nervous system function, hand tremor was found to be of borderline significance, with the Ranch Hands faring slightly worse than the Comparisons. A borderline significant group interaction (Ranch Hand hand tremor by insecticide exposure) may have had biological and operational significance. Overall, substantially fewer neurological abnormalities were detected at the followup examination than at the Baseline examination. The exposure analyses showed only occasional statistically significant results, although no consistent pattern with increasing exposure was evident. In the longitudinal analysis of the Babinski reflex, a significant change over time was observed. This was due to a nonsignificant finding in the Ranch Hands at the followup, which differed from the significant adverse finding at Baseline. The covariates of age, alcohol history, and diabetes showed classical effects with many of the neurological measurements. Overall, the followup examination results were quite similar to the Baseline findings. 39 �Psychological Assessment The reported and verified data on lifetime psychological illnesses showed no significant differences between groups. Distributional tests of the 14 Minnesota Multiphasic Personality Inventory (MMPI) scales, stratified by occupation, revealed that only 2 of the 42 results approached significance. For the total Cornell Medical Index (CMI), separate distributional tests were conducted with stratification by age, race, occupation, education, and current drinking status; a significant difference was found for one statum of each of the covariates. In all cases, the mean of the Ranch Hand distribution was greater than the mean of the Comparisons. The analysis of the 14 MMPI scales showed that there was a significant difference between the two groups for denial and masculinity/femininity, with more abnormalities in the Comparisons than the Ranch Hands. The results of the analyses for hysteria were of borderline significance, with more abnormalities in the Ranch Hands. There were more abnormalities in the Ranch Hands than the Comparisons for social introversion, which was of borderline significance. Differences in the to.tal CMI 'and A.-H area subscore were found to be significant, with more abnormalities in the Ranch Hands. There was no significant difference between the two groups on the Halstead-^Reitan Battery impairment index, a measure of the functional integrity of the CNS. The exposure index analyses did not reveal any pattern consistent with a dose-response relationship. As expected, the effects of age, educational level, and alcoholic history showed profound effects on many of the psychological measurements. Gastrointestinal Assessment Although the followup gastrointestinal assessment disclosed more statistically significant findings than the Baseline examination, the abnormalities were distributed equally between the two groups, and there was no clinical, statistical, or exposure pattern consistent with an herbicide-related effect on health. No historical or biochemical evidence was found to suggest an increased likelihood of porphyria cutanea tarda (PCT) in the Ranch Hand group. Only sparse and nonsignificant liver disorders were reported for the interval between Baseline and followup. Also, for the lifetime history of liver disorders, there were no significant differences between groups. Further, there were no significant group differences in reported lifetime peptic ulcer disease. A review of digestive system mortality showed a relative excess in the Ranch Hands but a relative lack of malignant neoplasms. The results of the physical examination showed a borderline increase'of hepatomegaly in the Ranch Hand group. There was a significantly lower mean serum glutamic-pyruvio transminase (SGPT) level, a greater mean alkaline phosphatase level, and a lower mean uroporphyrin level in the Ranch Hand group. The analysis of coproporphyrin was of borderline significance, with the mean of the Ranch Hands in excess of the mean of the Comparisons. No group differences were found for serum glutamic-oxaloacetic transminase (SCOT), gamma^glutamyl transpeptidase (GGTP), total and direct bilirubin, lactic dehydrogenase (LDH), cholesterol, or triglycerides. The numerous grouprbycovariate interactions did not disclose any consistent subgroup patterns detrimental to the Ranch Hands. These findings were generally consistent with the results of the 1982 assessment. The longitudinal analyses for SCOT, SGPT, and GGTP showed no significant differences between results by group over time. 40 �Dermatologies! Evaluation No significant group differences were identified in the dermatological evaluation. None of the questionnaire data showed an increased likelihood of past chloracne, as determined by anatomic patterns of acne, and no cases were diagnosed in the physical examination. Analyses were conducted on six dermatologic disorders (comedones, acneiform lesions, acneiform scars, inclusion cysts, depigmentation, and hyperplgmentation) and on a composite variable of 16 other minor conditions (the latter not generally associated with chloracne). Exposure index analyses did not reveal consistent patterns suggestive of a'dose-response relationship. The longitudinal analysis, based on a composite dermatology index, showed no significant differences between the results over time. Substantially more dermatologic abnormalities were detected at the followup examination than at the Baseline examination. In general, however, the followup results were consistent with the findings at Baseline. Cardiovascular Evaluation Overall there was remarkable similarity in the cardiovascular health of the Ranch Hands and the Comparisons. Of the 27 cardiovascular variables, there was a significant difference for only one, verified heart disease, with an excess in the Ranch Hand group. This finding was largely unsupported by other cardiac measurements. The cardiovascular assessment was based on reported and verified heart disease; the measurement of central cardiac function by systolic blood pressure, abnormal heart sounds, and EGG findings; and the evaluation of peripheral vascular function by diastolic blood pressure, funduscopic examination, presence of carotid bruits, and detailed manual and Doppler measurements of five peripheral pulses. Doppler recordings of five peripheral pulses were similar in both groups,, a finding which was in marked contrast to the Baseline examination that found significant pulse deficits in the Ranch Hand group. This change was most likely due to a required 4-hour abstinence from tobacco prior to the pulse measurements. Overall, the exposure analyses were unsupportive of any meaningful doseiresponse relationship. The longitudinal analyses confirmed the change in pulse abnormalities in the Ranch Hand group over time, but showed no significant group change in overall EGG findings between the examinations. Hematological Evaluation The hematological evaluation found that neither group manifested an impairment of the hematopoietic system, consistent with similar findings at the Baseline. The evaluation was based on eight peripheral blood variables: red blood cells (RBC), white blood cells (WBC), hemoglobin (HGB), hematocrit concentration (HCT), corpuscular volume (MCV), corpuscular hemoglobin (MCH), corpuscular hemoglobin concentration (MCHC), and platelet count (PLT). Both the discrete and categorical analyses revealed no significant group differences. The covariate effects of age, race, occupation, and smoking history were highly significant for many of the variables. Two group-iby-covariate interactions in the analyses of mean differences did not appear to have a meaningful interpretation. The exposure index analyses did not support any plausible dose'-response relationship. The longitudinal �analyses of MCV, MCH, and PLT found significant differences only for PLT between the Baseline and the followup, with the Ranch Hands exhibiting a slight decline in mean level from Baseline and the Comparisons showing an opposite change. Renal Assessment None of the six renal variables of reported kidney disease, urine protein, occult blood, urine white blood cell count, blood urea nitrogen, and urine specific gravity showed a significant difference between the two groups based on the unadjusted analyses. In the adjusted analyses of the laboratory variables, however, there were significant group-by-covariate interactions that did not yield a consistent pattern to suggest a renal detriment to either group. The finding of group equivalence for past kidney disease was in contrast to the Baseline examination, which found significantly more reported disease in the Ranch Hand group. The difference in findings is more likely due to a change in questionnaire wording than to a true change in renal health. Like the Baseline findings, the exposure index analyses showed very little evidence of a dose-response relationship. In the longitudinal analyses of blood urea nitrogen, there were no significant changes between the examinations by group. Endocrine Assessment In general, the endocrine health status of the Ranch Hands and the Comparisons was reasonably comparable. The examination found no significant differences between the two groups for past thyroid disease, or thyroid and testicular abnormalities determined by palpation. In the analyses of the seven laboratory values (To % Uptake; thyroid stimulating hormone (TSH); testosterone; initial, second, and differential cortisol; and postprandial glucose), significant differences were found for TSH and testosterone, with higher mean levels in the Ranch Hands. These analyses were not supported by the categorical analyses. The thyroid test results were conflicting with respect to an assertion of hypothyroidism in the Ranch Hands (a possible dioxin effect). Mean levels of testosterone were significantly elevated in the Ranch Hand group as contrasted with the Comparisons in the 10-25 percent body fat category. The effects of personality score and percent body fat on the differential cortisol levels were not fully expected. Although tests of 2-hour postprandial mean values showed no significant group differences, comparable categorical tests revealed that significantly fewer Ranch Hands had impaired glucose levels, but conversely, had more (nonsignificant) diabetic levels of glucose. Analyses of the composite diabetes indicator (history plus 2*hour postprandial results) did not disclose significant group differences. The exposure index analyses suggested that the enlisted flyers in the medium exposure level were significantly different from those in the low exposure level for differential cortisol, postprandial glucose, and testosterone. The corresponding high to low contrasts were not significant. The longitudinal analyses were based on T, % Uptake, TSH, and testosterone, and revealed only symmetrical and nonsignificant changes in the Ranch Hand and Comparison groups over the time interval. �Immunological Evaluation Overall, there were no significant group differences or any indication of impaired immunological competence in either group based on comprehensive cell surface marker and functional stimulation studies. Six cell surface markers (total T cells, helper T cells, suppressor T cells, B cells, monocytes, HLA-DR cells, and a constructed helper/suppressor ratio variable) and three functional stimulation studies (PHA, pokeweed, and mixed lymphocyte culture) were conducted on 4? percent of the study population. No significant differences were revealed for five of these variables, in the analyses of the other five variables, there were significant group-by^covariate interactions, but no discernible pattern was identified to suggest a detriment in any subgroup of either group. Skin test assessments of delayed hypersensitivity were characterized by interpleader variation and shifting diagnostic criteria for anergy. The skin test data were judged invalid and were not subjected to statistical testing for group differences. No consistent pattern of immunological deficits could be associated with increasing levels of herbicide exposure in the Ranch Hand group. Pulmonary Disease The pulmonary assessment did not reveal any statistically significant differences between the Ranch Hand and Comparison groups that were suggestive of an herbicide-related disease. The analyses consisted of group assessments of respiratory disease incidence, physical examination abnormalities, and the current prevalence of x-ray abnormalities. There were no significant differences between the Ranch Hands and Comparisons for history of asthma, bronchitis, pneumonia, or for six of seven clinical variables (excluding rales) determined by x-ray or auscultation. Analyses of history of pleurisy, history of tuberculosis, and rales showed significant but inconsistent group-by-covariate interactions. These findings did not indicate any patterns suggesting a different disease experience in the two groups. The exposure index analyses did not reveal any consistent pattern suggestive of an increasing dose response. CONCLUSION The results of the first followup study in 1985 have shown a subtle but consistent narrowing of medical differences between the Ranch Hands and Comparisons since the Baseline Study in 1982. The 1985 examination results provide reassuring evidence that the current state of health of the Ranch Hand participants is unrelated to herbicide exposure in Vietnam. Continued close medical surveillance of these military populations is strongly indicated. This followup report concludes that there is not sufficient plausible or consistent scientific evidence at this time to implicate a causal relationship between herbicide exposure and adverse health in the Ranch Hand group. FUTURE DIRECTIONS The scope and complexity of the AFHS has required gradual refinement and correction to meet the challenges of changing technology and scientific direction, and to ensure continued participation of all enrolled members. �This chapter outlines some of the changes incorporated in the fifths-year followup examination and identifies several areas of future work expected to significantly augment the study. FIFTHr-YEAR FOLLOWUP EXAMINATION Since the fifth-year followup examination was initiated prior to the full analysis of the data from the third-year examination, most modifications were founded upon quality control issues and the desire to make the clinical content of the examination more responsive to the medical needs of the participants. Clinical quality control enhancements were made to improve measurement techniques. The digit preference noted in systolic and diastolic blood pressure readings led to the use of automated blood pressure recording; all other parameters of the blood pressure readings (e.g., sitting position, three recordings, nondominant arm a.t heart level) were not changed. The problem in skin test reading was met by a rigorous quality control plan that included the following elements: refresher training for readers; a required reading of the four skin tests of all participants by both readers, each blind to the results of the other; a required reread of 10 percent of all tests by each of the readers, each blind to the previous reading; and a required weekly report citing numbers and proportions of participants with possible anergy, reversal of induration"erythema measurements, and untoward skin reactions or other reading problems (e.g., participant refusal). In addition, new skin test forms were developed to facilitate accurate recording and transcription; specific clinical criteria were formulated to require consultation by an allergist; and the skin test measurement criterion for possible anergy, consistent with current World Health Organization guidelines, was adopted for the clinical interpretation of all skin test readings. It is anticipated that this clinical quality control program will standardize both readings and interpretations, and will produce a uniformly superior data set. EXPOSURE INDEX REFINEMENTS Since the development of the Study Protocol and the analysis of the 1982 Baseline data, there has been concern among some scientists and the principal investigators over the accuracy and validity of the exposure estimates. It is unclear whether statistically significant differences in some variables between the Ranch Hand and Comparison groups, unsupported by dose?-response estimates, have been due to chance, or whether true differences are obscured by an inadequate exposure index or group misclassification. In mid-1986, strong correlations between dioxin levels in fat tissue and serum were demonstrated by the CDC and other institutions. Because of these results, the Air Force is currently engaged in a collaborative study with CDC to determine whether serum dioxin levels vary significantly in the Ranch Hand population. Approximately 200 AFHS volunteers have supplied a pint of blood to be analyzed for dioxin at the CDC laboratories. If clear and meaningful �exposure findings are evident from this study, several additional studies are feasible: testing can be expanded to the entire study population and a meaningful exposure index based on total current TCDD body burden may be developed; and by means of archived AFHS serum samples from the Baseline study, it may be possible to calculate a reasonably precise half-life of TCDD in humans. These expanded studies will allow the estimation of body burdens of TCDD at the time of departure from SEA (assuming the absence of intervening vocational and recreational exposures). If, in fact, these potential studies become reality within the next 2 years, the fifths-year followup study data will be statistically analyzed using a more appropriate exposure index. In anticipation of this advance, the AFHS is currently collecting 280-^350 ml of blood from all volunteers attending the fifth-year followup study. ADDITIONAL ANALYSES AND STUDIES As in the 1 98M Baseline Report, not all of the measured dependent variables were subjected to statistical analysis (e.g., prothrombin, leutinizing hormone, follicle stimulating hormone), largely because they were not within the bounds of the Air Force^prescribed analyses. Exploration of many of the unanalyzed variables is contemplated as time and resources permit. Similarly, many analytic opportunities to define possible symptom-clinical sign clusters or syndromes by multivariate analysis of variance techniques were passed over due to time and charter. Particularly challenging as an area of future work may be the changing relationships of some immunological variables over time and the biological impact of these changes on the induction of diseases such as cancer. Likewise, future efforts to define shifting cardiovascular disease patterns are a logical extension of the rich longitudinal data base of the AFHS. Such efforts await future analysis and publication. The assessment of possible selection and participation bias has been addressed in a comprehensive manner in this report (see Chapter 5). The analyses and discussion suggest that statistical use of the total Comparison group (versus the Original Comparison group) is justified in this report, and that the impact of selection and participation biases have been minimal. As the followup studies continue, it is anticipated that a wealth of data on compliance-participation factors will be available for continued comprehensive bias analyses. In particular, it is hoped that more complete data will exist to examine the true differences in current health status between refusals and their replacements. As the data set grows over time, the bias analyses will become more complex and will have to deal with changing motivations of the participants to continue in this study. Such bias analyses and assessments will always be of great importance to this study as they ultimately set the bounds for an inference on herbicide causality. � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 059 Folder The folder containing the original item. 1584 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. William H. Wolfe Joel E. Michalek Judson C. Miner Michael R. Peterson Stella G. Machado Theodore G. Karrison William D. Grubbs Wanda F. Thomas Date A point or period of time associated with an event in the lifecycle of the resource 1987-10-01 Title A name given to the resource Air Force Health Study: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Summary, First Followup Examination Results, January 1985-September 1987 Subject The topic of the resource Air Force Health Study morbidity trends ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/bbf35a99b55abf8c88934112df9a7feb.pdf 42ccd8f79282d93f8852c63a66d31427 PDF Text Text Item ID Number °1578 Author Young, Alvin L. Corporate Author U.S. Department of Defense Report/Article TitlO Position Paper: Criteria for Estimating Exposure Levels of Military Personnel to Dioxin and Herbicide Orange during the Vietnam War Journal/Book Title Year 1979 Month/Day December12 Color Number of Images 13 Descripton Notes Wednesday, May 23, 2001 Page 1579 of 1608 �POSITION PAPER Advisory Committeeon Health-Related Effects of Herbicides Question #10: Can criteria be established for determining the level of exposure of military personnel to dioxin during the Vietnam war based on spraying tapes and unit histories? Coordinator: COL J..W. Thiessen, M.D. Position paper submitted by the US Department of Defense to the US Veterans Administration Advisory Committee on Health-Related Effects of Herbicides, Washington, DC, 12 December 1979 �CRITERIA FOR ESTIMATING EXPOSURE LEVELS OF MILITARY PERSONNEL TO DIOXIN AND HERBICIDE ORANGE DURING THE VIETNAM WAR* Any attempt to determine exposure levels of military personnel to Herbicide Orange and its associated dioxin must be predicated on events that occurred at least 10 years ago. Since there were no routine occupational or environmental sampling programs associated with the handling or dissemination of the herbicides in South Vietnam, a quantitative detenni nation of exposure can only be subject to speculation. In addition, since specific no-effect criteria for comparison with actual or derived values do not exist, the calculation of theoretical exposure levels might provide data in the absence of a means to assess their significance. The approach taken in this document is to develop data points for determining "relative" exposures to Herbicide Orange and dioxin (TCDD). The population at risk certainly did not include all military personnel who served in South Vietnam. Moreover, within the military population at risk, the range in magnitude of exposure must have been great. Therefore, it is important to evaluate those factors which would have influenced the potential for a given individual to be "at risk" and those which would have influenced the magnitude of that exposure. The following factors for determining relative exposures are proposed: Time When was the individual in South Vietnam? What job(s) did the individual perform? What was the situation at the time of exposure? What aircraft/vehicle was involved in the exposure? How did the exposure occur? Each of these questions will be discussed and available data will be provided in order to evaluate the magnitude of exposure. * Prepared by Major Alvin L. Young, Ph.D. and Lt Colonel William H. Wolfe, M.D., USAF School of Aerospace Medicine/Epidemiology Division, Brooks Air Force Base, Texas 78235. �I. WHEN WAS THE INDIVIDUAL IN VIETNAM? This issue of time is very important. Not all of the herbicides used in South Vietnam were used throughout the entire 10 years (1962-1971) encompassed by the Department of Defense (DOD) defoliation program. In addition, 2,4,5-T formulations used early in the program are believed to have contained higher levels of TCDD than did the formulations used in the later years. The three time periods shown in Table 1 can be differentiated on the basis of specific herbicides used and the mean dioxin content. TABLE 1. THE DIFFERENTIATION OF THREE TIME PERIODS DURING THE US MILITARY DEFOLIATION PROGRAM IN SOUTH VIETNAM* Herbicides Used January 1962 June 1965 Purple, Pink, Green Blue Mean Dioxin Content ~32f 0 July 1965 June 1970 ~2 § 0 July 1970 October 1971 * t t § Orange White, Blue White, Blue 0 Source: Young et al.3 Found only in 2,4,5-T containing formulations. Value based on analyses of five samples. Value based on the analyses of 488 samples. Herbicide Orange was the most extensively used herbicide in South Vietnam. Orange accounted for approximately 10.7 million gallons of the total 17.7 million gallons of herbicide used (Table 2). It was used from mid-1965 to June 1970. However, as noted in Table 2, Orange was not the only 2,4,5-T containing herbicide used in the defoliation program. Small quantities of Purple, Pink, and Green, all containing 2,4,5-T were used from 1962 through mid-1965. In subsequent sections of this document, the term "Herbicide Orange" will refer to all of the 2,4,5-T containing herbicides used in Vietnam (Purple, Pink, Green, and Orange). �TABLE 2. NUMBER OF GALLONS OF MILITARY HERBICIDE PROCURED BY THE US DEPARTMENT OF DEFENSE AND DISSEMINATED IN SOUTH VIETNAM DURING JANUARY 1962 - OCTOBER 1971* Code Name " Herbicide Orange White 2,4-D; 2,4,5-T 2,4-D; Picloram 10,646,000 5,633,000 1965-1970t 1965-1971f Blue Cacodylic Acid 1,150,000 1962-1971t Purple Pink Green 2,4-D; 2,4,5-T 2,4,5-T 2,4,5-T Total 145,000 123,000 ^0 20 17,705,200 1962-1965 1962-1965 1962-1965 * Source: Young et al.3 t Last fixed-wing mission of Orange 16 April 1970; last helicopter mission of Orange 6 June 1970. t Last fixed-wing mission 9 January 1971; all herbicides under US control stopped 31 October 1971. II. WHAT JOB(S) DID THE INDIVIDUAL PERFORM DURING HIS TOUR(S) IN SOUTH VIETNAM? There were relatively few military operations that involved the handling of herbicides by military personnel. It is, thus, appropriate to examine both the functions, or jobs, where individuals would have been at risk, and to estimate the size of the population at risk. A review of operations involving Herbicide Orange in South Vietnam from January 1962 to April 1970 revealed that there were essentially three groups of US military personnel potentially exposed to Herbicide Orange and its associated dioxin contaminant. These three groups were: 1. "Operation RANCH HAND" personnel actively involved in the defoliation program. This group included aircrew members and maintenance and support personnel directly assigned to the RANCH HAND squadrons. 2. Personnel assigned to selected support functions that may have resulted in exposure to Herbicide Orange. This group included, for example, personnel who sprayed herbicides, using helicopters or ground application equipment; personnel who may have delivered the herbicides to the units performing the defoliation missions; aircraft mechanics who were specialized and occasionally provided support to RANCH HAND aircraft; or, personnel who may have flown contaminated C-123 aircraft, but were not assigned to RANCH HAND (e.g., during the Tet Offensive, all RANCH HAND aircraft were reconfigured to transport supplies and equipment, and were assigned to non-RANCH HAND squadrons). �3. Ground personnel who may have been inadvertently sprayed by defoliation aircraft or who, during combat operations, may have entered an area previously sprayed with Herbicide Orange. k- Pppu 1 ati0" j,stimates. The total number of US military personnel exposed to Herbicide Orange is not known. Approximately 1,200 RANCH HAND personnel were exposed in direct support of the defoliation operations; however, there are no data on the number of non-RANCH HAND personnel who may have been exposed. The actual number of people may be in the thousands since at least 100 helicopter spray equipment units were used in South Vietnam, and most military bases had vehicle-mounted and backpack spray units available for use in routine vegetation control programs. The number of military ground personnel who may have inadvertently been sprayed by RANCH HAND aircraft, or who may have entered areas recently sprayed with Herbicide Orange during combat operations is not known. Approximately 10 percent of South Vietnam was sprayed with herbicides, and most of this area was contested and/or controlled by enemy forces. As estimated frequency of occurrence for selected exposure scenarios is given in Table 3. TABLE 3. ESTIMATED FREQUENCY OF EVENTS WHERE MILITARY GROUND PERSONNEL MAY HAVE BEEN EXPOSED TO HERBICIDE ORANGE JEve_nt_ _____ Direct application of herbicide on ground troops E.reguency_ Rare Ground troops moving into area treated within 24 hours Seldom Ground troops entering a defoliated area (1 month or more after herbicide application) Frequent Discussions with RANCH HAND aircrew members confirmed that in at least one instance in 1967, direct application of herbicide onto a Marine patrol did occur. The basic concept for the major use of the defoliation program, i.e., the use of chemicals to remove foliage to enhance visibility, supports the contention that it was unlikely that troops would be in areas to be treated, or would move into the areas immediately after treatment since the desired effect would not be evident until 3 to 6 weeks after the herbicides were applied. However, the occurrence of the first two scenarios in.Table 3 cannot be ruled out. �III. WHAT WAS THE SITUATION AT THE TIME THE INDIVIDUAL WAS EXPOSED? There are a number of exposure scenarios in which an individual was more likely to have been significantly exposed to a specific herbicide or even another pesticide, including: 1. Guards at a base perimeter. 2. An individual at a Special Forces camp in the inland forest. 3. An individual on combat patrol in the Rung Sat Special Zone. 4. An individual repairing contaminated aircraft. 5. A supply clerk or depot aide handling drums of chemicals. These different situations could have exposed individuals to varying amounts of different herbicides and insecticides since the use patterns of these chemicals differed markedly. Each of the three major herbicides (Orange, White, and Blue) had specific uses. Ninety-nine percent of Herbicide White was applied in defoliation missions. It was not recommended for use on crops because of the persistence of Picloram in soils. Because the herbicidal action on woody plants was usually slow, full defoliation did not occur for several months after spray application. Thus, it was an ideal herbicide for use in the inland forests in areas where defoliation was not immediately required, but where it did occur it would persist longer than if the area were sprayed with Orange or Blue. Herbicide Blue was the herbicide of choice for crop destruction missions involving cereal or grain crops. Approximately 50 percent of all Blue was used in crop destruction missions in remote or enemy controlled areas with the remainder being used as a contact herbicide for control of grasses around base perimeters. Ninety percent of all Herbicide Orange was used for forest defoliation and it was especially effective in defoliating mangrove forests. Eight percent of Herbicide Orange was used in the destruction of broadleaf crops (beans, peanuts, ramie, and root or tuber crops). The remaining 2 percent was used around base perimeters, cache sites, waterways, and communication lines. Table 4 shows the number of acres in South Vietnam within the three major vegetational categories. �TABLE 4. THE NUMBER OF ACRES TREATED IN SOUTH VIETNAM, 1962-1971, WITH MILITARY HERBICIDES WITHIN THE THREE MAJOR VEGETATIONAL CATEGORIES Vejetatjonal ,. Areas Treated* Inland forests Mangrove forests Cultivated crops Total * Areas receiving single or multiple coverage. 2,670,000 318,000 260,000 3,248,000 Source: Certain portions of South Vietnam were more likely to have been subjected to defoliation. Herbicide expenditures for the four Combat Tactical Zones of South Vietnam are shown in Table 5. These data were obtained from the HERBS tape2 and total volume is not in complete agreement with the actual procurement data shown in Table 2 because volume was calculated via spray line data (an estimate of rate of application and area sprayed). TABLE 5. US HERBICIDES EXPENDITURES IN SOUTH VIETNAM, 1962-1971: A BREAKDOWN BY COMBAT TACTICAL ZONE* Herbicide Expenditure (gallons) Orange White Combat Tactical Zones Blue CTZ I 2,250,000 363,000 298,000 CTZ II 2,519,000 729,000 473,000 CTZ III (includes Saigon) 5,309,000 3,719,000 294,000 CTZ IV 1,227,000 435,000 62,000 Ili305J3°° 5-l2461000 lj.iZTjQQQ Subtotals Grand total 17^678^000 * Source: HERBS tape2 In addition to the herbicides, numerous other chemicals were shipped to South Vietnam in 55-gallon drums. These included selected fuel additives, cleaning solvents, cooking oils, and a variety of other pesticides. The �insecticide Malathion was widely used for control of mosquitoes and at least 400,000 gallons of it were used from 1966 through 1970. In addition, much smaller quantities of Lindane and DDT were used in ground operations throughout the war in Southeast Asia. The distribution of the herbicides within Vietnam after their arrival did not occur randomly. About 65 percent was shipped to the 20th Ordnance Storage Depot, Saigon, and 35 percent was shipped to the 511th Ordnance Depot, Da Nang. IV. WHAT MILITARY AIRCRAFT/VEHICLE WAS INVOLVED IN THE EXPOSURE? Numerous aircraft were used in the air war in Vietnam, but only a few of these aircraft were used for aerial dissemination of herbicides. The "work horse" of Operation RANCH HAND was the C-123/UC-123, "Provider." This cargo aircraft was adapted to receive a modular spray system for internal carriage. The module (the A/A 45 Y-l) consisted of a 1,000-gallon tank, pump, and engine which were all mounted on a frame pallet. An operator's console was an integral part of the unit, but was not mounted on the pallet. Wing booms (1.5 inches in diameter, 22 feet long) extended from the outboard engine nacelles toward the wing tips. A short tail boom (3 inches in diameter, 20 feet long) was positioned centrally near the aft cargo door. Each aircraft normally had a crew of three men: the pilot, co-pilot (navigator), arid flight engineer (console operator). During the peak activity of RANCH HAND operations (1968-1969), approximately 30 C-123/UC-123 aircraft were employed. However, many other squadrons of non-RANCH HAND C-123 aircraft were routinely used throughout South Vietnam in transport operations. The control of malaria and other mosquito-borne diseases in South Vietnam necessitated an extensive aerial insecticide application program in order to control these vector insects. From 1966 through 1972, three C-123 aircraft were used to spray Malathion, an organophosphate insecticide. These aircraft could be distinguished from the Herbicide-spraying aircraft because they were not camouflaged. These aircraft routinely sprayed insecticide adjacent to military and civilian installations, as well as in areas where military operations were in progress, or about to commence. Approximately 10 to 12 percent of all herbicides used in South Vietnam was disseminated by helicopter or ground application equipment. Generally, helicopter crews were not assigned to herbicide spray duties on a full-time basis and rotated the spraying duties with other mission requirements. The military UH-1 series of helicopters, deployed by the Air Force, the Army, and Navy units, generally sprayed the herbicides. The most common spray system used was the AGRINAUTICS unit. This unit was installed in or removed from the aircraft in a matter of minutes because it was "tied down" to installed cargo shackles and aircraft modifications were not required for its use. The unit consisted of a 200-gallon tank and a collapsible 32-foot spray boom. The unit was operated by manual controls to control the flow valve and a windmill brake. Generally, each helicopter had three crew members. �A summary of the aircraft used in herbicide and insecticide operations is shown in Table 6. Ground crews that maintained these aircraft were also at risk for exposure to the herbicides and insecticides. TABLE 6. US MILITARY AIRCRAFT USED IN THE DISSEMINATION OF HERBICIDES AND INSECTICIDES IN SOUTH VIETNAM* _Q5in2y.£Ii9^1___ £!i^i£§!JP IJLS ^oilr"a t§^ C-123/UC-123 Yes All Herbicides C-123 Helicopter Air Force UH-F Army UH-1B/UH-1D No Malathion Yes Orange, Blue Navy UH-1E * Source: Young et Various ground delivery systems were also used in South Vietnam for control of vegetation in limited areas. Most of these units were towed or mounted on vehicles. One unit that was routinely used was the Buffalo turbine. It developed a wind blast with a velocity up to 150 mph at 10,000 ft^/minute volume. When the herbicide was injected into the air blast, it was essentially "shot" at the foliage. The Buffalo turbine was useful for roadside spraying and applications of perimeter defenses. The herbicides of choice in these operations were Blue and Orange. V. HOW DID THE EXPOSURE OCCUR? As previously noted, the population at highest risk was the RANCH HAND group since these individuals were exposed to herbicides on a daily basis. Non-RANCH HAND support personnel who handled herbicides and performed secondary level maintenance were also at risk. Beyond these limited populations, the likelihood of other individuals being heavily exposed to herbicides was significantly less. The exposure of personnel could have occured by essentially three routes: 1. Percutaneous absorption and inhalation of vapors/aerosols by direct exposure to sprays. 2. Percutaneous absorption and inhalation of vapors by exposure to treated areas following spray application, and 3. Ingestion of foods contaminated with the material. 8 �As previously discussed, the use of Herbicide Orange in South Vietnam was for the purpose of denying the enemy the cover of dense jungle foliage. The areas normally sprayed were remote, unpopulated, forested areas where very few, if any, US military personnel were located and the exposure to direct spray of Herbicide Orange would have been unlikely. In addition, because of the dense canopy cover, the target of the defoliation operation, the amount of herbicide penetrating to the forest floor would have been small. The chemical and physical characteristics of Herbicide Orange and the spray, as it would have occurred following dissemination from a C-123, are important factors in assessing relative exposures to the Herbicides and TCDD. Table 7 reviews the pertinent chemical and physical characteristics of Herbicide Orange. Table 8 reviews both the application parameters of the spray system used in the C-123 aircraft and the characteristics of the spray itself. Generally, herbicides were sprayed in the early morning or late afternoon, so as to minimize the effects of air movement on particle dispersion. TABLE 7. PERTINENT CHEMICAL AND PHYSICAL CHARACTERISTICS OF HERBICIDE ORANGE Formulation Concentrated (8.6 Ib ai/gal)* Water Insoluble Density = 1.28 Vapor Pressure 3.6 x 10-4 mm Hg at 30°C NBEt 2,4-D : 1.2 x 10-4 NBE 2,4,5-T : 0.4 x 10-4 TCDD : 1 x 10-4 Viscous 40 centipoises at 20°C Noncorrosive to metal Deleterious to paints, rubber, neoprene Long shelf life * Pounds active ingredient (2,4-D and 2,4,5-T) per gallon. t NBE = Normal butyl ester �TABLE 8. APPLICATION PARAMETERS AND SPRAY CHARACTERISTICS OF THE C-123 MODULAR INTERNAL SPRAY SYSTEM Aircraft speed 130 KIAS* Aircraft altitude 150 ft Tank volume 1,000 gal Spray time 3.5-4 min Particle size: 1.9% 100-50Q)j 76.2% >500y 21.9% 87% impacted within 1 min 13% drifted or volatilized Mean particle volume 0.61 ^1 Spray swath 260 +20 ft Mean deposition 3 gal/acre Total area/tank 340 acres * Knots indicated air speed. Ground combat forces normally would not have been expected to have entered a previously treated area for several weeks after treatment, during which time numerous environmental factors would have reduced the potential for exposure to military personnel. Young et al.3 have conducted an indepth review of the environmental fate of Herbicide Orange and TCDD. The following is a summary from that report: . . . Available data indicate that the vast majority of the phenoxy herbicides would impact forest canopy, the intended target. Rapid uptake (e.g., within a few hours) of the ester formulations of 2,4-D and 2,4,5-T would occur. Most of the herbicide probably would undergo rapid degradation (weeks) within the cellular matrix of the vegetation. However, some of herbicide may remain unmetabolized and would be deposited on the forest floor at the time of leaf fall. Soil microbial and/or chemical action would likely complete the degradation process. 10 �Herbicide droplets that impacted directly on soil or water would probably hydrolyze rapidly (within hours). Biological and nonbiological degradative processes would further occur to significantly reduce these residues. Some violatilization of the esters of 2,4-D and 2,4,5-7 would occur during and immediately after application. The volatile material most likely would dissipate within the foliage of the target area. Phctodecomposition of TCDD would minimize the amount of biologically active volatile residues moving downwind of the target area. Accumulation of phenoxy herbicides in animals may occur following ingestion of treated vegetation. The magnitude of this accumulation would likely be at nontoxia levels. Herbicide residues in animals would rapidly decline after withdrawal from treated feed. Most TCDD sprayed into the environment during defoliation operations would probably photodegrade within 24 hours of application. Moreover, recent studies suggest that even within the shaded forest canopy, volatilization and subsequent photodecomposition of TCDD would occur. Since translocation into vegetation would be minimal, most TCDD that escaped photodegradation would enter the soil-organic complex on the forest floor following leaf fall. Soil chemical and microbial processes would further reduce TCDD residues. Bioconcentration of the remaining minute levels of TCDD may occur in liver and fat of animals ingesting contaminated vegetation or soil. However, there are no field data available that indicate that the levels of TCDD likely to accumulate in these animals would have a biological effect. The environmental generation of TCDD from 2,4,5-T residues, through thermal or photolytic processes, would be highly unlikely and of no consequence.... VI. CONCLUSIONS. While a precise determination of herbicide exposure cannot be achieved, the five factors discussed in this document might permit both a characterization and a relative estimate of the magnitude of the exposure. In the preparation of a total exposure for a given individual, answers to the five questions must be determined for jeach exposure incident, and a summary exposure estimate developed. 1.1 �LITERATURE CITED 1. Committee on the Effects of Herbicides in South Vietnam 1974. Part A: Summary and Conclusions. National Academy of Science, Washington, DC, 398 p. 2. HERBS tapes. 1971. Computer data on herbicide missions in Vietnam, 1965-1971. Department of Defense, Washington, DC. 3. Young, A. L., J. A. Calcagni, C. E. Thai ken, and J. W. Tremblay. 1978. The Toxicology and Environmental Fate, and Human Risk of Herbicide Orange and its Associated Dioxin. Technical Report OEHL-TR-78-92. USAF Occupational and Environmental Health Laboratory, Brooks Air Force Base, TX 78235, 247 p. 12 � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 059 Folder The folder containing the original item. 1578 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Young, Alvin L. William H. Wolfe Description An account of the resource <strong>Corporate Author: </strong>U.S. Department of Defense Date A point or period of time associated with an event in the lifecycle of the resource December 12 1979 Title A name given to the resource Position Paper: Criteria for Estimating Exposure Levels of Military Personnel to Dioxin and Herbicide Orange during the Vietnam War Subject The topic of the resource Air Force Health Study Ranch Hand crew herbicide exposure HERBS database ao_seriesIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 059 Folder The folder containing the original item. 1577 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. Patricia M. Moynahan Richard A. Albanese William H. Wolfe Description An account of the resource <strong>Corporate Author: </strong>United States Air Force School of Aerospace Medicine, Aerospace Medical Division (AFSC), Brooks AFB, Texas Source A related resource from which the described resource is derived Banbury Report 18: Biological Mechanisms of Dioxin Action Date A point or period of time associated with an event in the lifecycle of the resource 1984 Title A name given to the resource An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides Subject The topic of the resource Air Force Health Study Ranch Hand crew ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/ed8c1fec7593a8ef31278d2cec0a7ecf.pdf d01b274a9f366bc3df58bc723762ccd2 PDF Text Text Item ID Number 01 576 Lathrop, George D. United States Air Force School of Aerospace Medicine ( RupOrt/ArtlClO TitiB An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Mortality Update-1984 Journal/Book Title Year 1984 Month/Day December 10 Color Number of Images n so Descrlpton Notes Wednesday, May 23, 2001 Page 1577 of 1608 �PROJECT RANCH HAND II AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES MORTALITY UPDATE -1984 10 DECEMBER 84 William H. Wolfe, Colonel, USAF, MC Joel E. Michalek, PhD Richard A. Albanese, M.D. George D. Lathrop, Colonel, USAF, MC Patricia M. Moynahan, Colonel, USAF, NC Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 EPIDEMIOLOGY DIVISION DATA SCIENCES DIVISION USAF SCHOOL OF AEROSPACE MEDICINE (AFSC) BROOKS AIR FORCE BASE, TEXAS 78235 �Project Ranch Hand TI Mortality Update - 1984 EXECUTIVE SUMMARY BACKGROUND The purpose of the Ranch Hand II Study is to determine whether those individuals involved in the aerial spraying of herbicides in Vietnam during the Ranch Hand operation have experienced any adverse health effects as a result of their participation in that program. The study evaluates both mortality (death) and morbidity (disease) in these individuals over a 20-year period of time after the studies were initiated. The- baseline mortality study was released in June 1983 and the baseline morbidity study in February 1984. Neither study demonstrated health effects which could be conclusively attributed to herbicide or dioxin exposure. The reader is referred to reports of the studies for further details (1, 12). METHOD The present study report describes the second mortality analyses. Deaths in the 1256 Ranch Hand and 6171 comparison subjects were determined, using the data sources of the Air Force, Veterans Administration, Social Security Administration, Internal Revenue Service, and personal contacts. As of 31 December 1983, 54 Ranch Handers and 265 comparison subjects had died. Death certificates were obtained on all subjects. Autopsies were conducted on 157 of the individuals who had died. Results have been obtained for 104 of these autopsies to confirm the death certificate findings. Autopsy reports for the 53 others have been requested, but have not yet been obtained. Extensive statistical analyses were accomplished, as detailed in the report, to compare the death experience in the Ranch Hand population with the comparison group. In addition, death experience in these groups was compared to the 1978 U.S. White Male Mortality experience, the 1978 Department of Defense Nondisability Retired Life Table, the mortality experience of the West Point Class of 1956, the USAF active duty personnel, and the active U.S. Civil Service population. RESULTS As was the case in the first mortality report, the current mortality analyses did not reveal any statistically significant differences in mortality between the exposed and comparison groups. The percentages dead in each major category are summarized below. Percent Deaths Ranch Hand Comparison Rank Officers Enlisted 3.2 4.9 4.0 4.5 Occupation F l y i n g 3 . 6 Ground 5.1 4.7 3.9 Note: None of the above differences between the Ranch Hand and Comparison groups are statistically significant. i �Ranch Hand Total Overall 4.3 Comparison 4.3 As was reported in the baseline mortality study, the Ranch Hard officers had a nonstatistically significant though slightly lower death rate than their comparisons, Ranch Hand flyers had a nonstatistical1v significant though slightly lower death rate than comparisons, and Ranch Hand ground personnel had a slightly higher but nonstatistically significant death rate than the comparisons. The herbicide/dioxin exposure index described in the morbidity report was applied to the data, and no relationship between exposure and mortality experience was identified. As was also noted in the baseline mortality study, analyses consistently demonstrated significantly better survival in the Ranch Hand officers than Ranch Hand enlisted members, as was the case with comparison officers and comparison enlisted personnel. Cause-specific analyses did not demonstrate any increased Ranch Hand mortality for accidents, suicide, homicide, malignancy or circulatory system disease. No unusual patterns of malignancy were observed in either the Ranch Hand or comparison groups, a finding which would be expected from the small number of deaths to date. When compared to the 1978 U.S. White male population, the Ranch Hand officers, comparison officers, and comparison enlisted are living significantly longer than expected. Although Ranch Hand enlisted are also living longer, the difference is not significant. A similar pattern was seen in analyses using the DOD retired population. All groups had a mortality experience similar to the civil service population. As would be expected from the fact that individuals in the active duty population who develop severe chronic disease are medically retired, all groups in this study had an increased mortality when compared to the Air Force population currently on active duty. Both Ranch Hand and comparison officers had mortality similar to the West Point group. CONCLUSION AND RECOMMENDATION Continued mortality surveillance is recommended, since the study groups are still relatively young and healthy. While sufficient time may have elapsed for some clinically significant conditions to occur, additional time is necessary for other conditions, which may possibly be attributable to herbicide exposure, to develop. At this time, however, there is no evidence of increased mortality «is a result of herbicide exposure in those individuals who accomplished the Ranch Hand spray operations in Vietnam. ii �Table of Contents Page Executive Summary i Table of Contents ill 1 . In t roduct i on 1 2. 2 Ranch Hand Versus Comparison Group Analyses 3. Within Group Analyses of Mortality 12 4. 13 Cause-Specific Analyses 5. Noncause-Speclfic Comparisons with External Population . . 20 . 6. Comparisons with the West Point Group 33 7. Further Covariate Adjustments 38 8. Future Commitments 40 9. Summary and Conclusions..... 40 References 42 Principal Investigators 43 Appendix - Mortality by Year of Birth 44 iii �Project Ranch Hand II Mortality Update - 1984 1. Introduction This report updates the findings of the baseline mortality report (1) released on June 30, 1983. The reader is referred to the baseline report for information regarding the study design, statistical procedures, the mortality determination process and previous findings. Nine newly identified Ranch Banders have been added to the data file since the baseline report. One of these, a non-Black Enlisted ground crew member, died in 1981 of circulatory disease. Summary counts of the population at risk and the number of deaths in each of the three groups are shown in Table 1. The analyses in this report are based on this data and the data in Table 4. Table 2 contains the counts of new deaths in the population since the last report. Table 3 in this report corresponds to Table 3 in the baseline report and contains summary counts and death rates by job, race and group. These counts reflect cumulative mortality as of 31 December 1983 (certified as of 8 June 1984). Table 1 Summary Counts of Death by Rank and Occupation Ranch Hand At Risk Dead Rate (%) Rank Officers Enlisted Comparison At Risk Dead Rate ( ) % 466 790 15 39 0.032 (3.2) 0.049 (4.9) 2278 3893 91 174 0.040 ( . ) 40 0.045 (4.5) Flying Ground 646 610 23 31 0.036 (3.6) 0.051 (5.1) 3163 3008 149 116 0.047 (4.7) 0.039 (3.9) Total 1256 54 0.043 (4.3) 6171 265 0.043 (4.3) Occupation Table 2 Deaths During 1983 by Rank and Occupation Ranch Hand Rank Officer , Enlisted Comparison At Risk 1983 Deaths At Risk 1983 Deaths 451 754 0 3 2190 3731 3 12 3023 2898 9 6 Occupation Flying^ Ground 624 581 At risk count does not include the newly identified Ranch Hander who died prior to 1983. 1 �Table 3 Occupational and Race Specific Mortality Ranch Hand Comparisons At Risk Dead Rate At Risk Dead Ra_te_ Race Occupation Non-Black 350 Officer-Pilot 82 Officer-Nav 25 Officer-Other Enlisted-Fit Eng 191 532 Enlisted-Other 6 Officer-Pilot 2 Officer-Nav 1 Officer-Other Enlisted-Fit Eng 15 Enlisted-Other 52 Black Total 1256 0.034 0.024 0.040 0.037 0.053 0.000 0.000 0.000 0.133 0.038 1740 390 123 935 2628 13 10 2 75 255 74 14 3 51 101 0 0 0 10 12 54 0.043 6171 265 0.043 12 2 1 7 28 0 0 0 2 2 0.043 0.036 0.024 0.055 0.038 0.000 0.000 0.000 0.133 0.047 2. Ranch Hand Versus Comparison Group Analyses. Survival contrasts were made using linear rank procedures, survival curves, relative risk estimation and standardized mortality ratios. Survival curves were estimated via the product-limit estimate of Kaplan and Meier ( ) 2 . Linear rank testing was carried out using the logrank test and Prentice's censored data extension of the Wilcoxon test ( ) All linear rank tests were carried out with 3. matched sets merged when Ranch Banders differed by less than one year relative to date of birth. Within each stratum of job and race, these merged matched sets were used as separate strata for testing purposes. The matched data relative risk procedure, due to Ejigou and McHugh (4) is applied only to the 1241 Ranch Handers with matched comparisons and the stratified relative risk or SMR estimate is applied to all 1256 Ranch Handers. �Group contrasts were made on officers, enlisted personnel, flying personnel, ground personnel and the total group. Summary counts are shown ir\ Table 4. Table 4 Summary Counts by Rank, Occupation and Group Flying Personnel Groups Ranch Hand Comparisons At Risk 440 2153 Officer Dead Rate 14 88 0.032 0.041 Enlisted At Risk Dead Rate 206 1010 9 61 0.044 0.060 At Risk 646 3163 Total Dead 23 149 Rate 0.036 0.047 Ground Personnel Groups Ranch Hand Comparisons Officer At Risk Dead Rate 26 125 1 3 0.038 0.024 At Risk 584 2883 Enlisted Dead Rate 30 113 0.051 0.039 Total At Risk Dead Rate 610 3008 31 116 0.051 0.039 Survival curves were estimated only for officers, enlisted, flying, ground personnel and all personnel in Ranch Hand and the comparison groups. There is a substantial degree of overlap between these subgroups, with 96% of both the Ranch Hand and comparison ground personnel being enlisted. The enlisted category includes both ground support and flying enlisted personnel. Survival curves for the overall Ranch Hand and comparison groups are shown in Figure 1. The curves for officers, enlisted personnel, flyers and ground personnel are shown in Figures 2 through 5. �Figure 1 Survival Curve Estimates for All Ranch Banders and All Comparisons 1 0.9 0.8 COMP I °'? i—< > 0.8 OT o.e o »-« 0.4 o u_ 0.2 0.1 10 90 30 40 BO 60 70 flGE Figure 2 Survival Curve Estimates for Ranch Hand and Comparison Officers 1 0.9 0.8 O z 0.7 0.8 CO 0.5 . z o O cr 0.3 a; u_ 0.2 _ O.I - 0 10 20 30 40 flOE A 80 60 70 �Figure 3 Survival Curve Estimates for Enlisted Ranch Banders and Comparisons 1 0.9 COMP 0.8 cb > 0.6 o: » z o «- 0.6 0.4 U cr 0.3 a: u_ 0.2 0.1 10 20 _L 40 J_ 30 60 80 70 flQE Figure 4 Survival Curve Estimates for Ranch Hand and Comparison Flyers 1 0.8 0.8 RH Cb S °' 7 > > 0.6 CO 0.6 z CD •- »-«, 0 0.4 0.2 0.1 _ 0 to 20 30 40 flOE 5 60 60 �Figure 5 Survival Curve Estimates for Ranch Hand arid Comparison Ground Personnel 1 0.9 0.8 > r> w 0.6 0.61- z o *-* 0.4 _ O 0.3 0.2 . 10 20 40 60 70 fl&E The patterns qualitatively evident in these graphs are seen quantitatively in subsequent statistical analyses. Linear rank procedures were carried out on the same four subgroups and on all personnel to assess death patterns by time. These procedures are designed so that the statistic will be positive when Ranch Handers are dying before comparison subjects and negative when comparisons are dying prior to Ranch Handers. The results are shown in Table 5 (Table 6 in the baseline report). The rank statistic used is a fair measure of group difference only when this difference occurs consistently within each tested stratum. Since the strata in these analyses were formed by date of birth, occupation and race, the rank statistic is fair only when the group difference in death times does not change with date of birth, race and occupation. As will be shown later, there is an indication that there is an effect of date of birth on relative risks in the officer subgroup. Thus, rank statistics on officers must be interpreted carefully. Further, since there is an indication that mortality contrasts change with rank and occupation, the overall logrank value and p-value, shown in Table 5 are not valid summarv statistics. �Table 5 Test Results and P-values for Noncause-Specific Survival Group Officer Enlisted Flying Ground Logrank (value) P-value (-0..682) 0.495 ( 0,640) 0.522 (-1.144) 0.253 ( 1.303) 0.192 Wilcoxon (value) P-value (-0.771) 0.441 ( 0.575) 0.565 (-1.228) 0.220 ( 1.235) 0.217 Total ( 0.076) 0.939 ( 0.009) 0.993 Table 5 suggests that ground personnel in the Ranch Hand group are dying sooner than their matched comparisons (logrank = 1.303), but the difference is not statistically significant (p=0.192). The negative values of the logrank and Wilcoxon statistics for officers (logrank = -0.682) and flying personnel (logrank « -1.144) suggest that Ranch Banders in this group may be living longer than their matched comparisons. Similar analyses on the same subgroups (officer, enlisted, flying, ground and total) were carried out on data from non-Black subjects only. The results are shown in Table 6. Table 6 Test Results and P-values for Noncause-Specific Survival Non-Black Ranch Handers and Non-Black Comparisons Group Logrank (value) P-value Wilcoxon (value) P-value Officer Enlisted Flying Ground (-0.668) ( 0.686) (-1.229) ( 1.436) 0.504 0.492 0.219 0.151 (-0.755) 0.450 ( 0.626) 0.531 (-1.305) 0.192 ( 1.360) 0.174 Total ( 0.101) 0.919 ( 0.037) 0.970 The findings in Table 6 clearly parallel those of Table 5, as would be expected from the small size of the Black cohort in this study. Relative risk estimates, the associated 95% confidence intervals, two-sided p~values for testing the null, hypothesis of relative risk equal to unity, and power for detecting a relative risk of 2 in these data are shown in Table 7. These estimates are based on a matched data algorithm and summarize the relative prevalence of death in the Ranch Hand and comparison groups. As with the rank tests, the estimated relative risks are unbiased only when the relative risks can be assumed to be constant across date of birth strata. There is indication that this assumption is not met in the officer cohort so that their estimated relative risks must be viewed with caution. On the other hand, the assumption appears to be well met in the flying, ground and enlisted subgroups so that these relative risk estimates appear to be reliable. Since there is an indication that relative risk changes with rank and occupation, the overall relative risk, 0.965, is not a valid summary statistic. 7 �Table 7 Relative Risks, 95% Confidence Intervals, P-Values and Power for Noncause-Specific Deaths to Date (1241 Ranch Handers Versus 6171 Matched Comparisons) Group Rel Risk Conf Int 1.134) 1.475) 1.052) 1.793) P-yalue Powejr 0.275 0.692 0.174 0.274 0.886 0.980 0.968 0.928 Officer Enlisted Flying Ground 0.715 1.077 0.718 1.259 (0.295, (0.679, (0.385, (0.724, Total 0.965 (0.666, 1.264) 0.823 0.998 Table 7 shows that Ranch Hand flyers are experiencing fewer deaths than their matched comparisons (relative risk = 0.718), but this group difference Is not statistically significant (p=0.174). The Ranch Hand ground personnel have experienced more deaths (relative risk = 1.259) than their matched ground comparisons, but, again, this excess is also not statistically significant (p=0.274). Year-of-birth specific mortality rates are given in Tables 8 through 12, with the corresponding standardized mortality ratios (SMR) and associated p-values (5). In each analysis, the comparison group is the internal standard. The SMR will accurately estimate the relative risks within each stratum in these analyses if the year-of-birth specific relative risks are equal. A likelihood ratio test for the hypothesis of equal year-of-birth specific relative risks was carried out for each analysis, and its p-value is denoted by PI. In addition, the hypothesis that the relative risk is unity, given that relative risk is constant across strata, was tested; its p-value is denoted by P2. The SMR and both p-values are given for each contrast. Additional post hoc analyses are presented at the end of this section to show that the hypothesis of equal year-of-birth specific relative risks may not be met in the officer and flying cohorts. Table 8 Year-Of-Birth Specific Mortality Rates (1256 Ranch Handers Versus 6171 Comparisons) (SMR = 1.008, PI = 0.258. P2 - 0.983) Ranch Hand At Risk Dead Rate 1905-1914 1915-1919 1920-1924 1925-1929 1930-1934 1935-1939 1940-1944 1945-1954 Total 5 17 48 84 305 210 210 377 2 5 3 2 17 7 5 L3 1256 54 0.400 0.294 0.063 0.024 0.056 0.033 0.024 0.034 Comparison At Risk Dead Rate 14 96 241 501 1389 1020 1096 1814 3 12 24 44 73 36 23 50 6171 265 0.214 0.125 0.100 0.088 0.053 0.035 0.021 0.028 �Table 9 Officer-Specific Mortality Rates by Year-Of-Birtb (SMR =0.797, PI = 0.236, P2 - 0.404) Birth Year Ranch Hand At Risk Dead Rate 1910-1924 1925-1934 1935-1939 1940-1944 1945-1949 41 194 95 91 _45 3 4 4 2 2 Total 466 15 Comparison At Risk Dead Rate 205 930 458 495 190 17 51 12 6 5 2278 0.073 0.021 0.042 0.022 0.044 91 0.083 0.055 0.026 0.012 0.026 Table 10 Enlisted-Specific Mortality Rates by Year-Of-Birth (SMR = 1.105, PI - 0.663. P2 = 0.590) Ranch Hand At Risk Dead Rate 1905-1914 1915-1919 1920-1924 1925-1929 1930-1934 1935-1939 1940-1944 1945-1954 4 9 16 41 154 115 119 332. 2. 2 3 2 13 3 3 11 Total 790 39 Comparison At Risk Dead Rate 12 54 80 211 749 562 601 1624 3 8 11 24 42 24 17 45 3893 0.500 0.222 0.188 0.049 0.084 0.026 0.025 0.033 174 0.250 0.148 0.138 0.114 0.056 0.043 0.028 0.028 Table 11 Flying-Specific Mortality Rates by Year-Of-Birth (SMR =0.751, PI = 0.765. P2 = 0.186) Ranch Hand At Risk Dead Rate 1910-1924 1925-1934 1935-1939 1940-1944 1945-1949 44 272. 145 121 64 4 9 6 2 2 Total 646 23 0.091 0.033 0.041 0.017 0.031 Comparison At Risk Dead Rate 220 1316 698 653 276 23 78 24 14 10 3163 149 0.105 0.059 0.034 0.021 0.036 �Table 12 Ground Specific Mortality Rates by Year-of~Birth (SMR = 1.306, PI « 0.604. P2 = 0.203) Ranch Hand 1905-1914 1915-1919 1920-1924 1925-1929 1930-1934 1935-1939 1940-1944 1945-1954 5 8 13 31 86 65 89 313 2 1 3 2 8 1 3 U_ Total 610 Comparison 0.400 0.125 0.231 0.065 0.093 0.015 0.034 0.035 31 14 51 66 151 423 322 443 JL538 3 6 7 19 20 12 9 _40 3008 0.214 0.118 0.106 0.126 0.047 0.037 0.020 0.026 116 When year-of-birth is dichotomized (1905-1934, 1935-1954) and survival status (alive, dead) is analyzed by group (Ranch Hand, comparison) and rank (officer, enlisted), a significant four-way interaction is evident (p=0.024). That is, the survival status by birth year by group relationship changes with rank. The officer and enlisted relative risks are 0.50 and 1.23 in the 1905-1934 year-of-birth stratum and 1.72 and 0.97 in the 1935-1954 birth-year stratum. There were no three-way interactions in this analysis. When rank is replaced by flying status (flying, ground) in this four factor analysis, no four-way interaction is seen (p=0.250), and no significant group by flying status by birth-year interaction (p=0.790) Is observed. Further, when the officer, enlisted, flying and ground subgroups are analyzed separately on survival status, group and birth year, there is no three-way interaction for enlisted (p=0.480), flying (p=0.265) or ground personnel (p=0.634) but there is a significant three-way interaction for the officers (p=0.027). That is, the survival status by group relationship changes with year of birth in the officer cohort. Taken together, these log-linear analyses indicate that date of birth is affecting the relative risk estimate (and thus the SMR and rank tests) in the officer category. Specifically, the overall death experience of the Ranch Hand officers appears to compare favorably with the comparisons. However, these diminished death rates appear to be found in the Ranch Hand officers born before 1935, while Ranch Hand officers with later birth dates evidence a rate equal to or exceeding that of the comparisons (as seen in Table 14). A summary Table 13. of logrank, relative risk 10 and SMR results obtained is shown in �Table 13 Noncause-Speclfic Statistical Summary Age at Death Logrank Group Value P -value -0.682 Officer 0.495 0.640 Enli sted 0.522 -1.144 Flying 0.253 Ground 0.192 1.303 0.076 Total 0.939 Ground Deaths to Date SMR Relative Risk RR P-value SMR P-value 0.715 0.275 0.797 0.404 0.692 1.077 1.105 0.590 0.718 0.174 0.751 0.186 0.274 1.306 0.203 1.259 Total 0.965 Group Officer Enlisted Flying 0.823 1.008 0.983 The data in Table 13 show reasonable consistency. The ground cohort displays excess death in the Ranch Hand group in contrast to the comparison group, but this group difference is not statistically significant. The officer cohort evidences less death in the Ranch Hand group in contrast to the comparison group, but, again, this group difference is not statistically significant. However, as discussed above and shown in Table 14, this favorable mortality experience occurs in those individuals born before 1935, while Ranch Hand officers born after 1935 have experienced the same or greater death rate than their comparisons. 11 �Table 14 Death Rates ~by Group, Rank, Occupation and Year-Of-Birth Year of Birth Ranch Hand Death Rate Comparison Death Rate Relative Risk Before 1935 Enlisted 0.030 0.060 0.50 After 0.034 0.020 1.72 1935 Before 1935 0.100 0.080 1.23 After 1935 0.030 0.031 0.97 Flying Before 1935 After 1935 0.041 0.030 0.066 0.030 0.62 1.00 Ground Before 1935 After 1935 0.112 0.032 0.078 0.026 1.44 1.23 Occupation The favorable, though not statistically significant, survival experience of Ranch Hand flying personnel, relative to the matched comparison flyers is shown in Figure 4, where the survival curves for Ranch Hand and comparison flyers are drawn on the same scale and coordinate system. In contrast, the relatively poorer, but not statistically significant, survival experience of the Ranch Hand ground personnel is illustrated in Figure 5, wherein the Ranch Hand and comparison ground personnel survival curves are drawn on the same coordinate system. 3. Within Group Analyses of Mortality Within group year-of-birth adjusted contrasts by occupation and rank via SMR's are summarized in Table 15. The corresponding SMR analyses are shown in the Appendix. Table 15 Summary of Within Group SMR Analyses Subgroups SMR PJ^ P^ Officers versus Enlisted Ranch Hand Comparison 0.483 0.663 0.204 0.811 0.031 0.003 Flying versus Ground Ranch Hand Comparison 0.548 0.926 0.376 0.607 0.052 0.782 12 �Table 15 shows that Ranch Hand officers are having significantly fewer deaths (SMR=0.483, p=0.031) than Ranch Hand enlisted personnel, after adjustment for yearof-birth. This officer versus enlisted differential is also significant and in the same direction in the comparison group (SMR=0.663, p=0.003). The table demonstrates the favorable mortality experience of Ranch Hand flyers and adverse mortality of the Ranch Hand ground personnel in that Ranch Hand flyers are experiencing significantly fewer deaths than Ranch Hand ground personnel (SMR=0.548, p=0.052). This flyer versus ground differential is not apparent in the comparison group (SMR=0.926, p=0.782). As discussed before and as displayed in Table 14, the favorable Ranch Hand officer and flyer mortality experience is confined to the group born before 1935. **• Cause-Speel fic Analyses Table 16 shows death counts by cause and subgroup (flying officer, ground officer, flying enlisted and ground enlisted). Counts are shown for all 1256 Ranch Handers and the 6171 comparisons. The distribution of new deaths in the Ranch Hand and comparison groups are presented in Table 17 and age adjusted relative risks for these data are shown in Table 18. Relative risks are calculated using a matched data algorithm, hence, only the 1241 Ranch Handers having matched comparisons are used. Of the 15 unmatched Ranch Handers, two have died; a flying officer died of an accident and a ground airman died of circulatory system disease. Since these data are sparse, relative risks are only calculated on officer, enlisted, flying and ground subgroups, as well as on all personnel combined. Three cells in Table 18 contain two p-values for the significance of the relative risk estimate. The first is calculated using a null variance of the estimated relative risk and the second, within parentheses, is calculated using the group non-null variance estimate. A null variance is defined as a variance that requires knowledge of the true value of the estimated parameter, and that value is set equal to the value specified In the null hypothesis. The question of which variance estimate to use, and hence, which p-value to report is a point of research in theoretical statistics. We have chosen to use the null variance when computing p-values because of analogies with other testing situations and because our power studies have shown the resulting test to be more powerful than the test using the general non-null estimate. Unfortunately, the non-null variance must be used in computing 95% confidence intervals for the relative risk, making the p-value and confidence interval sometimes incompatible. Whenever this kind of disparity occurs, both p-values are given. 13 �Table 16 Deaths by Cause and Subgroup G round Flying C RH C RH C RH £ RH C Accident 8 32 0 1 4 26 7 35 19 94 Suicide 0 4 1 0 ] 3 1 9 3 16 Homicide 0 0 0 0 0 1 2 3 2 4 Parasitic infection 0 2 0 0 0 0 0 2 0 4 Malignant neoplasm 0 13 0 1 1 11 5 18 Uncertain neoplasm 0 1 0 0 0 0 0 1. 0 2 Endocrine 0 1 0 0 0 0 1 0 1 1 Mental disorder 0 0 0 0 0 0 0 1 0 1 Nervous System 0 1 0 0 0 0 0 1 0 2 Circulatory System 4 27 0 0 1 14 12 34 17 75 Respiratory System 0 1 0 0 0 2 0 2 0 5 Digestive System 2 4 0 1 1 3 2 5 5 13 Genitourinary System 0 1 0 0 0 0 0 2 0 3 Congenital anomalies 0 0 0 0 0 0 0 0 0 0 111 defined 0 1 0 0 1 1 0 0 1 2 14 88 1 3 9 61 30 113 Flying Cause RH Total Ground 14 Total 6 43 54 265 �Table 17 New Deaths by Cause Comparison 1 1 1 1 Total Lung Cancer Stomach Cancer Undetermined Circulatory 4 Circulatory 2 Digestive 4 Cancer 1 Homicide 1 Parasitic Infection 1 Respiratory _7_ Suicides 15 _ 4 * The newly identified deceased Ranch Hander. 15 �Table 18 Cause-Specific Age Adjusted Relative Risks by Group (1241 Ranch Banders versus 6171 Comparisons) Statistic Accident Suicide Cause Homicide Malignancy Circulatory Digestive Officer Rel Risk Conf Int P-Value Power 0.994 (0.161, 1.827) 0.989 0.601 Enlisted Rel Risk Conf Int P-Value Power 0.844 0.833 (0.279, 1.410) (0.000, 2.081) 0.624 0.814 0.780 0.374 Flying Rel Risk Conf Ir.t P-Value Power 0.935 (0.324,1.546) 0.841 0.767 Ground Rel Risk Conf Int P-Value Power 0.803 1.111 (0.090, 1.516) (0.000, 2.814) 0.633 0.892 0.621 0.334 3.333 ( . 0 , 9.297) 000 0.099 0.246 1.235 1.633 (0.000, 2.486) (0.480, 2.786) 0.675 0.144 0.461 0.592 1.667 (0.000, 4.334) 0.505 0.291 Total Rel Risk Conf Int P-Value Power 0.937 0.937 (0.456, 1.418) (0.000, 2.094) 0.804 0.919 0.899 0.425 2.500 (0.000 , 6.743) 0.221 0.262 0.684 1.043 (0.095, 1.272) (0.459, 1.627) 0.397 0.883 0.681 0.832 1.923 ( . 0 , 3.907) 000 0.174 0.387 0.400 ( . 0 , 0.972) 000 0.221 (0.412) 0.526 2.500 ( . 0 ,6.743) 000 0.221 0.262 1.258 (0.445,2.072) 0.474 0.701 1.875 (0.000,4.363) 0.312 0.320 0.208 (0.000,0.625) 0,113(0.000) 0.515 0.714 (0.000,2.211) 0.758 0.306 0.998 (0.108,1.889) 0.997 0.565 0.385 (0.000,0.836) 0.117(0.008) 0.652 2.143 (0.000,5.041) 0.217 0.306 �Tables 16 and 18 must be Interpreted with care since the data are very sparse in some categories. The Ejigou-McIIugh relative risk estimate uses a variance weighting scheme of relative risks. The variance expression used is correct only for large aggregates of such matched sets. Since matched sets with large numbers of comparison deaths are rare, but tend to occur iri the older subject cohorts, it must be anticipated that relative risks from the older cohorts may not be properly weighted In the relative risks shown here. Most disease information resides in the categories of malignant neoplasm and circulatory system deaths. Digestive system mortality by ICD code is shown in Table 19, site-specific malignant neoplasm mortality is shown in Table 20 and the morphology of neoplasms is shown in Table 21. There was one case of soft tissue sarcoma in a comparison individual, but none in the Ranch Hand group. Table 19 Digestive System Mortality Deaths Ranch Hand Comparison TCD Code Pancreatitis (5770) Alcoholic cirrhosis (5712) Nonalcoholic cirrhosis (5715) Nonalcoholic fatty liver (5718) Chronic liver disease (5728) Alcoholic liver disease (5711) Duodenal ulcer (5325) Peptic ulcer (5334) Hepatocellular disease (573a) 2 4 3 1 2 0 1 0 _0 5 Total 1 0 3 0 0 1 0 0 _0 13 These codes were based on death certificate data; more detailed etiologic information has been requested but not yet received for the nonalcoholic cirrhosis and fatty liver deaths. 17 �Table 20 Site-Specific Malignant Neoplasm Mortality Site I CD Code Ranjch Hand Lip, oral cavity, pharynx (140-149) Digestive organs, peritoneum (150-159) Respiratory, intrathoracic (160-165) Bone, connective tissue, skin, breast (170-175) Genitourinary organs (179-189) Brain (191-192) Lymphatic and hematopoietic tissue (200-208) No site specification (199) Comparison 18 4 9 17 0 1 0 t 3 3 0 _!_ 5 J_ 6 Total 0 1 3 43 �Table 21 Morphology of Neoplasms ICD Code 9th Ed. M800 Deaths Ranch Hand Comp Nomenclature Neoplasms not otherwise specified (NOS) 0 Brain M801-804 1 Bronchus and Lung Colon Esophagus Pancreas Intestinal Tract Unspecified site Epithelial neoplasms (NOS) 0 1 8 0 Kidney 1 Nasopharynx 0 0 Pancreas M872-879 M905 M938-948 M959-963 M964 M965-966 M986 Prostate Unspecified site Colon Papillary and Squamous Cell Nasal Sinus Lip Tongue Lung Tonsil Adenomas and Adenocarcinomas Appendix Bronchus and Lung Colon Kidney Stomach Pancreas Nevi and Melanomas Skin (NOS) Mediaetinal Trunk Mesothelioma Bronchus and Lung Gliomas Frontal Lobe Brain (NOS) Lymphomas NOS and Diffuse Lymphomas (NOS) Reticulosarcoma Malignant lymphoma histiocytic, (NOS) Hodgkin's disease Hodgkin's (NOS) Myeloid Leukemias Acute Myelocytic Leukemia Total 19 0 1 1 1 1 1 0 2 0 0 0 0 0 0 0 0 0 1 1 Stomach M814-838 1 0 0 0 0 0 Bronchus and Lung Esophagus M805-808 ] 5 1 0 0 0 0 0 0 1 1 1 0 1 1 2. 1 2 1 0 0 1 0 0 0 1 0 0 0 0 1 0 2 J) JL 6 43 �-*• Noncause-Specific Comparisons with Ex t erna 1 Pgpu 1 a t i ons It is important to know not only how Ranch Handers and their matched comparisons relate to each other, but also how their mortality rates compare with other military and civilian populations in the United States. These contrasts are used in an attempt to place the study groups in perspective with the overall mortality experience of known populations. Given the selection factors Involved for entry to and retention in the military service, it is anticipated that the study groups would exhibit lower mortality than the U.S. White male population but poorer mortality than the active duty Air Force population. Similarly, they might be expected to be more equivalent to the DOD retired personnel or occupational cohorts such as the U.S. civil service. In this report, the mortality experience of Ranch Handers and their matched comparisons is compared with the expected death rates with reference to the 1978 U.S. White Male Life Table ( ) the 1978 Department of Defense 6, period life tables for nondisability retired military officer and enlisted personnel ( ) 1979 active duty Air Force officer and enlisted personnel life, tables (8) and a 7, 1974 U.S. active male civil service life table (9) and the West Point class of 1956 (10). 5 ] • In Tables 22 and 23, Ranch Hand officers and comparison group officers are contrasted to a 1978 DOD nondisability retired officer life table and in Tables 24 and 25, Ranch Hand and comparison group enlisted personnel are compared with a 1978 DOD nondisability retired enlisted life table. In each table, the column labeled "At Risk" lists the number of subjects entering each five-year age interval, the column labeled "Deaths" tabulates the number of deaths in the age intervals and the column labeled "Expected Deaths" gives the expected number of deaths in the age intervals of the study subjects if they had experienced the same death rates as those specified by the DOD table. The value of the test statistic for testing the null hypothesis of equality of the study and referenced life table is denoted by T; its two-sided p-value is denoted by P. While each table summarizes the findings with five-year age intervals for ease of presentation, one-year age intervals were used for the actual computation of the statistic T. A negative value of T means that the study cohort has lived longer than expected relative to the reference population. All contrasts are unadjusted for race since the DOD tables are not race-specific. All analyses are conditioned on survival to age 35, since the DOD tables do not begin until that age., The totals in Tables 22 through 25 do not, therefore, agree with Table 1. 20 �Table 22 Ranch Hand Officer Versus DOD Nondisability Retired Officer Life Table (T=-4.494, P<0.001) Age At Risk Deaths Expected Deaths 36-39 40-44 45-49 50-54 55-59 60-64 65-69 459 386 309 209 2 1 1 1 1 2 0 3.568 4.469 5. .021 .847 3. 2.320 1.212 0.141 71 36 5 20.578 Total Table 23 Comparison Officers Versus DOD Nondisability Retired Officer Life Table (T=-3.288, P<0.001) Age At Risk 35-39 40-44 45-49 50-54 55-59 60-64 65-69 2264 1924 1448 Deaths Expected Deaths 12 13 24 14 10 4 _0 77 988 367 170 33 Total 21.630 21.892 23.808 19.291 11.860 6.144 1.158 105.783 Table 24 Ranch Hand Enlisted Personnel Versus DOD Nondisability Retired Enlisted Life Table (T=-0.220, P=0.826) At Risk 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-71 Total 735 432 311 182 54 23 9 2 Deaths Expected Deaths 7 5 6 6 2 3 0 562 999 163 228 2.553 1.774 0.779 0.118 30 31.176 21 �Table 25 Comparison Enlisted Personnel Versus DOD Nondisability Retired Enlisted Life Table (T=-3.731, P<0.001) Age At Risk Deaths £2E£££S.(jL5£atlLs 35-39 3611 21 37.166 40-44 2117 20 29.397 45-49 50-54 55-59 60-64 65-69 70-74 1534 907 258 116 46 7 36 17 14 4 2 0 35.350 25.533 12.545 9.005 3.638 0.908 0 0.065 114 153.607 75-76 2 Total Tables 22 and 23 show highly favorable mortality experiences for Ranch Hand and comparison officers. Conditioned on survival to age 35, they are living significantly longer than expected using the DOD death rates (p<0.001 and p<0.001, respectively). Tables 24 and 25 show that Ranch Hand enlisted personnel are experiencing mortality patterns similar to the DOD retired enlisted population (p=0.826), and comparison enlisted are living significantly longer (p< 0.001) than the DOD nondisability retired enlisted population (conditioned on survival to age 35). This, together with the nonsignificant logrank value for Ranch Hand versus comparison enlisted personnel shown in Table 5 (p=0.522), suggests that the Ranch Hand versus comparison officer and enlisted contrasts change with age at death. A view of this is seen in Table 26, which shows linear rank test results, comparing Ranch Handers and comparisons conditioned on survival to age 35 (analogous to Table 5). Table 26 Test Results and P-values for Noncause-Specific Survival Conditioned on Survival to Age 35 Group Logrank (value) P-value Wilcoxon (value) P-value Officer Enlisted Flying Ground (-1.741) ( 1.379) (-1.331) ( 1.535) 0.082 0.168 0.183 0.125 (-1.879) ( 1.345) (-1.440) ( 1.491) Total (-0.033) 0.974 (-0.110) 0.913 22 0.060 0.179 0.150 0.136 �Categorical analyses reveal the interaction suggested by the Ranch Hand versus U.S. White male contrasts. These are shown in Tables 2.7 and 28 where survival status (alive, dead) is analyzed as a function of group (Ranch Hand, comparison) and rank (officer, enlisted) on deaths under 35 years of age and separately, on deaths over 35 years of age. Table 27 Death Before 35, Ranch Hand Versus Comparisons (Group By Rank By Status Interaction: P=0.043) Status Rank Officer Grout Alive Dead Total Ranch Hand 459 7 466 Comparison 2264 14 2278 Ranch Hand 781 9 790 Comparison 3833 60 Relative Risk 3893 2.44 Enlisted 0.74 Table 28 Death After 35, Ranch Hand Versus Comparisons (Group By Rank By Status Interaction: P=0.019) Status Rank Officer Group Alive Dead Total Ranch Hand 451 8 459 Comparison 2187 77 2264 Ranch Hand 705 30 735 Comparison 3497 114 361] Relative Risk 0.51 Enlisted 1.29 23 �In Table 28, the Ranch Hand versus comparison contrast in the officer category is significantly different from the corresponding contrast in the enlisted category. This suggests that, among those surviving to age 35, Ranch Hand officers are experiencing fewer deaths (relative risk =-- 0.51) than their matched comparisons while the Ranch Hand enlisted are experiencing more deaths than their matched comparisons (relative risk = 1.29). Death rates are shown in Table 29. The rate that is most apparently different is the low Ranch Hand officer death rate for those officers who survived to age 35. This low rate may parallel the favorable mortality experienced by those Ranch Hand officers born before 1935. Further analyses will attempt to clarify these patterns, with specific attention to cause of death. Table 29 Death Rates as a Function of Age at Death Death Rates Ranch Hand Officers En.listed Before Age 35 After Age 35 0.015 (N=466) 0.017 (N=459) Comparison Officers 0.006 (N=2278) 0.035 (N-2264) 0.014 (N=790) 0.041 (N=735) Enlisted 0.016 (N=3893) 0.033 (N-3611) 5.2 Comparison with Active Duty Air Force Life Tables The mortality experience of the Ranch Handers and their matched comparisons is contrasted with the total active duty 1979 Air Force life table, unadjusted for race, in Tables 30-35. Officers and enlisted personnel in the Ranch Hand and comparison cohorts are contrasted with active duty officer and enlisted Air Force life tables in Tables 32-35. Since the active duty Air Force life tables were accurate to only three significant figures, the expected deaths shown in Tables 30-35 are computed to three significant figures. In the active duty Air Force, individuals found to have major health deficits are quickly removed from the population by medical discharge or disability retirement. Hence, this external population is biased toward excellent health and favorable mortality. These contrasts are conditioned on survival to age 20 and death up to age 50. The totals in Tables 30 through 35 do not, therefore, agree with Table 1. Table 30 Ranch Handers Versus 1979 Active Duty Air Force Life Table (T-3.99, P 0.001) Age At Risk Deaths Expected Deaths 20-24 25-29 30-34 35-39 40-44 45-49 1256 1254 1247 1194 818 620 2 7 7 9 6 __7 5.04 5.06 3.7? 5.12 4.41 4.17 38 27.52 Total 24 �Table 31 Comparison Versus 1979 Active Duty Air Force (T-7.41, P<0.001) Ag£ At Risk 17-19 20-24 25-29 30-34 35-39 40-44 45-49 Deaths 6171 6169 6151 6122 5875 4041 2982 Deaths Expected 2 18 29 25 33 33 60 200 Total 18, 24, 24.8 153.20 Table 32 Ranch Hand Officers Versus 1979 Active Duty Air Force Officer Life Table (T=4.43, P<0.001) Age At Risk 25-29 30-34 35-39 40-44 45-49 Deaths 466 463 459 386 309 Expected Deaths 2.34 1.40 0.859 1.06 1.32 3 4 2 1 _L 11 Total 6.979 Table 33 Comparison Officers Versus 1979 Active Duty Air Force Officer Life Table (T=8.37, P<0.001) At Risk 25-29 30-34 35-39 40-44 45-49 Total 2278 2269 2264 1924 1448 Deaths Deaths Expected 9 5 12 13 24 11.4 6.86 4, 26 5. 15 6.25 63 33.92 25 �Table 34 Ranch Hand Enlisted Versus the 1979 Active Duty Air Force Enlisted Life Table (T=3.30, P<0.001) ASS. — M®™ 20-24 25-29 30-34 35-39 790 788 784 735 40-44 45-49 ^??J;!1JL 432 311 2. 4 3 7 Expected Deaths 3.17 3.18 2.31 3.57 5 __6 2.67 _JL-li 27 18.49 Total Table 35 Comparison Enlisted Personnel Versus the 1979 Active Duty Air Force Enlisted Life Table (T=6.42, P<0.001) Age At Risk 17-19 20-24 25-29 30-34 35-39 40-44 45-49 3893 3891 3873 3853 3611 2117 1534 Total Deaths Expected Deaths 2 18 20 20 21 20 _36 11.7 15.6 15.6 11.4 17.5 13.1 17.8 137 102.70 As expected, the central death rates for the active duty Air Force population are lower than those for the DOD nondisability retired population. In addition, it is expected that the Ranch Banders and their comparisons should He somewhere between these two reference populations, for reasons such as the healthy worker effect and the medical retirement of unfit individuals from the active force. This is, in fact, the case for Ranch Hand officers, comparison officers and comparison enlisted personnel. All three of these groups are living significantly longer than expected from the DOD life table, but are dying significantly sooner than expected relative to the active duty Air Force life tables. The exception to this pattern is seen in the Ranch Hand enlisted personnel who are experiencing mortality only equivalent to the DOD enlisted life table (p=0.826). They, like the other groups, are also having a significantly worse than expected mortality experience relative to the active duty Air Force enlisted life table (pcO.OOl). 26 �5 3 • Comparisons with the U.S.^Actiye Male Civil Service Life Table To further place the Ranch Handers and their matched comparisons in perspective, Ranch Handers, comparisons, and officer and enlisted personnel are 9 . These contrasted with the 1974 male active U.S. civil service life table ( ) contrasts are shown in Tables 36 through 41. There is no adjustment for civil service grade in these analyses. Therefore, socioeconomic factors may not be fully equivalent, especially in the analyses of the officer and enlisted subgroups. In future mortality updates, attempts will be made to account for the grade structure of the civil service population. Table 36 All Ranch Handers Versus U.S. Male Civil Service (T=0.140, P-0.889) Age At Risk 21-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-71 1256 1254 1247 1194 818 620 391 125 59 14 2 Deaths Expected Deaths JL 6.773 5.998 5.679 6.495 7.830 8.853 5,.907 3.,176 1,,758 0.463 0.054 54 52.997 2 7 7 9 6 7 7 3 5 0 Total Table 37 Comparison Versus U.S. Male Civil Service (T=-0.957, P-0.339) Age At Risk Deaths Expected Deaths 19-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-73 6169 6167 6149 6120 5873 4039 2980 1893 623 284 77 5 2 18 29 25 33 33 60 31 24 8 2 0 10.52.3 43.093 29.444 27.912 31.995 38.333 42.793 29.220 15.906 8.797 2.355 0.168 265 280.549 Total 27 �Table 38 Ranch Hand Officers Versus U.S. Male Civil Service ( — 1.728, P=0.084) T At Risk 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 466 463 459 386 309 209 71 36 5 Deaths Ex£ect«jd_De_aths 2.226 2.118 2.885 3.821 4.46] 3.239 1.886 0.917 0.099 15 Total 3 4 2 1 1 1 1 2 0 21.652 Table 39 Comparison Officers Versus U.S. Male Civil Service (Comparisons: T=-1.658, P=0.097) Age 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 Total At Risk 2278 2269 2264 1924 1448 988 367 170 33 Deaths 9 5 12 13 24 14 10 4 0 10.910 10.418 14.261 18.710 2.1.152 91 106.788 28 16.237 9.648 4.635 0.817 �Table 40 Ranch Hand Enlisted Personnel Versus U.S. Male Civil Service (1=1.661, P=0.097) A_g_e_ At Risk Deaths 21-24 25-29 790 788 2 4 30-34 784 3 3.561 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-71 735 432 311 182 54 23 9 2 7 5 6 6 2 3 0 _!_ 3.610 4.009 4.392 2.668 1.289 0.841 0.364 0.054 39 28.828 Total Expected Deaths 4.258 3.772 Table 41 Comparison Enlisted Personnel Versus U.S. Male Civil Service (T=1.528, P=0.127) Age At Risk 19-19 20-24 25-29 30-34 3891 3889 3871 3871 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-73 Total Deaths Expected Deaths 2 18 20 20 6.637 27.158 18.535 17.494 3609 21 17.733 2115 1532 905 256 114 44 5 20 36 17 14 4 2 0 19.623 21.641 12.983 6.258 4.162 1.538 0.168 174 153.930 The Ranch Banders and their matched comparisons are statistically quite close to the male civil service population. In these contrasts, the healthy worker effect is roughly equivalent although there is no adjustment for socioeconomic status. The contrasts of officers and enlisted personnel in the Ranch Hand and comparison cohorts with the male civil service reveal that the Ranch Hand and comparison 29 �officers are experiencing a slightly, but not significantly better mortality tban the civil service, with the Ranch Hand officers faring somewhat better than the comparison officers. Ranch Hand and comparison enlisted personnel are experiencing more mortality than the civil service with the Ranch Hand enlisted personnel faring slightly worse than the matched comparison personnel, but none of these observations are statistically significant. All of these findings are consistent with the linear rank testing shown in Table 5, the relative risks in Table 6 and the SMR's In Tables 8, 9, and 10. 5.4 Comparisons With the U.S. 1978 White Male Life Table Finally, the mortality experience of the non-Black Ranch Handers and their matched comparisons is contrasted with the 1978 U.S. White Male Life Table. Table 42 Non-Black Ranch Handers Versus the 1978 U.S. White Male Life Table (T=-4.828, P<0.001) Age At Risk 21-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-71 1180 1178 1172 1121 779 592 379 124 59 14 2 Total Deaths Expected Deaths 2 6 7 8 5 7 6 3 5 0 9.073 9.858 9.596 10.022 11.028 13.424 10.093 5.763 3.699 0.959 0.11.0 50 83.635 30 �Table 43 Non-Black Comparisons Versus the 1978 U.S. White Male Life Table (T=-12.286, P<0.001) At Risk 19-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-76 5816 5815 5799 5772 5537 3857 2846 1831 618 286 79 7 2 Total Deaths 1 16 27 23 31 29 53 31 22 8 2 0 0 Expected Deaths 10.325 55.444 48.592 47.336 49.594 54 . 105 64.837 49.932 28.956 18.756 5.228 0.845 0.063 430.324 243 Table 44 Non-Black Ranch Hand Officers Versus the 1978 U.S. White Male Life Table (T=-5.438, P< 0.001) Age At Risk 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 457 454 450 381 306 208 71 36 5 Total Dead Expected Deaths 3 4 2 1 1 1 1 2 0 3.819 3.735 4.620 5.585 6.981 5..633 3. .429 1.919 0.205 15 35.926 31 �Table 45 Non-Black Comparison Officers Versus the 1978 U.S. White Male Life Table (T—9.141, P<0.001) 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 9 5 12 13 24 14 10 4 0 Total 18.880 18.530 22.997 27.325 33.096 28.249 17.513 9.725 1.699 91 22.53 2244 2239 1899 1433 980 367 170 33 150.689 Table 46 Non-Black Ranch Hand Enlisted Personnel Versus the 1978 U.S. White Male Life Table (T=-1.585, P=0.113) Age At Risk 2.1-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-71 723 721 718 671 398 286 171 53 23 9 2 Total Dead Expected Deaths 2 3 3 6 4 6 5 2 3 0 1 5.556 6.039 5.861 5.402 5.443 6.443 4.460 2.334 1.779 0.754 0.110 35 44.181 32 �Table 47 Non-Black Comparison Enlisted Personnel Versus the 1978 U.S. White Male Life Table (T—6.393, P 0.001) Age At Risk Dead Expected Deaths 18-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-75 3563 3562 3546 3528 3298 1958 1413 851 251 116 46 7 2 1 16 18 18 19 16 29 17 12 4 2 0 _Q 6.325 33.938 29.713 28.806 26.597 26.780 31.741 21.683 11.443 9.031 3.529 0.845 0.063 152 230.494 Total The healthy worker effect Is an expected phenomenon In these data since Air Force veterans have been selected for active duty on the basis of health and technical ability. This effect is clearly evident in the overall contrasts shown in Tables 42. Both Ranch Banders and comparisons are seen to be living far longer than expected relative to the general U.S. White male population. The same effect is seen in both Ranch Hand and comparison officers (Table 44) and in comparison enlisted personnel In Table 47. The Ranch Hand enlisted personnel, however, are seen to be similar to the U.S. White male population (T--1.585, p=0.113); they are living longer than expected but not significantly so, In contrast to the other groups. The healthy worker effect is less evident in the Ranch Hand enlisted group, and this suggests that they are faring less well against the U.S. White male population than their matched comparisons. It is also important to note, in view of the poorer survival experience of Ranch Hand ground personnel, shown in Tables 5 and 6, that this group is closer to the U.S. White male population than the enlisted (Appendix Table 5), with an observed to expected death ratio of 0.883, based on 557 non-Black Ranch Hand ground personnel. Further, the corresponding finding for non-Black Ranch Hand enlisted ground personnel (Appendix Table 6), with an observed to expected ratio of 0.908 (based on 584 Ranch Hand enlisted ground personnel) suggests that the enlisted ground personnel may be experiencing adverse mortality that, while not significant relative to their matched comparisons, deserves close attention in fnture updates. 6. Comparisons with the West Point Study Group The mortality experience of Ranch Hand and comparison officers is also contrasted with the West Point Study Group. Although the West Point group is too small for all but very crude statistical comparisons, it does provide a useful benchmark for general mortality contrasts. 33 �The West Point Study Group consists of 474 members of the West Point Class of 1956. These men have been followed since that time for morbidity and mortality. All members of that class were, or still are* officers in the U.S. Armed Forces. The purpose of the West Point Study is to investigate the relationship between blood lipid levels and cardiovascular disease. Each study subject is physically examined biennially and blood samples are obtained for lipid and lipoproteln analyses at the USAF School of Aerospace Medicine (11). 6.1 Noncause-Specific Comparisons of Ranch Hand and Comparison Subgroups with the West Point Study Group No new deaths have occurred in the West Point (10) group since the baseline report and prior to 31 December 1983. The number of West Point deaths, therefore, remains at 36. For the purpose of these mortality comparisons, 15 of the 36 known West Point deaths occurring on or before 31 December 1983 were deleted; 9 of the 15 were killed in action, 1 was killed in 1959 in the line of duty and 5 were killed in automobile crashes prior to 1962. The rationale for these deletions is identical to that used for deaths of personnel killed In action from the Ranch Hand and comparison groups. Noncombat or accidental deaths prior to 1962 were deleted because death prior to 1962 would have precluded membership in the Ranch Hand or comparison group. In addition, one West Pointer is also a Ranch Hander and was removed from the West Point data base. That individual was alive on 31 December 1983. A summary of the remaining 21 deaths among the 458 West Point subjects used in these analyses is given in Table 48 and by age In Table 49. Table 49 lists the number of West Pointers at risk in each age group, the number alive on 31 December 1983, and the number dead. Table 48 West Point Deaths by Year-Of-Birth Year of Birth At Risk Dead 1930 1931 1932 1933 1934 1935 20 59 90 136 141 12 0 2 6 8 4 J_ Total 458 21 34 �Table 49 West Point Deaths by Age Alive 25-29 30-34 35-39 40-44 45-49 50-52 458 456 451 448 446 290 Dead 0 0 0 0 148 289 437 Total 2 5 3 2 8 _!_ 21 In this analysis, non-Black Ranch Hand and comparison officers are compared, without regard to cause of death, with the West Point Study group (all of the West Point subjects are non-Black). Non-Black Ranch Hand officers were matched, one-to-one, by year-of-birth, to West Point subjects. Due to the relatively small number of Ranch Hand officers and the limited year-of-birth range imposed by the age of the class of 1956, only 297 of the 458 West Point subjects received a matched Ranch Hander. Matched sets with West Pointers having the same year-of-birth were then merged to create six matched sets, corresponding to the six years-of-birth, 1930 through 1935. To compare West Pointers with comparison officers, 1368 non-Black comparison officers were matched to the 458 West Point officers, and these were then merged to six single-year-of-birth strata. Logrank tests were carried out on these matched data sets and the results are summarized in Table 50. In these analyses, survival time is age at death. SMR analyses, with the West Pointers being the standard, are shown in Table 51. Table 50 Non-Black Study Group Versus West Point Group Logrank P-values Contrasts P-value Ranch Hand officers versus West Point Comparison officers versus West Point 35 0.273 0.944 �Table SMR Comparison of Non-Black Ranch Hand and Comparison Officers With West Point (SMR=0.490) Ranch Hand Birth Year At Risk 1925-31 1932 1933-34 1935-40 98 35 60 107 Total Dead (SMR=1.00) West Point (SMB-0.790) Comparison Rate At Risk Dead Rate At Risk Dead Rate 0.020 0.029 0.017 0.037 500 164 257 _555. 38 7 6 JL3 0.076 0.043 0.023 0.023 79 90 2.77 12 2 6 12 _L 0.025 0.067 0.043 0.083 1476 64 458 21 300 2 1 1 4 In Table 51, the test for constant relative risk across year-of-birth strata has a p-value of 0.134, and a likelihood ratio test suggests that these groups are not different (p=0.306). The analyses shown in Tables 50 and 51 indicate that there is no significant difference between non-Black Ranch Hand and comparison officers and the West Pointers. 6.2 Cause-Specific Comparisons The cause-specific death counts for the West Point study group are given in Table 52. Table 52 West Point Mortality by Cause Deaths Cause Accidents Infectious Diseases Malignant Neoplasms Circulatory Digestive Genitourinary Ill-Defined Conditions 6 5 1 1 Total 21 Cause-specific comparisons are carried out for cancer (malignant neoplasms), other diseases, and nondiseases (accidents, suicides, homicides and ill-defined conditions), with an adjustment for year-of-birth by stratification on year-ofbirth. Relative risks are calculated using the method of Mantel and Haenszel (11). The results, based on the counts in Tables 53 and 54, are shown in Table 55. 36 �Table 53 Cause-Specific Comparisons Non-Black Ranch Hand Officers Versus West Point Ranch Hand Dead At Risk West Point Dead At Risk Cause Birth Year Mondisease 1925-1933 1934-1940 169 131 1 4 305 153 Cancer 1925-1930 1931 1932 1933 1934 1935-1940 74 24 35 36 24 107 0 0 0 0 0 0 20 59 90 136 141 12 0 1 3 1. 1 0 Other diseases 1925-1934 1935-1940 193 107 2 1 446 12 8 1 Table 54 Cause-Specific Comparisons Non-Black Comparison Officers Versus West Point Cause Birth Year Comparison At Risk Dead West Point At Risk Dead 1925-1931 1932 1933 1934-1940 500 164 148 664 18 2 1 6 79 90 136 153 1 2 2 1 Cancer 1925-1931 1932 1933 1934-1940 500 164 148 664 4 2 1 3 79 90 136 153 2 3 1 1 Other diseases 1925-1932 1933 1934 1935-1940 664 148 109 555 19 1 1 6 169 136 141 12 1 5 2 1 Nondisease 37 �Table 55 Cause-Specific Relative Risks, P-values and 95% Confidence Intervals for Relative Risk 95% Conf Int Rvalue Cause Comparison Re1 Risk Nondisease RH vs WP Comp vs WP 1.250 1.192 (0.257, (0.361, 6.072) 3.939) 0.782 0.774 Cancer RH vs WP Corap vs WP 0.551 (0.156, 1.949) 0.355 Other diseases RH vs WP Comp vs WP 0.446 0.951 (0.027, (0.131, 7.278) 6.916) 0.571 0.961 All causes RH vs WP Comp vs WP 0.475 0.882 (0.198, (0.189, 2.279) 4.100) 0.352 0.872 The Ranch Hand versus West Point cancer comparison cannot be assessed using the Mantel-Haenszel procedure due to the lack of cancer deaths in the Ranch Hand officer group. The overall and cause specific equivalence of these study groups and the West Pointers suggest that these analyses do not contribute enough to this study to warrant yearly reporting. 7' Further Covariate Adjustments Some of the contrasts shown in previous sections in this report are further analyzed here using information about the Vietnam experience for Ranch Handers and comparisons. These analyses are motivated by the need for clarification of previous contrasts and should be viewed as preliminary to more complete analyses which will be presented in future reports. The information used here consists of (1) tour length and (2) a measure of cumulative exposure to dioxin. Tour length is defined as the cumulative time, in months, spent on assignment to Ranch Hand units by a Ranch Hander and to C-130 cargo units in SEA by a comparison. Cumulative exposure to dioxin, termed the "exposure index," is defined in the baseline morbidity report (12) and is proportional to the dioxin content of the herbicides being sprayed and inversely proportional to the number of persons sharing the workload with the subject to whom it is applied. Ranch Hand Tour^Length The effect of tour length on mortality will be investigated in detail in future reports. In this report, some descriptive statistics on tour length are presented, and tour length is used as a factor in some exposure analyses. Table 56 shows the 5, 50 and 95 percentiles of tour length in months for flying and ground personnel, and officers and enlisted personnel in Ranch Handers and the comparisons. 38 �Table 56 Tour Length Percentlies (In Months) for Ranch Handers and Comparisons Flying Status Percentiles 50% 95% Sample Size Flying Ground 13 13 19 16 439 26 Enlisted Flying Ground 4 5 12 13 22 19 206 582 Officer Flying Ground 12 11 20 17 48 44 2939 152 Enlisted Comparison 5 5 Flying Ground 10 10 20 19 52 48 1412 3767 In general, the comparisons had longer tour lengths than Ranch Handers. This is the result of longer tours of duty at noncombat zone bases (comparisons) relative to combat area bases (Ranch Hand). 7. 2 Ranch Hand Exposure Analysejs The effect, of exposure on mortality was assessed on the 1230 Ranch Handers having exposure information in a log-linear analysis based on survival (dead, alive), rank (officer, enlisted), year-of-birth (1905-1934, 1935-1954) and exposure (light, medium, heavy). These data are shown in Table 57. Table 57 Ranch Hand Mortality Adjusted for Year-Of-Birth, Rank and Exposure Exposure Rank Light Officer Enlisted Medium Officer Enlisted Heavy Officer Enlisted Birth Year 1905-1934 1935-1954 1905-1934 1935-1954 1905-1934 1935-1954 1905-1934 1935-1954 1905-1934 1935-1954 1905-1934 1935-1954 Survival Status Alive Total Dead 62 80 62 173 80 66 55 274 78 60 81 106 2 3 8 6 2 2 9 5 3 2 5 6 39 64 83 70 179 82 68 64 279 81 62 86 112 Death Rate 0.031 0.036 0.114 0.034 0.024 0.029 0.141 0.018 0.037 0.032 0.058 0.054 �There is no four-way Interaction in these data (p=0.304), there is no three-wav interaction involving survival and exposure and the two-way survival by exposure interaction is not significant (p=0,691). The survival hy year-of-birth by rank interaction is marginally significant (p~0.0627) and the year-of-blrth hv r«nk by exposure interaction is very significant (p 0.001). Both of these observations are expected from previous analyses of these data. Tn summary, survival is not affected by exposure, with or without adjustment for rank and year-of-birtb. A restriction of survival, exposure and adjusted for birth year a restriction to deaths new findings. the analysis to officers shows no relationship between birth year (p-0.967) or between survival and exposure (p=0.907) or unadjusted for birth year (p=0.905). Finally, after 35 years of age in non-Black Ranch Handera yields no A restriction of the analysis to enlisted personnel shows a significant survival by exposure by birth year interaction (p=0.044), indicating that the survival by exposure relationship within the 1905-1934 birth year cohort is significantly different from that of the 1935-1954 cohort. Classic dose-response patterns are not seen here so that a herbicide effect cannot be reliably inferred at this time. Future work will attempt to evaluate mortality patterns as a function of occupational subgroup in the ground cohort. This effort will require simulation studies and additional interviews to delineate differential exposure between occupational subgroups. Flight line duties and herbicide contact will be ascertained objectively along with additional medical risk factors, occupational exposures and socioeconomic factors. These analyses will be increasingly meaningful as the population ages and mortality rates permit use of more incisive statistical tools. Finally, joint morbidity-mortality analyses, adjusting for relevant covariates will be carried out. Future research will be directed at the development of statistical procedures which take the repeated testing aspect of these updates into account. The feasibility of using comparison data from the entire 1:8 design will also be studied. 9, u m m r v ^ Conclusion Evaluation of summary counts of death by rank and occupation did not reveal any statistically significant differences between the Ranch Hand and comparison groups. Other mortality analyses described in this report have revealed some differences in death experience between the herbici.de/dioxin exposed group, their matched comparisons and other external comparison groups. Overall mortality of the Ranch Rand group is nearly Identical to that of the comparison group, being 4.3%. Ranch Hand officers born between 1905 and 1935 have experienced fewer deaths than comparison officers born during the same era. On the other hand, Ranch Hand officers born after 1935 have experienced more death s than their comparisons. Although these differences within birth year strata are not statistically significant, this change in the group by survival status relationship with birth year is statistically significant (p=0.0?7). Additionally, Ranch Hand officers experienced fewer deaths after ape .35 years than did comparison officers, while �Panch Hand officers experienced more deaths before age 35 years than did comparisons. Further research will investigate whether there is any association between birth year and age of death and mortality patterns in these officer cohorts. At this time, Ranch Hand ground and enlisted personnel have experienced more mortality than their comparisons, but these differences are not statistically significant. Preliminary analyses using exposure indices have indicated no association between herbicide exposure in either the officer, enlisted, flying or ground Ranch Hand subgroups. Both Ranch Hand and comparison officers have experienced less mortality than Ranch Hand or comparison enlisted personnel. Ranch Hand flying personnel have experienced less mortality than Ranch Hand ground personnel, while comparison flying and ground personnel have experienced similar mortality patterns. Examining causes of death, Ranch Hand officer and flying groups have experienced fewer deaths from cardiovascular disease and cancer than have the comparisons, but this difference is not statistically significant. No apparent specific disease excesses were noted in the Ranch Hand ground or enlisted groups relative to their comparisons. All Ranch Hand cohorts are elevated in the category of digestive system deaths, but this difference is not statistically significant. There was a single case of soft tissue sarcoma in the comparison group and no cases occurred in the Ranch Banders. The Ranch Hand and comparison groups have been contrasted with five comparison groups. Ranch Hand and comparison officers are experiencing significantly less mortality than U.S. White males and DOD retired officers. Comparison enlisted personnel are similarly experiencing significantly less mortality than U.S. White males and DOD retired enlisted. Ranch Hand enlisted personnel have experienced a mortality rate not statistically distinguishable from U.S. White males or DOD retired enlisted personnel. The Ranch Hand and comparison groups taken together have experienced a mortality pattern not statistically different from civil service employees. However, all Ranch Hand and comparison groups are experiencing significantly more mortality than the active duty Air Force, as would be expected by active duty Air Force health qualification standards. Finally, no significant differences between Ranch Hand or comparison officer death rates and those of West Point officers from the class of 1956 were detected. In conclusion, summary counts of death by rank and occupation did not reveal any statistically significant differences between the Ranch Hand and comparison groups. However, Ranch Hand officers born between 1905 and 1935 have experienced favorable mortality relative to their comparisons while the converse is true for officers born after 1935. Analogous patterns are seen in officers conditioned on age at death. Although Ranch Hand ground personnel have experienced unfavorable mortality relative to comparisons irrespective of date of birth or age at death, this difference is not statistically significant. Exposure index analyses indicate that these mortality rate differences cannot be attributed to herbicide exposure at this time. These analyses have identified several findings of interest, which will be further evaluated In future mortality updates. 41 �References 1. Lathrop, G. D., Moynahan, P. M., Albanese, R. A., Wolfe, W. H. (1983). An Epidemiologlc Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Baseline M o r t a l i t y Study Results. (NTIS Order Number: AD-A130 793) 2. Kaplan, E. L. and Meier, P. (1958). Nonparametric estimation from incomplete observation. JoujJ1^L_2l _?nJr Amer1can. Statist teal Association 53:457-481. "~ ~ 3. Prentice, R. L, (1978). Linear rank tests with right censorc-ci data. Biometrika 65:167-179. 4. Ejigou, A. and McHugh, R. (1981). Relative risk estimation under multiple matching. Biometrika 68:85-91. 5. Gail, M. (1978). The analysis of heterogeneity for indirect standardized mortality ratios. Jovrnal._of _the_ RojaJ^_S£aJtJ_sjtjk:jil Society, A, 141:224-234. 6. Vital Statistics of the United States, 1978 Vol II - Section 5, Life Tables; US Dept of Health and Human Services, DHHS Publication No (PHS) 81-1104; Hyattsville, Maryland; 1980. 7. Evaluation of the Military Retirement System FY 1980. Office of the Actuary, Defense Manpower Data Center, 300 North Washington Street, Alexandria, Virginia 22314. 8. Servicemen's and Veterans Group Life Insurance Programs, Sixteenth Annual Report, Year Ending June 30, 1981. VA Regional Office and Insurance Center, Philadelphia, PA 19101; 1981. 9. Board of Actuaries of the Civil Service Retirement System, Fifty Seventh Annual Report. US Government Printing Offices; 1980. 10. Clark, I)., Allen, M. and Wilson, F. (1967). Longitudinal study of serum lipids - twelve year report. American Journal of Nutrition 20:743-752. 11. Mantel, N. and Haenszel, W. (1959). retrospective studies of disease. Institute 22:719-748. 12. Statistical aspects of data from Journal of the National Cancer Lathrop, G. D., Wolfe, W. H., Albanese, R. A., Moynahan, P. M. (1984). An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Baseline Morbidity Study Results. (NTIS Order Number: AD-A138 340) 42 �PRINCIPAL INVESTIGATORS George I). Lathrop, MD, MPH, PhD, FACPM Colonel, USAF, MC Chief, Epidemiology Division Will!am H. Wolfe, MD, MPH, FACPM Colonel, USAF, MC Chief, Epidemiology Division Richard A. Albanese, MD, GM-15 Chief, Momathematical Modeling Branch Data Sciences Division Patricia M. Moynahan, BSN, MS Colonel, USAF, NC Chief, Occupational Epidemiology Section Epidemiology Division Joel E. Michalek, PhD, GS-13 Mathematical Statistician Data Sciences Division Retired 1 October 1984 �Appendix Table 1 Ranch Hand Officers Versus Ranch Hand Enlisted Mortality by Year-Of-Birth (SMR - 0.483, PI = 0.204, P2 = 0.(M1) Birth Year Ranch Hand Officers At Risk Dead Rate 1905-1924 1925-1934 1935-1939 1940-1944 1945-1954 41 194 95 91 45 3 4 4 2 2 Total 466 15 Ranch Hard Enlisted At Risk Dead Rate 29 195 115 119 332 7 15 3 3 11 790 0.073 0.021 0.042 0.022 0.044 0.241 0.077 0.026 0.02.5 0.033 39 Appendix Table 2 Comparison Officers Versus Comparison Enlisted Mortality by Year-Of-Birth (SMR = 0.663, PI = 0.81.1, P2 = 0.003) Birth Year 1905-1919 1920-1924 192.5-1929 1930-1934 1935-1939 1940-1944 1945-1954 Total Officers At Risk Dead Rate 44 161 290 640 458 495 190 4 13 20 31 12 6 5 2278 At Risk Enlisted Dead 66 80 2.11 749 562 601 1624 11 11 24 42 24 17 45 3893 91 0.091 0.081 0.069 0.048 0.026 0.012 0.026 Rate 174 0.167 0.138 0.114 0.056 0.043 0.028 0.028 Appendix Table 3 Ranch Hand Flying Personnel Versus Ranch Hand Ground Personnel Mortality by Year-Of-Birth (SMR - 0.548, PI - 0.376, P2 = 0.052) Birth Year At Risk Flyers Dead 1905-1924 1925-1934 1935-1939 1940-1944 1945-1954 44 272 145 121 64 4 9 6 2 2 Total 646 2.3 0.091 0.033 0.041 0.017 0.031 44 Ground Dead Rate 26 117 65 89 313 6 10 1 3 11 0.231 0.085 0.015 0.034 0.035 610 Rate 31 At Risk �Appendix Table 4 Comparison Flying Versus Comparison Ground Personnel Mortality by Year-Of-Blrth Within Comparison Group (SMR - 0.926, PI - 0.607, P2 - 0.782) At Risk Flyers Dead 6 17 25 53 24 14 10 1905-1919 1920-1924 1925-1929 1930-1934 1935-1939 1940-1944 1945-1954 Total 3163 Rate Ground At Risk Dead 0.021 65 66 151 423 322 443 0.036 1538 9 7 19 20 12 9 40 3008 Rate 116 0.133 0.097 0.071 0.055 0.034 149 0.138 0.106 0.126 0.047 0.037 0.020 0.026 Appendix Table 5 Non«-Black Ranch Hand Ground Personnel Versus the 1978 U.S. White Male Life Table ( - -0.728, P - 0.466) T Age At Risk Dead 21^24 25^29 30*3^ 35*39 40*44 45>-49 50A-54 55*59 60«64 65*69 70*-71 557 555 552 504 255 179 117 45 22 10 2 2 3 4 4 3 4 3 2 3 0 J_ Total 29 45 Expected Deaths 3.236 32.824 �Appendix Table 6 Non~Black Ranch Hand Enlisted Ground Personnel Versus the 1978 U.S. White Male Life Table (T - ^0.549, P - 0.583) Age At Risk Dead Expected Deaths 21*2»» 25*29 30-31 35^-39 40M4 15-M9 50»5H 55"59 60^61 6S--69 70K71 Total 532 530 527 180 235 169 110 ^1 21 9 2 2 3 3 1 3 4 3 2 3 0 J_ 28 1.085 4.137 *».281 3.^92 3.193 3.927 2.991 1.881 1.671* 0.75^ 0.110 30.828 46 � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 059 Folder The folder containing the original item. 1576 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. Patricia M. Moynahan Richard A. Albanese William H. Wolfe Joel E. Michalek Description An account of the resource <strong>Corporate Author: </strong>United States Air Force School of Aerospace Medicine (AFSC), Epidemiology Division and Data Sciences Division, Brooks AFB, Texas Date A point or period of time associated with an event in the lifecycle of the resource December 10 1984 Title A name given to the resource An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Mortality Update - 1984 Subject The topic of the resource Air Force Health Study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/5d9cf1d1d4bbd644387fc1c0fc067587.pdf e6a4c318a390c44a4b2e0659f5f9e40b PDF Text Text Item ID Number °1573 Author Lathrop, George D. Corporate Author United States Air Force School of Aerospace Medicine, Report/Article HUB ^n Epldemiologlc Investigation of Health Effects In Air Force Personnel Following Exposure to Herbicides: Baseline Morbidity Study Results Journal/Book Title Year 1984 Month/Day February 24 Color Number of Images D 362 Descriptor! Notes Wednesday, May 23, 2001 Page 1574 of 1608 �AIR FORCE HEALTH STUDY (PROJECT RANCH HAND II) AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES BASELINE MORBIDITY STUDY RESULTS 24 FEBRUARY 1984 Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 George D. Lathrop, Colonel, USAF, MC William H. Wolfe, Lieutenant Colonel, USAF, MC Richard A. Albanese, M.D., GM-15 Patricia M. Moynahan, Colonel, USAF, NC Approved for public release; distribution unlimited. USAF SCHOOL OF AEROSPACE MEDICINE A e r o s p a c e Medical Division (AFSC) Brooks Air Force Base, T e x a s 7 8 2 3 5 �UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE (When Data Entered) REPORT DOCUMENTATION PAGE 1. REPORT NUMBER 2. GOVT ACCESSION NO 4. TITLE (and Subtitle) 7l READ INSTRUCTIONS BEFORE COMPLETING FORM 3. RECIPIENT'S CATALOG NUMBER 5. TYPE OF REPORT & PERIOD C O V E R E D An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides. Baseline Morbidity Study Results . ., fteo'r^D. Lathrop, Colonel, USAF, MC William H. Wolfe, Lt Colonel, USAF, MC Richard A. Albanese, MD, GM-15 Patricia M. Moynahan, Colonel, USAF, NC Interim 1979-1982 6. P E R F O R M I N G ORG. REPORT N U M B E R 8. CONTRACT OR G R A N T NUMBER(» 10. PROGRAM ELEMENT. PROJECT, TASK A R E A & WORK UNIT NUMBERS 9. P E R F O R M I N G O R G A N I Z A T I O N N A M E AND ADDRESS USAF School of Aerospace Medicine (EK) Aerospace Medical Division (AFSC) Brooks Air Force Base, Texas 78235 12. REPORT D A T E 11. CONTROLLING OFFICE NAME AND ADDRESS February 1984 The Surgeon General United States Air Force Washington, D.C. 20311* 13. N U M B E R OF P A G E S T4. MONITORING AGENCY NAME 4 ADDRESS^/ different from Controlling Office) 15. SECURITY CLASS, (of this report) Unclassified 15«. DECLASSIFICATION/DOWNGRADING SCHEDULE 16. DISTRIBUTION STATEMENT (of this Report) Approved for public release; distribution unlimited. 17. DISTRIBUTION STATEMENT (of the abstract entered in Block 20, It different from Report) 18. SUPPLEMENTARY NOTES 19. KEY WORDS (Continue on reverse aide It necessary and identity by block number) Dioxin Ranch Hand Air Force Health Study Epidemiologic Investigation Phenoxy Herbicides Herbicide Orange Morbidity 20. ABSTRACT (Continue on reverse aide It necessary and Identify by block number) In 1979 the United States Air Force (USAF) made the commitment to Congress and to the White House to conduct an epidemiologic study of the possible health effects from chemical exposure in Air Force personnel who conducted aerial herbicide dissemination missions in Vietnam (Operation RANCH HAND). The purpose of this epidemiologic investigation is to determine whether long-term health effects exist and can be attributed to occupational exposure to herbicides. This study uses a matched cohort design in a nonconcurrent prospective setting, incorporating mortality, morbidity, and follow-up studies. This DO , ™NRM73 1473 UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE (When Data Entered) �UNCLASSIFIED SECURITY CLASSIFICATION OF THlS PAGfefWhen Data Entered) report presents the results of health Information on 2706 Ranch Handera and comparison individuals obtained by questionnaire and 2269 Ranch Handers and comparison individuals undergoing an extensive physical examination. This baseline report concludes that there is insufficient evidence to support a cause and effect relationship between herbicide exposure and adverse health in the Ranch Hand group at this time. The study has disclosed numerous medical findings, mostly of a minor or undetermined nature, that require detailed follow-up. In full context, the baseline study results should be viewed as reassuring to the Ranch Handers and to their families at this time. UNCLASSIFIED SECURITY CLASSIFICATION OF THIS �EXECUTIVE SUMMARY BASELINE MORBIDITY STUDY The Ranch Hand II epidemiologic study uses a matched cohort design in a nonconcurrent prospective setting, and incorporates mortality, morbidity, and follow-up studies. The purpose of this report is to present the baseline morbidity study. The morbidity study design matched each living Ranch Hander (by age, job, and race) to the first living and compliant member of a randomly selected comparison mortality set of 5 individuals, producing a 1:1 contrast. The comparison group was formed from numerous flying organizations which transported cargo to, from, and within Vietnam, but were not involved in aerial spray operations of Herbicide Orange. Of the potential study participants, 99.5$ were located.. Early in the physical examination phase of the study, it was discovered that 1856 of the entire comparison group was ineligible to participate because of inappropriate selection. Thereafter, study eligibility was certified only after a hand-review of personnel records. Next-in-line compliant comparisons entered the study as replacements after fully completing the questionnaire and physical examination. Statistical analyses of these replacement individuals later showed that they differed from the original comparisons in a variety of subtle and often opposite ways. As a conservative measure to avoid possible bias by the inclusion of the replacements in the analyses, a management decision was made to base the statistical tests in this report primarily upon contrasts of the Ranch Hand group to the original comparison group. The preponderance of data was obtained from the in-home interviews and the physical examination, each conducted under contract to the Air Force by Louis Harris and Associates, Inc., New York NY, and the Kelsey-Seybold Clinic, P.A., Houston TX, respectively. All contacts with the participants were carried out with utmost professionalism and sensitivity. Other morbidity data sources included reviews of medical records, military personnel documents, and birth certificates; in-home questionnaires and telephone questionnaires of the study participant's wives, former wives and, occasionally, their next-of-kin. All aspects of the study were voluntary. As a contract requirement, data collection personnel were blind as to the exposure status of the participants. Ninety-seven percent of the Ranch Handers and 93% of the comparisons participated in the in-home interview. For the physical examination, 87% of the Ranch Handers and 76$ of the comparison group participated, a total of 2,272 individuals. This differential attendance at the examination may have introduced a potential participation bias that, in a military population predominantly engaged in flying duties, is multifactorial and complex. All study phases were monitored by stringent quality control standards. Statistical analyses of the data consisted primarily of log-linear models, logistic regression techniques, generalized linear models, matched covariate analyses, and Kolmogorov-Smirnov, chi-square, and t tests. The physical examination and the in-home questionnaire data were analyzed by major organ system. In terms of general health, more Ranch Handers perceived themselves to be in fair or poor health than did their comparisons. No �group difference^ were detected for hewatocrit or percent body fat determinations. Unadjusted group differences in sedimentation rate were not observed; however, significantly more young comparisons had abnormalities in sedimentation rate than did their Ranch Hand counterparts, There were no statistically significant differences in the occurrence of malignant or benign systemic tumors between the groups. One cage of soft tissue sarcoma was found in a comparison member, Significantly more nqnmelanQtic gkin cancer was noted in the Ranch Hand group, but these analyses have not yet considered (adjusted for) sunlight exposure, the prime etiology of these cancers. Such nonmelanotic skin cancer (predominantly basal cell carcinoma) is the most common neoplasm in the White population of the United States, Up to the statistical limits of the study there were no consistent data that showed that the Ranch Handers were developing uncqmmpn cancers, or> cancer in unusual sites, or at an unusual age. Measures of fertility and reproductive outcome showed mixed results. It is emphasized that the fertility and reproductive results are preliminary at this time as they are based largely upon subjective self reports that await full medical record and bir^h certificate verification. Four measures of fertility: number of childless marriages, couples with the desired number of children, the infertility index and the fertility index, showed no difference between the Ranch Hand and comparison groups, A semen specimen obtained from those willing and able to provide one showed no group differences with respect to total sperm count or percent abnormal sperm. There were no significant findings in conception outcomes for miscarriages, stillbirths, induced abortions, or live births. For live birth pytcpmps no differences ..were observed for prematurity, learning disability, or infant deaths. There, was no significant disparity between groups for the classificat.ions; of severe or moderate birth defects. By parental history, however, Ranch Hand offspring showed significantly more minor birth defects (birth marks, etc). Reported neonatal deaths and physical handicaps were also significantly excessive in the Ranch Hand group when contrasted to the tot.al comparison group. All fertility and reproductive findings in the Ranch Hand: group showed inconsistent relationships to the herbicide exposure index. Medical records and birth, certificates are currently being chronicled for cpmplete verification of all historical findings. A comprehensive neurological examination showed: no consistent abnormalities in the cranial nerves, peripheral nerves or central nervous system function of the Ranch Handers. As expected, there was a profound influence of diabetes and alcohol in both groups upon numerous neurological tests. Detailed psychologic data were obtained on all participants at both the in-home interview and the physical examination. It is emphasized that, the majority of psychological data was derived from self reported responses during interview and has not been fully assessed for the effect of differential reporting. A variety of subjective deficits (fatigue, anger, fear, anxiety, etc) were Significantly more common in the high school educated Ranch Handers,. Educational level, significantly and consistently influenced rapst subjective test results. In sharp contrast, more objective performance testing by the Halstead-Reitan. battery and IQ testing did not reveal any significant iritergroup differences.. The roles of overreporting and the Post Vietnam; §tres,s Syndrome in these, analyses have not as yet been assessed. Li.ver function tests and clinical history data showed mixed results. Ranch. Handers had. some elevated liver enzyme tests, and lower cholesterol levels. More Ranch Handers were found to have hepatomegaly and verified histories of prior hepatic disease, than their counterpart comparisons. Exposure to alcohol, degreasing chemicals, and industrial chemicals in general, influenced ii �the liver test results. Ranch Handers reported significantly more symptoms resembling porphyria cutanea tarda than the comparisons, but these data have not been verified by medical record reviews nor were they substantiated by laboratory testing or by physical examination. Exposure index analyses were essentially negative. In the dermatologic evaluation, no cases of chloracne were diagnosed clinically or by biopsy. A thorough questionnaire analysis of acne showed that the incidence, severity, duration, and anatomic location did not differ between groups, and suggested that the historical occurrence of chloracne was highly unlikely in the Ranch Handers. Evaluation of the cardiovascular system showed equal proportions of abnormalities in blood pressures, electrocardiograms, past electrocardiograms, and heart sounds in both groups. Ranch Handers are not having premature heart attacks or generalized heart disease. However, the Ranch Handers showed significant deficits in 2 specific peripheral leg pulses and all leg pulses as a group. These puzzling findings were highly correlated with age and smoking patterns, and verified past heart disease. The assessment of the immune system by laboratory testing was compromised by excessive test variability. An independent review committee determined which test data were suitable for statistical analysis. As an unexpected finding, the test data were significantly influenced by the age and smoking history of the participant; no group differences were detected after adjustment for these factors. A hematologic test battery revealed.three red cell abnormalities in the Ranch Hand group, but these were difficult to place into a clinical or epidemiologic context. Evaluation of renal, pulmonary, and endocrine functions generally disclosed small and inconsistent proportions of abnormalities between groups, and were deemed clinically unimportant. An unrefined assessment of all summed and weighted organ system abnormalities by group did not show an aggregation of multisystem disease or malfunction. Any interpretation of these study data, in whole or in part, must carefully consider the methodical steps required for a proper inference of causality. It is specifically pointed out that many group differences were largely based upon subjective data, and that a subtle effect of differential reporting is suggested but has not been fully evaluated. For objective data, group differences were generally within normal ranges and were not correlated to the herbicide exposure index, nor fell within the expected latency periods following Vietnam service. The proposed clinical end points of dioxin exposure, chloracne, soft tissue sarcoma, and porphyria cutanea tarda, were not found in the Ranch Hand group (study power limitations recognized). Overall, substantial credence is given to the objective study findings, particularly after observing the consistent duplication of the classical effects of risk factors such as age, smoking, alcohol, etc., in almost all clinical areas. Additional work with these baseline data is still required in the areas of data base refinement, statistical testing and bias analysis, exposure index refinement, establishment of the follow-up examination requirements, and collaboration with other dioxin research studies. This baseline report concludes that there is insufficient evidence to support a cause and effect relationship between herbicide exposure and adverse health in the Ranch Hand group at this time. The study has disclosed numerous medical findings, mostly of a minor or undetermined nature, that require detailed follow-up. In full context, the baseline study results should be viewed as reassuring to the Ranch Handers and their families at this time. iii �PREFACE In October 1978, the United States Air Force (USAF) Surgeon General made the commitment to the Congress and to the White House to conduct an epidemiologic study of the possible adverse health effects arising from the herbicide exposure of Air Force personnel who conducted aerial dissemination missions in Vietnam (Operation Ranch Hand). The purpose of this epidemiologic investigation is to determine whether long-term adverse health effects exist, and whether they can be attributed to occupational exposure to herbicides and their contaminants. The study protocol for this effort incorporates a matched cohort design placed in a nonconcurrent prospective setting. The study approach includes mortality, morbidity, and follow-up elements linked tightly in time in order to produce the most data in the shortest time. The study addresses the question: Have there been, are there currently, or will there be any adverse health effects among former Ranch Hand personnel caused by repeated occupational exposure to 2,1,5-Trichlorophenoxyacetic acid (2,4,5,-T) containing herbicides and the contaminant, 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)? At the request of the Principal Investigators (see Appendix I) the study protocol was extensively and independently reviewed. The review agencies included: The University of Texas School of Public Health, Houston TX; the USAF Scientific Advisory Board; the Armed Forces Epidemiological Board; and the National Research Council of the National Academy of Sciences. In 1980, the Science Panel of the Agent Orange Working Group was created as an additional peer review agency. This group, redesignated as the Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants, has consented to the oversight responsibility of the Ranch Hand study and continues to monitor the conduct of this epidemiologic investigation (see Appendix II). The Air Force Health Study (Ranch Hand II) protocol emphasizes the suboptimal statistical power of the mortality study. The mortality study was motivated by the desire to use a full spectrum epidemiologic approach to the herbicide question. Additionally, the investigators were scientifically obliged to pursue the mortality study because of previous and emerging studies (some with small sample sizes) which suggested the possibility of a soft tissue sarcoma end point (Honchar, 1981; Hardell, 1979; Erikson, 1979). Within the inherent sample size limitation Of the Ranch Hand population, detection of such a rare condition will be missed unless there is marked case clustering and correspondingly high relative risks. Also, because of sample size limitations as well as the myriad of proposed clinical end points, a case-control design was not entertained. In the morbidity phase of the study, the investigators have attempted to enhance statistical power and analytic sensitivity where possible by using (1) precise matching procedures with a replacement strategy to maintain statistical power while averting a loss-to-study bias, (2) exacting quality control procedures, (3) mortality-morbidity linkages, (4) a lengthy follow-up study, (5) state-of-the-art statistical methodology, (6) continuously distributed physical examination variables, and (7) data collection focused on verifiable end points. iv �The mortality analyses have not revealed any adverse death experience in the herbicide/dioxin exposed cohort. The results of the analyses were consistent: at this time, there is no indication that Ranch Hand personnel have experienced any increased mortality or any unusual patterns of death in time or by cause. They are not dying in increased numbers, at earlier ages, or by unexpected causes. The fact that only a relatively small number of Ranch Hand, deaths were available for analysis is reassuring in itself. However^ the fact that adverse effects have not yet been detected does not imply that an effect will not become manifest at a future time or after covariate-adjusted analyses. For this reason, further analyses are intended and mortality in the ' study population will be ascertained annually for the next 20 years. The morbidity portion of the study was conducted in two phases; an in-home, face-to-face interview, and a comprehensive physical and psychological examination. Both phases were conducted by civilian organizations under contract to the Air Force, using materials and procedures prescribed by the contract. One thousand, one hundred seventy four (97?) of the Ranch Hand group and 1,156 (93?) of the initially selected comparison group participated in the questionnaire. An additional 376 comparison subjects were interviewed as replacement subjects, bringing the total number of comparison participants to 1f532. Two thousand, seven hundred eight current and former wives of the study participants were interviewed. One thousand forty five (87?) of the Ranch Hand group participated in the physical examination, and 936 (76?) of the initially selected comparison subjects participated. Two hundred eighty-eight replacement subjects also participated in the examination process, giving a total of 2,269 participants, resulting in 1,024 matched pairs for analysis. The first chapter of this report is devoted to a discussion of the background of the study and the next seven chapters present a summary of the methodology used in gathering, analyzing, and interpreting the data. The results and discussion of these analyses, organized by organ system and/or disease end point, are contained in the remaining chapters. This report assumes that readers are familiar with statistical and epidemiologic techniques. It also assumes that the reader has a familiarity with the herbicide/dioxin issue and a detailed knowledge of the protocol of the Air Force study, the baseline questionnaire, and the baseline mortality results. In the interest of brevity, the reader is referred to the protocol published as US Air Force School of Aerospace Medicine Technical Report 82-44, the baseline questionnaires published as US Air Force School Aerospace Medicine Technical Report 82-42, and the Baseline Mortality Study Results, 30 June 1983. These reports are available from the National Technical Information Service, 5285 Port Royal Road, Springfield, Virginia 22161. �ACKNOWLEDGMENTS The Principal Investigators gratefully acknowledge the outstanding support given to this project by: The Ranch Hand Association and its elected officers, for sustained encouragement of the study, assistance in population ascertainment, and camaraderie and patriotism which contributed to unparalleled participation rates. Our peer review groups, the University of Texas, School of Public Health, the Air Force Scientific Advisory Board, the Armed Forces Epidemiological Board, the National Research Council, and our Advisory Committee for their scientific contributions which have facilitated the conduct of this study and enhanced public credibility. The Air Force Research Management Structure at the Aerospace Medical Division, Brooks AFB TX; 33°3<i Contracting Squadron, Air Training Command, Randolph AFB TX; Office of the Command Surgeon, Air Force Systems Command, Andrews AFB MD; and the Air Force Surgeon General's Office, Boiling AFB DC, for program advocacy and provision of resources. The over 100 professionals, consultants, technicians, military and civilian, whose dedication and hard work over the past five years have made this report possible. vi �TABLE OF CONTENTS Executive Summary Preface Acknowledgments Table of Contents I II Ill IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII XIX Background and Study Design Population Questionnaire Methodology Physical Examination Methodology Study Selection and Participation Bias Quality Control Procedures Statistical Methods Exposure Index Development General Physical Health Malignancy . Fertility and Reproductive Outcomes Neurological Assessment Psychological Assessment Evaluation of Hepatic Status. Dermatologic Evaluation Evaluation of Other Organ Systems 1. Cardiovascular Evaluation 2. Immunology 3. Hematological Variables 1*. Pulmonary Function and Disease. 5. Renal Disease and Function 6. Endocrine Function Individual Health Assessment. Future Commitments Interpretation of Study Results and Conclusions.. •••••• References Appendixes I II III IV V VI Principal Investigators and Key Personnel Advisory Committee on Special Studies Relating to the Possible Long-term Health Effects of Phenoxy Herbicides and Contaminants Contract Management Kelsey-Seybold Normal Value Report Blood Chemistry Definition of Birth Defects, Learning Disabilities, and Physical, Mental or Motor Impairments Physical Examination Forms �VII VIII IX X XI XII 1 XIII XIV XV XVI XVII XVIII XIX XX Examination Parameters and Abnormality Weights Used in Assessing Individual Health Total Mortality and Morbidity Study Site Specific Malignant Neoplasms General Health Analyses Using Data From All Comparisons Fertility and Reproductive Analyses; Ranch Handers versus All Comparisons Introductory Letters Occupational Category and Race of the Fully Compliant Population in Percent and Counts Self-Reported Reasons for Noncompliance to Questionnaire Self-Reported Reasons for Noncompliance to Physical Examination Coefficient of Variation for Tri-Level Controls Specific Rules for Entry Into the Morbidity Study Percent Compliance by Flying Code and Military Status of the Ranch Hand and Comparison Population Non-Black Officers Relative Risks (RR) and Mean Shifts (Y) for Selected Clinical End Points Spouse and Participant Reported Birth Defects Not Meeting Study Criteria Observed Cancer Versus SEER Data Expected in 117^ Partially Compliant Ranch Handers and 956 Original Comparisons �Chapter I BACKGROUND In January 1962, President John F. Kennedy approved a program to aerially disseminate herbicides in the Republic of Vietnam (RVN). This program, code named Ranch Hand, was conducted in support of tactical military operations and had 2 missions: defoliation and crop destruction. During the 9-year duration of the operation, approximately 19 million gallons of herbicides were sprayed on an estimated 10-20$ of South Vietnam (Young, 1978; Buckingham, 1982). Of the 6 herbicides used, Herbicide Orange was the primary defoliant, and approximately 11 million gallons were dispersed. Because of the controversial nature of the mission and enemy propaganda which raised political sensitivity to chemical warfare charges, the Ranch Hand operation was subjected to intense scrutiny from the start. Initial concerns were focused on the military, political, and ecological ramifications of the spray operations (Buckingham, 1982). Since 1977, the issue has shifted to a health concern. Numerous U.S. military personnel from all services have claimed exposure to herbicides, particularly Herbicide Orange and its dioxin contaminant, during their duty in the RVN. These possible exposures, coupled with claims of attributable adverse health, have resulted in class action litigation and substantial controversy within the Government, Veterans' groups, the scientific community, and the public. The U.S. Air Force Medical Service expressed its concern for the health of Air Force personnel exposed to herbicides in October 1978, when the Deputy Surgeon General, Major General Garth M. Dettinger, told the U.S. House of Representatives' Veterans Affairs Committee that the USAF would evaluate the health of Ranch Hand personnel. An epidemiologic study design was prepared by the USAF School of Aerospace Medicine to meet this commitment. Following extensive peer review, a final study protocol was published, (Lathrop, Wolfe, Albanese, Moynahan, 1982) and the epidemiologic study was initiated. Since 1978, numerous governmental agencies, universities, and industrial firms have planned or launched additional animal and human studies. An immediate scientific issue was identified in these studies, specifically, the characteristics of the RVN exposure. Succinctly, these questions are: (1) Who was exposed to which herbicide? (2) By what means can these individuals be accurately identified for study? (3) How much, or to what degree, were they exposed (route of administration, influence of personal hygiene measure, etc.)? These areas merit careful consideration because the process of population or exposure estimation may generate substantial misclassification errors that would call for inordinate sample sizes in a contemplated study. Government and civilian scientists and the Congress have recently inquired of the Air Force Health Study as to whether it might clarify the exposure controversy in ground personnel. The answer is a qualified yes. The dose-response principle suggests that if the Ranch Hand population was more exposed to herbicides and dioxin than ground personnel, then the Ranch Handers should manifest stronger and/or earlier indications of adverse health, if they have occurred or will occur in the future. This principle is constrained by statistical power but, as noted in Chapter VII, the Ranch Hand 1-1 �morbidity study has substantial power in some clinical areas. The fact is that the average Ranch Hander was substantially exposed to the herbicides and dioxin (relative to other military personnel in RVN) on almost a daily occupational basis. Exposure calculations have estimated that an average Ranch Hander in his tour received, at a minimum, 1000 times more exposure to Herbicide Orange than would an average unclothed man, standing in an open field directly beneath a spraying aircraft. Unfortunately, the relative degree of Ranch Hand exposure vis-a-vis ground personnel has been consistently undervalued, and even reversed by various advocacy groups and the media. It is our firm belief that the Ranch Hand population is the most herbicideexposed military cohort to have served in the RVN. The fact of the unequivocal exposure in a totally ascertained population, when matched to an equally clear-cut nonexposed cohort, provides as ideal an epidemiologic setting as possible from a wartime environment. Findings of adverse health, or lack thereof, in the Ranch Hand group should serve as a significant epidemiologic pointer to the health effects issue in exposed ground personnel. STUDY DESIGN This study uses a matched cohort design in a nonconcurrent prospective setting, incorporating mortality, morbidity, and follow-up studies. A detailed population ascertainment process has identified 1269 Ranch Hand personnel who served in the RVN during the period 1962-1971. A comparison group was formed by identifying all individuals assigned to selected Air Force organizational units with a mission of flying cargo to, from, and in the RVN during the same period. Complete details on the selection of the comparison population are cited in the study protocol. By a computerized nearest neighbor selection process, up to 10 comparison individuals were matched to each Ranch Hander by job category, race, and age to the closest month of birth. An average of 8.2 comparison individuals for each Ranch Hander were determined by record review to be fully suitable for study. From each matched comparison set, 5 individuals were randomly selected for the mortality study (1:5 design). Results of the Mortality Study were released to the public on 30 June 1983. Each living Ranch Hander and the first living member of his comparison set were selected to participate in a morbidity study consisting of an in-home interview and a comprehensive physical examination. Data collection for both the questionnaire and physical examination was accomplished by contract. The follow-up study consists of mortality and morbidity components. Every Ranch Hander and his set of comparisons will be the subjects of annual mortality updates for the next 20 years, so that any emerging mortality patterns or disease clusters may be detected with maximal sensitivity. In addition, follow-up questionnaires and physical examinations will be offered to all participants in subsequent years 3, 5, 10, 15, and 20, in order to bracket the latency periods associated with possible attributable disease. 1-2 �Chapter II POPULATION The exposed population, termed "Ranch Hand", was defined as those individuals who were formally assigned to the USAF organizations responsible for the aerial dissemination of herbicides and insecticides in the Republic of Vietnam from 1962 through 1971. These individuals were identified from historical data sources (morning reports, military personnel records, and historical computer tapes) at the National Personnel Records Center (NPRC), St. Louis, Missouri and the USAF Human Resources Laboratory, Brooks Air Force Base, Texas. A total of 126*1 Ranch Hand personnel were identified through this initial process. The comparison population was defined as those individuals who were assigned to a variety of cargo-mission organizations throughout Southeast Asia during the same time period. Cargo-mission aircrew members and support personnel were selected because of sufficient population size, similar training and military background experiences, and psychological similarities to the Ranch Hand group. The comparison population was not occupationally exposed to herbicides or insecticides in the Republic of Vietnam. Identification of this population (2*1,971 individuals) was completed using the same historical data sources as were used to identify the Ranch Hand population. 1. Original Match Before matching the Ranch Hand and comparison populations, all individuals killed in action (KIA) were removed from the data base. The rationale for their removal is the assumption that combat death in the Ranch Hand group was independent of herbicide exposure. Twenty-two Ranch Handers were identified as KIA. KIA's were also removed from the comparison group for comparability purposes. The remaining Ranch Hand population was matched to the comparison population with an iterative nearest-neighbor computer program (Lathrop, Wolfe, Albanese, Moynahan, 1982). This procedure attempted to match 10 comparison individuals with each Ranch Hander to the closest month of birth* race (Black versus non-Black), and occupational code (1-officer—pilot, 2-officer—navigator, 3-officer—nonflying, H-enlisted—flyer, and 5-enlisted—ground). Table II-1 presents the total number of study participants by occupation code, and race. �Table II-1 DISTRIBUTION OF THE INITIALLY MATCHED STUDY POPULATION BY OCCUPATION AND RACE Number Occupation Code Ranch Hand Comparisons Non-Black 1 2 3 4 5 3318 780 250 1871 5277 Officer-Pilot Officer-Navigator Officer-Nonflying Enlisted-Flyer Enlisted-Ground Subtotal 1167 11,496 Black 1 2 3 4 5 6 2 1 15 51 Officer-Pilot Off icer-Navigator Of ficer-Nonflying Enlisted-Flyer Enlisted-Ground Subtotal 75 1242 TOTAL The total Ranch Hand population consists of 37% officers and 63$ enlisted personnel. Seventy-seven percent of the total Ranch Hand officer population are pilots, 17$ navigators, and 6% other officers; 26$ of the total Ranch Hand enlisted population are flight engineers and 74$ are enlisted nonflying personnel. Following the match, the majority of Ranch Handers had 10 comparisons. The exceptions were the non-Black pilots who had a mean of only 9.5 comparisons per exposed individual due to the extreme ages of several individuals, and the Black pilots and other Black officers who had means of 2.7 and 5.0 comparisons, respectively. Six percent of the exposed population was found to be Black and88$ of this population was enlisted. Of these enlisted personnel 77$ were occupational code 5, Enlisted - Other. All subjects are males. The mean age of the study subjects is approximately 45 years. 2. Ineligibility In December 1981, the USAF Principal Investigators were advised by the questionnaire contractor that several comparison subjects had reported no experience in Southeast Asia, suggesting that inappropriate selection of some comparison subjects had occurred. Manual review of the comparison populations II-2 �military personnel records revealed that 18$ of the 12,193 comparison individuals in the original match were indeed ineligible for study. The inadvertent inclusion of several non-Southeast Asia military organizations had resulted in the selection of these inappropriate individuals. The percent loss to the total 1:1.0 matched comparison population due to ineligibility by occupation code, race, and average age is presented in Table II-2. Table II-2 PERCENT INELIGIBLE BY OCCUPATION CODE AND RACE, WITH AVERAGE AGE OF INELIGIBLES BY OCCUPATION CODE Race Non-Black Black Total Percent Loss and Occupation Code Counts of Ineligible Comparisons 4 2 TOTAL 1 5 3 (12$) 414 (12$) 90 (34$) 84 (12$) 230 (24$) 1254 (18$) 2072 (13$) 2 (5$) 1 (60$) 3 (10$) 15 (23$) 115 (20$) 136 (12$) 416 (11$) 91 (34$) 87 (12$) 245 (24$) 1369 (18$) 2208 Average Age in 48 Years (as of Nov 83) 48 46 48 42 44 Table II-2 shows that of the 18$ loss to the total matched population 18$ occurred in the non-Black and 20$ occurred in the Black population subsets. Thirty-four percent of all participants in occupation code 3 (nonflying officer) and 24$ in occupation code 5 (nonflying enlisted) were lost due to ineligibility. The losses from occupation code 5 clearly exceed the losses in the other 4 categories. The nonflying enlisted individuals were on average the youngest (42 years) while the flying officer and flying enlisted categories were on average the oldest (48 years). A full log-linear analysis (see chapter VII) with all three matching variables included simultaneously was not performed because of the many small cell counts involved. A log-linear model fitted to the three-way frequency table based on eligibility, occupation code, and race, revealed a significant association of eligibility with occupation code (P<.001, adjusted), but not with race (P=.41, adjusted). Because the comparison ineligibility problem was identified after the morbidity study questionnaire and physical examination contracts had been implemented, the ineligible comparisons were removed from the matched cohort and the remaining comparison matrix was collapsed to fill the vacancies created by these removals. This process is characterized in Figure II-1. II-3 �Figure II-1 REMOVAL OF THE INELIGIBLE COMPARISONS AND THE SHIFT LEFT RANCH HAND RH . COMPARISONS X X" ! . V f r2 C X* 4-t. V" X ^ > This figure shows a hypothetical Ranch Hander (RH) and his 10 comparison subjects (CT-CIQ). The C^, C^ and C7 were found to be ineligible and removed. All remaining eligibles were then shifted to the left, i.e., C2 became C-|, Cij became C2, etc. Following the removal of all ineligible subjects, the study was reduced to a 1:8 design. The ineligible selection, the shift left and the subsequent comparison population reduction was presented to the Advisory Committee in 1982. This group felt that the impact of group ineligibility on the study design was negligible; however, subsequent analysis demonstrated a potential impact on inferential reliability (See Chapter V, Compliance and Bias). Statistical considerations required that the shifted population be flagged and analyzed independently of the original comparisons. The data in this report have been primarily analyzed using the original comparisons in an attempt to best describe potential herbicide effects. Wherever possible, analyses using the entire comparison population are also included. During the conduct of the initial morbidity study 5 additional Ranch Handers were identified through personnel record sources and Veterans Administration Education Benefits and Financial Records. These 5 individuals had not been identified earlier because the majority of their military personnel records had been destroyed in a fire at the NPRC in St. Louis. Three of these 5 were newly discovered Ranch Handers and 2 were comparisons who were subsequently identified as Ranch Handers. Ten additional Ranch Handers were identified following the completion of the morbidity study. These individuals will be included in the follow-up study. No attempt was made to select comparisons for these new Ranch Handers. During the removal of ineligible subjects, 1 Ranch Hander, a Black officer pilot, lost his only comparison and remains unmatched, giving a total of 16 unmatched Ranch Handers, of which 6 are in this study. At the time of morbidity study implementation there were 1,211 Ranch Handers matched to 1,026 original and 212 shifted comparisons. Three eligible shifted comparisons were deleted following data collection. The comparison population (C-j) eligible for data collection for the baseline morbidity effort is presented in Table JI-3 by occupation group and nature of the comparison group, i.e., original or shifted. II-lJ �Table II-3 COMPARISON POPULATION ELIGIBLE FOR THE MORBIDITY STUDY BY. OCCUPATION CODE AND NATURE OF COMPARISON GROUP I.E., Occupation Code ORIGINAL OR SHIFTED (C-|) Original Comparisons (0) Shifted Comparisons (S) Total Non-Black 1 2 3 307 72 13 169 Subtotal 6 12 18 122 966 199 1165 5 2 1 15 0 0 0 0 13 5 2 1 15 50 60 13 73 1026 212 1238 Black 1 2 3 Subtotal TOTAL Sixty-four percent of the shifted comparison population is in occupation code 5 (Enlisted-ground). All Black shifted comparisons are in this group, as well. The study protocol estimated that 39? of the entire Ranch Hand population would complete the physical examination portion of the morbidity study. This initial estimate of compliance was based on an estimate of the influences of status (military active duty, military retired, separated and flying) on the individual who could not be guaranteed confidentiality of medical findings. Status also influenced locatability. Active duty and military retired personnel are located through military data sources, while separated individuals must be located through civilian sources. The status and the flying category of the Ranch Hand and comparison population are presented in Tables II-4 and II-5. II-5 �Table II-4 STATUS OF THE RANCH HAND AND MATCHED MORBIDITY COMPARISON POPULATION (C<\) Status Comparison Original Shifted Ranch Hand Active Duty Retired From Military Separated TOTAL 185 576 442 157 510 359 27 85 100 1203* 1026 Total 212 1238 *39 Ranch Hands were deceased at the initiation of the morbidity study. Table II-4 demonstrates that 4856 of the population is retired from the military; 15% remain on active duty; and 37% are separated. Those individuals currently holding military or civilian flying certificates are presented in Table II-5. Table II-5 COUNTS OF THE INDIVIDUALS HOLDING MILITARY AND CIVILIAN FLYING CERTIFICATES, THE RANCH HAND AND MATCHED COMPARISON POPULATION (Cj) Status Original Ranch Hand Military Flying Federal Aviation Admin Certificate TOTAL Comparison Shifted Total 82 128 78 128 12 16 90 144 210 206 28 234 This table shows that 17/6 (210/1203) of the Ranch Handers and 19$ (234/1238) of the total C-\ population presently have military aviation codes or Federal Aviation Administration (FAA) certificates that define active participation in aviation. Twenty percent (206/1026) of the original and ^3% (28/212) of the shifted comparison population hold FAA certificates. 3. Study ^Selection The study protocol defines the morbidity population as all living Ranch Handers and their first randomly selected, alive and compliant comparison. The selection procedure for the questionnaire and physical examination is presented in Figure II-2. II-6 . �Figure II-2 SELECTION PROCEDURE FOR THE QUESTIONNAIRE, PHYSICAL EXAMINATION, AND FOLLOW UP STUDY COMPARISON INDIVIDUALS (RANDOMLY ORDERED) RANDOMLY SELECTED MORTALITY CONTROLS MATCHED RANCH HAND I I I II 1 _. * t « ** t f DEAD — UNWILLING * VOLUNTEER * * REPLACEMENT CANDIDATES In this example, the first randomly ordered comparison was found to be dead. The second was contacted but unwilling to participate, and the third vol-^ unteered to participate in the morbidity study. This process resulted in a third comparison subset, the replacement population. As shown in Figure II-2, this population resulted from the refusal of the original and shifted comparisons to participate in the morbidity study. The study protocol required that the replacement comparisons be matched to the noncompliant individuals on health perception and that they be treated separately in the statistical analyses. In actuality, they were not matched on health perception but were the first volunteers in the randomly ordered mortality sets following original comparison refusals. Because the health perception of the replacement was not matched to the original, comparison subject data analyses and inferences based on these analyses will only be reported for the original and total comparison populations. In this design, deceased Ranch Handers cannot be replaced for physical exam, while deceased comparisons can be replaced due to the one-many matching. This disparity could lead to inferential bias if cause-specific death rates differ in the two groups. Thus far, these rates are not significantly different. II-7 �This epidemiologic study was designed as a matched cohort design. There were 1211 Ranch Handers matched to comparisons by age, race and occupational category at the initiation of the morbidity study. The matched comparison population consisted of 1026 original and 212 shifted comparisons. Three ineligible shifted comparisons; were deleted following data collection. The shifted group resulted from inappropriate selection, removal, and shifting left of the comparison population. Additionally there were 16 Ranch Handers who could not be matched, Ninety-four percent (1171/1247) of the study population is non-Black. The average age of the population is 45 years and 15$ (185/1203) remain on active dutyT Eighteen percent (210/1203) of the Ranch Handers and 19% (234/1238) of the total comparison group have either military flying duties or FAA certificates that denote active participation in, aviation. ,,There were 39 known deceased Ranch Handers. As a study requirement, all morbidity study comparisons were alive at the initiation of the morbidity effort. In summary, 1208 living Ranch Handers and 1238 original and shifted comparisons were entered into the morbidity study. II-8 �Chapter III QUESTIONNAIRE METHODOLOGY 1. Introduction The purpose of the extensive questionnaire was to collect data that could be analyzed for the subjective presence of adverse health effects that might be related to herbicide exposure. The study protocol required that all living exposed subjects and their primary comparisons be offered a comprehensive personal and family health questionnaire administered in the subject's home by a civilian contractor experienced in survey research. The personal nature of the information, peer review recommendations, and the study protocol called for face-to-face interviewing techniques (Herman, 1977; Fry, 1958). In addition to the study participants, the contractor was also required to interview the participant's current and former wives, as well as the first order next-of-kin of deceased individuals to obtain complete morbidity data. Whenever individuals, their spouses, or next-of-kin would not consent to participate in a face-to-face interview, attempts were made to collect the information by telephone (Colombotos, 1969). For the individual who absolutely refused to participate in this data collection process an abbreviated or noncompliant telephone interview format was designed and its use was attempted (Simon, 1971*). This chapter discusses the development and the implementation of the questionnaires used in the study. 2. Questionnaire Development The data collection instruments for the morbidity study were developed and implemented by three separate contracts. The first of, these, awarded to Research Statistics, Inc of Houston, Texas in 1979, developed a statement of work (SOW) which described, in survey research terms, the questionnaire requirements to support the effort. This SOW was used as the basis for the questionnaire development contract which was later awarded to the National Opinion Research Center (NORC) of New York, New York. The questionnaire instruments were developed by NORC in cooperation with the Principal Investigators and included questions concerning a broad range of health effects. The choice of specific effects included in the instruments was based on scientific studies of humans and animals exposed to phenoxy herbicides and dioxins. Hypothetical health effects based on studies in biochemical and biological systems were also included. In addition, veterans' complaints and the public's perception of the health effects of these chemicals were also considered. Questions were designed to allow the maximum degree of data verification by physical examination and medical and personnel record reviews. At the suggestion of NORC, portions of previously field-tested questionnaires were incorporated in the study instruments to maximize the validity of the questionnaires. The sources of the field-tested questionnaires are presented in Table III-1. �Table III-1 SOURCES OF QUESTIONNAIRE ITEMS Section of USAF Health Study Questionnaire Field Tested Questionnaires Marital History The Lives of Women in American Society (Institute of Human Reproductive Studies; Columbia University School of Public Health, Denise B. Kandel) Pregnancy outcomes The Lives of Women in American Society Conception difficulty National Survey of Family Growth Cycle, (National Center for Health Statistics; Vital and Health Statistics, Series 2, #76 January 1978 William F. Pratt) Education General Social Survey (National Opinion Research Corporation, Roper Public Opinion Research Center, University of Conneticut 1981, James A. Davis) Occupation General Social Survey Health outcomes Procedures and Questionnaires of the National Medical Care Utilization and Expenditure Survey (National Center for Health Statistics; Series A, Methodological Report #1, 1980 Robert R. Fuchsberg) Smoking, drugs Drug Abuse Reporting Program (Institute of Behavioral Research, Texas Christian University, 1976 Saul B. Sells) Drinking Drug Abuse Reporting Program Erosion of cognitive abilities Drug Use Vietnam Veteran 11972; Resurvey of Vietnam Veterans 197 * (Washington University. Department of Psychiatry Lee I. Robbins; Special Action Office Monograph, Series A #1, April 1973) Aggression Stressful Life Events and Their Contexts (Rutgers University Press 1981; Barbara Snell and Bruce T. Dohrenwend) Isolation Young Adults Survey, New York State Drug Study (Columbia University School of Public Health. Longitudinal Research on Drug Use 1978, Denise B. Kandel Fatigue Young Adults Survey Social Desirability response set Health Insurance Study 1975-1982 (Rand Corporation; Santa Monica, CA Dec 1979 John E. Ware, Jr.) III-2 �Anxiety Health Insurance Study Depressive episode Diagnostic Interview schedule (Dr. Lee Robbins, Washington University, St. Louis, MO) Panic disorder Diagnostic Interview Schedule An acceptability pretest of the developed questionnaires was conducted in May 1981. Twenty study subject, 18 spouse, and 2 next-of-kin interviews were completed. Following minor modifications, these instruments became the final questionnaires for the implementation contract. They were not publicly released prior to implementation. 3. Questionnaire Implementation; Contract Award and Administration Louis Harris and Associates, Inc (LHA) was competitively awarded a 9-month implementation contract in September 1981. The purpose of this contract was to collect baseline data on the study population through the use of the developed questionnaires. The specific elements of each questionnaire are presented in Table III-2. Table III-2 ELEMENTS OF THE QUESTIONNAIRES Type Questionnaire Elements Study Subject Demographic, educational, occupational, medical, compliance, toxic exposures,and reproductive experience Spouse (present and former) Comprehensive reproductive history Next-of-kin Modification of study subject questionnaire Noncompliant (Telephone) Perception of health, use of prescribed medication, medical conditions, work absenteeism, income and reasons for noncompliance LHA first reviewed the questionnaire and reformated the instruments from a horizontal to a longitudinal format to better suit their interviewing style. The contractor's management personnel selected interviewers, scheduled training programs, and defined procedures to be used in the conduct of the contract. Ninety interviewers were selected and trained in a series of 11 training sessions held throughout the United States and Europe. All training sessions were taught by either the LHA Vice-President for Research Services, or the Project Director. All LHA interviewers (84 women and 6 men) had a minimum of 1 year prior experience in interviewing, with at least 1 experience with health data collection. Race matching of interviewers and respondents was accomplished in III-3 �the majority of cases in order to enhance rapport and accuracy of data (Hyman, 195*0. Interviewer bias was additionally limited through a review of the interviewer* s military experience and background. Several potential interviewers were excluded because they were spouses of USAF personnel or personally knew some of the study participants. The LHA staff was not informed of the exposure status of any individual in this study before or after the completion of the contract. LHA interviewers reported to the Project Director in the New York office on a weekly basis. The first two interviews of each interviewer were critiqued by this staff prior to allowing further interviewing. Additionally, the USAF received weekly reports from the Project Director on all aspects of the contract. An interactive relationship between the USAF and LHA staff was essential throughout this contract. In addition to data collection, LHA also contracted to locate the study population, obtain signed medical release forms, assess the intent of the subject to participate in the physical examination phase of the morbidity study, and to attempt to convert those individuals who absolutely refused all data collection attempts. 4. Questionnaire Implementation; .....Location Initial contact with the Ranch Hand and the original comparison population occurred in November 1981. At this time each potential participant was sent certified introductory letters and a fact sheet. These letters were signed by the Secretary of the USAF and the USAF Surgeon General. They defined participation as voluntary and explained the limited confidentiality of positive medical findings diagnosed during the physical examination portion of the Morbidity study. Examples of these materials are presented in Appendix XI. LHA followed the USAF letters with their own introductory letters. The assigned interviewer then contacted the potential study participant by phone for scheduling the in-person questionnaire. Initial contact with the shifted population was also completed by this series of letters and telephone contact. Letter mailing and identification of this group to LHA was completed by April 1982. Initial contact with the replacement comparison group occurred by letter followed by LHA phone contact until the final questionnaire administration contract extension, i.e. November 1982. From November 1982 all initial contact with replacement comparisons was by the USAF by telephone. For this small group, questionnaire administration was scheduled by the USAF interviewers in conjunction with the physical examination. Introductory USAF letters were sent after the replacement comparison agreed to complete the physical examination. LHA letters were, of course, not sent to this population. Therefore, within the replacement subset of comparison participants there are individuals whose interview was completed by the USAF at the physical examination site and not in their home. Table III-3 presents the algorithm developed for locating study participants during the questionnaire administration contract. �TaMalll-3 ALGORITHM OF THE LOCATION PROCESS OF LHA AND USAF DURING THE QUESTIONNAIRE ADMINISTRATION CONTRACT LHA Location Process Location •«—[ USAF Location Process TOTAL POPULATION USAF CERTIFIED LETTER nontocation Location-4- LHA LETTER USAF RECORD SEARCHES: FAMILY CONTACTS nontocttion Location nontocation POSTAL SEARCH nontocation location-4- INTERNAL REVENUE SERVICE (1980 DATA BASE) LOCAL PHONE DIRECTORY SEARCH nontocation nontocation Location • VETERANS ADMINISTRATION FACILITY USE DATA BASE MOTOR VEHICLE REGISTRY SEARCH y nonlocatton nontocation REMAINS UNLOCATABLE Location GEOGRAPHIC SEARCH nontocation Location • REMAINS UNLOCATABLE COMPARISON REPLACEMENTS This algorithm demonstrates the multiple sources used to locate study participants. This process was completed for all study subjects forwarded to LHA (Ranch Hand; original, shifted, and replacement comparisons). For a small number of replacement comparisons (23) not forwarded to LHA because of contract termination, the majority of the USAF location process was completed while the LHA process was not completed. Replacements for the original and shifted nonlocatable comparisons were not identified to LHA until the location algorithm was complete. 5. Questionnaire Implementation; Data Collection Once the study participant was located, an individual LHA interviewer was assigned. The interviewer initially contacted the participants by phone or by telegram -if his phone number was unlisted. The participant was informed of the length of the interview (average 1.5 hrs; range 30 minutes to 3 hours) and scheduled the in-home questionnaire at his convenience. Whenever possible, interviews of current spouse were scheduled for the same day and followed the study participants interview. These interviews were conducted privately in Order to obtain independent reproductive histories. If the participant refused to participate in the interview, his name was forwarded to the central office III-5 �and conversion attempts were made by the LHA central office. Noncompliant telephone questionnaires were administered to the refusals by the central office. The telephone administration system was implemented in April 1981. At the time of the in-persbn questionnaire, all participants read and signed a privacy act statement and completed a Life Events Chart. This chart acted as a recall guide to the chronology of events discussed in the questionnaire. Interviewers were required to ask questions exactly as written, were not allowed to interpret questions, or inject personal commentary, nor were they allowed to skip between sections of the questionnaire. They were also instructed to probe "don't know" answers at least once. At the conclusion of the interview, medical record release forms were signed for those physicians and medical facilities reported in the questionnaire, and the study participant was also asked whether or not he would agree to participate in a physical examination. The respondent was also asked to give the' current name and address for each former spouse listed in the questionnaire, so that spouse interviews could be scheduled and conducted with these individuals. Medical permission forms for medical record data of spouses and children were inadvertently omitted at the time of interview. A system to obtain these data was initiated following the USAF receipt of questionnaires. Due to high and favorable participation rates, patient flow and logistic difficulties in both the questionnaire and physical examination portions of the morbidity study, it was necessary to extend the LHA contract through November 1982 and the examination contract to 15 December 1982. Because the contracts did not overlap experienced USAF interviewers were required to complete questionnaire administration to participants at the physical examination site. 6. Questionnaire Implementation:^Data^Processing All completed interviews were sent to the LHA central office following initial field editing by the responsible interviewer. Each completed questionnaire was repeatedly edited by the LHA Project Director's staff. To ensure that every question was answered, participants were recontacted to provide missing data. This staff also coordinated and supervised the coding, keypunching and key verification of all completed interviews as they were translated to computer tape. Classifications and coding schemes used included the International Classification of Diseases, 9th Revision, Bureau of Labor Statistics, Standard Industrial Classification, and specific USAF codes for job and aircraft designation. LHA reported that it took an average of 2 hours to fully edit and code each interview. All keypunching was 10056 verified. Discrepancies were reconciled by review of the hard copy interview. A set of data cleaning programs was developed by the LHA data processing staff to locate and identify errors and inconsistencies in the data set on tape. These programs were reviewed and approved by USAF data processing personnel. In addition, the USAF developed additional programs to further cleanse the data. In neither case were programs used that would force data to meet inner consistency checks. The objective of data editing was to ensure that the final data set accurately represented the respondent's information. A total of 6 data tapes were delivered to the USAF from LHA. A copy of the data tapes was sent directly from LHA III-6 �to the Advisory Committee on Special Studies Relating to the Possible Long-term Health Effects of Phenoxy Herbicides and Contaminants. The data tapes were delivered at least 3 months later than the original contract established dates. 7. USAF Data Process! ng The questionnaire data collected during contract extensions and at the physical examination site were edited but not keypunched. These data were delivered in hard copy to the USAF. The USAF coded, verified, keypunched and entered the data on computer tape. Because of late data delivery and the volume of unkeypunched data, systematic review and comparison of all (LHA and USAF) hard copy questionnaires to the data tapes was not accomplished as planned. A comparison of 25 hard-copy questionnaires to data entered on the tapes was accomplished by USAF data processing personnel. The findings of this keypunch review are presented in Chapter VI, Quality Control Procedures. Morbidity coding was reviewed; however, because of incorrect and missing morbidity codes the USAF receded all reported medical conditions. Additionally, the LHA data tapes did not include all data collected by the interviewer in the supplemental recording book. These data were required to form the link between the parents, their children, and all medical provider data (the basis of medical verification procedures). The USAF therefore developed systems and hired personnel to support the entry of these data in preparation for analyses. 8. Summary Questionnaire methodology includes the development and implementation of multiple questionnaire instruments through civilian contractors. The NORC developed and LHA administered the instruments. Both contractors worked closely with the USAF. These close interactions resulted in the participation shown in Table Table SUMMARY OF QUESTIONNAIRE PARTICIPATION Type Questionnaire Ranch Hand - Study Subject Counts of Participants Comparison Original Shifted Replaced Air Force TOTAL 1171 956 200 346 30 1208 - Spouse (Current & Former) - Telephone 10 Noncompliant 962 200 333 5 1500 34 8 7 20 69 1532 Medical record release forms were obtained by the contractor during data collection. These permission forms are the basis of the medical record verification program presently in process for data collected by questionnaire. Data delivery to the USAF from the contractor was delayed. Medical coding was reaccomplished and data linkage systems were developed by the USAF to make the most efficient use of the data collected. III-7 �Chapter IV PHYSICAL EXAMINATION METHODOLOGY .Subsequent to the administration of the questionnaire, a voluntary comprehensive physical examination was offered to all individuals in both the exposed and comparison groups. The primary prerequisite for entry into the examination phase of the study was the completion of the baseline questionnaire. In the event that the initially selected comparison chose not to participate in both the questionnaire and the physical examination, a replacement was selected from among the other comparisons in the matched set, as depicted in Chapter II, Figure 11-1. The two and one-half day examination was conducted in Houston, Texas by the Kelsey-Seybold Clinic, P.A. At the time of evaluation, an extensive physical examination, medical history with a review of systems, and indepth laboratory analyses were conducted. A concise Examiner's Handbook in the Air Force Health Study Protocol placed strong emphasis on quality assurance and was used to minimize variability and insure comparability of data over the 12-month duration of the physical examination contract. Strict compliance with this document was required. Physical examinations were performed at the earliest practical time following the completion of the questionnaire, since close sequencing would limit the development of major symptoms or diseases in the interval between the questionnaire and the examination. Physical examinations were performed at a single location and all contractor personnel evaluated the participants without knowledge of their exposure status. The number of examiners and the turnover of staff membersiwas kept to a minimum to limit between-examiner variability. A more detailed discussion of the physical examination quality control program is contained in Chapter VI. All laboratory tests were subjected to rigid quality control, and laboratory and physical examination data were measured on a continuous scale whenever possible to improve statistical power in the analysis. An Air Force physician was present at the examination site throughout the duration of the contract to act as a liaison between the subjects, the contractor and the Air Force, and to insure that the examination protocol was scrupulously followed. Although the on-site monitors closely observed each examiner and technician, the monitors remained unobtrusive during the examinations, and were not permitted to confirm, criticize or otherwise influence the examiners' findings. The components of the physical examination were specifically selected to address those medical end points known or suspected to be caused by phenoxy herbicides and dioxin (Crow, 1970; Kimbrough, 1980). The question of whether significant chronic effects are produced in humans is a controversial issue (Homberger et al, 1979; Reggiani, 1980; Wolfe and Lathrop, 1983). Reviews of physical chemistry data, animal toxicity data, human exposure case reports, and epidemiologic studies have been relatively unsuccessful in identifying specific and objective medical end points for the chronic effects of exposure (Jirasek et al, 1973; Jirasek et al, 1974; Poland, 1979; Young, 1978). The list of known IV-1 �or suspected acute and subacute effects following TCDD exposure is extensive, and many of the end points are highly subjective and extremely difficult to evaluate (Oliver, 1975; Poland et al, 1979). While chloracne appears to be a consistent, chronic effect of moderate to heavy exposure, the implication of this condition on long-term health is unknown (Young et al, 1978). At best, a list of potential organ systems which should be carefully evaluated can be developed. Ideally, one would like to have a sensitive and specific examination or laboratory procedure to detect the effects of these chemicals in human tissues. Unfortunately, there is a lack of clearly defined end points in the scientific literature, and, other than chloracne, distinct clinical syndromes or unique effects indicative of chronic illness have not been identified. The signs and symptoms currently attributed to exposure are confounded by age and other causes, and the effect, if present, may be lost in common symptoms from other causes of disease (in contrast to conditions such as diethyl-stilbestrolinduced vaginal adeno-carcinoma and angiosarcoma of the liver caused by vinyl chloride exposure). In the absence of sensitive and specific indicators of exposure, a comprehensive examination format was developed around these target organ systems listed in Table IV-1. The complexity and the length of the evaluation and the invasiveness of each examination procedure were all key factors in the final choice of the examination components since all of these factors have a significant impact on the compliance behavior of the individuals considering participation in the study. Table IV-1 TARGET ORGAN SYSTEMS/CONDITIONS Dermatologic Hepatic Neoplastic Neurological/Psychiatric Endocrine/Reproductive Immunologic Hemopoietic A general summary of the major components of the examination is presented in Table IV-2, and examples of the examination forms are included in Appendix VI. The laboratory procedures conducted on each subject are listed in Table IV-3. For each participant 20 cc of serum, 100 cc of urine, and all remaining semen were aliquoted and stored at ~70°C for future analyses. When technology developments identify additional analytic procedures which will IV-2 �assess the health effects of phenoxy herbicides and dioxin, these specimens will then be tested. The slides used in the 10,000 white blood cell differential and the semen analysis were also preserved. Table IV-2 RANCH HAND II PHYSICAL EXAMINATION General Physical Examination (Internist) Neurological Examination (Neurologist) Dermatological Examination (Dermatologist) Electrocardiogram (Resting, H-Hour Fasting) Pulmonary Function Study (1 Second Forced Expiratory Volume, Vital Capacity) , Chest X-ray (Ulnar, Peroneal, Sural) Nerve Conduction Velocities Psychological Evaluation Minnesota Multiphasic Personality Inventory (MMPI) Cornell Wechsler Memory Scale I Wechsler Adult Intelligence Scale (WAIS) Wide Range Achievement Test (WRAT) Halstead-Reitan Neuropsychological Battery (Internist) (PhD Psychologist) Patient Outbriefing and Discussion of Individual Results IV-3 �Table IV~3 LABORATORY PROCEDURES Chemistry Panel: Blood Urea Nitrogen (BUN) Creatinine Cholesterol High-Density Lipoprotein Triglyceride Serum Oiutamic Oxaloacetic Transaminase (SCOT) Serum Giutamic Pyruvic Transaminase (SGPT) Gamma Glutyrl Transpeptidase (GGTP) Lactic Dehydrogenase (LDH) Creatine Phosphokinase (CPK) Blood Alcohol Total Bilirubin Direct Bilirubin Alkaline Phosphatase Glucose i } Fasting and 2 Hour Cdrtisol Hormone Assay: Leutenizing Hormone (LH) Triiodothyronine (T3) Follicle Stimulating Hormone Total Thyroxine (T4) (FSH) Free Thyroxine Index (FTI) Testosterone Hematology Panel: Erythrocyte Sedimentation Rate Prothrombin Time Serological Test for Syphilis (RPR) White Blood Cell Count (with 10,000 cell differential) Red Blood Cell Count Hemoglobin Hematocrit Red Cell Indices Platelet Count Urinalysis 2JJ-Hour Urine: Volume Delta Amino Levulinic Acid Coproporphyrins Uroporphyrins Porphobilinogen Creatinine Semen Analysis: Volume Count Abnormal Forms Hepatitis B Testing: Surface Antigen Antibody to Surface Antigen Core Antibody �Under special circumstances, additional laboratory procedures were carried out on selected participants. Those individuals with a history of having fathered children with birth defects had' blood drawn for a determination of karyotype. The serum of participants with a medical history or review of systems indicating the possibility of an immune system deficiency was evaluated by immunoelectrophoresis. Antinuclear antibody determinations were performed on individuals with a history suggestive of connective tissue disorders. In addition, all individuals with a past history of hepatitis were tested for antibody to hepatitis A virus. After 20 April 1982, all participants whose study identification number ended in either 1, 3» 6 or 9 were selected for special immunologic testing. Blood from these individuals was drawn and sent to a subcontractor for the evaluation of B and T cell counts, enumeration of T cell subpopulations, and studies of B and T cell function following mitogen stimulation. In all, 592 randomly selected subjects took part in this portion of the evaluation. Since human sensitivity and compassion could seriously enhance participation in the follow-up phases of the study, every opportunity was taken by the contractor and the Air force to make the experience enjoyable, relaxing and rapport building. Study participants were housed in a comfortable motel, and transportation, meals and a modest stipend were provided. Family members were encouraged to accompany the participants, but at no expense to the government. Any emergency medical care required by the participants during their stay, in Houston was provided by the contractor and paid for by USAF. Additionally, any diagnostic procedures necessary to clarify potentially lifethreatening conditions were also performed (computerized tomography, cardiology consultation, etc.). Detailed in-briefings were provided to ail participants (and optionally to accompanying family members), in order to explain the background and nature of the study as well as the routine medical requirements for the fasting status laboratory procedures. During waiting periods between examination phases, participants were encouraged to become acquainted with other participants and ask any questions they had about the examination, its rationale or the Air Force Health Study. The normal tension associated with psychological testing was relieved by frequent breaks. Any individual problems were quietly and diplomatically managed by the contracting staff and the site monitor. Over 95% of the participants expressed praise for the quality and thoroughness of the examination and pledged to return to the next examination. Subjects arrived in Houston on either a Sunday or a Tuesday afternoon. A 1-hour briefing was given to each group of participants by the Air Force monitor and a Kelsey-Seybold physician. During this briefing, the purpose of the study and a detailed explanation of the examination content and schedule were discussed. The next 2 days (Monday/Tuesday or Wednesday/Thursday) were spent in the examination. Upon arrival at the clinic on the first morning, all participants were met by two Kelsey-Seybold staff members: the Patient Coordinator and the Program Director. After the day's events were explained, medical history and other forms were completed and blood specimens were drawn. All participants on active flying status with the Department of Defense or FAA had their blood drawn while reclining. Others had the option of sitting or lying. IV-5 �All fasting blood specimens were obtained following a minimum of seven hours without alcohol, food or cigarettes. Participants were requested to consume a 250-gram carbohydrate diet for the 3 days prior to their arrival to prepare for the fasting and 2-hour postprandial glucose testing. All alcoholic beverages were to be avoided as well. Compliance with these requirements and the 21* hour urine collection was determined. Breakfast followed the blood draw and postprandial specimens were then obtained at appropriate times. One-half of each group underwent physical examination on the first day while the other half were in psychological testing. On the second day, the schedule was reversed. During the final half-day, each participant received detailed briefings from a PhD psychologist and one of two Internal Medicine specialists. During these briefings, the results of all portions of the physical examination performed at the Kelsey-Seybold Clinic were discussed with the subject, any questions he had were answered, and suggestions for medical treatment or follow-up were made when indicated. If immediate follow-up was indicated, direct contact with the participant's personal physician was made, and appropriate treatment was arranged. The results of those laboratory procedures performed at subcontracting laboratories and the results of the MMPI were not discussed. Payment of expense vouchers and the provision stipend checks were delayed until after the completion of the debriefing to encourage attendance at these sessions. IV-6 �Chapter V STUDY SELECTION AND PARTICIPATION BIAS 1. Introduction The main emphasis in the design and conduct of any epidemiologic study is comparability of the groups under study (Monson, 1980), and the strength of epidemiologic inference is directly associated with group comparability. In this study, Ranch Hand and comparison group comparability was assured by design since strict criteria were used to define the exposed (Ranch Hand) and thf nonexposed (comparison) cohorts and since replacement comparisons were to be matched to original comparisons by perception of health. The cohorts were matched on the variables of age, race, and occupation group to minimize confounding and assure comparability in these variables. Within the nonexposed cohort, however, 4 subgroups resulted from the original match, the removal of ineligibles, replacement for noncompliance, the termination of the questionnaire and physical examination contracts, and the lack of data to match replacements to original comparisons. These groups are: original comparisons (0), shifted comparisons (S), replacement comparisons (R), and those replacement comparisons questioned by experienced Air Force interviewers (A). Because of logistic limitations, scheduling opportunities differed somewhat for each of these groups. Since compliance with this study was voluntary, the occurrence of differing scheduling options could have resulted in inadvertent selection bias (Cook and Campbell, 1979). The purpose of this chapter is to present the factors known to influence study participation, describe and analyze the responses of the Ranch Hand and the comparison groups to the opportunity to participate and to assess the potential bias of differential compliance. The analytic, and inferential implications of self-selection and potential participation bias will also be discussed. Participation is described in terms of location and compliance. A total of 1208 Ranch Hands and 1669 comparisons were the potential participants in this morbidity study. 2* Factors Known to Influence Study Participation The study protocol estimated that 65% of the Ranch Handers would participate in the questionnaire and that 6Q% of these subjects would also participate in the physical examination. One major reason for these low estimates was the recognition of the negative influence of employment in flying occupations on compliance to physical examination. This negative influence was reinforced _in the press and the subsequent advice of the Airline Pilots Association to their members not to participate in this study. This difficulty was anticipated by the principal investigators and is discussed in section VIII of the study protocol. Table V-1 presents a list of factors that could affect study participation. Those components of each factor that are considered in the study protocol for data collection are identified with an asterisk. V-1 �Table V-1 FACTORS POTENTIALLY AFFECTING STUDY PARTICIPATION Factors Components Health Bias *Self perception Current Use Long Term Care Abortion Pattern *Absenteeism *Current Medications Fertility History Current Family Health Familial History Severity of Past Disease Pending Retirement Bias Death Logistic Factors *Time Away From Family *Time Away From Job Distance to Exam Site *Income *Active Pilot (FAA) Other Factors Flying Status (USAF) Officer/Enlisted Age Race Current Status: AD/Sep Stipend Employment Status Dissatisfaction with Military "Operational Factors" Manner of Study Contact Scheduling Window Interviewer Bias Publicity Bias Motivational Bias Compensation Bias The factors and the outlined components of each factor suggest the complexity of the compliance/noncompliance decision made by each study participant, Ranch Hander or comparison. The importance assigned to each component by the individuals in the Ranch Hand and comparison groups is most likely not equivalent. The Ranch Hand group was actively encouraged by the Ranch Hand Association to participate while no such organization exists for the comparison group. V-2 �3- Location Mailing addresses, fpp e@ch study subject were determined through multiple military and civilian sourpes. Study subject location was initially identified by a certified mailing to these addresses/. Current mailing addresses could not be identified for the nQnlpc§tabJe population. Two-tenths percent of the Ranch Hand and 0.5? p? the total comparison group were nonlocatable. This, is well above the 99? location r§te estimated in the stu^y protocol. Table V~2 presents the counts pf the. Ipcated/npnlpcated population by Ranch Hand and type of comparison. Table V-2 COUNTS AND PERCENT OF LOCATABJ,E/NQN-L.OCATABLE ALIVE. STUD! SUBJECTS BY RANCH HAND AND 'NATURE OF THE COMPARISON GROUP Comparison Ranch Ha.nd Locate 1206 (99.8$) NonLocate 2 (0.2?) 1208 Original Shifted Replacements* Total 1023 (99.7$) 212 (10056) 425 (98.6?) 1660 (99.5?) 3 (0.3?) 6 (1.130 9 (0.5?) 1026 212 431 1669 *Includes those indiyiduia.ls interviewed by USAF interviewers (A). The two unlocated Ranch Hajnd individuals were separated from the military, and both had been nqnflying enlisted personnel when on active duty. One was Black and the other was npn-Black;. Three of the 9 unbeatable comparisons were in the originally selected cohort. These 3 individuals were separated from the military, enlisted when on active duty (1 was a flying enlisted while the other 2 were nonflying enlisted) and all were non-Black. The locate algorithm was not completed on the replaced comparison "cannot-locate" population. Five of these 6 individuals were non-Black, The, Black individual was?, separated and had served in an enlisted nonflying capacity. One other separated nonflying enlisted individual was non-Black.. The remaining 4 replaced npnlocated comparisons were non-Black pilots. Two pf these;, were separated, 1 was on reserve status and the other was retired. Overall, nonlocation did npt impact data collection in this study. The 11 npnlocataple subjects are assumed, to be alive and location will be attempted fpr the fpilpw.-up phases of the study. The replacement comparison group nonlocatable rate of 1.4? is of borderline significance when contrasted with th,e rajbe in the originally selected group (P = 0.06). This test was performed pn the proportions using the normal approximation to the binominal, this difference., was a, result pf the termination of the questionnaire contract prior to completion pf the exajnijia.tion process. The names of 3 P? the 6 replacement individuals we,re npt sent to the questionnaire contractor while the 3 others were sent only 1 month prior to contract termination. The V-3 �replacement strategy as de.signed in the study protocol could not be implemented due to termination of the questionnaire contract prior to the completion of the physical examination contract. 4. Study Participation; Compliance Study participation was characterized as being either fully compliant ( FC) (completed the physical examination and the questionnaire); partially compliant (PC) .(completed only the questionnaire) or noncompliant (NC) (refused the physical examination and the in-home questionnaire). Within the noncompliant group are those who completed an abbreviated telephone questionnaire. Figure V-1 shows that of the 1206 locatable Ranch Handers alive at the initiation of the morbidity study, 1045 were fully compliant to the physical examination and an additional 129 completed the questionnaire but refused the physical examination. Ten of the 32 noncompliant Ranch Handers completed the telephone questionnaire. Figure V-1 RANCH HAND PARTICIPATION Original Population n ,1264 .. . ,1 Matched Population n=1242 22 Added Population to Morbidity Study ' \ Living? n=1247 / \ N Deceased n=39 ° Potential Morbidity Study Participants n=1208 ' % * f Is Study Participant Locatable? n=1208 J \ No Non-locat.!* / 1 / Complied with ^ < Questionnaire \ n=1174 TUT /'Complied Witlr, / Physical v \ Examination / \ n=1045 / \ ——,.-- ---• TT Fully Compliant n =1045 No Partially Compliant n=129 �Figure V-2 describes the compliance patterns for the original, shifted and replaced comparison population. Of the 1023 locatable eligible original comparisons, 773 were fully compliant, 183 were partially compliant and 67 were noncompliant. Thirty-four of the noncompliant individuals completed the short telephone questionnaire. FHure V-2 COMPARISON POPULATION PARTICIPATION P /ait . . "° Took \ \ __ /No s LCan't Locate 3 Noncompkant n«67 y 7io" J ^ PartiaNy Compliant n--183 IBS Fully Compliant n=773 1 Completed Telephone n=8 Partialy Phy$ic81 J n=3? _._ No \ t FuNy Compliant n*163 ... . .... ... V fteplacements-for-ineligible Shifted Comparisons V-5 NoncOfflp) •nt n=49 48* 1 Completed Telephone n=27 Yes / Took A "* > \ Physical /"" Yes \ Original Comparisons / Took \ N \ < Questionnaire >~~ Yes »\ V n-137 t t n- 302 j i iMWMM —[canl Locate 6 | — ' -_i» ml29 Replacemerit Compairinn n-431 , j j.- ai _._-_. ^e--.* _ / Took '» rWoconipiini "*" n»12 No\ / Tetephone n«34 ;. Yes y> > p Contact End Omitted 17 1 Shifted »on + 2 Lociti / * i 102 No ( Y«. / \n» 1 / p • 1*\ \" V- Mtital Matched Compailioni .••1241 ^ Yes FuNy Compliant 1 1 t Partialy 1 CompKant t 75* n«88 4 * Number differencenext replacement stopped because prior comparisnn compked after refusing 1 V Replacements-for-NoncdmpNant comparisons �Data collected by the noncompliant telephone instrument was delivered to the United States Air Force in written format following the implementation of the replacement strategy. The telephone questionnaire was not administered to the noncompliant replacement candidates prior to selection for the study, and therefore, the data necessary to match the original and replacement comparisons by similar perception of health status was hot available (Lathrop, 1982). The next living individuals in the designated matched sets were selected as replacements. The data collected in the noncompliant instrument will be discussed in future publications. Figures V-1 and V-2 are summarized in Table V-3, in which Ranch Hand and comparison participation is presented. Table V-3 FULL, PARTIAL, NONCOMPLIANCE OF THE RANCH HAND AND COMPARISON POPULATION BY NATURE OF THE COMPARISON GROUP, i.e., ORIGINAL (0), SHIFTED (S), REPLACED (R), AIR FORCE INTERVIEWERS (A) lp_ RH Fully Compliant (FC) Partially Compliant (PC) NonCompliant (NC) TOTALS 773 183 67 1206 1023 :S Comparisons _A I 258 163 88 37 49 H 212** 395 Total r» 1224 308 128 30 1660 3P *4 individuals were interviewed at the Physical Examination site by USAF interviewers. **3 Additional shifted comparisons were removed due to ineligibility identified following data collection. V-6 �The mean age of the population by compliance, group is presented In Table V-4, Table V-4 MEAN AGE OF THE RANCH HAND ANP COMPARISON POPULATION BY NATURE OF THE COMPARISON GROUP (0, S, R) AND TYPE OF COMPLIANCE (NC, PC, FC) Type Compliance Comparison Mean Age S 0 Ranch Hand Mean Age Non-Black 41 43 44 41 42 45 39 39 43 40 41 41 39 39 41 NC PC FC 39 43 42 35 39 42 34 38 40 Black NC PC FC ^Includes those individuals interviewed by USAF interviewers U). Table V-4 indicates that the nonoompliant group is on the average younger than either the partially or fully compliant in both Black and non-Black strata. The compliant population i? further described by race in Table V-5. This data is abstracted from Appendix XII, Occupational Category and Race of the Fully Compliant Population in Percent and Counts. Table V-5 PERCENT FULLY COMPLIANT OFFICER/ENLISTED CATEGORIES BY RACE RANCH HAND AND COMPARISONS (0, S, R) Comparison ' Shifted Replacements Ranch Hand Original Non-Black Officers Enlisted 35* 88* 73* 77% 78* 77* 61* 74* Black Officers Enlisted 67% 90% 88% 75% * # 69* 62* No individuals in this category. V-7 �This table suggests that Ranch Hand enlisted personnel complied at higher rates than officers and that Ranch Hand non-Black officers complied more than Black officers. The number of Black participants is very small and is therefore not included in the following analyses but is included in Appendix XII. Appendix XVII was used to construct the data in Table V-6. Flying status is presented as flying/nonflying which includes both military and civilian information. Military status is categorized as active duty, retired, and separated/reserve. Table V-6 PERCENT FULLY COMPLIANT OFFICERS BY FLYING STATUS AND MILITARY CATEGORY (NON-BLACK ONLY) Ranch Hand n=372 NonFlylng Flying Comparison Original Shifted Replacements n=46 n=283 n-113 NonNonNonFlying Flying Flying Flying Flying Flying Active Duty (A) 77.8 96.3 58.9 76.2 87.5 75.0 57.9 88.9 Retired (R) 86.0 93-5 86.0 86.5 100.0 96.0 83.3 77.1 Separated/ Reserve (SV) 51.9 87.0 39.3 62.9 37.5 61.5 32.4 63.0 The flying separated/reserve category in this data set complied less than any other strata (P<0.01), and flying status contributed significantly to the compliance decision (P<0.01). As illustrated in Table V-6, a complex set of interactions was involved in compliance. A log-linear model which was fitted to the three-way frequency table based on flying/military status, compliance, and group membership, revealed a three-way interaction (P=.07) in these data, rendering interpretations based on simpler models misleading. Since age and race are also related to flying/military status, tests of association between these factors and compliance need to be studied in the context of the many interactions present. These more complex relationships will be explored in future reports. V-8 �A summary of compliance is presented in Table V~7. Table V-7 PERCENT OF THE STUDY POPULATION COMPLYING TO THE QUESTIONNAIRE AND PHYSICAL EXAMINATION Ranch Hand Questionnaire 97% (1174/1206) Physical Examination 87% (1045/1206) Original Comparison Shifted Replacements 92% (956/1023) 94* (200/212) 88% (376/425) 76$ (773/1023) 77% (163/212) 68% (288/425) Ranch Hand personnel participated in the questionnaire at a rate higher than all comparison groups. This participation was 32% greater than the original protocol estimate of Ranch Hand compliance. Differential compliance to questionnaire did occur in the comparison groups with the original and shifted group complying 5% more than the replaced comparison group (unadjusted; P=0.003)« Table V-7 shows that differential compliance also occurred between the Ranch Hand and the original comparison group in their compliance to physical examination (unadjusted; P<0.001) as well as within the comparison groups with the original and shifted comparison groups complying 8-9% more than the replaced group (unadjusted; P<0.001). 5. Noncompliance The reasons given by study participants for noncompliance were compared. Appendixes XIII and XIV display all reasons given. These data were collected in a nonstandard manner by Louis Harris and Associates, the Kelsey-Seybold Clinic, and USAF personnel. The responses were then allocated to the categories presented in the appendix. They describe that the majority of the reasons given for noncompliance were "no time-no interest" and passive refusal. Table V-8 shows the percent of refusals in the Ranch Hand and comparison groups implying these disinterest reasons. V-9 �Table V-8 PERCENT OF REFUSALS CATEGORIZED AS REFUSALS FOR REASONS OF DISINTEREST Comparison Shifted Ranch Hand Questionnaire Physical Examination Original 86% 67$ 91$ 49$ 50% 58% 54$ 58% Replacements These data indicate that the noncompliant replacement comparisons were passive refusals less often than were the other comparison groups. The percent refusals due to job commitment and confidentiality are described in Table V-9. Table V-9 PERCENT OF QUESTIONNAIRE REFUSALS CATEGORIZED AS JOB COMMITMENT AND CONFIDENTIALITY Original Ranch Hand Job Commitment Confidentiality/ Active Duty TOTAL Comparison Shifted Replacements 3% - 24$ 5% IM - 24$ 5% 17$ - 48$ Forty-eight percent of the replaced population stated that they refused to participate in the questionnaire because of a job commitment or the issue of confidentiality. 6. Scheduling Opportunity The names of the Ranch Hand and original comparison groups were provided to the questionnaire contractor in November 1981. The contractor was given the shifted comparison population in April 1982 and the replacement population continued to be identified to the contractor through 15 Nov 1982. Physical examination scheduling was contingent upon completion of the questionnaire. Therefore, while the Ranch Handers and the original comparisons had 1 year to schedule and complete the study, the shifted comparisons had a maximum of 9 months, and the replacement comparisons were afforded a more limited scheduling opportunity. V-10 �Figure V-3 PERCENT COMPLETED PHYSICAL BY CALENDAR DATE JAN FEB MAR RANCH HAND ALL CONTROLS — APR MAY ORIGINAL KIN V-11 JUL AUG SEP OCT NOV DEC �Figure V- PERCENT COMPLETED PHYSICAL BY CALENDAR DATE 100 80 ^A 60 40 X r X 20 X JAN FEB ALL CONTROLS X MAR APR MAY JUN SHIFTED JUL AUG SEP OCT NOV DEC REPLACEMENTS Figures V-3 and V-1! show the cumulative percent of the Ranch Hand and comparison groups (original, shifted and replacement) completing the physical examination by time. Figure•V-3 shows the similar time pattern of the Ranch Hand and original comparison group completing the physical examination. Figure V-ty shows that the shifted and replacement comparison groups were restricted in scheduling by the nature of the implementation of the design and contract time limitations. The overall comparison group cumulative completion of physical examination by calendar date is shown on both Figure V-3 and V-4. Fifty percent of the Ranch Hands and the original comparisons had completed their physical in May 1982, 50$'of the shifted group had completed in July 1982, while 50% of the replaced group did not complete until October 1982. V-12 �7. Bias Assessment of Replacement Comparisons From the above discussions and that in Chapters II and III, 2 questions are forthcoming which are of interest to inferential reliability. First, "Are the shifted and replaced comparisons valid for use without special statistical treatment?" Secondly, "What is the bias, if any, associated with the differential compliance to the physical examination?" The following sections deal with these 2 questions in turn. 8. Evaluation of the Replacement Comparison Participants^ Since the replacements used in the study, whether S, R or A, were simply the next individual in the randomized match set involved, the appropriate test for replacement bias is the test for 0, S, R or A group differences while conditioning on the variables of age, occupation and race. Specifically, if S, R and A are unbiased groups they should appear to be random samples drawn from the same population as yielded the original (0) set, after adjustment for matching variables. Tests of replacements against original comparisons were accomplished in accordance with procedures set out in the Study Protocol. Following the proto^col, replacements for comparisons were tested first in terms of 3 primary variables to be ascertained oh all participants: (a) subjective health assessment, (b) current utilization of long-term health care, and (c) recent work absenteeism pattern. Statistical testing of these 3 primary variables and•of additional questionnaire and physical examination variables was done in a prespecified manner. First, group A was tested against group R to determine if these groups could be combined. If R and A could be combined, the R + A group was tested against group S to determine if these groups could be combined. If R + A and S groups could be combined, 0 was tested against R + A + S. All testing was done at the 0.05 level. If the test for combination was not met at any stage, appropriate subtesting was performed. When the dependent variable was categorical, testing was performed with log-linear models adjusting by occupational category and age, with age dichotomized as less than 40 years and greater than or equal to 1)0 years providing groups of roughly equal sizes across occupational categories. When the dependent variable was continuous, analysis was performed with a general linear models program adjusting for occupational category and age as with the log-linear models. All of this testing was done to ascertain whether the S, R and A groups could be viewed as drawn from the same population as yielded the 0 group. Thus, the problem is one of hypothesis testing. Careful estimation of the magnitude or directionality of effects noted was not attempted. However, the reader can evaluate magnitude by reviewing data presented in the following paragraphs. In reporting their health status, participants were allowed to use the categories: "excellent," "good," "fair" and "poor." Because of small sample sizes, the "fair" and "poor" responses were combined in the analysis of the data. Table V-10 provides a view of the data, collapsed across occupational V-.13 �categories and age. No statistically significant differences between the S, R and" A groups were found in either the partially compliant or fully compliant groups. However, when taken together, the fully compliant S, R and A groups appeared statistically different from the fully compliant original comparisons (P < 0.001). Additionally, the fully compliant 0 and S groups were found to be statistically different (P = 0.01), as were the fully compliant 0 and R groups (P - 0.0045). No statistically significant differences were noted among those individuals who took the questionnaire only. Table V-10 SELF-ASSESSMENT OF HEALTH STATUS (NON-BLACK PARTICIPANTS ONLY) Participants Who Took Questionnaire Only Fair or Poor Status •* Excellent Good Group •*• N Participants Who Took Questionnaire & Physical Examination Fair or Excellent Good Poor N 0 50.935 34.7$ 14.5$ 173 38.0$ 48.0$ 14.0$ 727 S 61.8* 26.5$ 11.8$ 34 36.4$ 40.3$ 23.4$ 154 R 51.3$ 10.5$ 76 49.6$ 34.3$ 16.1$ 242 A _ 38.2$ _ _ 0 46.7$ 43.3$ 10.0$ 30 36.4$ 11.0$ 118 38.4$ 41.4$ 20.2$ 976 Ranch Hand 0 S R A 52.556 - Original Comparison » Shifted Comparison - Replacement Comparison = Air Force Interviewed Comparison Use of long-term health care was assessed by inquiring about regular use of medications for heart, kidney, thyroid, renal and other disease states. No statistically significant differences were found between the 0, S, R and A groups regarding regular use of medications. Table V-11 provides a view of the data collapsed across occupational categories and age. V-14 �Table V-11 MEDICATION USE (NON-BLACK PARTICIPANTS ONLY) Group Participants Who Took Questionnaire Only (PC) Percent with Chronic Medication Use N Participants Who Took Questionnaire and Physical Examination (FC) Percent with N Chronic Medication Use 0 23.6$ 174 28.3$ 728 S 14.755 34 27.9$ 154 R 19,7$ 76 30.2$ 242 A - 0 16.7$ 30 118 29.4$ 979 Ranch Hand 0 S R A 14.4$ * Original Comparison - Shifted Comparison * Replacement Comparison - Air Force Interviewed Comparison Work absenteeism was assessed by a consideration of reported time loss from work during the 6 months prior to interview. No statistically significant differences were noted between the 0, S, R and A group on this parameter (relevant data provided in Table V-12). V-15 �Table V-12 WORK LOSS (NON-BLACK PARTICIPANTS ONLY) Group 0 s R A Ranch Hand 0 S R A Participants Who Took Questionnaire Only (PC) Percent with Work Loss N 16.8* 14.7* 12.0$ 18.81 Participants Who Took Questionnaire and Physical Examination (FC) Percent with Work Loss N 20.556 707 21.1* 18.6$ 152 237 23.356 20.356 173 34 75 0 112 30 955 » Original Comparison - Shifted Comparison - Replacement Comparison - Air Force Interviewed Comparison Thus, for the 3 basic variables emphasized for test by the study protocol, the replacement comparisons (S+R+A) were found to be statistically significantly dissimilar from the originals on 1 variable, self-assessment of health. To more fully assess replacement-original differences, 9 additional variables from the questionnaire were examined: (1) household income, (2) participant education (high school or less, greater than high school), (3) participant anger scale, (4) participant psychoneurological erosion scale, (5) participant anxiety scale, (6) participant depression, (7) reported liver ailments, (8) spouse miscarriage rate, and (9) occurrence of acne. The fully compliant nonBlack replacements (S+R+A) were observed to be statistically significantly different from the fully compliant original comparison participants as regards education (P = 0.04), anxiety level (P =0.02), and psychoneurological erosion (P = 0.02). With respect to education 48.856 of the fully compliant replacement comparisons report more than a high school education, while 43.756 of the original comparisons report more than a high school education. Original fully compliant comparisons reported more moderate to severe anxiety than did the replacements (56.9/6 versus 55.656 respectively). Reported psychoneurological erosion addresses difficulties with mental tasks such as arithmetic work. The replacement comparisons reported erosion more commonly (37.256) than did the original comparisons (30.256). These measures of psychological status were not validated as truly measuring their intended end points and they are not necessarily statistically independent of one another, nonetheless, a picture of differences between the comparisons subsets is evident. Thus, of 12 variables drawn from the questionnaire, 4 variables (reported health status, education, anxiety level and psychoneurological erosion) distinguish the replacement comparisons(S+R+A) from the original comparisons testing V-16 �at the 0.05 level. The differences observed are not only statistically significant but may also reflect clinically meaningful differences if the self-reporting is accurate. Analyses of bias have also been conducted using physical examination data end points to obtain a firmer evaluation, and these analyses are described in the following paragraphs. Five laboratory variables have also been examined for evidence of differences among the comparison groups: white blood cell count (WBC), hemoglobin concentration (HGB), total bilirubin (TBIL), serum glutamic oxalic transaminase (SCOT) and lactic dehydrogenase (LDH). This testing is summarized in Table V-13- The analyses were performed with a general linear models program, operating on WBC and HGB in natural units and TBIL, SCOT and LDH in logarithmic units. It is clear from Table TM3 that there is definite indication of comparison group differences. Table V-13 SUMMARY OF BIAS ASSESSMENTS OF REPLACEMENT COMPARISONS USING LABORATORY MEASURES (NON-BLACK PARTICIPANTS ONLY) Clinical Variable WBC HGB TBIL SGOT LDH Adjusted Mean For Original (0) Comparisons Adjusted Mean For All Replacements (S+R+A) P Value F<j>r Mean Differential 7 .78 15 .9 0 ,609 32 ,7 141.2 0.027 0.522 0.063 0.498 0.265 7.24 16.0 0.577 33.1 142.0 Lastly, 13 clinical variables from the physical examination itself were evaluated for 0, S, R, A comparison group differences. As summarized in Table V-14, statistically significant differences were found. V-17 �Table V-14 SUMMARY OF BIAS ASSESSMENTS OF REPLACEMENT COMPARISONS USING MEASURES FROM THE PHYSICAL EXAMINATION *1. Systolic Blood Pressure No differences detected *2. Diastolic Blood Pressure No differences detected *3. Posterior Tibial Pulse rS 1 statistically different from R + A 0 statistically different from R + A *4. Dorsalls Pedis Pulse No differences detected rS 1 statistically different from R + A 0 statistically different from R + A *5. EKG rS statistically different from R + A 0 not different from R + S + A 6. Vibration Sense 1 7. Tremor rs 1 statistically different from R + A 0 statistically different from S 8. Nerve conduction velocity above the elbow No differences detected 9. Nerve conduction velocity below the elbow No differences detected 10. Peroneal nerve conduction velocity No differences detected 11. Full Scale Intelligence Quotient No differences detected rS 1 statistically different from R + A 0 statistically different from R + A 12. MMPI Scale D No differences detected 13. MMPI Scale L *Black participants removed. Taken together the analyses described above imply very strongly that the S, R and A comparison groups are not random samples drawn from the same population as the original comparisons ( ) Since the comparison group differences are 0. not observed in all variables studied, a possible approach is to perform a prior test of significance (PTS) to test for appropriateness of replacement use, followed when possible by a Ranch Hand-all comparison test. This use of a PTS has been discussed with appreciable detail in the statistical literature (Bozivich et al, 1956; Bancroft, 1964; Kale and Bancroft, 1967; Arnold, 1970; Cohen, 197*0. Recommendations in this literature suggest a preliminary test for combination using an alpha level of 0.25> followed by a test of differences at an alpha level of 0.05. Calculations of study power with and without V-18 �the PTS have indicated that, given the sample sizes in this study, the PTS only provides partial protection against inferential bias. This result can be understood by reference to Figure V-5 where 2 power curves are given. Figure V~5 POWER CURVES FOR ALTERNATIVE ANALYTICAL METHODS Figure V~5. Curve 1: Power curve for Ranch Hand-original comparison tests on means. Curve 2: Power curve for Ranch Hand-comparison tests on means assuming replacement comparisons are unbiased. F is the symbol for ratios of Ranch Hand-comparison means. The lower power, curve (curve #1) is for a test of difference between the Ranch Hand group (N=1045) and the original comparisons (N=773)« The upper curve (curve #2) is for the same test of difference but between the Ranch Hand group and all comparisons (N-1224) assuming that the replacements are unbiased. These curves are drawn for a hypothetical clinical variable with ratio of standard deviation to mean being 0.200. The variable F is the ratio of the exposed mean to the comparison mean. The slight displacement of the 2 curves in the vertical direction (power) is easily negated by small degrees of bias in the replacement comparisons. V-19 �The Study Protocol reflects a strong concern for a variety of biases that may be operating in this study. The effect of the potential bias, by using the shifted and replacement members of the comparison group, was not uniformly viewed by the Principal Investigators. Because of time constraints, the Science Panel was not convened to address this complex issue. Instead, a management decision was made to base the primary clinical analyses upon a contrast of the Ranch Hand group and members of the original comparison group. For completeness of data descriptions, some chapters additionally contain analyses founded upon the entire comparison group. 9. Noncompliance Bias The data in the previous section suggest that a degree of self-selection did occur in association with compliance to the physical examination, indicating that the group who came to physical examination may be biased from the original sample. Since this report emphasizes analysis of data from fully compliant participants, selection biases associated with physical examination compliance are of importance. Table V<-15 displays differences between fully and partially compliant study participants. Table V-15 DIFFERENCES BETWEEN FULLY COMPLIANT (TOOK QUESTIONNAIRE AND PHYSICAL EXAMINATION) AND PARTIALLY COMPLIANT (TOOK QUESTIONNAIRE) STUDY PARTICIPANTS: P VALUES FOR TEST OF NO DIFFERENCE Ranch Hand Fully Compliant Original Comparison Fully Compliant Versus Par t i al 1 y_ Compl i ant _ Versus . Health Status Medication Use Work Loss Household Income Education Anger Anxiety Erosion Depression Liver Ailments Miscarriages Acne 0.006 <0.001 0.79 0.32 0.66 <0.001 0.020 <0.001 0.007 0.76 0.97 0.37 , 0.001 0.23 0.30 0.86 0.39 0.01 0.61 0.002 0.36 0.64 0.077 0.75 Eighty-seven percent of the Ranch Hand group were compliant to the physical examination while 76% of the original eligible comparisons attended. Let RRObS be the observed relative risk calculated from the physical examination data and RR be the actual relative risk of the originally drawn groups. Direct algebraic considerations provide the relationship V-20 �0.13 Ye + 0.8? RR - ———^—,_,-— 0.24 Yc + 0.76 RRQbs Equation #1 In this equation, Ye is the ratio of the prevalence of the finding in the Ranch Hand group noncompliant to physical examination, to the prevalence in Ranch Hand individuals who were examined; the term Yc is the same ratio for the comparison group. In other words, the values Ye and Yc are within-group noncompliant-to-compliant relative risks. The values of Ye and Yc are in fact not known so that RR can in fact not be known with exactness. Were RR0bS • 1.00 and were the finding ra,te 0.100 in the fully compliant comparison group, Ye and Yc could both range from zero to 10, indicating that RR could take values from 0.28 to 2.86. Thus, noncomplianee to the physical examination is a serious concern in the attempt to properly infer herbicide effects from group differences noted at physical examination. It is possible to develop an indication of the magnitude of the withingroup relative risks Ye and Yc using data from the questionnaire. From Table V-15, it is clear that in several instances (roughly 50%') the fully compliant replacements are not statistically different from the partially compliant or, approximately, Ye =* rc = 1.0. In these cases, an observed relative risk, RI *obs» is at least approximately equal to the actual relative risk, RR of the original sample. On the other hand, using the health status data, Ye is estimated to be 0.54 while Yc is 1.0*4 for .the categories "fair-poor" health, indicating (using Equation #1) that RR = 0.93 RRObs- Tnis result implies the possibility that the use of physical examination data can overestimate a relative risk by 7%.. On the other hand, for the erosion scale Ye is 0.52, while i c is 0.63, providing RR = 1.03 RRObs» vhich implies the possibility that the physi: cal examination could underestimate relative risk by 3%> These calculations of Ye and Yc use questionnaire data, and thus, the results are indications only of bias in the physical examination, due to the extrapolation from 1 data set to another. Nevertheless, the results do indicate a range of bias which is much smaller than the range obtained when no assumptions about Ye and YC are made. It is difficult to conceive of a partially compliant rate or proportion as being different from a fully compliant rate or proportion by more than a factor of 2. Thus it may be assumed that 0.5 ^ Ye $ 2.0 0.5 •$ Yc < 2.0 under this assumption 0.75 RR0bs S RR ^ 1.28 RRObs V-21 �An inequality such as the one above should be applied to each study result reported here to reflect the possible effect of selection bias. If the above inequality is used, the smallest observed relative risk that can be considered actually larger than 1 is 1.33 (=0.75~1) and the largest observed relative risk that can be considered actually smaller than 1 is 0.78 (=1.28~1). Or, as a simpler rule of thumb, full sample relative risks may be assumed to be within ±30% of observed relative risks. Of course, this measure of uncertainty due to noneompliance must be added to the uncertainty due to finite sample sizes, and to other sources of possible inferential error. It is not feasible to numerically evaluate the degree of bias in physical examination measurements of continuously distributed variables such as blood pressure, hemoglobin concentration or pulmonary volumes, using questionnaire data, as no analogous values were obtained from the questionnaire. An equation similar to Equation #1 holds for the ratio of group mean values for a continuous variable, namely: 0.13 Ye1 + 0.87 RAT RAT = obs Equation #2 0.24 Yc1 +0.76 In this equation, RATobs is the ratio of the Ranch Hand fully compliant mean to the comparison fully compliant mean, RAT is the ratio of the means of the complete original samples, Ye1 is the ratio of the partially compliant mean to the fully compliant mean in the Ranch Hand set and Y,J is the same ratio for the comparison participants. Estimates of Ye1 and Yc' are not available; however, it is difficult to conceive of a partially compliant mean as different from a fully compliant mean in the same group by more than 20$; whence, we assume: 0.80 ^ Ye1 S 1.20 0.80 S YC1 < 1.20 Under this assumption 0.93 RATobg < RAT ^ 1.08 RATObs that is, full sample ratios are anticipated to be within ±8% of observed sample ratios of means. The potential error in sample mean ratios portrayed above must be considered by the reader in the interpretation of mean shift data presented in this report. V-22 �10. Summary and Conclusion The comparison group in this study is divisible into 3 subgroups: original comparisons, shifted comparisons and replacements. Due to study implementation and contractual constraints, the shifted and replaced comparison groups were scheduled differently from the original comparison group for the study questionnaire and physical examination. The original comparisons were handled in a manner essentially identical to that of the Ranch Handers. Analysis has shown that replacements differ from original comparisons on compliance to questionnaire and physical examination; however, shifted comparisons are not statistically significantly different from originals on these parameters. Both shifted and replacement comparisons have been found to be statistically significantly different from the original comparisons on a variety of questionnaire and physical examination measures. This source of potential bias is completely avoided in this report through the primary use of the original comparisons in hypothesis testing. Differential compliance to the physical examination occurred with 87/6 of the Ranch Handers and 76% of the comparisons attending. This fact raises the concern for a second bias which cannot be avoided, and it could be a result of media and Ranch Hand Association support for this study. It is suggested, however, that this bias is not large. Worst-case estimates imply that observed relative risks are displaced from correct relative risks by no more than 30$ by noncompliance effects, and observed mean ratios are displaced by no more than 8%. V-23 �Chapter VI QUALITY CONTROL PROCEDURES Quality control aspects of the Air Force Health Study have been of major importance since the inception of the study design. The focus of quality control concerns has been 1) to ensure the highest quality and validity of this study, 2) to reduce variability and bias in all data, 3) to validate all statistical methods and enhance statistical power wherever possible, and 4) to protect government resources. The purpose of this chapter is to present a categorical overview of the quality control procedures and to present representative data, where appropriate. 1 • Prestudy Considerations The Study Protocol was formulated and refined in 1979-1980, during which time it underwent H independent peer reviews and a final review and approval by the Science Panel of the Agent Orange Working Group. Knowledge gained from visits to national and international herbicide dioxin experts was also instrumental in refining the Protocol. Initial contract management aspects were handled on a scientific business basis. The Principal Investigators developed comprehensive statements of work with specific evaluation criteria. All contract proposals were evaluated without reviewer knowledge of the proposer and then scored independently on their scientific and business merits. Contracts were awarded on the basis of scientific and medical quality; price considerations were secondary. Fixed-price competitive contracts were written where feasible. During the conduct of the contracts, numerous scientific and business meetings were held with the contractors in an attempt to ensure quality and timeliness of the data. Scientific concerns continued as the primary emphasis throughout the periods of contract performance. The population ascertainment process for both the Ranch Hand and comparison groups has continued for over 4 years. Extensive computer searches and a hand review of all available military personnel records have assured an almost conr plete and comparable identification mechanism. In addition, individual responses to the Ranch Hand Reunion Association and wide media coverage of the Agent Orange issue have greatly assisted both the ascertainment and addressupdate processes. A few potential study participants whose records were burned in the National Personnel Record Center remain uncategorized at this time. Both populations were subjected to a rigorous systematic location process (see Chapter III), resulting in a location efficiency of 99.556; this achievement has eliminated population selection bias and has afforded each individual a maximum opportunity to participate in the study. The computer technique to match each Ranch Hander to a comparison individual by job category, race, and age to the closest birth month was exceptionally rewarding, as about 70% of the matches were exact to birth month and year, as well as to job and race. Such precision has enhanced the analytic flexibility of the statistical techniques cited in this report. VI-1 �2. Questionnaire Data The quality of questionnaire data was enhanced by 2 distinct mechanisms: 1) all questionnaire instruments were designed by nationally recognized survey research organization; and 2) the instruments were administered in an in-home setting by another outstanding survey research firm. A minimum number of highly qualified interviewers were used to reduce data variability, and the interviewers were blind to the exposure status of the respondent. In addition, the interviewers were specially trained and then race matched to the study participants, where possible. Spouse fertility data was obtained independently of the male interview but within the same interview setting. The data collection verification process was conducted sequentially. The Louis Harris Associates Incorported (LHA) field interviewer completed a questionnaire thoroughness edit, followed by a Central Office thoroughness check and appropriate editing. Participants were recontacted by phone, when necessary. LHA trained the United States Air Force interviewers and project staff to complete the identical sequential process. A double blind key punch system was used for both the LHA and USAF collected questionnaire data. Range checks identified outliers, and discrepancies were resolved. The contractor randomly validated completed interviews by phone; however, these interviews have not been analyzed for this report. An early USAF sampling review of the data revealed key punch error rates in specific sections of the questionnaire that ranged from 0 - ].H%. The USAF systematic review and receding of all medical areas included in this report have reduced these error rates. Further, subsequent to the questionnaire, each participant's military personnel record was hand reviewed, in order to provide exact data in the time and location of military assignments. These data have been used in this report in lieu of the memory-dependent military duty information obtained by the questionnaire. Most study-participant questionnaire data were designed to be crossreferenced to review-of-systems data and physical examination findings. A no^ table exception, fertility birth defect data, will be validated by birth certificate or medical records, if retrievable. Female response data were used in all fertility/birth defect analyses, when available. In instances of multiple marriages and offspring, unexpected difficulty was often encountered in assigning a child to the correct spouse pair. Such discordant results were resolved by a hand review and computer input of the questionnaire .data. Thereafter, this system supported all offspring data for analyses herein. Next-of-kin interview data will be verified by cross reference to the deceased's medical records. No attempt was made to validate the abbreviated noncompliant questionnaire because of the individuals expressed disinterest in the study. 3» Physical Examination Data The bulk of scientific data of most concern to the public and veterans will stem from the physical examinations in this study. Consequently, great emphasis has been placed upon quality control of the physical examination and laboratory procedures. VI-2 �All examinations were conducted at a single site by a contract medical organization of unquestioned reputation. The contractor was required to provide board certified physicians for the examination. Dermatologists were required to attend a 1-day intensive training session on the diagnosis of chloracne. A minimum number of physicians and paramedical staff was used to reduce data variability. The credentials of each physician and senior psychol'ogist were submitted to the Air Force for approval. The contractor fulfilled the commitment to maintain a stable work force throughout the contract, best exemplified by the facts that (1) approximately 90% of the general physical examinations were conducted by one internist, (2) all electromyographic tests were performed by one technician using a single constantly calibrated machine, and (3) 90% of the final diagnostic assessments were made by 2 internists (master diagnosticians). All medical examiners were required to adhere strictly to the physical examination specifications as cited in the Study Protocol and were not permitted to evaluate a participant outside of his medical specialty area. Thus, each examiner was blind to examination findings outside his area of expertise, as well as to the exposure status of each participant. An Air Force physician, serving as an om-site physician monitor, conducted frequent inspections of all aspects of the physical, psychological, and laboratory examinations to ensure contract compliance and to approve further diagnos^ tic workups for those participants exhibiting serious medical findings. Further, the Air Force monitor was periodically supplemented by Air Force consultant physicians in the areas of internal medicine, cardiology, dermatology, psychiatry, psychology, immunology, and laboratory medicine. For study participants crossing 2 or more time zones, 1 to *» additional rest days were provided before the examination, in order to standardize psychological and laboratory parameters. All examination data were provided to the diagnostician who confirmed significant positive findings and formulated a diagnosis, if one was warranted. The diagnostician then carefully debriefed the participant and recommended follow-up medical action, if indicated. Electrocardiograms (ECG's) on all participants were sent to the Clinical Sciences Division, USAF School of Aerospace Medicine for cross-reference to the USAF ECG Repository. All data from the examination was collated and checked for completeness; this process was rechecked prior to submission to the data processors. Computer entry of all data was made by a single key-to-disk entry with hard copy verification; visual range checks were accomplished prior to transmittal. The Air Force data processors conducted a small sampling from the data set and detected sectional error rates ranging from 0.2 - 1.3$, with 6 of the 7 sectional rates ranging from 0.2 - Q.H%. Plausible ranges were established for most variables and all data outside this range were verified against the hard copy of the examination. All discordant transcription errors were corrected; otherwise, the data were accepted as correct. Inconsistent dates were corrected, where possible. All data sets or subsets were checked for reasonability and, in many cases, the information was verified by the hard copy of the examination. 1. Laboratory Procedures Because the thrust of the physical examination was to cast as wide a clinical net as possible, the importance and number of laboratory tests were substantially increased over an ordinary diagnostic or screening examination. Thus, all contract and subcontract laboratories were required to be licensed VI-3 �and certified by the College of American Pathologists or by the Centers for Disease Control under the Clinical laboratory Improvement Act of 1977. For the laboratory battery of 36 tests, each responsible contract or subcontract laboratory was required to maintain quality control data for audit. The bulk of nonradioassay procedures was accomplished at the contract clinic; a DuPont Automated Chemical Analyzer III (ACA) and Hemalogs 890 and D90 Automated Counters performed the majority of tests. For the ACA, reagents of the same lot number were used throughout the study period. Stringent research grade coefficients of variation (CV's) were required for most assays (see Appendix XV), often necessitating repeat runs to meet these standards. Where available for specific assays, trilevel controls were run at intervals of every 10th specimen, and 1 specimen set of every 15th was run in duplicate. These results were used to generate cumulative sum quality control charts to determine if test systems drifted significantly out of control over time since the CV's are relatively insensitive to trends over time. Of the 14 assays with CV requirement standards, 7 were significantly (P <.05) out of standard at 1 or more levels, On-site visits and detailed power calculations with respect to detecting differences between means showed that these variances would not substantially or biologically alter group comparisons or conclusions. Adjustment of study participant clinical values for drift and other variations in laboratory control levels was considered, but was determined unnecessary. This decision was made by evaluating participant and laboratory quality control values for High-Density Lipoprotein (HDL). Deviations were computed from each overall tri-level mean and these were substra.cted from each participant's value. The distributions with and without adjustment were then contrasted. The results are tabulated below: Table VI-1 HDL VALUES ON 2227 PARTICIPANTS (mg/100 ml) Qrig_lnalmValue Mean Standard Deviation 46.18 12.61 Adjusted Value 46.12 12.72 No increase in HDL precision is noted. In fact, a small increase in the standard deviation was found, clearly indicating that adjustment would not improve the ability to detect group differences. Immunologic assessments were performed by subcontract on 592 participants. Participants were randomly selected (terminal digit of their random study number) midway through the physical examination contract. The subcontractor was blind as to the exposure status and group membership of each individual. The functional capacity of lymphocytes to respond to mitogens or antigens and the number of T and B lymphocytes were measured in isolated peripheral blood. An Immunologic Peer Review Group (see Appendix I) was convened on-site to review technical procedures and to develop analytic strategies. This panel determined VI-4 �that 56 of the 592 samples were not processed due to technical errors in specimen handling. The procedure used for isolation of purified mononuclear cells was substandard. This resulted in cell populations which were depleted of adherent mononuclear cells and contaminated with polymorphonuclear leukocytes and red cells. Differential counts on purified cells were not accomplished so that the actual number of nomonuclear cells used for each assay was not determined. A number of the lymphocyte function assays had excessive variation, manifested by a coefficient of variation (CV) greater than 1556, as reflected in Table VI-2. Table VI-2 PERCENT OF GROUPED LYMPHOCYTE FUNCTION ASSAYS EXCEEDING A CV OF 15$ Functional Test Concanavallin A Phytohemagglutinin Tetanus Toxoid Pokeweed Antigen Percent 15.8 20.3 75.7 10.2 Although CV's were excessive, these variations appeared to be randomly distributed since there were no observed trends over time and there were no differences in error distribution between groups. Only 11 duplicate specimens were received (1 per 50 specimens). Intraspecimen reproducibility was impaired and several split samples varied by more than 50$. Similarly, intraspecimen reproducibility was reduced and represented sporadically within the data set. Further, 54/432 specimens (12.555 of the total) had a ratio of concanavallin A to phytohemagglutinin less than 0.30, indicating mitogen dysfunction rather than failure of lymphocytes to respond to mitogen. The low levels of stimulation observed in many tetanus toxoid-stimulated cultures additionally suggested that caution should be used in the interpretation of the functional results. Accordingly, the Immunology Peer Review Group recommended that the lymphocyte function data not be used clinically to determine the immune status of an individual participant. Further, the panel recommended that the functional data set be used only to evaluate differences, if any, between the Ranch Hand and comparison groups. The T and B lymphocyte enumeration studies demonstrated acceptable reproducibility and acceptable daily and long-term variations between the total T lymphocyte (T^) and the sum of lymphocyte subsets (Tij and T0). Criteria for exclusion of T and B lymphocyte data were (1) samples exhibiting greater than a 30% background fluorescence (11 samples or 2$), and (2) samples with a Tg or TH proportion of less than 10$ (7 samples or 1.3$). Although differential counts were not performed initially on the Ficoll-hypoque separated cells, sufficient paraformaldehyde-stored cells were available after conclusion of the contract to permit a 250 cell differential count on 525 of the 592 specimens. VI-5 �This count permitted the calculation of absolute T and B lymphocyte numbers. After application of acceptibility criteria, cell count data were available on 1*90 specimens. 5. In-Hpuse Data Col1 ec tion and _Statistical,Analysis The complexity and time constraints of this study have made it impractical to hire a series of contractors and expect them to accomplish integrated and timely work. Thus, the Air Force investigators and technical staff have assumed major roles in the areas of population ascertainment and location, verification of eligibility in the study, medical recbrd and personnel record validations, determination of replacements, examination scheduling, medical coding, repository formation, and statistical analyses. Where at all possible, in-house actions have been documented by coding schemes, decision rules, user manuals, and computer audit trails. It is our desire to submit duplicate unedited copies of all contractor data tapes to the Advisory Committee for storage and any possible later use. The data repository task has been monumental. All medical coding has been accomplished in duplicate with resolution of disputes. All in-house gathered data have been subjected to 100/f echo and consistency checking. Subsamples have been obtained to develop quality control error rates. Backup hard copies have been created for all data bases in the event of computer loss or malfunction. The statistical approach to this study consists of a preset state-of-theart framework. The statistical strategy was detailed before the data were reviewed or the group membership codes broken. Both external peer review and internal reviews (conducted by civilian consultants) have validated our approaches. Computer software have been extensively validated by using mock data sets. Vl-6 �Chapter VII STATISTICAL METHODS 1. Statistical Study Design Study data fall naturally into 3 classes: data addressing symptoms, as reported by the subject at questionnaire or in the medical history; data addressing medical signs, determined at physical examination or by review of medical records; and data addressing mortality. A fully expressed or overt herbicide effect would be characterized by increased mortality and more signs and symptoms in the Ranch Hand group as contrasted with the comparison group. These effects should increase with increasing exposure to herbicide. As defined in the study protocol, a subclinical herbicide effect should not be associated with increases in mortality or symptom reporting, but should be found as increases in abnormal findings on physical examination of exposed personnel. These abnormal findings should be more common in the subset of individuals most highly exposed. Symptom reporting is subjective by definition and, thus, subject to influences that could significantly impair proper inference. For example, a stoic and/or highly patriotic individual might unconsciously or consciously suppress the expression of symptoms. Similarly, anxiety associated with middle or older age could prompt elaboration of symptoms. Association of increased symptom reporting with increasing herbicide exposure is suggestive of a true herbicide effect but is not strongly confirmatory as exposed personnel are at least partially aware of the degree of their exposure and could be suppressing or elaborating symptoms in terms of their perceived exposure. The study design permits a specific check on the possibilities of overreporting or underreporting of symptoms. The technique involves contrasting the incidence of physical findings when symptoms are present, between the Ranch Hand and comparison groups. The policy followed in this report is that, if there are no group differences in the sign to symptom ratio, underreporting or overreporting is considered unlikely. If there are group differences in the sign to symptom relationship, underreporting or overreporting is possible, but medically, a real group difference may still exist. Overreporting can be assessed by contrasting reported illness with the results of the physical examination and by medical record reviews. However, this assessment is much more difficult for reported psychological symptoms, since a record of hospitalization, the most reliable indicator of verified illness, occurs only in the most severe forms of psychological illness. 2« The Need forAdjustment Procedures When samples are drawn from a very large or potentially infinite population of individuals, 2 samples of equal size rarely display the same number of diseased individuals. Thus, when comparing 2 groups of individuals, one must ascertain whether the differences are or are not compatible with differences VII-1 �due to random sampling. Two groups of individuals are said to be statistically significantly different when the differences between the groups cannot be accounted for by random sampling or chance mechanisms. If 2 groups are statistically significantly different and 1 of the groups has experienced a specific exposure, this is suggestive that the exposure and the disease may be causally related. However, great care must be exerted in this setting since other unevaluated factors may be the true cause of the observed group differences, and group difference is only 1 element in the causal chain. Adjustment procedures are those statistical procedures which allow objective treatment of intervening variables which can distort the true herbicide effect, if one is, in fact, present. Failure to deal with an important intervening variable can either, induce a false effect or obscure a bona fide effect. Statistical procedures for ascertaining statistical significance and for adjustment used in this report are briefly outlined in a subsequent section of this chapter. The presence of intervening variables occurs either because the sampling procedure used was not completely random or because, by chance, widely different cohorts have been drawn. Matching is a statistical procedure which can partially protect against intervening variables. In this study, the exposed and comparison cohorts were matched on age, race and military occupational category. Intervening variables are also called covariables, risk factors, or subr stitution variables, depending on the literature consulted. There currently exists no objective method for ascertaining that all relevant intervening variables have been accounted for. When all known intervening variables have been examined, there is some degree of comfort that observed relationships are correct. Small sample sizes can, however, markedly inhibit study of intervening variables. A type of intervening variable that is of special interest is the confounding variable (Kleinbaum et al, 1981; Anderson et al, 1980). A confounding variable is an intervening variable that is associated both with the disease under consideration and the exposure categories being used in the study. Failure to adjust for the confounding variable means that the estimated exposuredisease association may be biased. Nonconfounding intervening variables, on the other hand, affect the precision of estimated exposure-disease associations. In the context of intervening variables or covariables, the concept of interactions is important (Kleinbaum et al, 1982). Interaction occurs when the statistical distribution of a random variable (such as a relative risk, or the difference between group sample means) is a function of a second variable (such as age or weight). The study of interactions in a data set is important for it may lead to the discovery of subpopulations at increased or decreased risk from the population taken as a whole. Confounding and interaction can occur together or separately. VII-2 �The use of 1 or more measures of exposure (exposure indices) is an extremely useful addition to the study of group differences. Supplementing the analysis of group differences, the use of exposure indices looks within the exposed group to determine whether the more highly exposed individuals tend to exhibit more disease or abnormalities. The use of exposure indices provides a potentially tighter assessment of herbicide exposure. However, by working with the Ranch Hand group, primarily, sample size limitations also impact this technique. Also, use of exposure indices does not obviate the need to be concerned with confounding and other intervening variables. The construction of exposure indices for the Ranch Hand II study is described in another section of this report. 3. Overview of Specific Statistical Methods In this report, log-linear models have been used when the dependent variable under consideration was categorical or made categorical. Covariables that are intrinsically continuous were stratified for use as adjusting variables in the analysis. Most of the analyses presented in this report are unpaired analyses and, thus, do not fully exploit the paired design of the study. Prior to performing a paired analysis that collapses over matching variables, it is important to determine that the matching variables do not interact with the exposure variable in affecting the dependent variable. The tests presented in this report include these assessments of interaction and, thus, are the early stage of a full paired analysis, as well as being useful for inference in their own right. When unpaired analyses are performed on paired data, there is a consequent loss of test power and less of a chance of detecting a herbicide effect, if one exists. However, an unpaired analysis can actually be more powerful than a paired analysis if study noncompliance or other causes of missing data have resulted in large numbers of broken pairs (Bishop, et al, 1975). The software package used to perform the log-linear analyses in this report is BMD-P4F. In all analyses, the hierarchical modeling procedure was used which starts by examining all covariates and collapses across covariates only when relevant interactions are noted to be null. Whenever the dependent variable was a continuous variable and the covariables were a mixture of categorical and continuously distributed values, regression, multiple regression and/or general linear models were used (e.g., GLM of the Statistical Analysis System). In these analyses in the report, the covariables were always entered as linear terms only. Also, .unless otherwise noted, all group-by-covariate terms (interaction terms) were used in all models. When group comparisons were made without adjusting for intervening variables, simple parametric tests were used, such as the statistic assuming underlying normal distributions. When it was judged that parametric assumptions were not reasonable, the hypothesis of no difference between Ranch Hand and comparison distributions was tested by the Kolmogorov-Smirnov Two-Sample Test (Gibbons, 1971). VII-3 �In this study, a very large amount of data has been collected on each par>ticipant. In this report more than 190 dependent variables were tested. Testing at the 0.05 level means that in 5 out of 100 instances where there has actually been no association, an association will be falsely inferred. The picture is more complex in this report, since as with many epidemiologic studies, measures are not independent but are highly associated. Those variables thought to be most associated with one another have been grouped into clinical categories and these are used for reporting; e.g., general health, psychology, neurology, etc. However, it cannot be assumed that the clinical categories are completely independent from one another. Within each clinical category, whenever possible, summary indices have been developed to provide an overall view of participant status and lessen the likelihood of false inference. Another important concept which protects against false attribution of herbicide effect is careful consideration of the pattern of statistically significant results. If a herbicide effect is being falsely inferred, it might be in a direction opposite to that expected from prior reports. On the other hand, if a test is found significant with a high degree of confidence, its credibility must be considerably enhanced. The inverse of falsely attributing a herbicide effect is the problem of failing to detect an effect when one actually exists. This involves the questions of study power. Power is addressed at length in the study protocol but an overview is provided in this chapter. Under the condition of equal Ranch Hand and comparison group sizes, and assuming unpaired analyses, Table VII-1 provides the approximate sample sizes needed to detect specific relative risks with approximate probability 0.80 (<* = 0.05). The present study is able to detect (with probability 0.80) those relative risks enclosed below the heavy line drawn through the table. Study power for continuous variables is shown in Table VI1-2. The mean shift refers to the displacement of the Ranch Hand mean relative to the control. The variables considered are normally distributed; and unpaired testing is assumed in the table. The present study has approximately an 80$ chance of detecting mean shifts below the heavy line drawn through the table. One thousand forty-five Ranch Handers complied to the physical examination in this study. With this size group, disease states with a cumulative incidence in the group of 1/500 or less have a 1058 chance or greater of no cases at all being encountered. More detail on this point is given in Table VII-3 where the probability or seeing no cases at all is provided for other cumulative incidence values. Another view of study power can be obtained through use of the P values reported in this volume. These observed probabilities permit a direct evaluation of study power against the alternative hypothesis defined by the observed statistic. For example, in categorical tables, the chi-square statistic can be inferred from the cited P value. This observed chi-square statistic can be used as the alternative hypothesis to the null hypothesis of statistical independence. Taking the observed chi-squared statistic as the noncentrality parameter in the appropriate chi-squared distribution, a calculation of study power against the observed effect is possible (Johnson and Kotz, 1970). Table VII-ij �VII-4 provides a short summary of P-value power relationships. Using Table VII-1, if a P-value of 0.10 is reported from a 2X3 table categorical analysis, it may be inferred that study power against the observed effect was H7% (using the two degrees of freedom column in the table). This implies that, if the groups are really as different as they appear from the data, this difference would be detected as statistically significant 4? times out of 100 hypothetical repetitions of this study. Table VII-i| can also be used to approximately assess the power of linear model analyses. The test statistic in these analyses is an F distribution associated with Y} and Y2 degrees of freedom. The degrees of freedom, Yg associated with dependent variable mean squared error is usually quite large in this study. Thus the FCYi.Yg) distribution can be usually well approximated by a X2(Yi) distribution. The degree of freedom, YI , will be 1 when equality between 2 variables such as slopes or group means is under test, and will be the number 2 when equality between 3 variables is under test, as in the trilevel exposure index case. VII-5 �Table VlI-1 NEEDED SAMPLE SIZES 10 fiETECf EXPOStJRE EFFECTS IN TWO SAMPLE TESTING ASSUMING EQUAL SAMPLE SIZES* DATE Of DISEASE IN CONTROL POP = P CONTROL MULTIPL[ESMCtCft IN EXPOS!ED GROUP •RELATI I/ERISK 1.25 1.50., 1 10000 1 ,408,64? 388,536 1 5000 704,244 194,244 2.00 3.00 4.00 5,00 6.00 7.00 8.00 9.00 10.00 114,381 36,618 19,623 12,843 9,339 7,244 5,869 4,905 4,196 57,182 18,306 9,809 6,420 4,668 3,621 3,656 1,958 1,281 1 931 722 585 489 418 1,824 1 977 1 '" 464 639 360 291 243 208 2,933 2,451 2,097 1 ioob" 1 140,722 500 70,282 19,331 5,703 13,930 3,838 1,127 359 6,886 1,895 555 176 38,810 11,423 1 TOO r 192 125 90 70 56 47 40 94 61 44 34 27 22 19 1 50 *This study has unequal sample sizes; therefore these tabled values are underestimates. VII-6 �Table VII-2 NEEDED SAMPLE SIZES TO DETECT EXPOSURE EFFECTS IN TWO SAMPLE TESTING ASSUMING EQUAL SAMPLE SIZES* *This study has unequal sample sizes; therefore underestimates. these tabled values are Table VII-3 PROBABILITY OF ZERO CASES AS A FUNCTION OF CUMULATIVE INCIDENCE Probability of Finding Zero Cases in a Group of 1045 Participants Piease Prevalence .901 .811 .593 .351 .123 .005 1/10,000 1/5,000 1/2,000 1/1,000 1/500 1/200 VII-7 �Table VII-4 STUDY POWER AGAINST OBSERVED EFFECTS OBSERVED PROBABILITY (P- VALUE) DEGREES OF FREEDOM .001 .01 .05 .908 .730 .500 .924 .780 .583 .938 .816 .642 .948 .845 .689 .10 .25 ,376 .210 .470 .300 .536 .367 .590 .425 Study power can be severely influenced by the analytical or statistical method brought .to. bear on the data. For example, in an evaluation of blood pressure, very small differences in group mean blood pressure can be detected using parametric or nonparametrlo testing of measures of location; however, if group differences in hypertension prevalence are analyzed, a lesser or no group difference might be found using categorical .statistical methods such as loglinear models. In general, there is less power to detect a group difference in specific medical diagnoses of a disease state with categorical procedures, than with the underlying continuous variable. However, even in the absence of statistically significant differences in disease rates, group differences in means and variances are still indicative of differences in disease rates that might be detected if sample sizes were larger. Because of these considerations, analyses in this report of continuous variables and the associated normal-abnormal categories are both provided wherever possible. 4. Verification By Medical Records and Interpretive Precision This report contains a retrospective morbidity element since both the questionnaire and physical examination inquire about illnesses or medical conditions that may have occurred in the participant prior to this study. These reports of illness are currently being verified by medical record. The study plan additionally includes verification of negative responses. In this report, some reported conditions have been verified by medical record but no verification of negative responses is currently available. This correction of false positives improves the hypothesis testing only if the false negative rate can be assumed negligible, perhaps a reasonable assumption in a military population. If the false negative rate is not negligible, significant bias and loss of precision remains in the hypothesis test. VII-8 �Chapter VIII EXPOSURE INDEX DEVELOPMENT A potential link of clinical end points with herbicide exposure can be tested within the Ranch Hand cohort by using a measure of exposure (exposure index). In general one would search for increasing indications of illness at higher levels of exposure. However, exceptions to this assumption of a consistently increasing dose-response curve are possible through a variety of biomedical mechanisms. The exposure index used in this report relates to the TCDD-containing herbicides: Herbicide Orange, Herbicide Purple, Herbicide Pink and Herbicide Green. Archived samples of Herbicide Purple suggest that the material had a mean TCDD concentration of approximately 33 ppm and that Herbicide Orange had a mean concentration of 2 ppm. Herbicides Pink and Green contained twice the TCDD of Herbicide Purple and therefore have been estimated to contain TCDD at a concentration of approximately 66 ppm. The index used TCDD ! in this report is written below: Gallons of TCDDContaining Herbicide Ej - {weighting} x Sprayed in the RVN Factor Theater During the Ith Subject's Tour 1 x Number of Airmen with Subject's Duties in the Vietnam Theater during the ith Subject's Tour The TCDD Weighting Factor is 24.0 or 1.0, depending on whether the material sprayed was sprayed before or after 1 July 1965. The weighting factor of 1 is used for the period after 1 July 1965, as the HERBS TAPE and other documentation (Young et al, 1978) show only Herbicide Orange being disseminated by Air Force-flown, fixed-wing aircraft at that time. Prior to 1 July 1965, procurement records and dissemination information show that a combination of Green, Pink and Purple was procured and sprayed by Air Force individuals in Vietnam. Using available data (Young et al, 1978) on gallons of Green, Pink and Purple procured and sprayed, a mean of 48.0 ppm was established for the time period prior to 1 July 1965. Dividing by 2 to normalize to Herbicide Orange, the weighting factor becomes 24.0 (i.e., 48/2 = 24/1). The dates of each subject's tour(s) in the Republic of Vietnam were determined by a manual review of military records. The HERBS TAPE was used along with Contemporary Historical Evaluation and Combat Operations (CHECO) Reports and quarterly operations reports to construct a table of gallons of TCDD-containing herbicide sprayed for each month during the operation. These data are shown in Table VIII-1. For Herbicide Orange missions actual gallons are shown; while for Herbicides Purple, Pink and Green the factor of 24.0 is already VIII-1 �Included making these effective Herbicide Orange or equivalent Herbicide Orange gallons (TCDD at 2 ppm). the CHECO Reports and quarterly operations reports were used in addition to the HERBS TAPE, as the HERBS TAPE currently available does not list all pre-1965 spray missions. Again, only fixed^wing spray missions are compiled in Table VIII-1, as Ranch Hand personnel were not involved with helicopter and other spraying (e.g., backpack). Also provided in Table VIII-1 are Ranch Hand manning in each occupational category by month, as derived from a review of military records. A computer program was written to address this table with each subject's tour dates to the nearest month, to calculate his exposure index in effective or equivalent Herbicide Orange gallons. The exposure index reflects the effective number of gallons of Herbicide Orange to which the airman was potentially exposed, where exposure to the higher TCDD-containing herbicides (Purple, Pink, Green) has been properly weighted to place them on the same footing as Herbicide Orange. As seen by examining the above index definition, the index developed should correlate with the individual's exposure but cannot be an exact measure of actual exposure or body burden. The index is an estimate only, since TCDD concentration is known to have varied across herbicide lots, and since the index does not reflect exceptional exposures such as aircraft hits by enemy fire or dumps (these events are essentially assumed equally distributed). Additionally, the index reflects potential exposure only and does not address specific and determining details of the actual contact. While the index certainly contains errors when applied to judge the exposure of a specific individual, in,studying groups of Individuals epidemiologically, as in this report, these individual errors are expected to balance out or statistically cancel to a great extent, providing some degree of useful inference. The numeric exposure index calculated by the procedure described above was subsequently categorized into 3 levels (Low, Medium, High) for use in statistical analyses; and, this categorization was accomplished in a different manner for each Ranch Hand occupational category in order to optimize study capability to detect a herbicide effect. Details of the exposure categorization are as follows. The study design called for 5 occupational categories: (a) officer-pilot, (b) officer-navigator, (c) officer^other, (d) enlisted-flying, and (e) enlisted-ground. For all exposure index analyses presented in this report, only 3 occupational categories are employed. Specifically all officers were combined into 1 class titled "officer". This combination was accomplished since navigators and pilots were exposed in the same manner, and since individuals in the "officer-other" category were administrators whose exposure was considered effectively zero. Additionally, in the enlisted-ground group, all administrative personnel were assigned a zero exposure value. Under these basic rules, the categorizations shown in Table VIII-2 were developed. A very balanced membership in each occupational category has been provided for each exposure level, optimizing statistical ability to detect a herbicide effect if one exists. VIII-2 �HERBICIDE ORANGE EQUIVALENT GALLONS AND RANCH HAND MANNING BY MONTH Gal Ions Sprayed 10/61 11/61 12/61 01/62 02/62 03/62 04/62 05/62 06/62 07/62 08/62 09/62 10/62 11/62 12/62 01/63 02/63 03/63 04/63 05/63 06/63 07/63 08/63 09/63 10/63 11/63 12/63 01/64 02/64 03/64 04/64 05/64 06/64 07/64 08/64 09/64 10/64 11/64 12/64 01/65 02/65 03/65 04/65 05/65 06/65 07/65 08/65 09/65 10/65 11/65 12/65 01/66 02/66 03/66 04/66 05/66 06/66 07/66 08/66 09/66 10/66 11/66 0 0 0 191426 324216 191426 0 0 0 0 0 334126 334126 0 90879 0 0 0 0 0 174024 259150 0 0 339588 377172 942630 121454 363758 755312 56799 152271 612709 282789 777669 1413945 1413945 1413945 1413945 1296116 1437510 730538 659841 1767431 0 942630 26500 44650 78850 106900 148525 152450 129150 135600 141050 183900 191830 112300 192050 213970 122040 164800 Pilot (Occ 1) 0 5 9 14 14 15 16 15 12 13 11 12 9 10 8 9 7 12 12 10 10 11 8 10 7 6 5 7 5 8 9 10 7 9 9 8 9 11 10 11 12 13 14 15 16 19 19 22 23 24 23 21 22 21 22 21 20 21 26 28 34 41 Navigator (Occ 2) 0 1 2 2 2 2 2 3 2 2 2 2 1 0 0 0 Other Officer (Occ 3) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2 3 3 3 3 3 3 3 4 5 4 3 4 4 4 4 4 4 6 5 6 6 4 5 6 6 8 8 9 8 8 Flying Enlisted (Occ 4) Other Enlisted (Occ 5) 0 6 7 7 7 7 6 6 5 5 5 5 5 5 4 5 5 5 5 5 4 8 8 9 9 10 6 7 7 5 6 5 5 6 5 4 4 4 6 6 6 6 6 6 7 7 7 6 6 10 11 0 14 20 23 23 20 14 13 7 4 5 6 6 5 5 4 4 6 6 7 7 6 4 4 6 6 6 5 4 4 2 2 2 3 3 2 2 1 1 1 1 1 2 2 4 3 3 3 6 12 12 16 26 32 37 38 41 45 46 62 85 104 to 2 2 2 2 2 2 2 3 4 VIII-3 10 10 10 9 10 9 11 12 16 18 �HERBICIDE ORANGE EQUIVALENT GALLONS AND RANCH HAND MANNING BY MONTH Ga11ons Sprayed 12/66 01/67 02/67 03/67 04/67 05/67 06/67 07/67 08/67 09/67 10/67 11/67 12/67 01/68 02/68 03/68 04/68 05/68 06/68 07/68 08/68 09/68 10/68 11/68 12/68 01/69 02/69 03/69 04/69 05/69 06/69 07/69 08/69 09/69 10/69 11/69 12/69 01/70 02/70 03/70 04/70 05/70 06/70 07/70 08/70 09/70 10/70 11/70 12/70 01/71 02/71 03/71 04/71 05/71 06/71 07/71 08/71 09/71 10/71 212100 202360 363830 285400 208300 251320 335860 253884 162895 298615 265335 372425 383605 333595 27450 48200 307740 336300 226325 258100 289160 216300 72250 189100 218750 264450 197450 356500 339800 353800 383533 287425 299100 206800 181000 205100 276900 186350 152100 153730 45700 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Pilot (Occ 1) 45 49 59 51 50 53 55 51 63 60 55 55 58 54 65 69 72 75 77 84 91 89 89 101 94 98 91 90 94 93 88 91 85 83 83 90 76 66 58 59 54 51 47 44 40 40 34 30 25 23 23 23 23 23 28 29 29 29 29 Navigator jOcc 2) 9 9 13 13 14 15 13 15 13 18 19 17 18 19 19 20 20 18 18 19 18 22 20 17 17 19 18 17 20 19 19 16 16 15 17 16 16 15 15 13 13 14 14 11 9 7 6 5 4 4 4 4 4 4 4 4 4 4 4 Other Officer (Occ 3) 5 5 5 4 4 4 3 3 4 5 5 6 6 6 6 5 6 6 6 7 9 8 8 7 8 7 5 5 6 6 7 6 6 6 6 6 5 5 5 5 5 5 3 2 Flying Enlisted (Occ 4) 28 28 28 28 33 34 36 37 32 33 36 33 34 33 35 34 36 32 37 42 45 44 49 53 51 51 51 53 54 54 57 55 55 61 61 60 52 54 41 39 37 29 18 16 14 13 14 15 13 14 14 14 14 14 14 14 14 14 14 Other Enlisted (Occ 5) 123 123 116 114 108 101 105 163 160 161 149 145 129 127 141 160 161 160 164 187 192 147 155 153 154 154 166 172 161 151 155 152 155 142 122 118 114 116 122 125 109 94 84 74 63 43 37 35 30 28 28 28 28 28 28 28 28 28 28 �Table VIII-2 EXPOSURE INDEX CATEGORIZATION Exposure Category Effective Herbicide Orange Gallons Corresponding to Exposure Category Number of Ranch Hand Participants in Exposure Category Officer Low Med High S 35,000 35,000 - 70,000 > 70,000 140 150 151 Enlisted-Flying Low Med High £50,000 50,000 *- 85,000 > 85,000 67 70 66 Enlisted-Ground Low Med High £ 20,000 20,000 i 27,000 > 27,000 185 186 207 Occupational Group VIIIr-5 �Chapter IX GENERAL PHYSICAL HEALTH Five general variables were used in the analyses of the general health status of the study participants. The individual's self-perception of health was obtained during questionnaire administration and reflects a personal and subjective evaluation of health. It is susceptible to varying degrees of bias, both conscious and subconscious. The physician's assessment of the presence of distress is a crude objective measure of general health status and is less biased. This assessment was made on initial observation by the examiner, prior to any direct examination. Thus, patients who appeared ill or in distress on this initial observation were generally quite ill. The examining physician also reported his assessment of the concordance between the subject's apparent age and his chronological age. Two other variables, percent body fat and the erythrocyte sedimentation rate, were also evaluated. There were 1045 Ranch Hand and 773 originally selected comparison participants included in the analyses in this chapter. Slight variations in these numbers occur occasionally due to missing data. Similar analyses were conducted using all compliant comparisons, regardless of replacement status. The results of these additional analyses were essentially no different from the results of the analyses with the originally selected comparisons presented in this chapter. Appendix IX contains representative results of these additional analyses. The relative risks and confidence intervals for the dependent variables analyzed in this chapter are included in Appendix XVIII. 1. Subjective Assessments The results of a log-linear analysis of the self-perception of health in the Ranch Hand and comparison groups with three covariates (age, race and occupational category) are discussed in this section and are shown in Table IX-1. Table IX-1 SELF-PERCEPTION OF HEALTH BY GROUP AND AGE .Age _<40 >i»0 Perception Ranch Hand Number Percent Comparisons Number Percent P value Excellent Good Fair/Poor 129 173 72 (31.5) (46.3) (20.9) 91 120 25 (38.6) (50.8) (10.6) P-.017* Excellent Good Fair/Poor 254 256 139 (39.1) (39.4) (21.il) 203 239 83 (38.7) (45.5) (15.8) P-.025** *Relative risk £40 - 1.82; 95% Confidence Interval (1.18 to 2.10) **Relative risk >40 = 1.35; 95? Confidence Interval (1.05 to 1.76) IX-1 �This analysis demonstrates a statistically significant difference between the two groups, with the Ranch Banders perceiving their health to be poorer than the comparisons. No significant three-factor interaction effects associated with self-perception and group were observed. However, age had a statistically significant association with health perception (P < 0.001) and with group membership (P -• 0.02), thus indicating confounding by age. Race was found to have no association with either group membership or perception of health (P values of 0.94 and 0.87, respectively). The examiner's initial assessment of the appearance of ill health or distress also paralleled the participants' self-perceptions, with more Ranch Handers appearing to be ill than comparison subjects. Although these illappearing individuals accounted for less than 1% of each group, there was borderline statistical significance as shown in Table IX-2. Table IX-2 EXAMINER'S ASSESSMENT OF ILLNESS OR DISTRESS BY GROUP Examiner's Assessment 111 Well Ranch Hand Numberv Percent 8 1,033 ( 0.8) (99.2) Comparison Number Percgnt 1 769 ( 0.1) (99.9) P - 0.056 This measure is somewhat more objective than the participant's selfperception of health but is nevertheless influenced by the participant's emotional status, and bias can thus still be a factor in this result. The participants' self-perception of health appeared to be worse than the examiner's assessment in both groups; however, as demonstrated in Table IX-3, the pattern of discordance does not differ between the two groups. When the examiner's estimates of the participant's apparent ages were contrasted to their chronological ages, 976 (93. W of the Ranch Handers and 737 (95.6$) of the comparisons were observed to appear as old as they actually were. Fifty-one (4.950 of the Ranch Handers and 19 (2.5%} of the comparisons appeared to be younger than their actual age while 18 (1.7?) and 15 (2.0?) respectively appeared to be older. This observation was statistically significant (P* 0.029) and demonstrated a tendency for the Ranch Handers to appear somewhat younger than their actual ages. IX-2 �Table IX-3 . ., • DISCORDANT SELF-PERCEPTIONS OF HEALTH Better than Examiner Ranch Hand Comparison Worse than Examiner 2 0 205 109 2. Objective Assessments Percent body fat and erythrocyte sedimentation rate were albo analyzed in the setting of general health status. While these measures are not indicative of specific diseases, they do indirectly reflect the general state of health. Body fat percentages were calculated from height (inches) and weight (Ibs) measurements (Hodgdon, 1983) using the formula. % Body Fat = (weight/height2) (1015.724) - (17.28460). Data were missing or unmeasurable (greater than 100$) for 7 participants (3 comparison and 4 Ranch Handera), and these individuals were excluded from the analysis. The distribution of these data is shown in Table IX-4 and Figure IX-1, where the percentage of participants falling in each grouping and the cumulative percentages are displayed. Figure IX-1 PERCENT BODY FAT DISTRIBUTION 34 x RH CUMULATIVE * COMP CUMULATIVE PERCENT BODY FAT | RANCH HANGERS (COMPARISONS IX-3 �The percent of body fat appeared to be reasonably normal in its distribution. No significant differences were detected between the variances (P = 0.34) or the means (P =• 0.67) of the two groups. Table IX-4 DESCRIPTIVE STATISTICS - PERCENT BODY FAT Number of Subjects Mean Std Dev 1,041 770 21.12 21.22 5.36 5.19 Ranch Hand Comparison In an effort to assess the extremes of obesity and leanness in the two groups of participants, individuals below 10$ or over 25? body fat were considered to be lean or obese, respectively. The distribution of subjects in three weight categories is shown in Table IX-5. Chi-square procedures revealed no significant differences between the Ranch Hand and comparison groups (P«0.89). Table IX-5 DISTRIBUTION OF BODY FAT PERCENT Lean <W%) Normal (10-25$) Number Percent Number Percent Ranch Hand Comparison 13 7 (1) (1) 824 607 (79) (79) Obese (>25) Number Percent 207 157 (20) (20) Total 1044 771 P= 0.89 The percent body fat and group membership relationship was further evaluated by covariance analysis using age, race and occupational category as covariates. Age and percent body fat were associated (P = 0.02), but this association was not affected by group membership; that is, there was no three-way interaction (P = 0.17). None of the sources of variation associated with race were found to be significant. Percent body fat was significantly different between the three occupational categories (P => 0.04), but this association was the same in both Ranch Hand and comparison groups. IX-4 �Sedimentation rate values presented a right skewed distribution for both groups. Table IX-6 presents the percentile values for each group. A two-sample Kolmogorov-Smirnov test revealed no significant differences in the two unadjusted distributions (P - 0.99). The normal range of sedimentation rate for males is less than or equal to 12 mm and only 5% of each group exceeded normal. Table IX-6 PERCENTILE DISTRIBUTION OF SEDIMENTATION RATE RESULTS 5* 25* JJ2* 75%. 95% Ranch Hand 0 1 2 4 12 Comparison 0 1 2 4 13 . Kolmogorov-Smirnov; P = 0.99 A multifactor log linear analysis of sedimentation rate by group membership, age (£40, >40), hematocrit «42, 42-52, or >52%) and the examiner's assessment of illness or distress was performed. The interaction of sedimentation rate, group membership, and age was significant (P = 0.002) as shown in Table IX-7. Ranch Handers 40-years of age or less had significantly fewer sedimentation rate abnormalities than did their comparisons, while no group difference was noted in individuals over the age of 40. Table IX-7 SEDIMENTATION RATE, AGE AND GROUP MEMBERSHIP Group Sedimentation Rate Abnormal Normal Number Percent Number Percent P Value £ 40 Ranch Hand Comparison 2 10 (0.5) (4.2) 372 227 (99.5) (95.8) 0.001 > 40 Ranch Hand Comparison 39 29 (5.8) (5.4) 628 504 (94.2) (94.6) 0.764 IX-5 �The sedimentation rate was found to have a significant association with hematocrit, the appearance of illness or distress, and percent ttody fat. Table IX-8 displays these data. Since these variables were unassociated with group membership, combined data for both groups are used. Table IX-8 SEDIMENTATION RATE HEMATOCRIT/DISTRESS/BODY FAT ASSOCIATIONS Sedimentation Rate Normal Abnormal Number (Percent) Number (Percent) P Value 13 66 1 < 42$ 42-52* > 52% Hematocrit 102 (11.3) ( 4.0) 1598 ( 3-D 31 (88.7) (96.0) (96.9) <0.001 2 78 (22.2) ( 4.3) 7 1724 (87.8) (95.7) 0.009 3 59 19 (15.0) ( 4.1) ( 5.2) 17 1372 348 (85.0) (95.9) (94.8) 0.049 Appearance of Illness or Distress 111 Well % Body Fat < 10 10-25 > 25 These findings are consistent since an increasing sedimentation rate, abnormal body weight, decreasing hematocrit, and an ill appearance are all traditional indicators of illness, and therefore should be related. The relationships between self-perception of health, sedimentation rate, and age were also explored. These significant relationships are shown in Table IX-9. IX-6 �Table IX-9 SELF-PERCEPTION OF HEALTH, AGE/SEDIMENTATION RATE ASSOCIATIONS Self-Perception of Health Excellent Good Fair/Poor P Value Sedimentation Rate Abnormal Normal 18 671 35 765 28 294 <0.001 S 40 > 40 224 465 294 506 97 225 0.06 Age These relationships were independent of group membership and are not unusual since illness generally increases with advancing age. 3. Herbici de Exposure Analysis The exposure index was applied to the variables in the general health analysis to determine whether a dose-response effect could be identified. As described in Chapter VIII, the index is expressed in equivalent-galIons of dioxin-containing herbicide potentially encountered by each individual during his Ranch Hand tour of duty. Three categories of exposure were used: low, medium, and high. The cutoff values for these categories were chosen so that statistical power could be maximized in the analyses. The interrelationship between a Ranch Hander's self-perception of health and exposure is shown in Table IX-10. Three occupational groupings were analyzed: officers, flying enlisted, and enlisted ground personnel. Nonflying officers were included in the analysis and were assigned to the low exposure category. Their jobs were primarily administrative in nature and involved relatively lower levels of exposure than the flying officers. IX-7 �Table IX-10 HEALTH PERCEPTION IN RANCH HANDERS BY OCCUPATIONAL GROUP AND EXPOSURE CATEGORY Counts Within Exposure Category Low Med High Occupational Group Perception Officer N = 361 Excellent Good Fair/Poor 65 Enlisted, flying N - 183 Enlisted, ground N - 472 P Value 68 42 .13 0.72 11 65 15 18 Excellent Good Fair/Poor 18 29 12 18 24 16 23 29 14 0.84 Excellent Good Fair/Poor 43 59 48 41 95 42 41 67 36 0.13 Total: 1016 These analyses revealed no significant association between exposure and perception of health. The P value of 0.13 among the enlisted ground personnel is of interest, but consistent trends are not seen in the data. Similarly, exposure was found to have no significant association with the examiner's assessment of distress or ill health. The occupational category analysis is shown in Table IX-11. Statistical testing of these data was not conducted due to the small number of individuals judged to be ill by the examining physician. IX-8 �Table IX-11 EXAMINER'S ASSESSMENT OF HEALTH JN RANCH HANDERS BY OCCUPATIONAL GROUP AND EXPOSURE CATEGORY Counts Within Exposure Category; Low Med High Illness or Distress Occupational Group Officer 111 Well 0 111 1 127 1 124 Enlisted, flying 111 Well 0 59 0 59 1 65 Enlisted, ground 111 Well 2 149 0 178 3 142 Similarly, the associations between exposure and apparent age and exposure and body fat were evaluated. These data are presented in Tables IX-12 and IX-13. Table IX-12 APPARENT AGE OF RANCH HANDERS BY OCCUPATIONAL GROUP AND EXPOSURE CATEGORY Occupational Group Counts Within Exposure Category Low Med High Apparent Age P Value Officer Younger Same Older 7 103 1 10 117 1 8 116 1 Younger Same Older 1 57 1 5 54 0 2 64 0 0.22 Younger Same Older 5 142 4 6 169 4 6 136 6 0.88 0.99 Enlisted-flying Enlisted-ground IX-9 �Table IX-13 PERCENT BODY FAT BY OCCUPATIONAL GROUP AND EXPOSURE CATEGORY Occupational Group Counts Within Exposure Category Low Med High % Body Fat P Value Officer 0 91 20 225% 1 103 21 0.57 1 52 6 0 51 15 0.34 2 14 35 10-25? 1 97 30 1 48 10 £10$ 136 39 3 115 30 0.95 Enlisted-flying 10-15$ Enlisted-ground 10-25$ £25$ It is evident from these data that levels of exposure had no relationship to the examiner's assessment of apparent age and percent body fat regardless of occupational category. 4. Summary Overall, the analyses of the general physical health of the study participants revealed classical associations between clinical measures of ill health such as sedimentation rate, obesity/leanness, age, hematocrit, self-perception and the appearance of distress. Statistically significant group differences between the Ranch Hand and Comparison groups were limited to the subjective measures of self-perception of health and the examiner's assessment of illness or distress. The Ranch Handers, as a group, perceived themselves to be in poorer health than did the comparison group. Similarly, the examiner felt that more Ranch Handers appeared ill than did the comparisons. However, ill appearing individuals accounted for less than 1$ of both groups. The analysis of these variables against the exposure index did not reveal any dose-response effects. Overall, the available evidence does not support the presence of such an herbicide effect operating at this time. IX-10 �Chapter X MALIGNANCY 1. Introduction Of all the health effects being attributed to dioxin, cancer is one of the most feared in the minds of the veteran groups, the media and the general public. Dioxin has been identified as a carcinogen or cocarcinogen in some strains of rats and mice (Toth, et al, 1979; Kociba et al, 1978, 1979; Kouri, 1978); however, its carcinogenic effects in humans are unclear. Epidemiologic studies of carcinogenic effects in humans have been generally limited to investigations of phenoxy herbicide exposure among soft-tissue sarcoma patients in Sweden (Hardell and Sandstrom 1979; Axelson, 1977) and studies among industrial groups involved in the production of trichlorophenol and 2,4, 5^T (Zack, 1980; Honchar, 1981). These studies have been contradictory and the issue is still being debated in scientific as well as public forums. The clarification of this important issue is a major focus of the Air Force Health Study. Questions concerning a history of cancer or tumor were asked during both the in-person questionnaire and the physical examination. Question 36a of the study subject questionnaire concerned cancer alone while other areas of the questionnaire focused on tumors or other major medical conditions. In addition, the physical examination subjectively identified additional participants with a history of cancer in the past medical history and objectively identified participants with evidence of prior or newly diagnosed cancer. Figure X-1 shows the algorithm used for data collection for cancer in the study population, as well as those reported cancers that were entered into the cancer verification process. In this algorithm 114 individuals (65 Ranch Handers and 49 comparisons responded "yes" to question 36a, 10 other individuals (3 Ranch Handers and 7 comparisons) responded yes to other questionnaire questions concerning tumors or other major conditions, while 92 additional individuals (50 Ranch Handers and 42 comparisons) reported or were diagnosed as having cancer or tumors during the physical examination. A total of 22 reported cancers occurred prior to the individual's Southeast Asia tour of duty, and these cancers were removed from all analyses. A total of 194 individuals reporting cancer were entered into the verification process (105 Ranch Handers and 89 comparisons). Cancer verification was completed by review of the individual's medical records and available pathology reports. Although cancers reported by all participants were entered into the validation process, only the data from the Ranch Hand group and the subset of originally selected comparisons who completed physical examination were fully analyzed statistically. The rationale X-1 �X-1 ALGORITHM OF COUNTS ON REPORTED CANCERS BY SOURCE OF DATA 1 Responded "Yes" to question 36a iiovo you ovor had cancer?" RH OC 32 55 scj RC.J Total 49 Data 1 Additional "Yes" Physical Examination Subjoctivo and/or objactivo report of cancer OCJfi SCJ2 RC to other Questions 1 scjj RCJJ Total_7 Totals RH 118 OCJSJ SCJ2 RCJ1 TotalJ|§ Removed from Analysis RH-Ranch Hand OC=Ori|inal Comparison SC -Shifted Comparison RC "Replaced Comparison RVN-Republic of Vietnam Tour Cancer occurred before qualifying RVN/SEA tour RH OC_g 13 SC 2 RC_J Total_9 Entered into cancer verification process for this restriction of the database is discussed in Chapter V, Study Selection and Participation Bias. Verification records were obtained with permission forms signed by the participants at the time the questionnaire was administered. The verification process was supported with a limited access computer software program. All reported cancers were classified as to behavior, type and morphology. In addition, cancers were classified as being skin or systemic due to the differing natures of these disease processes. The findings of the verification process are presented in Table X-1. X-2 �Table X~1 SUMMARY OF CANCER VERIFICATION PROCESS Location Behavior of Cancer Skin Malignant Ranch Hand . . ;':. .. 0 Comparison* S R Total 15 7 5 27 Benign 35,. 17 14 3 1 18 Diagnosis not supported 13 6 4 1 11 13 3 3 0 6 No record of treatment at facility as reported 1 0 0 1 Medical record not available 0_ J_ J_ _2 39 18 8 65 Differential Diagnosis at physical examination; individual declined follow-up TOTAL Systemic 79 Malignant 10 2* 2 14 Benign 8 10 0 0 10 Not supported 4 0 0 0 0 p_ p_ 0_ _Q_ 20 2 Medical record not available TOTAL 26 2 24 *Includes 1 Ranch Bander and 1 comparison who expired following interview 0 - Original S - Shifted R - Replacement X-3 �2. Skin Cancer Seventy-five percent (79/105) of all Ranch Hand and 73% (65/89) of all comparison-reported and verified neoplasms were cancer of the skin. Forty-four percent (35/79) of the Ranch Hand reported skin cancers were verified as malignant while H2% (27/65) of the reported total comparison skin cancers were verified as malignant (P - 0.74). All individuals with malignant skin cancer were non-Black. The occurrence of verified skin cancer in those participants who completed the questionnaire (regardless of their compliance to physical examination) was significantly higher in the Ranch Hand group when compared to the total comparison group (P-0.03) or to the subset of original comparisons (P=0.04). Table X-2 shows the distribution of verified malignant skin cancers by cell type. Table X-2 VERIFIED MALIGNANT SKIN CANCERS BY CELL TYPE; REPORTED BY FULLY AND PARTIALLY COMPLIANT PARTICIPANTS Comparisons Ranch Handera* 0 S R Total 31 11 5 5 21 Melanoma 3 1 1 0 2 Squamous Cell 1 3 0 0 3 0 J_ 0 1 15 7 5 Cell Type Basal Cell Fibrosarcoma TOTAL 35 ' 27 *1 Ranch Hander experienced 2 skin cancers, 1 melanoma and 1 squamous cell. He has been counted only once and placed under melanoma in this table. 0 = Original S - Shifted R - Replacement Nonmelanoma cancer accounts for 91$ (32/35) of the Ranch Hand and 93$ (25/27) of the comparison group skin cancers. This difference is not statistically significant (P =0.87). These findings are consistent with reported data that nonmelanoma cancer of the skin is the most common malignant neoplasm in the white population of the United States (Schottenfeld and Fraumeni, 1982). The distribution of these verified skin cancers by anatomic site is presented in Table X-3. X-4 �Table X-3 COUNTS OF SKIN CANCER BY ANATOMIC SITE Melanoma Nonmelanoma skin cancer RH Face, head and neck Comparison O S R Total 26 12* 5** 3 RH Comparison 0 S R Total 20 1 0, 0 0 0 Upper extremities 1 1 1 0 2 0 1 0 0 1 Trunk 5+ 1 0 2 3 2 0 1 0 1 Lower extremities 0 0 0 0 0__ 0 0 0 0 0 32 a 6 5 3 1 1 0 2 TOTAL 25 +Includes 1 Squamous cell *Includes 3 squamous cell **Includes 1 fibrosarcoma RH 0 S R - Ranch Hand » Original - Shifted - Replacement Nonmelanoma skin cancers arose on the face, head and neck in 81$ (26/32) of the Ranch Handers and in 80% (20/25) of all comparisons (P - 0.91). This distribution and the cell types of skin cancers is consistent with recently published information on the epidemiology of skin cancer (Schottenfeld and Fraumeni 1982). The occupational category of those individuals with verified skin cancer are presented in Table X**1*. The counts of these individuals with cancer are relatively small and all occupational categories contribute to the Ranch Hand increase. Followup reports will contain additional analyses of these data with detailed considerations of sample size and age in each of the occupational strata. Table X-4 COUNTS OF THE FACE, HEAD, AND NECK DISTRIBUTION OF NONMELANOMA SKIN CANER; RANCH HAND VERSUS TOTAL COMPARISONS Occupational Code Officers Flying Enlisted Nonflying Enlisted Ranch Hand Cases Rate/100 16 3 3.7 1.5 1.3 26 Total Comparisons Cases Rate/100 11 1 8i •I • ••! !! 20 X-5 1.9 O.H �While medical literature implicates ultraviolet radiation from the sun as the dominant risk factor in the development of nonmelanomic skin cancer (Scott et al 1974), it was not possible to fully evaluate the effects of sun exposure in the initial phase of this study. Information required for this analysis will be obtained in the follow-up phases of the effort. 3. Systemic Cancer A total of 50 systemic cancers (26 Ranch Handers and 24 comparisons) were reported and entered into the verification process (Table X-1). Of these, 14 Ranch Handers and 14 comparisons (10 Originals, 2 Shifted, and 2 Replacements) were verified as having had malignant systemic neoplasms. All individuals with systemic malignancy are non-Black.The site specific classification of these neoplasms is presented in Table X-5. Table X-5 MORBIDITY SITE SPECIFIC VERIFIED SYSTEMIC MALIGNANT NEOPLASMS Ranch Hand Site; ICD Code (9th Ed) 0 Comparison S R Total Lip, oral cavity, pharynx (140-149) 4 2 0 0 2 Digestive organ, peritoneum (150-159) - 4 0 1 5 Respiratory, intrathoracic (160-165) 3* 1 1* 0 2 Genitourinary organ (179-189) 6 2 1 0 3 Other & unspecified sites (190-199) 1 1 0 0 1 Lymphatic & hematopoietic tissue (200-208) - 0 0 1 1 14 10 2 2 14 Bone, connective tissue, skin, breast (170-175) TOTAL *Includes 1 Ranch Hander and 1 comparison who expired following interview 0 = Original S = Shifted R - Replaced Four Ranch Handers and 2 original comparisons were found to have had neoplasms of the lip, oral cavity and pharynx, and all of these individuals reported a history of cigarette and/or cigar smoking. X-6 �Six Ranch Handers and 3 comparisons were found to have had malignancies of the genitourinary organs. The (? Ranch Hand cancers included 1 prostate, 2 testicular, 2 bladder and 1 kidney neoplasm while the 3 comparison cancers included 1 of the prostate and 2 of the bladder. Both cases of testicular cancer were of a germ-cell morphology (one embryonal and one seminoma). Unadjusted statistical testing revealed no significant difference in total genitourinary cancer in the two groups (P = 0,42). Peak incidence rates of testicular cancer in the general population occur between the ages of 35 and 55, and bladder cancer has a peak age of onset between 50 and 70 years. All Ranch Hand bladder cancers, occurred prior to age 50 and all verified comparison genitourinary cancer occurred at age 55 or later. The Ranch Hand testicular cancers occurred at 35 and 38 years of age. These are observational data, and are based on very small sample size. Five comparisons were, found to have had verified malignancies of the digestive organs. There were no Ranch Hand cancers of this organ system. These cancers included 1 of the appendix, 1 of the pancreas, and 3 colon cancers. The annual incid.ence rate for colon cancer increases dramatically with increasing age after the age of 30. The ages at the onset of the colon cancers in the comparison group were 35, ^3, and 50 years. The occurence of gentourinary, oropharyngeal and digestive cancers in the study population was compared to the experience of the Surveil.lanc.e, Epidemiology and End Results program (SEER). Based on these tumor registry data, there is a 30% probability of observing two or more testicular cancer in the Ranch Hand group, and a 29$ probability of two or more bladder cancers. Similar contrasts revealed only a 3% chance of observing the 4 oropharyngeal cancers and a 2% chance of seeing a total absence of digestive cancers i,n the Ranch Hand group. The probabilities of finding the observed numbers of these malignancies in the comparison group were 32$ or greater.. Table X-6 shows the known morbidity and mortality of the Ranch Handers and comparisons from cancer to date. Appendix VIII shows the site specific distribution of bpth the morbidity and mortality study cancers. The mortality sections of these tables include only the first cohort of the comparison population from the Baseline Mortality Study (Lathrop, 1983). X-7 �Table X-6 TOTAL MORTALITY AND MORBIDITY STUDY MORPHOLOGY OF SYSTEMIC NEOPLASM ICD-0 CODES MORTALITY NOMENCLATURE RANCH HAND MORBIDITY* COMPARISON RANCH HAND COMPARISON 0 M800 Neoplasm not otherwise specified (NOS) Bronchus and Lung Intestinal Tract s H 0 0 1 1 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 1 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 M801-804 Epithelial neoplasms Appendix Bladder Bronchus and Lung Kidney Lip Nasopharynx Tongue Unspecified site Vocal Cord 0 0 0 0 1 0 1 0 1 0 0 M805-808 Papillary and Squamous Cell Lip Lung 0 0 0 0 2 1 2 0 0 0 M812-813 Transitional Cell Papillomas and Carcinomas Bladder 0 0 2 0 1 0 M814-838 Andenomas and Adenocarci nomas Bronchus and Lung Colon Kidney Prostate Pancreas 0 0 0 0 0 1 0 0 0 0 1 1 0 0 2 0 1 1 0 0 0 0 0 0 0 0 M850-854 Ductal, lobular, and medullary neoplasms Thyroid 0 0 0 1 0 0 M872-879 Nevi and melanomas Mediastinal 1 0 0 0 0 0 Mesothelioma Bronchus and Lung 0 1 0 0 0 0 Germ cell neoplasms Testicle 0 0 2 0 0 0 0 1 1 0 0 0 M905 M906-909 M938-948 Gliomas Frontal Lobe 1 0 0 1 0 1 1 0 1 X-8 0 1 �Table X-6 (Cont) TOTAL MORTALITY AND MORBIDITY STUDY MORPHOLOGY OF SYSTEMIC NEOPLASM ICD-0 CODES RANCH HAND M965-966 Hodgkins disease Hodgkins (NOS) M986 MORBIDITY* MORTALITY NOMENCLATURE COMPARISON RANCH HAND COMPARISON 0 s R 0 0 Myeloid Leukemias Acute myelocytic leukemia 0 0 0 1 0 0 ~r T 0 = Original S •- Shifted R * Replaced *Two morbidity study participants (1 Ranch Hand, 1 comparison) expired following interview. They are included in the mortality column of this Table because of their date of death. ^ • Covarlate JVn^alysis Group Membership The previous sections of this chapter contained descriptions of the cancer data on the occurrence of skin cancer and systemic cancer in the Ranch Hand and originally selected comparison groups. Except where noted, the remaining analyses in this chapter are based on the Ranch Hand and comparison population that had verified cancer and had completed the physical examination. Covariates used in these analyses included smoking habits and exposure to asbestos, industrial chemicals (yes, no), insecticides (yes, no), degreasing chemicals (yes, no), and nontnedical x-ray sources (yes, no). The results of the basic two-factor analysis are shown in Table X-7. X-9 �Table X-7 VERIFIED CANCER AND GROUP MEMBERSHIP Original Comparisons (N-773) Skin Cancer Yes No Ranch Hand (N~1045) 11 762 \ 35 1010 / \ 25 1169 / P = 0.07 13 1032 / \ 11 1183 / P = 0.46 P - <0.01 Systemic Cancer Yes No 8 765 \ P = 0.68 Total Comparisons* (N=1194) ' . * ^j * This total does not include the 30 participants interviewed by USAF interviewers. The group differences in skin cancer are statistically significant, in the original subset that completed physical examination, (P = < 0.01) and borderline in the total comparisons (P = 0.07), with an excess in the Ranch Hand group. The relative odds of skin cancer in the Ranch Handers are 2.35 and are 1.20 for systemic cancer, with confidence intervals of 1.16 to 4.;90, and 0.47 to 3.15 respectively. These broad intervals are due to the small numbers of cancers available for analysis. The analysis of skin cancer in the Ranch Handers and the original comparisons was repeated with months of agricultural/forestry/fisheries work as a covariable. Seventy-one (6.855) of the Ranch Handers and 66 (8.555) of the original comparisons had worked in these occupations; however, these statistical adjustments did not alter the significant difference between the groups. The P value after adjustment remained 0.01. These analyses are as yet incomplete since they have not accounted for the relationship between skin cancer and geographic area of residence or exposure to other potential skin carinogens. Geographic area of current residence in a mobile military population may not discriminate differences in ultraviolet radiation exposure. An attempt to collect data that will support analyses for geographic and ethnic background will be made at the time of the first follow-up examination. Three-factor analytic techniques were used to account for the possible confounding effects of the covariables listed above. Exposure to industrial chemicals, degreasing chemicals and smoking habits were not different in the Ranch Hand and comparison groups. The analyses of systemic cancer demonstrated an association between cancer and smoking which approached statistical significance (P » 0.07). However, there were no significant differences or suggestive trends between the groups for systemic cancer. X-10 �Significant group differential in exposure to x-ray (P <0.001), insecticides (P <0.001), and asbestos (P = 0.05) were also identified. More comparisons than Ranch Handers were exposed to asbestos and x-ray but more Ranch Handers had previously been exposed to insecticides, many during their tours of duty in RVN. Three-way interactions between variables were significant only for the systemic cancer by group by insecticide analysis (P - 0.01) and suggestive for the systemic cancer by asbestos by group analysis (P =0.16). The results of these analyses are displayed in Table X-8. Table X-8 RESULTS OF THREE-FACTOR LOG-LINEAR ANALYSES OF SYSTEMIC CANCER, GROUP MEMBERSHIP AND CHEMICAL EXPOSURE (P VALUES) Exposure rroup by Cancer Statistical Relationship Group by Cancer by Cancer by Exposure Exposure Exposure by Group Asbestos 0.72 0.04 0.33 0.16 Degreasing Chemicals 0.68 0.33 0.71 0.23 Industrial Chemicals 0.71 0.25 0.34 0.84 Insecticides 0.72 <0.001 0.89 0.01 Smoking 0.50 0.46 0.07 0.53 X-Ray 0.63 <0.001 0.46 0.86 X-11 �Table X-9 RESULTS OF THREE-FACTOR LOG-LINEAR ANALYSES OF SKIN CANCER, GROUP MEMBERSHIP AND EXPOSURE (P VALUES) Exposure Analysis Group by Cancer Group by Exposure Cancer by Exposure Asbestos 0.009 0.04 0.24 0.11 Degreasing Chemicals 0.009 0.37 0.20 0.47 Industrial Chemicals 0.009 0.30 0.03 0.58 Insecticides 0.02 <0.001 0.19 0.79 I Smoking 0.01 0.44 0.70 0.22 X-Ray 0.008 <0.001 0.86 0.51 Cancer by Exposure by Group As shown in Table X-9, analyses of skin cancers demonstrated a significant difference between the Ranch Hand and the original comparison group that completed physical examination. These data again demonstrate the, significant group differential in skin cancer. Even after covariate adjustment (asbestos, industrial chemicals, smoking, x-ray, insecticide and degreasing chemical exposure) the significant group difference in the occurrence of skin cancer remained. Significant between group differentials were noted for x-ray and, asbestos exposure, as previously seen in the systemic cancer analyses. A significant association between skin cancer and exposure to industrial chemicals. #as found (P - 0.03). Associations between the occurrence of skin cancer and exposure to degreasing chemicals and insecticides are also of interest, with g|H68estive P values of 0.20 and 0.19 respectively. 5« Exposure Index Analyses The group difference in cancer occurrence was further evaluated using the exposure index, divided into low, medium, and high degrees of exposure. These analyses used only data gathered on the Ranch Hand group. Table X-10 contains the data and results from the basic two-factor analysis (herbicide exposure versus cancer). X-12 �Table X-10 HERBICIDE EXPOSURE VERSUS CANCER Occupational Group Systemic Cancer Yes Jlo Exposure Level Skin Cancer Yes jto Flying Officers 1 1 3 Low Medium High 110 127 122 P - 0.48 7 5 8 104 123 117 P - 0.62 Flying Enlisted 0 2 1 Low Medium High 59 57 65 P = 0.35 3 1 0 56 58 66 P « 0.14 Ground Enlisted 2 .3 0 Low Medium High P - 0.31 149 176 148 2 5 4 149 174 144 P * 0.63 These analyses did not reveal a dose-response effect between herbicide exposure and the occurrence of either skin or systemic cancer in the Ranch Hand group; however, the number of cancers within each exposure level are very small. A "suggestive" negative association between herbicide exposure and skin cancer was noted among the enlisted flying group (P -• 0.14) with decreasing occurrence of cancer with increasing exposure; however, cell sizes were quite small. Three-factor analysis suggested the presence of interactive effects from insecticide and x-ray exposure, in the flying officers for systemic cancer, and industrial chemicals, degreasing chemicals, and insecticides among the enlisted ground personnel for skin cancer. The results of these analyses are Shown in Tables X-11, and X-12, X-13, X-14, and X-15. X-13 �Table X-11 THREE-FACTOR ANALYSIS: EXPOSURE, SYSTEMIC CANCER, AND INSECTICIDE EXPOSURE AMONG FLYING OFFICERS* Insecticide Exposure Yes Herbicide Exposure Systemic Cancer Yes N_o 1 1 0 low medium high 74 79 72 P = 0.62 No 0 0 3 low medium high 36 48 50 P =0.09 * Three-way interaction P value - 0.10 These data demonstrate confounding by insecticide exposure, with a borderline association between systemic cancer and herbicide (P » 0.09) in the noninsecticide-exposed group of officers. However, the validity of statistical testing in this instance is compromised due to the extremely small number of cases in the analysis. Similarly, this effect is seen with x-ray exposure (Table X-12). Tables X-13, X-14 and X-15 present the data for the herbicide exposure, cancer, industrial chemical, degreasing chemical and insecticide three-factor analyses for enlisted personnel. Confounding is again seen. X-14 �Table X-12 THREE-FACTOR ANALYSES: HERBICIDE EXPOSURE, SYSTEMIC CANCER, AND X-RAY EXPOSURE AMONG FLYING OFFICERS X-ray Exposure Herbicide Exposure Yes Systemic Cancer Yes N£ 1 1 0 low medium high 23 23 33 P - 0.49 No low medium high 0 0 3 87 104 89 P = 0.04 * Three-way interaction P value - 0.04 Table X-13 THREE-FACTOR ANALYSIS: HERBICIDE EXPOSURE, SKIN CANCER, AND INDUSTRIAL CHEMICALS EXPOSURE AMONG ENLISTED GROUND PERSONNEL* Industrial Exposure Yes Herbicide Exposure Skin Cancer Yes 14o 0 1 3 low medium high 79 96 73 P - 0.12 No 2 4 1 low medium high 70 78 71 P « 0.45 * Three-way interaction P value - 0.10 X-15 �Table X-14 THREE-FACTOR ANALYSIS: HERBICIDE EXPOSURE, SKIN CANCER, AND DECREASING CHEMICAL EXPOSURE AMONG ENLISTED FLYING PERSONNEL* Degreasing Chemical Exposure Skin Cancer Yes No Herbicide Exposure low medium high Yes 3 0 0 41 51 0.04 low medium high No 0 1 0 16 17 15 P = 0.42 * Three-way interaction P value - 0.17 Table X-15 THREE-FACTOR ANALYSIS: HERBICIDE EXPOSURE, SKIN CANCER AND INSECTICIDE EXPOSURE AMONG ENLISTED FLYING PERSONNEL* Insecticide Exposure Yes Herbicide Exposure Skin Cancer Yes No low medium high 3 0 0 30 36 41 0.03 No low medium high 0 1 0 26 22 25 P - 0.32 * Three-way interaction P value =0.13 While these data show some confounding for exposure to x-ray, insecticides, industrial chemicals and degreasing chemicals, stratified analysis reveals no evidence of a dose-related effect for exposure to the herbicides used by the USAF in the RVN and the occurrence of cancer. The validity of the statistical X-16 �testing in the exposure index analyses is compromised by the extremely small numbers of cancers available for analysis. Therefore, any Inferences based on these data must be made with caution. 6. Summary The analysis of these data revealed significantly more skin cancer in the Ranch Hand group than in the subset of original comparisons who completed physical examination. This finding was of borderline significance in all original comparisons and in the total comparison population; however, these data are not fully corrected for exposure to the sun and other skin carcinogens. There were no significant group differences for the occurrence of systemic cancer. A small increase in oropharyngeal cancers and a total absence of digestive cancers were observed in the Ranch Hand group. The exposure index analyses did not demonstrate a dose-response effect for either skin or systemic cancer. Of interest was a borderline significant association between systemic cancer and smoking in both groups, demonstrating the sensitivity of the analyses to the effects of this known carcinogen. X-17 �Chapter XI FERTILITY AND REPRODUCTIVE OUTCOMES 1. .Introduction The potential effects of Herbicide Orange exposure on reproduction, fertility., or the incidence of birth defects are highly emotional issues among Vietnam veterans and have received wide media coverage. Animal fertility studies in various species have shown variations in 2,4-D; 2,11,5-T and TCDD toxicity relative to age, dosage levels and routes of administration. TCDD exposed male mice when mated with unexposed females exhibited no abnormalities in mating behavior, fertility, sperm concentration, sperm motility, survival of offspring, or neonatal development (Lamb, 1980). Conversely, administering Herbicide Orange directly to pregnant mice resulted in three fetal effects: cleft palate, decrease in fetal weight, and fetal mortality (Courtney, 1970). The Australian Birth Defects Study of veterans serving in Vietnam showed no asspciation between birth defects of children from veterans and their Vietnam experience (Case Control Study, Australia 1983). Reports from the Seveso, Italy accident, where 220,000 people were potentially exposed to TCDD in 1976, have shown that the incidence of congenital malformations and abortions in exposed women was below expected values for the region. Of 31* aborted fetuses examined for defects, no fetal malformations were attributed to exposure to TCDD. Additionally, developmental abnormalities in children have not been exhibited (Regianni, 1980). A reproductive study of the wives of DOW Chemical Company workers exposed to 2,4,5-T/TCDD found no differences in fertility patterns, fetal wastage, or birth defects (Townsend and Badner, 1981). In 1979 the Administrator of Environmental Protection Agency declared an emergency suspension of 2,4,5,-T based on the Alsea, Oregon study finding of an increased incidence of spontaneous abortion .in 3 Oregon areas sprayed with the herbicides. This study's findings prepared by the Epidemiologic Studies Program, Human Effects Monitoring Branch, Benefits and Field Studies Division, Office of Pesticide Programs, Office of Toxic Substances, and The Environmental Protection Agency remain controversial. Data concerning fertility and reproductive events in this study were collected during the questionnaire and physical examination. Questions regarding reproduction, fertility/infertility, and offspring history were asked of study participants both in the in-home questionnaire and at the physical examination. In addition to the data collected from male respondents, questionnaires focusing on reproductive history were administered to all available spouses and partners. The data from the reconciliation of subject and spouse questionnaire responses constitute the data base described in this report. This reconciliation was based primarily on spouse data and study participant data only when spouse data was not collected. Analyses for this chapter are based on nonverified subjective questionnaire reporting. Analyses for this chapter are based on nonverified subjective questionnaire reporting. This report also contains data on children with defects and not defects per se. When a child was reported to have multiple birth abnormalities the most serious was analyzed. Sperm counts, and sperm abnormalities from the physical examination are also XI-1 �analyzed. Verification of reported fertility events is presently ongoing and the analyses presented here are based on interim unverified data. Seven thousand' three hundred ninety-nine conceptions are analyzed in this chapter. These represent 3293 Ranch Handers' or their spouses' reported conceptions and 4106 total comparison group or their spouses' reported conceptions. Comparison conceptions include 2669 original and 1437 shifted and replaced comparisons. The Ranch Hand and original comparisons' conceptions were analyzed considering 5 covariates: mother's smoking and drinking during each conception; mother's age; father's age; and the time of conception, i.e, before or after the father's military tour in Southeast Asia. Log-linear models were used to analyze the reproductive events of interest: miscarriages, still births, induced abortions, infant and neonatal deaths, and total numbers of live births. Live births were further analyzed for reported birth defects, learning disabilities and physical handicaps. Analyzed birth defects were those reported within a comprehensive range of ICD codes. Other reported birth defects included a broad range of pediatric conditions perceived by the parents as birth defects. Birth defects meeting ICD definition are further classified as to the severity of the defect. Fertility and reproductive outcomes were not analyzed by race for this report. These data will be presented in subsequent reports. Questionnaire collection of fertility and reproductive information was linked to reproductive events that occurred while the participant was married, living with a partner, or reported in the questionnaire as other pregnancies. Fertility and reproductive events were keyed to the specific relationship in order to reconcile the information with similar data collected from all available spouses and partners. Figure XI-1 presents an algorithm for the development of the fertility data base. XI-2 �Figure XI-1 ALGORITHM FOR THE DEVELOPMENT OF FERTILITY /REPRODUCTION DATA BASE otal Study Participant Total Spouse Reportet Reported Conceptions Conceptions 7204 6333 1 \ \ , \ v _ '.„,, .,„_ ,,._,_ ^ » Study Participant * Spouse Reported \ -—— __ Non-Live Birth Reported Live Births 6047 ^ Spouse Reported 1025 1 Non-live Births ""'""sso a'* i 1157 Respondent Data Base \ 1 Updated by Spouse x [ \ Merged Live Births 6085 i \ Twins and Triplets —•——— Counted as Single t Event .._l""~'l'" Live 6040 \ \ No Spouse Inf orma tion Response ""SriT" ,. , , -,- J<w ' i i L Non-Live 1359 1 1 Total Conceptions 7399 Of the 7204 total respondent reported conceptions shown in Figure XI-I 604? (84$) were reported as live births and 1157 (16$) were reported as nonlive births. The spouses reported 6333 total conceptions. These are shown in the upper right portion of the figure. Of the total conceptions reported by spouses as attributable to the male respondent, 5308 (84$) were reported as live births and 1025 (16$) were reported as nonlive births. Figure XI-1 shows that the spouse-reported births were matched to the respondent reported live births and 38 children were added to the respondent data base. Six thousand eighty-five live births were thus identified. The first born of multiple births were maintained in the data base and the remaining children were deleted yielding 6040 live births. Three hundred thirty-four nonlive births were added to the nonlive birth study subject file as a result of the match of the male respondent and spouse files. Seven thousand three hundred ninety-nine total conceptions are contained in the merge of the live and nonlive birth files. Xl-3 �The data in Figure XI~1 are based on unverified data. The data in the fertility file has not been fully cleansed of keypunch, editing or other potential sources of errors. The study participant data collection stressed natural children; but, inadvertently, data collection resulted in information on multi^ple adopted, step and natural children. Additionally, there was no data link between spouse, male respondent and children. Following receipt of data, a USAF computer system was created to define this link, but precise definition of total conceptions, live births and nonlive births must await verification by receipt of birth certificates and medical records. This processing is presently ongoing and will be finalized in future reports. Of the 7399 conceptions analyzed in this report 3293 were reported by Ranch Handers or their spouses and 4106 were reported by the total comparison group or their spouses. Comparison conception included 2669 in the subset of originally selected comparison individuals and 1^37 in the group of shifted and replacement comparisons. 2. Fertility/Infertility Data on the number of conceptions, number of marriages, duration of marital and nonmarital relationships, and the number of couples with the desired number of children were gathered during the in-home questionnaire. Three reproductive indices were derived from these data; the Infertility Index (number of childless marriages per total number of marriages), the Married Fertility Index (number of conceptions per years of marriage) and the Total Fertility Index (number of conceptions per years together). The Total Fertility Index includes time spent in nonmarital relationships. The data on fertility/infertility outcomes are presented in Table XI-1. XI- H �Table Xt-1 FERTILITY/INFERTILITY OUTCOMES FOR QUESTIONNAIRE COMPLIANT INDIVIDUALS Group PC P value; RH versus Originals All Variable RH AC Number of participants 1174 956 Number of Marriages 1456 1167 1860 Number of conceptions 3292 2668 4106 Number of participants with conceptions 1043 856 1359 1531 • Mean number of conceptions per participant 2.80 2.79 2.68 Mean number of marriages 1.24 1.22 1.21 - Number of childless marriages Infertility index Number of couples with children, having the desired number of children 385 283 0.264 0.243 708 560 - 448 0.241 891 0.32 0.23 0.67 0.73 Married fertility index 0.165 0.155 0.158 >0.25 >0.25 Total fertility index 0.163 0.154 0.157 >0.25 >0.25 RH = Ranch Hand OC = Original Comparisons AC = All Comparisons Although the crude numbers of conceptions and childless marriages differ between the Ranch Hand and comparison groups, the mean number of conceptions per participant and the proportion of marriages without children are not different. The percentages of couples with children who had the desired number of children, are not significantly different. Two hundred eighty-three of the 1045 Ranch Handers (27.1$) and 211 of the 733 originally selected comparisons (27.3$) attending the physical examination had vasectomies (P = 0.92). Seven hundred fifty-eight of the Ranch Handers (72.5$) and 561 of the comparisons (76.5$) submitted semen specimens. Of those participants willing and able to provide semen specimens, 186 Ranch XI-5 �Handers and 140 comparisons had vasectomies and/or orchiectomies (N = 6) and were therefore excluded from the statistical analysis of sperm counts. Six of these participants with a history of vasectomy were found to have sperm in their specimen and they were informed of these findings. The semen specimens from the remaining 993 participants were analyzed by general linear model techniques, using continuous variables of sperm count and the percentage of each participant's sperm which had abnormal morphology. The means, standard deviations and median values for the sperm counts and percent of sperm with abnormal morphology are displayed in Table XI-2. These analyses were adjusted for age and exposure to industrial chemicals, and revealed no significant group differences in sperm counts (adjusted P = 0.77), or in the percentage of abnormal sperm morphology (adjusted P = 0.71). Twenty-seven Ranch Handers and 19 comparisons had abnormal sperm morphology out of 560 and 409 analyzed specimens, respectively. Unprotected exposure to industrial chemicals (ever, never) had no significant effect in these analyses. However, age had a significant effect on sperm count (P = 0.0001), with sperm count increasing with age. The relevance of this observation is unclear since the counts may be biased somewhat by the differential compliance observed with increasing age. Compliance differed significantly with age (P < 0.001) but not by group (P = 0.78). This in sperm count increase was the same in both the Ranch Hand and comparison groups, with a slope of 1.69 in the Ranch Hand/original analysis, and 1.85 in the Ranch Hand/all analysis. These slopes were significantly different from zero (P = 0.0001). There was no significant association between age and abnormal sperm morphology (adjusted P = 0.57). The distribution of sperm counts in the two groups is presented in Figure XI-2, and the distribution of abnormal sperm morphology percentage is displayed in Figure XI~3. The patterns of compliance to semen specimen collection is shown in Figure XI-4. Table XI-2 DESCRIPTIVE STATISTICS OF SPERM VARIABLES BY GROUP Standard Deviation Mean P value Median Count (in million/ml) Original Comparisons Ranch Hand All Comparisons 111 .864 108.833 102.782 111.469 111.025 108.475 80 86 78 \ / \ 0.77 0.99 Percent Abnormal Sperm Original Comparisons Ranch Hand All Comparisons 9.614 9.705 9.643 5.182 5.525 5.946 XI-6 9 9 8 0.71 0.79 �Figure XI-2 DISTRIBUTION OF SPERM COUNTS BY GROUP 20-59 60-119 120-179 180-239 SPERM COUNT IN MILLION/ML XI-7 239 RANCH HAND COMPARISON �Figure XI-3 DISTRIBUTION OF ABNORMAL SPERM BY GROUP 0-4 5-9 10-14 15-19 20-24 PERCENT ABNORMAL FORMS XI-8 24 RANCH HAND COMPARISON �Figure XI- SEMEN SPECIMEN COMPLIANCE BY AGE AND GROUP 100 40 20 =35 35-39 40-44 45-49 AGE 50-54 x RANCH HAND * COMPARISON 3. Conception Outcomes In the evaluation of the outcomes of pregnancies fathered by study participants, analyses were conducted on all reported pregnancies in which the date of conception was known, and repeated on a subset of those in which information on maternal age, maternal smoking, and drinking habits was available from spouse questionnaires (complete data subset). There were an additional 95 conceptions in which data were too incomplete for analysis, and thus were deleted from the data base. There is no difference in the pattern of missing data between groups, as shown in Table XI-3. XI-9 the two �Table XI-3 COMPLETENESS OF CONCEPTION INFORMATION Group Complete Data Partial Data Incomplete Data Original Comparisons 2278 3t8'(13.0$) 42(1.6$) Ranch Hand All Comparisons 2781 (84.5$) 459 (13.9?) 3435 (83.75?) 599 (14.6$) (5W 8. \ 53 (1.6$) ( 72 (1.8$) / P Values • °'64 The occurrence of miscarriage was determined for each conception in which a date was reported. Similarly, outcomes of induced abortion, stillbirth and live birth were also determined. Adjustments for maternal factors of age (< 35, S 35), smoking (yes, no) and alcohol use (yes, no) and paternal age « 35, £ 35) could not be performed on these pregnancies with partial data, and no analysis was possible on those with incomplete data. In the covariate adjusted analyses, the primary statistical relationship of interest is the complex relationship between group outcome and time. Use of the pre-SEA conception experiences allows the Ranch Hand pre-SEA conceptions to serve as a standard for comparison with post-SEA conceptions. This is of special importance since 63-2$ of the Ranch Hand and 63.6$ of the comparison conceptions were pre-SEA events. Table XI-4 presents the data and the results of the analysis of these outcomes. Similar analyses using data from the entire comparison group are presented in Appendix X. The results of these additional analyses were essentially the same as those in Table XI-4. XI-10 �Table XI-4 ANALYSES OF CONCEPTION OUTCOMES, UNADJUSTED FOR MATERNAL COVARIABLES (COMPLETE AND PARTIAL DATA SUBSETS); RANCH HANDERS VERSUS ORIGINAL COMPARISON Yes Post-SEA Pre-SEA (50 No Yes No 190 (15.9) 130 (13.6) 1001 825 Miscarriage Ranch Hand Comparison (0) 295 (14.4) 205 (12.3) 1754 1467 0.06 0.13 Stillbirth Ranch Hand Comparison (0) 13 (0.6) 13 (0.8) 2036 1659 16 (1.3) 8 (0.8) 1175 947 P = 0 .27 0.60 Induced Abortion Ranch Hand Comparison (0) 13 14 (0.6) (0.8) 2036 1658 62 65 (5.2) (6.8) 1129 890 0.12 0.47 Live Birth Ranch Hand Comparison (0) 1723 (84.1) 1435 (85.8) 326 237 917 744 (77.0) (77.9) 274 211 0.62 0.15 These data demonstrate a borderline significant group difference in miscarriage (P = 0.06) prior to Southeast Asia duty and a suggestion of a difference (P = 0.13) post-SEA. However, inferences based on these analyses, unadjusted for key factors affecting pregnancy outcome, are of questionable value. Therefore, those conceptions in which full covariate information was known, were analyzed in greater detail. The data reflecting outcomes for both pre- and post-SEA conceptions are shown in Table XI-5, and the results of the adjusted analyses are displayed in Table XI-6. XI-11 �Table XI-5 CONCEPTION OUTCOMES (COMPLETE DATA SUBSET) BY GROUP MEMBERSHIP AND TIME; RANCH HANDERS VERSUS ORIGINAL COMPARISONS Yes Pre-SEA (%) Post-SEA (*) Yes No No Miscarriage Ranch Hand Comparison 239 172 (13.7) (11.9) 156 (15.0) 104- (12.5) 1505 1276 883 726 P = 0.12 P = 0.13 Stillbirth Ranch Hand Comparison 9 8 (0.5) (0.6) 12 8 1735 1140 P = 0.89 (1.2) (1.0) 1027 822 P - 0.69 Induced Abortion Ranch Hand Comparison 8 7 (0.5) (0.5) (3.6) (4.0) 37 33 1736 1441 1002 797 P = 0.61 P - 0.92 Live Birth Ranch Hand Comparison 1487 1258 (85.3) (86.9) 257 190 . 833 682 P = 0.19 (80.2) (82.2) P = 0.27 Table XI-6 RESULTS OF THE ANALYSIS OF CONCEPTION OUTCOMES; RANCH HANDERS VERSUS ORIGINAL COMPARISONS P value Relationship Miscarriage by Group by Pre/Post-SEA 0.76 Stillbirth by Group by Pre/Post-SEA 1.00 Induced Abortion by Group by Pre/Post-SEA 0.89 Live Birth by Group by Pre/Post-SEA 0.94 XI-12 206 148 �Although a group difference of 15/S versus 12.555 in post-SEA miscarriage is observed (P * 0.12), both groups had similar post-SEA conception outcomes relative to their own pre-SEA baseline experiences (P = 0.76). Ranch Hand miscarriages increased from 13-7? pre.-SEA to 15.054 post-SEA while comparison miscarriages increased from 11.9$ to 12.5%. Thus, while more Ranch Hand conceptions resulted in miscarriages than the comparisons, they started from a higher level before their herbicide exposures occurred, and in the overall analyses, there was no significant difference. These rates of miscarriage are comparable to estimates of 10-2055 for the general US population (Last, 1980). The rate of stillbirths in the US population is 0.98?, again comparable to the observed rates in this study. Similar analyses were conducted using data from all comparison individuals, and the results of these procedures were similar to those presented in Table XI-6. The data and analytic results of these additional analyses are shown in Appendix X. The effect of increasing maternal age was evident in all of these measures, with highly significant increases in miscarriage and induced abortion and decreases in live births associated with increasing age (P £ 0.01). The increase in induced abortions in both groups is unexplained, but is most likely the result of the altered legal status of induced abortion and its increased social acceptance. Exposure index analyses were performed in each of the three occupational categories (Officers; Enlisted, Flying; and Enlisted, Ground). The degree of exposure in each of these categories was stratified as low, medium or high (see Chapter VIII). Since the stratification by occupational category and exposure level and patterns of missing covariate data resulted in smaller groups, analyses had to be conducted using each covariate separately. A single analysis using all covariates would have resulted in unacceptably small cell sizes for meaningful analysis. The number of conception outcomes by occupational category available for each covariate analysis are presented in Table XI-7, and results of each covariate analysis are shown in Table XI"--8. XI-13 �Table XI-? NUMBER AND RESULT OF CONCEPTION OUTCOMES FOR EACH COVARIATE ANALYSIS BY OCCUPATIONAL CATEGORY Parameter Covariable Officers Yes No Category Enlisted Flying Yes No Enlisted Ground Yes No 102 102 122 122 Miscarriage Maternal Maternal Maternal Paternal Smoking Alcohol Age Age 34 34 44 44 225 225 241 250 19 19 22 22 100 100 119 119 Stillbirth Maternal Maternal Maternal Paternal Smoking Alcohol Age Age 2 2 3 4 257 257 282 290 2 2 2 2 117 117 139 139 Maternal Maternal Maternal Paternal Smoking Alcohol Age Age 17 17 18 24 242 242 267 270 6 6 9 9 113 113 132 132 14 14 23 29 630 630 707 710 Maternal Maternal Maternal Paternal Smoking Alcohol Age Age 205 205 219 219 54 54 66 75 92 92 108 108 27 27 33 33 521 521 576 576 123 Induced Abortion Live Birth XI-14 7 7' 8 9 542 542 608 617 637 637 722 730 123 154 163 �Table XI-8 RESULTS OF THE CONCEPTION/EXPOSURE INDEX ANALYSES Parameter Occupational Category Outcome/Exposure P Value, Adjusted for; Maternal Paternal Smoking Alcohol Age Age 0.04 0.30 0.54 0.04 0.26 0.50 0.07 0.19* 0.62 0.06 0.20 0.51 Officers Enlisted, Flying Enlisted, Ground 0.12 0.12 0.04* <0 .01 * 0.25 0.25 0.48 0.43* Officers Enlisted, Flying Enlisted, Ground 0.27 0.60 0.24 0.24 0.55* 0.23 0.57* 0.37* 0.29 0.59* 0.45 0.43 Miscarriage Officers Enlisted, Flying Enlisted, Ground Stillbirth Officers Enlisted, Flying Enlisted, Ground Induced Abortion Live Birth * Three-way covariate interaction is present. - Data too sparse for valid statistical analysis The only statistically significant findings observed are for miscarriage and for induced abortion among officers. Consistent patterns of increasing adverse outcomes of pregnancy with increasing herbicide exposure are not evident for other outcomes. In all four covariable analyses in the officer group, there was a significant association between miscarriage and exposure level (low, medium and high). 4. Live Birth Outcomes Those conceptions resulting in a live birth were further analyzed to determine the frequency of adverse events in those infants and children. As in the assessment of conceptions, unadjusted analyses were conducted on all reported live births in which a date of conception was known or could be estimated from the known date of birth. Analyses were repeated on those live births for which information on maternal age, maternal smoking, and maternal use of alcohol were available. Table XI-9 presents the distribution of live births within the subsets with complete and partial data. The difference in the proportion of the groups with only partial data are not statistically significant. Those births with inadequate data are omitted. XI-15 �Table XI-9 COMPLETENESS OF LIVE BIRTH DATA Complete Data Partial Data Total Original Comparisons 1940 (89.0$) 239 (11.0$) 2179 \ i Ranch Hand 2320 (87.8$) 320 (12.2$) All Comparisons 2922 (87.2$) 429 (12.8$) 2640 ( ) 3351 P Values \) • £ \ 0.43 Based on in-home questionnaire responses and respondent definitions of gestational age, there were no differences in the occurrence of prematurity, and postmaturity in the Ranch Hand and comparisons groups (PO.85). Further analyses of the incidence of prematurity based on objective criteria of birth weight will be conducted after birth certificate verification. Information concerning learning disabilities, physical handicaps, birth defects and the occurrence of neonatal and infant death was collected for each live birth. The information was obtained as a "yes" response primarily from the spouse questionnaire. Study subject responses were used when spouse data were unavailable. Data collection questions included: "Did (child) have any birth defects?"; "Does/Did (child) have a diagnosed learning disability?"; and "Does/Did (child) have any physical, mental, or motor impairments?" Yes responses to all 3 questions had been coded by the USAF from the ICD-9-CM based on the mother's or father's statement concerning the kind of birth defect, learning disability or physical, mental or motor impairment. For each defect reported for each child, the interviewer had the opportunity to document 3 statements within the question regarding the kind of birth or developmental problem. Therefore, each yes response had in some cases 3 ICD~9'-CH codes. A computer program was written to select defined birth defects, learning disabilities and physical, mental and motor impairments. For the child with multiple reported birth defects, he/she was counted only once for analysis. For children with multiple reported birth defects the most serious condition was analyzed. This report contains data on children with reported defects and not all reported defects; analyses of total reported defects will occur in a future report. A thorough review of the birth defect codes including key punch and code verification was accomplished prior to analysis of the merged data file. This review was not accomplished for reported learning disabilities or physical, mental and motor impairments, neonatal or infant death. The comprehensive definition of those reported defects within the definition for this report are presented in Appendix V. Reported birth defects not within the acceptable definition are presented in Appendix XIX. Counts of the total-reported and within-definition birth defects are presented in Table XI-10. Fifty-nine percent of the Ranch Hand and 64$ of the total comparison reported defects were within the acceptable defined range of birth defect. XI-16 �Table XI-10 COUNT AND PERCENT OF TOTAL REPORTED WITHIN-DEFINITION BIRTH DEFECTS Within Definition Total Reported Original Comparison Ranch Hand All Comparisons 218 292 137 172 212 ( W 6 334 P Values 0.37 0.24 The 5-6/6 difference in the perception of conditions which constitute a birth defect is not statistically significant. However, differential reporting of birth defects is of concern because media attention to hypothesized effects from exposure to the herbicide may affect parental reporting. In addition literature suggests the possibility that parents could perceive post-SEA births as "vulnerable" children (McCormick, 1982). Because of the above factors, all reported defects within range were categorized as severe, moderate, and limited (those of minor medical consequence) birth defects. This approach is based on a recent study (Christianson, 1981) which demonstrated that the incidence of reported congenital anomalies increased as children aged. Living children with reported defect average 23 years of age at the present time, with an age range of 2 through 39 years, and therefore, many years of parental observation have elapsed. The definition used for the collapsing of data into this system are as follows: Severe: Conditions which are life threatening or produce severe handicaps (e.g., physical, mental, motor). Moderate: Conditions which are not life threatening and handicaps which with medical care will not interfere with the individual's overall health or socio-economic progress. Limited: All conditions which without medical care would not interfere with the individual's health or socio-economic progress. Those reported birth defects without type of defect data were included in the limited category. Responses to birth defects which were unclear, incomplete or could be classified into more than one category were classified in the highest category applicable to the condition. Table XI-11 summarizes the reported birth defects categorized by level of severity system. XI-17 �Table XI-11 SUMMARY OF CHILDREN REPORTED WITH BIRTH DEFECTS BY LEVEL OF SEVERITY (SEVERE, MODERATE, LIMITED) RANCH HAND AND COMPARISON, PRE AND POST SEA TOUR Nature of Reported Defect Original Comparisons Counts % Ranch Hand Counts % Total Comparisons CountsT PRE-SEA Severe Moderate Limited 51 32 56. 35, 8 Ha., i •" 50 27 10 57 31 12 62 40 20 51 33 16 TOTAL 90 100 87 100 122 100 34 34 Ji 40 40 20 86 100 POST-SEA Severe Moderate Limited 32 22 26 TOTAL 80 18 20 10 100 48 37.5 111.5 21 100 TOTAL (PRE AND POST-SEA) Severe Moderate Limited TOTAL 83 54 33 49 32 19 68 47 20 50 35 15 96 74 38 46 36 18 170 100 135 100 208 100 This table shows that overall, 19% of the Ranch Hand, 15/t of the original and 18>6 of the total comparison group reported birth defects were classified as "limited." Ranch Handers reported &% limited pre-SEA and 32.5? post-SEA. Original comparisons reported '\2% pre^SEA and 2156 post-SEA and total comparisons reported 161 and 20%, respectively. These observations will be analyzed more fully in subsequent reports. Table XI-12 presents the analysis of the live birth outcomes for the partial and complete data subsets unadjusted for maternal factors of smoking, age and alcohol use. XI-18 �Table XI-12 ANALYSES OF LIVE BIRTH OUTCOMES, UNADJUSTED FOR MATERNAL COVARIABLES (COMPLETE AND PARTIAL DATA SUBSETS); RANCH HANDERS VERSUS ORIGINAL COMPARISONS Yes Pre~SEA (%) No Yes 1662 1373 77 51 Post-SEA (%) No Learning Disability Ranch Hand Comparison 61 62 (3.5) (4.3) P «= 0.26 (8.4) (6.9) 840 693 P = 0.24 Physical Handicaps Ranch Hand Comparison 144 112 (8.3) (7.4) 1579 1323 132 85 P = 0.57 (14.4) (11.4) 785 659 P = 0.07 Infant Death Ranch Hand Comparison 8 3 (0.5) (0.2) 1715 1432 4 (0.4) 3 (0.4) P = 0.23 913 741 P - 0.92 Birth Defects Ranch Hand Comparison 90 87 (5.2) (6.1) 1633 1348 80 48 P = 0.31 (8.7) (6.5) 837 696 P - 0.08 Neonatal Death Ranch Hand Comparison 25 17 (1,5) (1.2) P = 0.51 1698 1418 14 3 (1.5) (0.4) 903 741 P = 0.02 Live birth outcomes were not statistically different in the 2 groups prior to the participants tour of military duty in SEA. However, 3 of the 5 measures of outcomes after SEA duty demonstrated borderline or statistically significant differences between the Ranch Hand and comparison groups. The significant findings in neonatal deaths (P = 0.02), and the borderline significant XI-19 �finding for birth defects (P = 0.08) and physical handicaps (P = 0.07) were not adjusted for the effects of key covariables. Therefore, the data from those live births with full covariate information (complete data subset) concerning the maternal covariables were analyzed. Table XI-13 displays the pre-SEA and post-SEA data from this subset of births. Table XI-13 LIVE BIRTH OUTCOMES (COMPLETE DATA SUBSET); RANCH HANDERS VERSUS ORIGINAL COMPARISONS Parameter Pre-SEA (50 Post-SEA (50 No 75 47 (9.0) (6.9) 758 635 126 77 (15.1) (11.3) 707 605 No Yes 57 (3.8) 57 (4.5) 1430 1201 (9.0) (8.2) 1353 1155 Yes Learning Disability RH Comp Physical Handicap RH Comp Infant Death RH Comp 7 (0.5) 2 (0.2) 1480 1256 Birth Defects* RH Comp 78 (5.2) 80 (6.4) 1409 1178 76 44 Neonatal Death RH Comp 20 (1.3) 17 (1.4) 1467 1241 14 (1.7) 3 (0.4) 134 103 3 (0.4) 1 (0.1) (9.1) (6.5) 830 681 757 638 819 679 *Analysis includes 2 Ranch Hand birth defects which were double counted. Log-linear analyses, simultaneously considering all covariates (maternal age, maternal smoking, and maternal alcohol use, and paternal age) were accomplished. Table XI-14 confirmed the differences in birth defects initially seen in the unadjusted analyses of post-SEA live births. This finding was statistically significant (P = 0.04) after adjusted analysis. Suggestive associations were noted in learning disabilities (P = 0.19) and in neonatal deaths (P = 0.20). Incidence rates of neonatal death and infant death in the general US population are estimated to be 0.9958 and 1.4$, respectively (Last, 1980). The incidence rate of major birth defects in the general population is estimated to be 3-~5%, but varies, depending upon the criteria used to define the "defects." XI- 20 �Table XI-14 RESULTS OF THE ANALYSIS OF LIVE BIRTH OUTCOMES; RANCH HANDERS VERSUS ORIGINAL COMPARISONS Relationship P Value Learning Disability by Group by Pre/Post SEA Physical Handicap by Group by Pre/Post SEA 0.19 0.1(5 Infant Death by Group by Pre/Post SEA 0.81 Birth Defects by Group by Pre/Post SEA 0.04 Neonatal Death by Group by Pre/Post SEA 0.20 The distribution of reported post-SEA birth defects is presented in Table XI-15. This table clarifies the reported birth anomalies by level of medical consequence. Twelve congenital anomalies of the skin (ICD code 757) are present in the Ranch Hand data. This category of skin anomalies is quite broad, and includes simple birth marks, pigmentary changes, and more serious conditions. Reanalysis of the data concerning birth defects among live births in which full covariate data were available was accomplished with skin anomalies deleted. The birth anomalies included in the ICD category 757 are generally of minor medical consequences and their removal from analysis can be expected to provide a clearer understanding of group differences in birth defects of major health significance. This analysis revealed no significant group difference between Ranch Hand and comparison group live births for the remaining nonskin birth anomalies (P = 0.14). However, this weak association is still of interest. All reported birth defects are presently being validated by medical record reviews. Significant associations were noted (P < 0.05) between maternal smoking during pregnancy and learning disabilities, physical handicaps, infant deaths and birth defects. Maternal alcohol use during pregnancy was also associated with physical handicaps (P < 0.001). Future analyses of the birth defect data will also make use of the severity level classification. Live birth analyses using data from all of the comparisons were also conducted, and are contained in Appendix X. These analyses identified significant group differences in physical handicaps, birth defects and neonatal deaths. However, the influences of increased sample size and potential replacement group bias (differential reporting) have not been taken into consideration in these analyses. XI-21 �Table XI-15 COUNTS OF ANALYZED PQST-RVN BIRTH DEFECTS REPORTED BY RANCH HANDERS AND ORIGINAL COMPARISONS BY ICD CODE, LEVEL OF SEVERITY, AND AS STATED BY PARENT ICD-9-CM Codes 228 Ranch Harid Level of Severity S M L 1 Nomenclature Reported by Spouse/Study Subject Original Comparisons Level of Severity S M L 1 Blood tuner en nose Henagioma on left portion of head and face 1* 52* Micrcgnathia 5531 Umbilical hernia 711 1* 1* Spina bifida Open spine (severe case of Spina bifida) 712 1* 1* 1 Spinal cord and brain not connected Brain damage 1 713 744 1 Deaf in left ear (nerve underdeveloped) Malformed ear Bump en ear Missing small part of right earlobe 2 1 745 1 1* Septal defects Double cutlet right ventricle Heart murmur Foramen ovale was not totally closed 1 746 1 1 1 A congenital heart Heart valve Heart SV node, two nodes in heart Heart condition Blue baby 1 * 747 '3 1 748 Slightly, eye coordination 2* 1 Patent ductus Varicose vein in right groin Underdeveloped lungs, Premature Spot on lung XI-22 1 2 1 1 1 �Table XI-15 (Cont) COUNTS OF ANALYZED POST-RVN BIRTH DEFECTS REPORTED BY RANCH HANDERS AND ORIGINAL COMPARISONS BY ICD CODE, LEVEL OF SEVERITY, AND AS STATED BY PARENT ICD-9-CM Codes 749 Ranch Hand Level of Severity S M L 1* Cleft lip Cleft palate 2 Pyloric stenosis Skin growing across his esophagus Large bubble or abscess on throat TE fistula Tongue tied 2 1 750 751 Nomenclature Reported by Spouse/Study Subject Original Comparisons Level of Severity S M L 1 1 Couldn't eat her food Uhdescended testicle Hypospadia Opening for urinating lower than normal Vagina fused, had operation 752 3 1 753 Defective kidney Malformation of right kidney Infantile polycystic kidney disease 754 Talipes Club foot Dislocated hips Leg bowed in at birth required cast and then braces Chest cavity deformity Ankle bones deformed Foot turned in Toes turned in 2 2 Left hand had no fingers, has thumb Crooked femur bone Possible hip or feet or both developed later Deformed feet Two toes joined together Hip and foot defect, wore a brace Extra finger and toe 1 1 755 1- XI-23 3 1 2 1 2 1 �Table XI-15 (Cont) COUNTS OF ANALYZED PQST-RVN BOTH DEFECTS REPORTED BY RANCH HANDERS AND ORIGINAL COMPARISONS BY ICD CODE, LEVEL OF SEVERITY, AND AS STATED BY PARENT ICD-9-CM Codes Ranch Hand Level of Severity S M L 1 Nomenclature Reported by Spouse/Study Subject Original Comparisons Level of Severity S M L 3 1 Leg turned in, wore a cast for 3 months Bones from knees to ankles grew inward Webbed finger on hand Delta phalanges of index fingers Crooked foot or legs Leg problem, knees hurt as infant 1 Unusually tiny head Premature fusion of sagittal sutures Skull slightly deformed Bone deformity Small neck muscles from being in breach position Feet curved in at birth 1 1 1 5 1 1 Ichthyosis No finger or toe nails Skin pigmentation Skin discoloration Yellow color, disappeared in a week Birthmarks Two nipples on breast Skin tags Down's Syndrome 3 1 1 756 1 1 1 757 1 2 758 2 TOTAL 3D 18 26 = 74 1 1 1 1 1 19 19 *Child deceased. Table XI-15 relates the ICD codes to the level of severity to the reported statement of the spouse or study participant. Of the 7** post<-RVN Ranch Hand reported birth defects, 30 are of a severe and 18 of a moderate level of severity. Counts of reported birth defects pre-RVN and post-RNV by occupational category are presented in Table XI-16. Inspection of this table shows that the increase in reported birth defects postf-RVN are predominately from personnel in �the Ranch Hand and total comparison enlisted ground occupational category. However, these data have not yet been adjusted by the number of live births in each occupational category. Table XI-16 COUNTS OF REPORTED BIRTH DEFECTS PRE- AND POST-SEA BY OCCUPATIONAL CATEGORY (OFFICER, ENLISTED-FLYING, ENLISTED-GROUND) Occupational Category Officer Enlisted Flying Enlisted Ground TOTAL Ranch Hand Pre-SEA Post-SEA Counts Counts Original Comparisons PrerSEA Posfr-SEA Counts Counts Total Comparisons Pre-SEA Post-SEA Counts Counts 44 15 40 16 52 22 13 12 15 5 21 10 21_ _49 25 23 40 45 78 76 80 44 113 77 Exposure analyses were performed using the covariates of maternal age, maternal smoking, maternal alcohol use, and paternal age. Each covariable was analyzed separately. The number and result of live birth outcomes by occupational category available for each covariate analysis are presented in Table XI-17 and the results of each covariate analysis are shown in Table XI-18. XI-25 �Table XI-17 NUMBER AND RESULT OF LIVE BIRTH OUTCOMES FOR EACH COVARIATE ANALYSIS BY OCCUPATIONAL CATEGORY Parameter Learning Disability Covariable Officers Yes No Category Enlisted Flying No Yes Enlisted Ground Yes No Maternal Maternal Maternal Paternal Smoking Alcohol Age Age 15 15 16 16 190 190 203 203 8 8 8 8 84 84 100 100 52 52 53 53 469 469 523 523 Maternal Maternal Maternal Paternal Smoking Alcohol Age Age 26 26 26 26 179 179 193 193 12 12 13 13 80 80 95 95 81 81 86 86 440 440 490 490 Maternal Maternal Maternal Paternal Smoking Alcohol Age Age 1 1 1 1 204 204 218 218 1 1 1 1 91 91 107 107 2 2 3 3 519 519 573 573 Birth Defects Maternal Maternal Maternal Paternal Smoking Alcohol Age Age 12 12 12 12 193 193 207 207 11 11 12 12 81 81 96 96 50 50 53 53 471 471 523 523 Smoking Alcohol Age Age 3 3 3 3 202 202 216 216 4 4 4 4 88 88 104 104 6 6 6 6 515 515 570 570 Physical Handicap Infant Death Neonatal Death Maternal Maternal Maternal Paternal XI-26 �Table XI-18 RESULTS OF THE LIVE BIRTH/EXPOSURE INDEX ANALYSES Outcome/Exposure P Value, Adjusted for; Maternal Paternal Smoking Alcohol Age Age Parameter Occupational Category Learning Disability Officers Enlisted, Flying Enlisted, Ground 0.47 0.46 0.31 0.31* 0.92 0.94 0.89 0.85 Physical Handicap Officers Enlisted, Flying Enlisted, Ground 0.07 0.89 0.78 0.07 0.69 0.79* 0.06 0.47 0.76* 0.05 0.56 0.79 Infant Death Officers Enlisted, Flying Enlisted, Ground Birth Defects Officers Enlisted, Flying Enlisted, Ground 0.02 0.03 0.39 0.02 0.06 0.35 0.02 0.03 0.46 0.02 0.03 0.41 Neonatal Death Officers Enlisted, Flying Enlisted, Ground - Data too sparse for valid statistical analysis. * Significant three-factor interaction is present. These results demonstrate consistency across all covariates for each of the live birth outcomes; however, as noted in Table XI-18, the data are sparse in many instances, especially for officer and enlisted flying personnel. Birth defects are found to have a statistically significant association with herbicide exposure level in the officer and enlisted flying groups. However, there is not a consistent increase in defects with increasing exposure in the officer category. In the enlisted flying group the adverse outcome did increase consistently with increasing exposure. The pattern in the officer group demonstrated a two-fold rise in the medium level but the highest exposure group had the lowest proportion of children with defects (1.2$). Physical handicaps in children of officers demonstrated borderline significance. 5. Summary A summary of the findings of the fertility and reproductive analyses are displayed in Table XI<-19. XI-27 �Table XI-19 SUMMARY OF FERTILITY AND REPRODUCTIVE ANALYSES Parameter Infertility Sperm Count Sperm Abnormality Unadjusted 0 A NS P Values Exposure Analyses by Occupational Group Adjusted Enlisted Enlisted 0 A Officers Flying Ground NS NS NS NS NS NS NS NS NS NS NS NS NS Conception Outcomes Miscarriage Stillbirth Induced Abortion Live Birth 0.13 NS 0.12 NS 0.15 0.10 NS NS 0.04 0.19 NS 0.12 NS NS NS NS NS 0.05 NS NS NS 0.02 0.03 NS Live Birth Outcomes Prematurity Learning Disability Physical Handicap Infant Death Birth Defects Defects Excluding Skin Anomalies Neonatal Death NS NS 0.07 NS 0.08 0.05 <0.01 NS 0.04 0.19 NS NS 0.04 0.12 0.02 NS 0.02 0.02 <0.01 0.14 0.20 0.07 0.03 NS = Nonsignificant 0 = Original Comparisons A = All Comparisons The analyses in this chapter did not reveal any significant differences in fertility/infertility and sperm counts between the Ranch Hand and either comparison group. Conception outcomes of miscarriage, stillbirth, induced abortion and live births also were not found to differ significantly. Analyses unadjusted for known risk factors of pre-SEA conception history, maternal age, maternal smoking, and maternal alcohol use, and paternal age revealed a suggestive association for increases in miscarriage after the father's SEA service in the Ranch Hand group. However, this association and a borderline increase in post-SEA induced abortion in the original comparison group were not evident after consideration of these other risk factors. Analyses of these conception outcomes with the herbicide exposure index also did not reveal any evidence of herbicide effects. A statistically significant association between increasing herbicide exposure and miscarriage was identified in the officer group but this XI-28 �effect was not observed in the other occupational categories. Borderline significance was noted in officers for stillbirth and induced abortion, but these findings did not increase in occurrence with increasing exposure. Significant differences were reflected in the analyses of live birth outcomes. These differences were observed for birth defects after the analyses were adjusted for parental covariates. There appeared to be a clustering of birth anomalies of the skin in children of the Ranch Handers. There were no significant group differences for other birth defects, but a suggestive association remained (P = 0.1*0 after reanalysis with the skin anomalies excluded. Suggestive group differences between the Ranch Handers and original comparisons were also observed after adjusted analysis for learning disability and neonatal death. Exposure analysis identified several findings of statistical and borderline significance; however, the patterns were not consistent across occupational strata. Overall, birth defects demonstrated statistical significance in the adjusted intergroup analysis, and 2 of the 3 occupational group exposure analyses. A larger number of live birth outcome differences were observed in analyses comparing the Ranch Handers to the total comparison group; however, it is unclear whether these differences are true group differences, or are due to changes in sample size or replacement bias (differential reporting). The value of these analyses in making inferences is therefore limited at this time. The findings in this chapter do require further evaluation of the possible link between herbicide/dioxin exposure and birth defects. The analyses have relied heavily on unverified spouse reports, and the effect of differential reporting of conception and birth outcomes in pregnancies and in children who the parent might perceive as "special" or "vulnerable" has not been evaluated. This evaluation will be conducted using birth certificates and medical records so that an analysis of verified fertility/reproductive data can be included in the report of the first follow-up physical examination. XI-29 �Chapter XII NEUROLOGICAL ASSESSMENT 1. Introduction Neurological abnormalities have long been recognized as acute toxic effects following the exposure of humans to phenoxy herbicides and dioxin (Goldstein, 1959; Wallis, 1970; Berkley, 1963; Boeri, 1978). Signs and symptoms, such as hyporeflexia, a decrease in nerve conduction velocity, general muscular weakness and decreased sensation in the extremities have been noted. One study documented demyelination as a result of 2,4-D exposure (Dudley, 1972). While these effects have only been demonstrated acutely following heavy exposures, complaints of peripheral neuropathy are prominent among Vietnam veterans who have participated in the Veterans Administration Agent Orange Registry Program. Twelve percent of the 110,000 patients in the Registry had complaints compatible with symptoms of peripheral neuropathy. The recognized acute neurotoxicity of these chemicals and the prevalence of neurological complaints among veterans were primary factors in the decision to place a major emphasis on the neurological evaluation of participants in this study. During the administration of the questionnaire, each subject was asked to provide information on any major health conditions he may have experienced. All reported neurological conditions were coded using the ICD-9~CM and group analysis of the distribution of the conditions was performed. As revealed in Table XII-1, there were no statistically significant differences in reported neurological diseases between the Ranch Hand and comparison groups. Table XII-1 DISTRIBUTION OF REPORTED NEUROLOGICAL DISEASES BY GROUP MEMBERSHIP Disease Category Original Comparisons Ranch Hand 2 2 7 15 Inflammatory Diseases Hereditary and Degenerative Diseases Peripheral Disorders Disorders of the Eye Disorders of the Ear and Mastoid All Comparisons 3 1 7 3 3 11 21 21 23 P = 0.73 XII-1 0.69 �There were 1045 Ranch Handers, and 773 originally selected comparisons included in the analyses in this chapter. Where analyses were accomplished using the total comparison group, the data from 1194 comparisons were used. Some variation in numbers did occur due to missing data. In the analyses of the data obtained from the neurological evaluation, only those participants with a negative serological test for syphilis were included since chronic neurological disease can result from inadequately treated syphilis (5 Ranch Handers and no comparisons were found to have positive serological tests for syphilis.) In addition, data from 15 individuals found to have edema of the extremities on physical examination (8 Ranch Handers and 7 comparisons) were deleted from the analyses of the peripheral sensory nerve evaluation and nerve conduction velocities since edema can interfere with these clinical evaluations. Several covariables were considered in the analysis. The use of alcohol (dichotomized to ever/never); years of unprotected exposure to industrial chemicals (yes, no), insecticides (yes, no), and degreasing chemicals (yes, no); and 2-hour postprandial glucose levels equal to or greater than 120 mg/dl were used as covariates. 2. Cranial Nerve Status The functional integrity of all 12 cranial nerves was assessed during the neurological examination. The specific cranial nerves and the examination parameters used in their evaluation are listed in Table XII-2. XII-2 �Table XII-2 CRANIAL NERVE EVALUATION Cranial Nerve Parameter I Sense of smell Olfactory II Optic Visual fields Oculomotor Pupillary reaction to light Ocular movement IV Trochlear Ocular movement V Trigeminal Facial sensation Corneal reflex Clenching jaw III VI Abducens Ocular movement VII Facial Smile Palpebral fissure VIII Acoustic Balance (Romberg Sign) IX, Glossopharyngeal X Gag reflex Vagus Speech Tongue position XI Spinal Accessory Palate and uvula movement Neck movement XII Neck range of motion Hypoglossal Analysis of the examination data revealed no statistically significant differences in cranial nerve function between the Ranch Hand and comparison groups. No significant three-way interactions between the examination parameters, group membership and the covariables of glucose and alcohol were noted. These results are summarized in Table XII-3. Data from the entire comparison group are also presented. XII-3 �Table XII-3 ANALYSIS OF CRANIAL NERVE FUNCTION Cranial Nerve Parameter I Smell, left Smell, right II Group RH OC AC RH OC AC # Normal P Values; Ranch Hand Versus Original All # Abnormal Comparisons Comparisons 19 12 19 759 1172 1027 760 1171 0.67 0.68 17 11 17 1025 0.73 0.70 0.91* 0.87* Visual fields, RH OC left 1037 AC 1186 3 2 3 Visual fields, RH right OC 1038 768 1186 2 3 4 0.43* 0.51* 1031 763 1180 8 4 6 0.52 0.43 655 486 7*6 349 265 423 0.82 0.49 1035 769 1190 7 4 4 0.68 0.26 OC AC RH OC AC 1038 770 1191 4 3 3 0.99* 0,58* Corneal reflex RH 1043 772 1193 2 1 1 0.75* 0.49* OC AC RH OC AC 1042 773 1194 1 0 0 AC III Light reaction RH OC AC III-IV, Ocular movement RH OC VI AC V Sensation, left RH Sensation, right Jaw clench 768 _ XI I-4 _ �Table XII-3 (Cont'd) ANALYSIS OP CRANIAL NERVE FUNCTION Cranial Nerve Parameter Grour # Normal P Values; Ranch Hand versus Original All # Abnormal Comparisons Comparisons Smile RH OC AC 1035 767 1186 4 2 4 Palpebral fissure RH OC AC 986 731 1131 VIII Balance RH OC AC IX Gag reflex Speech VII 0.65* 0.85* 59 42 63 0.84 0.70 833 625 813 207 148 228 0.69 0.26 RH OC AC 1030 760 1180 15 13 14 0.67 0.58 RH OC AC 1041 770 1190 0 1 RH OC AC 879 662 1085 4 2 Palate and RH uvula movement OC AC 1042 771 1192 3 1 1 RH OC AC 1004 748 1158 41 Tongue in midline XI XI, XII Neck range of motion *P RH OC AC - 0.26* 3 0.63* 0.51* 0.48* 0.26* 0.44 0.24 3 25 36 values are of limited validity due to small cell sizes in these analyses = Ranch Hand = Originally selected comparison = All comparisons = Cells containing zeros; P values not valid XI I-5 �The 18 neurological parameters listed in Table XII-3 were again analyzed with regard to occupational group and exposure level. The exposure index, stratified into 3 occupational groupings and 3 levels of exposure, was applied to these cranial nerve data. These results are summarized in Table XI1-4. Fully adequate cell sizes were obtained in only 13 instances. In these analyses, in which no individuals in either group had abnormalities, statistical testing for significance was invalid, and P values are not given. Table XII-4 CRANIAL NERVE FUNCTION VERSUS EXPOSURE LEVEL WITH EACH OCCUPATIONAL CATEGORY Cranial Nerve Parameter I Smell, left 0/F E/F E/G 0.79 0.67 0.16 Smell, right 0/F E/F E/G 0.01 0.84 0.31 Visual fields, left 0/F E/F E/G 0.05 O.HO 0.41 Visual fields, right 0/F E/F E/G 0.06 0.40 0.11 Light reaction 0/F E/F E/G 0.32* II III Occupational Category P Value 0.28 III, IV, VI Ocular movement 0/F E/F E/G 0.21* 0.33* 0.47* V Sensation, left 0/F E/F E/G 0.32 0.12 0.72 Sensation, right 0/F E/F E/G 0.64 0.34 0.35 Corneal reflex 0/F E/F E/G 0.55 XII-6 �Table XII-4 (Cont'd) CRANIAL NERVE FUNCTION VERSUS EXPOSURE LEVEL WITH EACH OCCUPATIONAL CATEGORY Cranial Nerve Parameter Occupational Category Jaw clench 0/F E/F E/G 0 .64 0/F E/F E/G 0 .64 0 .57 Palpebral fissure 0/F E/F E/G 0 .97* 0 .14 0 .12* VIII Balance 0/F E/F E/G 0 .89* 0 .25* 0 .44* IX Gag reflex 0/F E/F E/G 0 .99 0 .84 0 .20 Speech 0/F E/F E/G 0 .38 0.34 0 .11 Tongue in midline 0/F E/F E/G 0 .07* 0 .30* 0 .40* Palate and uvula movement 0/F E/F E/G 0 .64 0/F E/F E/G 0 .67* 0 .78 0.46 VII XI XI, XII Smile Neck range of motion 0/F = Officer, flying E/F = Enlisted, flying * = Cell sizes of 5 or less i- » Cells containing zeros; P values not valid XII-7 P Value - - .0 .43 E/G = Enlisted, ground �3. Peripheral Nerve Status The variables used in the assessment of peripheral nerve function were analyzed with the covariates of 2-hour postprandial glucose in excess of 120 mg/6, history of alcohol use and unprotected exposure to industrial chemicals, insecticides and degreasing chemicals. There were statistical interactions between group membership (Ranch Hand and comparison) and insecticide exposure, and between insecticide exposure and the other covariables. Since these relationships have no impact on the primary question being addressed by this study, further statistical analyses of these interactions will not be undertaken at this time. Analysis of the data pertaining to the peripheral nervous system is summarized in Table XII-5. Data from the entire comparison group are also presented. With the exception of a borderline association between group and Babinski reflex in the originals and a significant association in the entire comparison group, these analyses did not demonstrate statistically significant differences in neurological functions between the 2 groups. Matched pair analyses were performed on the Babinski reflex and the vibration sense data, using the Breslow matched logistic regression technique. A P value of 0.18 was found for the Babinski reflex and a nonsignificant P value of O.1!? was found for vibration sense. Significant interactions were, however, detected between postprandial glucose levels and several of the examination parameters. The association between abnormal glucose metabolism and peripheral neurological disease is well recognized (Scientific American, 1983) and its demonstration in this study reflects a degree of confidence in the quality of the neurological data collection process. These glucose by neurological disease associations are shown in Table XII-6. A positive history of alcohol use had borderline significance with pin prick (P = 0.07). In this analysis, a continuing effect of abnormal glucose is seen for vibration (P = 0.0005), patellar reflex (P = 0.03), Achilles reflex (P = 0.04), and light touch (P = 0.03). Alcohol use also had a borderline significant effect on pin prick (P = 0.07). XII-8 �Table XI1-5 ANALYSIS OF THE PERIPHERAL NERVOUS SYSTEM Parameter Group # Normal P value; Ranch Hand versus All Original # Abnormal Comparisons Comparisons Pin prick RH OC AC 934 691 930 97 73 101 0.94 0.76 Light touch RH OC AC 958 707 953 73 57 78 0.78 0.67 Muscle Status (strength, bulk) RH OC AC 1003 745 1009 37 28 32 0.94 0.62 Vibration RH OC AC 954 698 941 78 67 91 0.38 0.30 Patellar Reflex RH OC AC 1034 766 1003 4 5 5 0.45 0.74 Achilles Reflex RH OC AC 995 746 1005 39 26 35 0.62 0.62 Biceps Reflex RH OC AC 1030 767 1032 8 H 8 0.53 1.00 Babinski Reflex RH OC AC 1024 770 1039 9 2 2 0.10 0.03 RH = Ranch Hand OC = Original comparisons AC = All comparisons XII-9 �Table XII-6 POSTPRANDIAL GLUCOSE ABNORMALITIES VERSUS NEUROLOGICAL FINDINGS (RANCH HANDERS VERSUS ORIGINAL COMPARISONS) Parameter Examination Status Glucose Status # Normal # Abnormal P Value Light Touch Normal Abnormal 1406 100 259 30 0.03 Vibration Normal Abnormal 1402 106 250 39 0.0005 Patellar Reflex Normal Abnormal 1514 5 286 4 0.03 Achilles Reflex Normal Abnormal 1463 48 273 17 0.04 Pin prick Normal Abnormal 1369 137 256 33 0.23 The data from the Ranch Hand group were also analyzed against the exposure index. As shown in Table XII-7, there were no three-way interactions between occupational group, herbicide exposure and the neurological parameters evaluated. No statistically significant results were found in the analysis of exposure versus examination parameters. Borderline associations were noted for vibration in the enlisted flying group (P = 0.10) and for Babinski Reflex in the enlisted ground personnel (P == 0.09). The relevance of these findings, in the face of the other negative results, is unclear at this time. There were no distinct patterns of increasing abnormality with increasing exposure. Table XII-7. PERIPHERAL NEUROPATHY BY EXPOSURE ANALYSES: SUMMARY OF P VALUES Occupational Group Parameter Pin prick Light Touch Muscle Status Vibration Patellar Reflex Achilles Reflex Biceps Reflex Babinski Reflex Officer Enlisted Flying Enlisted Ground 0.78 0.40 0.43 0.94 0.50 0.35 0.49 0.57 0.99 0.83 0.96 0.10 0.57 0.53 0.57 0.53 0.47 0.81 0.65 0.96 1.00 0.60 0.91 0.09 XII-10 �4. Evaluation of Central Functioning A brief evaluation of central nervous system coordination processes was accomplished, focusing on the presence of muscle tremor, finger-to-nose coordination, gait and balance as assessed by the modified Romberg Sign. These analyses are shown in Table XII--8. As in the analysis of the peripheral nerves, there were no significant interactions of these findings with chemical exposures or group membership; however, abnormal glucose metabolism was associated with abnormal balance (P = 0.0002) and the presence of tremor (P = 0.00*1). Alcohol also had a significant effect on the presence of tremor (P = 0.05) and a borderline effect on balance (P = 0.09). Breslow matched pair analysis of the tremor and coordination data revealed nonsignificant P values of 0.21 and 0.31 respectively. Table XII-8 ANALYSIS OF CENTRAL FUNCTION P values; Ranch Hand versus Original All # Normal # Abnormal Comparisons Comparisons Parameter Tremor RH OC AC 985 742 995 55 31 46 0.19 0.36 Coordination RH OC AC 992 7^3 998 48 30 43 0.44 0.59 Romberg Sign RH OC AC 833 625 813 207 148 228 0.64 0.26 Gait RH OC AC 24 14 22 0.47 0.76 758 1018 RH = Ranch Hand OC = Original comparisons AC = All comparisons Exposure analysis was performed on these parameters as well. Three-factor analysis of parameter by exposure level by occupational group again demonstrated no significant interactions. In these analyses, the herbicide exposure/coordination analysis yielded a suggestive association (P = 0.10). Again, there was a statistically significant association between an abnormal Romberg Sign and abnormal glucose metabolism (P = 0.002). Two-way analysis results are shown in Table XII-9. XII-11 �Table XII-9 HERBICIDE EXPOSURE VERSUS ABNORMALITY OF CENTRAL FUNCTIONING SUMMARY OF P VALUES Parameter Tremor Coordination Romberg Sign Gait Officers P Values Enlisted Flying 0.50 0.07 0.89 0.54 0.76 0.16 0.25 0.38 Enlisted Ground 0.20 0.63 0.44 0.11 5. Nerve Conduction Velocity Nerve conduction was evaluated using a continuous measurement and analyzed using a general linear model technique for maximal statistical power. Velocities were measured from 2 locations in the ulnar nerve and from 1 P°s~ ition in the peroneal nerve. Covariables in these analyses included history of alcohol use (measured in drink-years), abnormalities in postprandial glucose levels (equal to or greater than 120 mg/dl), and unprotected exposure to industrial chemicals, insecticides and degreasing chemicals. No associations between the chemical exposures and conduction velocities were identified on covariate analysis; however, highly statistically significant associations were noted in both the Ranch Hand and comparison groups between alcohol use and glucose and conduction velocity. This association held for both measurements of the ulnar nerve (P ts 0.01) with the velocity decreasing as the drink-years of alcohol increased. Glucose was found to be associated with conduction velocity in the peroneal nerve (P = 0.002) and both ulnar velocities (P = 0.001) with velocity decreasing as glucose level increased. These analyses did not demonstrate any significant intergroup differences in velocities in either nerve. The unadjusted and adjusted means and their respective P values are presented in Table XII-10. Similar analyses, using data from the entire comparison group, were performed with similar means and results. Table XII-10 NERVE CONDUCTION VELOCITY (M/SEC) AND GROUP MEMBERSHIP Nerve Ulnar (above the elbow) Group (N) Unadjusted Mean P Value Adjusted Mean P Value R (1035) C (769) 55.88 56.15 0.30 55.89 56.12 0.38 Ulnar R (1042) (below the elbow) C (771) 60.50 60.73 0.39 60.52 60.71 0.48 Peroneal 48.22 48.14 0.74 48.23 48.93 0.66 R (1041) C (769) XII-12 �Herbicide exposure analyses were performed using the covariates of occupational group serum glucose and history of alcohol use. These results are shown in Table XII-11. Table XII-11 ADJUSTED MEAN NERVE CONDUCTION VELOCITY (M/SEC) AND EXPOSURE Nerve Low Exposure Med-High High P Value Officers Ulnar (above elbow) 55.77 55.66 55.97 0.90 Ulnar (below elbow) 60.54 60.60 61.10 0.70 Peroneal 47.69 47.76 47.87 0.96 Ulnar (above elbow) 54.54 55.72 55.35 0.53 Ulnar (below elbow) 58.31 60.68 60.83 0.03 Peroneal 48.22 48.28 48.29 0.99 Ulnar (above elbow) 55.53 56.60 56.33 0.24 Ulnar (below elbow) 59.96 60.74 60.69 0.96 Peroneal 48.34 48.31 49.00 0.14 Enlisted Flying Enlisted Ground These exposure analyses have not demonstrated any consistent trends in conduction velocity and increasing exposure either within or between occupational categories. A single significant result (P = 0.03) was found in the distal ulnar nerve velocity in flying enlisted personnel, but there was no corresponding finding in the same nerve when measured over a larger distance above the elbow (P = 0.53). The borderline significance in the peroneal nerve velocity of ground enlisted personnel (P = 0.14) was not evident in the other occupational categories. Again, significant associations with glucose were noted, with P values falling between 0.06 and-0.005. XIl-13 �6. Summary As summarized in Table XII-12, detailed analyses of the neurological examination data pertaining to the status of the cranial nerves, peripheral nerves and central functioning were performed. Table XII-12 SUMMARY OF NEUROLOGICAL STATUS Parameter Group Analyses JP Values) Exposure Off Enl Fly Enl Gnd Cranial Nerves 1 2 3 4 0.01 0.05 NS NS NS NS NS NS 0.16 0.11 NS NS 5 NS NS 0.12 NS 6 7 • NS NS NS NS NS NS NS NS NS 0.14 NS 0.12 8 9 10 11 12 NS NS NS NS NS NS NS 0.07 NS NS NS NS NS NS NS NS NS 0.11 NS NS NS NS NS NS NS NS NS 0.10 NS NS NS NS NS NS NS NS NS NS NS 0.10 NS NS NS NS NS NS NS NS NS NS NS 0.09 0.19 NS NS NS NS 0.07 NS NS NS 0.16 NS NS NS NS NS 0.11 NS NS NS NS 0.03 NS NS NS 0.14 Peripheral Nerves Pin Prick Light Touch Muscle Status Vibration Patellar Reflex Achilles Reflex Biceps Reflex Babinski Reflex Control Function Tremor Coordination Romberg Gait Conduction Velocity Proximal Ulnar Distal Ulnar Peroneal NS NS NS NS = Nonsignificant XII-14 �With the exception of a borderline increase in the proportion of Ranch Handers with a positive Babinski reflex, there were no significant differences detected between the Ranch Hand and comparison groups with respect to neurological parameters. The Babinski reflex, however, did not show a significant relationship to past herbicide exposure. There were no consistent findings of increasing abnormality with increasing herbicide (dioxin) exposure. The relative risks and confidence intervals for the dependent variables - analyzed in this chapter are included in Appendix XVIII. Thus, it appears at this time, that there are no neurological abnormalities in the Ranch Hand group that can be attributed to herbicide exposure in Vietnam. The evaluation of neurological status among the participants in this study has demonstrated the ability to identify classical interactions between abnormal glucose metabolism and alcohol use and evidence of neurological abnormalities. These findings lend confidence to the validity of the negative findings of a chronic herbicide (dioxin) effect on the neurological system. XII-15 �Chapter XIII PSYCHOLOGICAL ASSESSMENT Since 1961 , psychological abnormalities have been ascribed to acute phenoxy herbicide exposure (Bauer, 1961). Subsequently, a wide range of psychological symptoms, including anxiety, depression, emotional instability, and asthenia have been reported following exposure (Monarca and di Vito, 1961; Kramer, 1974; Poland et al, 1971). Since many Vietnam veterans have expressed concern that their exposure to the defoliants during the war caused them to experience psychological and behavioral problems, the psychological functioning of the study participants was assessed in both the questionnaire and physical examination phases of the study. Overall, the responses of 1045 Ranch Handers, 1230 comparisons, and a subset of 773 originally selected comparisons were analyzed. Slight variations in these numbers occurred in some analyses due to missing data. Except where indicated, all analyses reported in this chapter used the data from the subset of originally selected comparisons. Each participant was asked whether he had ever experienced psychological illness. Additionally, six specific psychological dimensions were explored in detail in the questionnaire: depression, anxiety, erosion of skills, social isolation, fatigue, and aggressive or impulsive behavior. The questions used were selected from an extensive test battery, previously developed and validated (Robbins, 1982). More standardized measurements of psychological performance were obtained during the physical examination by the use of several standardized tests. The Cornell Index, the Minnesota Multiphasic Personality Inventory (MMPI), the HalsteadReitan Battery and the Wechsler Adult Intelligence Scale (WAIS) were the primary testing instruments. Throughout much of this chapter, educational level (high school versus college) and rank (officer versus enlisted status) received special attention in all analyses. These variables are widely recognized as having major influences on psychological testing performance (Dalstrom, 1960) and their importance in the setting of the Air Force Health Study was very apparent. Dependent variables were stratified by education and rank, and in log-linear techniques, they were used as covariables. Table XIII-1 displays the education and rank distributions of the Ranch Hand and original comparison groups. Table XIII-1 EDUCATION AND RANK DISTRIBUTION OF RANCH HAND AND ORIGINAL COMPARISON GROUPS Ranch Hand High School College Officers Enlisted Original Comparisons High School College 54 (14.3?) 324 (85.7?) 53 (18.2?) 239 (81.8?) 521 (80.8?) 124 (19.2?) 377 (79.4?) 98 (20.6?) XIII-1 �Regardless of statistical technique or procedure, the analytic results of all psychological testing from the high school group closely mirrored those of the enlisted group, and college results matched those of the officer group, since, in general, the attainment of a college degree is a prerequisite for commissioning as an officer. However, 121! of the Ranch Hand enlisted and 98 of the original comparison enlisted personnel have college degrees as well. The similarities between these groups are graphically demonstrated in Figure XIII-1, where full scale IQ scores are compared. Since the variables of rank and education had identical impact on the analyses of psychological data, only the data from the educational analyses will be presented. The results of the rank analyses parallel those of education, and their presentation in this report would not further clarify the herbicide/dioxin issue. Figure XIII-1 COMPARISON OF EDUCATIONAL ACHIEVEMENT AND RANK FULL SCALE 10 (RANCH HAND) FULL SCALE 10 (COMPARISON) 71 7M4 I OFFICERS FULL SCALE 10 (RANCH HAND) <70 7014 IS-11S 116-140 IQ SCORE H-111 11H40 »140 IOSCME FULL SCALE IQ (COMPARISON) »140 CD <70 BleH SCHO l ° XIII-2 7044 (5-115 10 SCORE 116-149 »14t O HKH SCHOOL �1. Analysis of Questionnaire Data a. Past History of Emotional or Psychological Illness Detailed information concerning reported emotional or psychological illnesses was sought and, wherever possible, these illnesses were coded to the ICD-9-CM, 1980 edition. The unadjusted chi-square analyses of these data are presented in Table XIII-2. It is evident from these analyses that there were no statistically significant differences in the type of reported psychological illnesses between the Ranch Hand and either the entire comparison group or the subset of original comparison individuals. Table XIII-2 DISTRIBUTION OF REPORTED PSYCHOLOGICAL ILLNESS BY TYPE OF ILLNESS Type of Illness Original Comparisons Ranch Hand Entire Group Comparison Psychoses i» 6 4 Alcohol Dependence 2 5 7 Anxiety H 9 5 Other Neuroses 6 16 9 \ / \ / \ / \ / P = 0.91 P = 0.59 b. Psychological Indices A further comparison of the responses to the psychological subsections of the questionnaire was performed. Responses to the questions addressing each psychological dimension were combined in an index equal to the number of positive responses for each dimension. Group differences in the distribution of questionnaire responses were tested by the Kolmogorov-Smirnov two-sample test, and the results are tabulated in Table XIII-3 and XHI-il. The isolation Index was analyzed in .a discrete fashion, adjusted for educational level. The data for this index are presented in Table XIII-5. When the responses to the isolation scale are dichotomized as equal or greater than 14 or less than 1*1, a relative risk of 1.97 is seen, with a 95% confidence interval of 1.11 to 3.58. The number of individuals analyzed in the depression index is reduced, since this is primarily an index of severity, and those individuals not reporting depression were excluded from the analysis. XIII-3 �Table XIII-3 QUESTIONNAIRE PSYCHOLOGICAL INDICES (HIGH SCHOOL EDUCATION) Standard Deviation Kolmogorov Smirnov P Value Group N Mean Score Fatigue Ranch Hand Comparison 573 430 15.33 13.64 6.24 5.52 < 0.001 Anger Ranch Hand Comparison 573 430 11.27 9.99 4.74 3.64 0.002 Erosion Ranch Hand Comparison 572 429 22.34 20.00 7.90 6.70 < 0.001 Anxiety Ranch Hand Comparison 555 419 24.62 21.91 8.67 7.73 < 0.001 Depression (Severity) Ranch Hand Comparison 141 60 5.79 5.30 3.15 2.85 0.89 Index Table XIII-4 QUESTIONNAIRE PSYCHOLOGICAL (COLLEGE EDUCATION) INDICES Group N Fatigue Ranch Hand Comparison 447 335 12.79 12.83 4.55 4.45 0.88 Anger Ranch Hand Comparison 447 335 9.55 9.46 3.09 3.08 0.71 Erosion Ranch Hand Comparison 448 336 20.12 19.90 5.80 5.54 0.94 Anxiety Ranch Hand Comparison 437 328 21.23 20.51 6.74 5.96 0.63 Depression (Severity) Ranch Hand Comparison 60 39 5.22 4.46 2.80 2.11 Index *Data too sparse for valid analysis XIII-4 Standard Deviation Kolmogorov Smirnov P Value Mean Score * �When an unadjusted analysis of reported depression (yes, no) was performed, there was a statistically significant group difference (P=0.002) with the Ranch Handers reporting more depression then the comparisons. This is not necessarily inconsistent with the analysis of severity (P=0.89). Table XIII-5 ISOLATION INDEX, ADJUSTED FOR EDUCATION Index Score Group S5 6-7 8-9 10-11 12-13 £14 Total Ranch Hand 16 81 535 269 91 48 1040 Comparison 3 75 425 200 49 18 770 0.002 The questionnaire responses to the questions concerning fatigue, anger, erosion, anxiety, and depression were analyzed with the exposure index, using a general linear model. When Blacks and non-Blacks were combined, the anger index was observed to be suggestively associated with exposure (P = 0.13) in officers but not in either of the enlisted occupational strata. All other exposure analyses had P values in excess of 0.40. Educational level is a major influence on responses to the psychological assessment portion of the questionnaire. The responses to these questions did not differ between college educated Ranch Handers and comparisons, but all indices except depression did differ significantly in the high school educated participants. These variables were all subjectively measured, and the specific subsets of questions were not validated. It is unclear from these data whether these differences reflect a herbicide effect unique to the largely high school educated enlisted group or an educationally related response to a highly emotional public issue. This difference may also be a reflection of postVietnam stress in the frontline Ranch Hand personnel in contrast to the reduced stress in the comparison group stationed in support areas of SEA. 2- Physical Examination Parameters During the physical examination, the Cornell Index, the Minnesota Multiphasic Personality Inventory (MMPI), the Halstead^Reitan Battery and the Wechsler Adult Intelligence Scales were used to assess psychologic functioning. Again, results were comparable whether using rank or educational attainment as stratification variables, and only the educational analyses are presented. XIII-5 �a. Cornell Index The Cornell Index is a subjective 10 to 15 minute self-administered inventory of neuropsychiatric symptoms and complaints. It has been standardized and is a widely used testing instrument. Grading of the responses to the Cornell results in an overall index and separate indices for each of the ten subelements of the instrument. A total index score of 8 or less is considered to be normal. The overall index scores for the Ranch Hand and comparison groups were contrasted using the Kolmogorov-Smirnov technique after stratification for educational level (Table XIII-6). High school educated participants demonstrated a highly significant group differential (P <0.001) but the index scores in the college, groups were not different. Table XIII-6 ANALYSIS OF CORNELL INDEX BY GROUP (KOLMOGOROV-SMIRNOV TWO-SAMPLE TEST) Educational Level Mean Score Group Standard Deviation P Value High School Ranch Hand Comparison 9.21 6.44 10.35 7.79 < 0.001 College Ranch Hand Comparison 3.66 3.44 5.13 4.58 0.59 The subelement scores were analyzed by log-linear techniques using 6 categories of response. These results are displayed in Table XIII-7( and the results of a similar analysis, using data from all available comparisons, are included as well. These results were all adjusted for educational level, since education was found to affect test scores in a highly significant manner (P <0.0001). Categorical analysis of the subelements revealed significant group differences between the Ranch Handers and the original comparisons in all areas except depression and the neurocirculatory system (NCS). This finding in depression on the Cornell Index is inconsistent with the significant observation noted in the responses to the in-home questionnaire, and may reflect the presence of differential reporting. The NCS scores were suggestive of group differences with a P value of 0.12. Analysis of the entire comparison group revealed similar findings. XIII-6 �Table XIII-7 CATEGORICAL ANALYSIS OF GROUP DIFFERENCES IN THE CORNELL INDEX (ADJUSTED FOR EDUCATION)* P Value: Ranch Hand Versus Original Comparisons All Comparisons Parameter Fear and Inadequacy Depression Nervousness and Anxiety Neurocirculatory System Startle Psychosomatic Hypochondria Gastrointestinal System Sensitivity Troublesomeness 0.02 0.39 0.002 0.12 0.004 0.002 0.05 0.01 0.08 0.06 0.06 0.16 0.009 0.14 0.04 0.002 0.12 0.01 0.29 0.06 * All of these parameters were significantly affected by education level (P <0.0001) Analysis of the Ranch Hand group's overall Cornell Index by degree of exposure was performed, using log-linear techniques. The Cornell Index was compared with exposure level (low, medium, and high) and education (high school and college) after stratification for occupation. In each occupational category, the index was clearly influenced by educational level but not by degree of herbicide exposure. Table XIII-8 contains the results of these analyses. Table XIII-8 EXPOSURE ANALYSIS OF THE CORNELL INDEX (ADJUSTED FOR EDUCATIONAL LEVEL) Occupational Category P Value Cornell Versus Exposure Cornell Versus Education Officer 0.91 0.09 Enlisted, flying 0.53 0.05 Enlisted, ground 0.26 0.04 Analysis of the overall Cornell Index identified significant group differences among high school-educated individuals (P <0.001), with the Ranch Handers having a significantly higher mean (abnormal) score. However, this XIII-7 �finding was not observed among the college educated individuals. Log-linear analyses of the Ranch Handers and original comparisons, adjusted for education, revealed significant differences in 6 of the 10 subscales of the index (P £ 0.05) and borderline or suggestive findings in three others (P S 0.12). Despite these group differences, education adjusted exposure analysis of the overall Cornell Index did not identify any association between level of exposure and Cornell Index. b. Minnesota Multiphasic Personality Inventory (MMPI) The MMPI, a standardized set of 566 subjective self-administered questions concerning various aspects of behavior and personality, was completed by 1023 Ranch Handers, 76? original comparisons, and 1194 total comparisons. Scoring was performed by machine, using the standard criteria for normality of 30-70. The comparison of the distributional characteristics of the responses to each of the subelements of the MMPI are shown in Tables XIII-9 and XI11-10. The effect of educational level on psychological scores is again seen, with more suggestive and/or significant differences between groups appearing in the high school stratum. The validity scale was not different between Ranch Handers and comparisons in either educational stratum; however, the high school comparisons exhibited a greater degree of denial (K scale) than the high school Ranch Handers. Depression (P = 0.16), paranoia (P = 0.19) and hysteria scales (P = 0.12) were suggestive of group differences in the high school stratum and significant differences were noted in the masculinity/femininity, hypochondria, mania/hypomania, and social introversion scales, with comparisons faring better than the Ranch Handers. The college stratum demonstrated borderline significance in the masculinity/femininity scale (P = 0.09) and a significant difference (P = 0.04) in social introversion. The masculinity/femininity scale is heavily influenced by the range of interests held by the participants. As individuals increase their education and broaden their interests beyond traditional "male" activities, the score tends to rise (Lachar, 1974). This is demonstrated by the means of 57.87 to 59.15 in the college stratum and means of 54.85 to 55.94 in the high school group. The consistent finding of significance in social introversion, with the Ranch Handers being more inwardly directed, is striking, but its clinical relevance is unclear. The percent of the Ranch Handers and comparisons exhibiting abnormal MMPI scores (greater than 70 or less than 30) are shown in Table XIIIH1 for those scales with suggestive or significant findings. The increased score on the denial (K) scale of the MMPI for the enlisted comparison group may be an indication of a relative differential in reporting between the two groups. When considered in the light of an increased enlisted Ranch Hand hypochondria scale on both the Cornell Index and the MMPI, overreporting in the Ranch Hand group is indicated. XIII-8 �Table XIII-9 ANALYSIS OF MMPI TESTING IN HIGH SCHOOL-EDUCATED PARTICIPANTS (RANCH HAND N - 575; COMPARISON N = 430) Parameter Validity Group Ranch Hand Comparison Defensiveness (L Scale) Ranch Hand Comparison Ranch Hand Comparison Ranch Hand Comparison Ranch Hand Comparison Ranch Hand Comparison Consistency (F Scale) Denial (K Scale) Hypochondria Depression Hysteria Mean Score 1 .85 1.73 51.99 52.03 51.95 50.65 53.95 55.63 59.74 57.22 60.47 58.39 60.12 58.90 56.38 55.89 55.94 54.85 51.72 50.68 Ranch Hand Comparison Ranch Hand Comparison Ranch Hand Comparison Ranch Hand Comparison Ranch Band Comparison Ranch Hand Comparison 57.53 55.97 Ranch Hand Comparison 56.03 54.49 Ranch Hand Comparison Comparisons greater than Ranch Hand 52.31 50.80 Psychopathic/Deviate Masculini ty/Feminini ty Paranoia Psychasthenia (Anxiety) Schizophrenia Mania/Hypomania Social Introversion XIII-9 57.27 55.59 Standard Deviation 4.54 4.07 7.84 8.15 9.29 7.16 8.86 8.12 13-36 10.95 13.98 11.96 KolmogorovSmirnov P Value 0.99 0.98 0.44 0.03* 0.05 0.16 9.96 8.23 11.00 10.52 0.12 8.32 8.94 0.01 8.66 8.33 12.23 10.07 13.42 9.71 10.36 10.31 10.38 9.50 0.19 0.86 0.47 0.45 0.01 0.006 �Table XIII-10 ANALYSIS OF MMPI TESTING IN COLLEGE-EDUCATED PARTICIPANTS (RANCH HAND N - 448; COMPARISON N = 337) Group Mean Score Standard Deviation KolmogorovSmirnov P Value Validity Ranch Hand Comparison 1.48 1.95 4.14 4.49 0.47 Defensiveness (L Scale) Ranch Hand Comparison 50.26 50.33 7.68 7.29 0.99 Consistency (F Scale) Ranch Hand Comparison 18.74 48.44 5.84 5.36 0.99 Denial (K Scale) Ranch Hand Comparison 58.46 58.41 7.53 7.64 0.99 Hypochondria Ranch Hand Comparison 55.42 54.65 9.34 8.45 0.96 Depression Ranch Hand Comparison 55.34 54.57 10.77 9.98 0.99 Hysteria Ranch Hand Comparison 59.75 59.32 7.38 7.01 0.98 Psychopathic/Deviate Ranch Hand Comparison 55.21 55.66 9.33 8.90 0.68 Masculinity/Femininity Ranch Hand Comparison 59.15 57.87 8.72 8.98 0.09 Paranoia Ranch Hand Comparison 53.62 53-26 6.96 6.64 0.63 Psychasthenia (Anxiety) Ranch Hand Comparison 53.62 54.18 8.04 8.36 0.84 Schizophrenia Ranch Hand Comparison 54.70 54.89 7.94 7.88 0.79 Mania/Hypomania Ranch Hand Comparison 55.22 54.05 9.55 10.03 0.51 Social Introversion Ranch Hand Comparison 46.83 47.50 8.67 7.98 0.04 Parameter XIII-10 �Table XIII-11 MMPI ABNORMALITY BY GROUP Level MMPI Scale Denial Ranch Hand Comparison 0.0 0.0 1.7 3.7 Ranch Hand Comparison 0.0 0.0 18.1 10.9 Depression Ranch Hand Comparison 0.2 0.0 18.1 12.2 Hysteria Ranch Hand Comparison 0.0 0.0 14.1 7.9 Masculinity/ Femininity Ranch Hand Comparison 0.0 0.0 4.5 5.6 Paranoia Ranch Hand Comparison 0.0 0.0 2.4 Mania/Hypomania Ranch Hand Comparison 0.3 0.2 8.5 8.6 Social Introversion Ranch Hand Comparison 0.0 0.0 6.8' 4.9 Masculinity/ Femininity Ranch Hand Comparison 0.0 0.0 11 .6 11.0 Social Introversion College % Below 30 Hypochondria High School Grout».. • . . Ranch Hand Comparison 0.0 0.3 1.6 1.8 . —, % Above 70 1.9 Log-linear analysis of the MMPI data, using dichotomous (normal/abnormal) responses was also conducted (Table XIII-12). Educational level was again found to exert a highly significant influence in all scales, with P values all less than 0.01. XIII-11 �Table XIII-12 LOG-LINEAR ANALYSIS OF THE MMPI SCALES BY GROUP (ADJUSTED FOR EDUCATION) P Value of Group Difference Scale Hypochondria Depression Hysteria Psychopathic/Deviate Masculinity/Femininity Paranoia < 0.001 0.02 0.002 0.39 0.84 0.26 Psychasthenia Schizophrenia Mania/Hypomania Social Introversion 0.21 0.007 0.52 0.32 Several of these analyses appear to be inconsistent with the results of the Kolmogorov-Stnirnov testing, making inference more difficult. Most of the statistically significant group differences found in the distributional analyses were in the high school group, but the log-linear analysis revealed highly significant group differences (P = 0.02) between the Ranch Hand and comparison groups after adjustment for education. Matched pair analyses, using the original comparison subset, were conducted on the hysteria, hypochondria, and masculinity/femininity scales, with respective P values of 0.02, 0.02, and 0.66. These results mirror those of the log-linear analysis in Table XIIIH2. The initial group analyses of the MMPI were performed without consideration for the variable of race. A repeat analysis of MMPI scores was also conducted for the 63 Ranch Handers and 45 originally selected comparisons, who were Black. The results of this analysis are presented in Table XIII-13. Wherever the sample size permitted, the analyses were adjusted for education; however, sparseness of data prevented adjustment in the analysis of the psychasthenia, schizophrenia, and masculinity/femininity scales and prevented any analysis for the paranoia and social introversion scales. The borderline significant finding in the schizophrenia scale (P = 0.07) is somewhat parallel to the significant P value for schizophrenia (P = 0.007) in Table XIII-12. These findings do not suggest that the factor of race is at all responsible for the overall differences in MMPI scores between the Ranch Hand and comparison groups. XIII-12 �Table XIII-13 MMPI ANALYSIS AMONG BLACK PARTICIPANTS Adjusted for Education, Scale P Value of Group Difference Hypochondria Depression Hysteria Psychopathic/Deviate Mania/Hypomania Yes Yes Yes Yes Yes 0.15 0.91 0.31 0.73 0.70 Psychasthenia Schizophrenia Masculinity/Femininity No No No 0.20 0.07 0.31 Paranoia Social Introversion N/A N/A Exposure analysis of the Ranch Hand group, using log"-linear techniques revealed a mixed pattern of significant, borderline and suggestive findings. These results are summarized in Table XIII-14. Education remains a significant factor, but consistency across occupational groups is not evident, since stratification by occupational group mirrored stratification by education. Table XIII-15 displays the exposure index data, and the percentage of abnormal MMPI scale results, for the exposure analyses with P values of concern. Only the hysteria scale in the officers attending college and the psychopathic deviate scale in both high school and college officers showed consistent increases in abnormality with increasing exposure. However, the number of abnormal scores in all of these scales was quite low and inferential accuracy is compromised. Table XIII-14 P VALUES OF THE MMPI/EXPOSURE ANALYSES (ADJUSTED FOR EDUCATION) P Value Parameter Hypochondria Depression Hysteria Psychopathic Deviate Masculini ty/Feminini ty Paranoia Psychasthenia Schizophrenia Mania/Hypomania Social Introversion Parameter Versus Exposure Enlisted Officer FlyingGround 0.21 0.70 0.21** 0.001* 0.09 1.00 0.89 0.09 0.32 0.39 : 0.97 0.11 0.76 1.00 0.81 0.64 0.05 0.12 0.13 0.33 0.02 0.16 0.0005 0.15 0.09 0.53 0.48 0.73 0.29 0.78 *Significant confounding by education present **Signifleant three-way interaction present XIII-13 P Value Parameter Versus Education Enlisted Officer Flying Ground 0.18 0.46 0.34 0.17 0.28 0.72 0.29 0.43 0.86 0.77 0.10 0.12 0.62 0.20 0.04 0.83 0.56 0.50 0.81 0.93 0.03 0.27 0.04 0.16 0.005 0.20 0.07 0,03 0.41 0.02 �TabU) XI 1 1 - 1 5 DOSE RESPONSE PATTERNS Exposure Leve I Number Normal Number AbnormaI (*) En I Isted Ground Low Med i um High 110 153 119 38 25 29 (57) 2.* (14.0*) (19.6*) En I isted Flying Low Med i um High 48 41 55 10 18 11 (17.2*) (30.5*) (16.7) En Iisted Ground Low Medium High 111 148 119 37 30 29 (25.0*) (16.9*) (19.6*) Officers (High School) Low MedIum High 10 14 24 0 5 0 (0*) (26.3*) (0*) Officers (College) Low Medium High 97 104 91 3 5 9 (.* 30) (.* 46) (9.1*) En 11sted Ground Low Med i um High 115 163 132 33 15 16 (22.3*) (.* 84) (10.8*) Officers (High School) Low Med i um High 10 19 23 0 0 1 (0*) (0*) (.* 42) Officers (College) Low Medium High 100 102 90 0 7 10 (0*) (.* 64) (10*) En Itsted Ground Low Medium High 127 164 131 21 14 17 (14.2*) (.* 79) (11.5*) Low Med i um High 105 113 111 5 15 13 (4.5*) (11.7*) (10.5*) En I Isted Ground Low Med i um High 135 172 136 13 6 12 (.* 88) (.* 34) (8.1*) Psychasthenia Enlisted Flying Low Med t um High 54 48 62 4 11 4 (.* 69) (1.9*) (6.1*) Schizophrenia Officers Low Med i um High 108 119 121 2 9 3 (1.8*) (.* 70) (2.4*) Enlisted Flying Low Med i um High 55 49 59 3 10 7 (.* 52) (16.9*) (10.6*) En IIsted Flying Low Medium High 53 50 63 6 9 3 (10.2*) (15.3*) (.* 48) Parameter Group Hypochondria Depression Hysteria* Psychopath Ic/DevI ate* Masculinity/Femininity Officers Mania/Hypomanla *Data are presented by educational level when the education/exposure Interactions are statistically significant. XIII-14 �Analysis of the MMPI data from the Ranch Hand and original comparison groups revealed significant group differences in the hypochondria, depression and hysteria scales (P S 0.02), after adjustment for education. Stratified analysis based on level of education revealed statistically significant group differences for the hypochondria and masculinity/femininity scales (P £ 0.05). However, there were no statistically significant group differences among college-educated individuals, and only in the masculinity/femininity scale was borderline significance reached (P - 0.09). Exposure analyses did not reveal any consistent patterns of statistical significance between occupational categories, level of exposure and MMPI scores. c - Haistead-Reitan The Haistead-Reitan Neuropsychological Test Battery was administered to each participant to assess the functional integrity of the central nervous system. An impairment index for each participant was calculated based upon the scores of the category, tactual performance, speech^sounds, Seashore rhythm, and finger-tapping portions of the battery. The impairment index ranged from zero to seven, based on the number of sub tests in which the participant scored abnormally. Impairment was declared if the index equalled or exceeded three. Larger numbers of participants were deleted from these analyses; since seven distinct tests contributed to the impairment index. The absence of any one made calculation of the index impossible. Analysis of dicotomous variables (normal/abnormal), adjusted for education, revealed no overall group differences (P = 0.74). A categorical analysis, unadjusted for educational level, was performed. The data and the results of the unadjusted analyses of the Ranch Hand group, the entire comparison group and the subset of original participants are presented in Table XIII-16. Table XIII-16 UNADJUSTED HALSTEAD-REITAN SCORES BY GROUP Impairment Index Original Comparisons N - 559 0 1 2 3 4 5 or more Ranch Hand N - 771 85 162 66.5%* 125 77 60 50 \ / x2 - 3.18 P = 0.67 * Cumulative % for Impairment Index 0,1,2 XIIIr-15 All Comparisons N - 883 124 226 141 66.5%* 248 194 134 85 81 163 126 68 64 \ / 2 » 1.35 X P = 0.93 66.0%* �Analyses adjusted for education were carried out on the Ranch Handers and the original subset of comparisons (Table XIIIH?). Education was again seen to be a significant factor (P < 0.0001). Table XIII-17 HALSTEAD-REITAN ANALYSIS BY GROUP AND EDUCATION Educational Level High School College Degree of Impairment 4 2 5 or Greater 1 3 Ranch Hand 54 56 45 108 88 80 Comparison 29 69 69 49 38 37 Ranch Hand 8 79 118 75 46 14 Comparison 56 28 22 13 93 56 P Value, adjusted for education - 0.57 0 An exposure index analysis was also accomplished on the data from the Ranch Hand group. As shown in Table XIII<~18, educational level was a significant covariable in the officer and enlisted flying groups, but there were no significant relationships between herbicide exposure and Halstead-Reitan performance. Table XIII-18 HALSTEAD-REITAN IMPAIRMENT AND EXPOSURE Occupational Group Adjusted P Values Halstead-Reitan Halstead-Reitan Versus Education Versus Exposure Officers 0.88 0.002 Enlisted Flying 0.44 0.05 Enlisted Ground 0.82 0.62 d. Wechsler Adult Intelligence Scale (WAIS) WAIS testing was completed on 1022 Ranch Handers and 733 original comparison individuals. The test was administered and scored in the standard manner by certified clinical psychologists and psychological technicians. As noted previously, intelligence scores (IQ) by rank were equivalent to IQ scores by education. The distributions of verbal, performance and full-scale IQ scores, by educational level and group, are shown in Figure XIII-2. XIII-16 �Figure XIII-2 FREQUENCY DISTRIBUTION IQ SCORES BY EDUCATIONAL LEVEL AND GROUP VERBAL 10 (COLLEGE) VERBAL M - (HMD WHCOL) PERFORMANCE ID (COLLEGE) PERFORMANCE 10 (HtflH SCHOOL) flU. SCALE Nt(COLLEGE) RU KALE M (MM SCHOOL) The IQ scores demonstrated consistent patterns within each educational stratum. A slight increase in the proportion of both Ranch Hand and comparison college graduates, with performance IQ's between 85 and 115, was noted. These distributions were tested for group differences by the Kolmogorov-Smirnov procedure. Suggestive but nonsignificant differences were noted for performance and full-scale IQ's in the high school stratum, but no differences were found among the college^educated group. These data are shown in Table XIII-19. XIII-17 �Table XIII-19 DISTRIBUTIONAL ANALYSIS OF IQ SCORE:, Group Mean Score Standard Deviation High School Ranch Hand Comparison 110.61 101.73 10.65 11.34 0.39 College Ranch Hand Comparison 117.00 116.84 12.97 13.73 0.73 High School Ranch Hand Comparison 102.40 104.14 11.38 11 .86 0.14 College Ranch Hand Comparison 113.70 112.37 12.62 13.33 0.50 High School Ranch Hand Comparison 101.18 102.74 10.71 11.32 0.15 College Ranch Hand Comparison 117.30 116.59 12.96 13.82 0.37 Scale Education Verbal Performance Full Scale P Value The distributions were observed to identify outliers, and the percentage of participants with scores in the abnormal range (below 85) was determined. These results are shown in Table XIII-20. Table XIII-20 ABNORMAL IQ SCORE BY GROUP AND EDUCATIONAL LEVEL College Grout Verbal Ranch Hand Comparison 3.7 3.3 9.8 13-7 Ranch Hand Comparison ,5.4 3.7 14,3 18.8 Full High School Scale Performance Educational Level Ranch Hand Comparison 4.0 3.5 10.6 15.1 Verbal Ranch Hand Comparison 0.9 0.3 58.8 54.1 Performance Ranch Hand Comparison 1.1 1.8 43.9 41.1 Full Ranch Hand Comparison 0.7 0.3 61.1 56.2 XIII-18 % Below 85 % Above 115 �Analysis of the WAIS testing scores of the Ranch Hand group, by level of herbicide exposure, revealed no consistent differences in IQ scores. The P values derived from these analyses are presented in Table XIII-21 and show only one statistically significant association (P = 0.0^). Table XIII-21 RESULTS OF IQ SCORES BY EXPOSURE ANALYSIS Occupational Group P Value Officers Enlisted Flying Enlisted Ground 0.99 0.31* 0.82 Performance Officers Enlisted Flying Enlisted Ground 0.99 O.OH 0.18 Full Scale Officers Enlisted Flying Enlisted Ground 0.99 0.23 0.25 2. Summary In this chapter, a large number of variables were analyzed using several techniques and multiple assessments. Consistent differences between high school-educated Ranch Handers and high school-educated original comparisons are seen throughout these analyses. With the exception of a single statistically significant result for social introversion (P = 0.04), these group differences are not apparent in the college educated stratum. Unstratified but educationally adjusted analyses of the MMPI scores did, however, reveal group differences which were more like those of the high school stratum. Exposure analyses did not reveal any patterns suggesting any association between psychological testing results and level of herbicide exposure. The relative risks, confidence intervals, and shifts in means for the dependent variables analyzed in this chapter are included in Appendix XVIII. Xlll-19 �Table XIII-22 PSYCHOLOGICAL ANALYSIS SUMMARY (RANCH HAND VERSUS ORIGINAL COMPARISON GROUP) Analytic Strategy (P Values) Adjusted for Education Parameter Questionnaire Indices Fatigue Anger Erosion Anxiety Isolation Depression (Severity) Stratified Analysis High School College <0.001 0.002 <0.001 <0.001 Exposure Analysis Off Enl Fly Enl Gnd NS* NS NS NS 0.002 0.89 Cornell Index NS NS NS NS NS 0 .11 NS NS NS NS NS NS 0 .05 0 .12 0 .13 NS 0.02 0.16 0.001 0.15 0.09 NS NS NS NS NS NS NS NS NS NS NS 0 .04 NS NS NS 0.18 NS <0.001 Fear and Inadequacy Depression Nervousness and Anxiety Neurocirculatory Startle Psychosomatic Hypochondria Gastrointestinal Sensitivity Troublesomeness NS NS 0.05 0.16 0.12 NS 0.01 NS NS NS NS 0.09 NS NS NS NS 0.006 0.04 0.02 NS 0.002 0.12 0.004 0.002 0.05 0.01 0.08 0.06 MMPI Hypochondria Depression Hysteria Psychopathic Deviate Masculinity/Femininity Paranoia Psychasthenia Schizophrenia Mania/Hypomania Social Introversion Halstead-Reitan <0.001 0.02 0.002 NS NS NS NS 0.007 NS NS 0.01 0.19 NS NS NS 0.001 0.09 NS NS 0.09 NS • IQ Scores Verbal Performance Full Scale NS 0.14 0.15 *Nonsignificant; P > 0.20 XI11-20 NS NS NS �The results of the analyses of the psychological data are summarized in Table XIII-22, and demonstrate a greater degree of statistically significant group differences in the more subjective measurements (questionnaire and Cornell Index) than are observed in the more objective assessments (HalsteadReitan and WAIS). The effect of differential reporting in this evaluation is as yet difficult to assess. However, the high school-educated Ranch Handers did have higher scores on the hypochondria scale of the MMPI and the psychosomatic portion of the Cornell Index than did the appropriate comparisons. Addi"tionally, the high school-educated comparisons scored higher on the MMPI K Scale (denial). These findings suggest that differential reporting may be influencing the analytic results of the in-home questionnaire and the Cornell Index. There may also be a differential response to the intense media interest in the herbicide/dioxin issue between the high school and college strata in this study. The role of "Post Vietnam Stress" in these findings is also unclear at this time. Further clarification of these factors and their impact must await analysis of the data from the follow-up phase of the study. Based on the psychological data collected during the initial in-home questionnaire and physical examination, there is no convincing evidence suggesting the presence of an adverse effect on emotional health caused by herbicide exposure. XIII-21 �Chapter XIV. EVALUATION OF HEPATIC STATUS 1. Introduction A very broad spectrum of hepatic phenomena has been reported in association with acute, subacute and chronic administration of TCDD to animals. Significant response differences between species occur, however. Serum enzyme changes (SCOT, SGPT, GGPT, LDH) have not been prominent, although SGPT levels were elevated in at least 1 study (Schantz et al, 1979). Elevated alkaline phosphatase levels have been observed with increased direct bilirubin levels (Kociba et al, 1976). Decreased serum cholesterol levels have also been noted after sublethal exposures (Schantz et al, 1979). TCDD interferes with hemoglobin metabolism affecting delta-aminolevulinic acid synthetase activity (Goldstein et al, 1973) and possibly other enzyme activities, providing, at sufficient doses, signs and symptoms of porphyria. Motivated by the literature reports of hepatotoxicity, signs and symptoms of hepatic dysfunction were sought in the participants in this study. In this qhapter, enzyme levels, bilirubin levels and lipid values are presented, along with determinations reflecting porphyrin metabolism. Clinical history data are also analyzed, along with hepatomegaly determined at physical examination. 2. Bipphemical DeterminatIons a. Analyses Overview In this section 9 biochemical determinations are studied: SCOT, SGPT, GGPT, alkaline phosphatase (Alk. Phos.), total bilirubin (T. Bill), direct bilirubin (D. Bili), lactic dehydrogenase (LDH), cholesterol (Choi) and triglycerides (Trig). These 9 variables are listed in Table XIV~1, along with the normal-abnormal ranges used in the reported statistical analyses. These ranges were adapted from Kelsey-Seybold laboratory normal ranges. In the analyses of these 9 variables, adjustments were made for 4 covariates: current alcohol ingestion (ALC), days of exposure to industrial chemicals (1C), days of exposure to degreasing chemicals (DC), and presence or absence of antibody to hepatitis B surface antigen (anti-HBsAg). The current alcohol use covariate was taken from the personal medical history administered at the time of the physical examination and is in units of average drinks per day (see Appendix VI, page 2). Current alcohol ingestion was selected as an adjusting variable over the drink years measure developed from the questionnaire, since preliminary testing indicated it correlated better with hepatic endpoints. The industrial chemical and degreasing chemical exposures were derived from the in^home questionnaire (total unprotected exposure). XIV-1 �The data analyzed were from the entire Ranch Hand cohort compliant to the physical examination (N = 1045) and the original comparisons compliant to the physical examination (N = 773). Ten Ranch Handers and 2 comparisons were removed from the analysis because of body temperature of 100°F or more, and the effect of fever on hepatic variables. Individuals whose blood contained hepatitis B surface antigen (HBgAg) were also removed from the analysis (8 Ranch Handers and 7 comparisons). b. Group Analyses Three sets of analyses were run: (1) Continuous-continuous analyses (CC): In these evaluations both the dependent variables and adjusting covariates, except anti-HBsAg which is dichotomous, were used as continuous variables in an analysis of covariance. (2) Continuous-discrete analyses (CD): In these analyses all 4 covariates were used as dichotomous variables while the dependent variables were maintained as continuous. (3) Discrete-discrete analyses (DD): All variables were analyzed in dichotomous form using the log-linear model for discrete data. In all 3 analysis settings, group-by~covariate interactions were examined. In addition, the continuous-continuous and continuous-discrete analyses models were fit without interaction terms to provide discussion of appropriate tests when dependent variable relationships with the covariates are the same in both groups. In the continuous-continuous and continuous-discrete analyses the dependent variable was normalized by using a logarithmic (base 10) transformation. Table XIV-1 NORMAL - ABNORMAL LEVELS OF NINE BIOCHEMICAL DETERMINATIONS REFLECTING HEPATIC FUNCTION Determination 1. 2. 3. 4. 5. 6. 7. 8. 9. SGOT SGPT GGPT Alkaline Phosphatase Total Bilirubin Direct Bilirubin Lactic Dehydrogenase Cholesterol Triglycerides Normal < 41 :£ 45 £ 85 < 9.7 S 1.2 SS 0.36 S200 <240' 5150 XIV-2 Abnormal > 41 > 45 > 85 > 9.7 > 1 .2 > 0.36 >200 >240 >150 �Table XIV-2 provides unadjusted means, adjusted means, and percent abnormality by groups for the 9 hepatic-related variables. A summary of the 3 classes of analyses is provided in Table XIV-3. The results in this table provide P values for Ranch Hand-comparison group differences. Table XIV-2 UNADJUSTED MEANS, ADJUSTED MEANS AND PERCENT ABNORMALITY FOR NINE LIVER-RELATED VARIABLES Variable SCOT Group RH Unadjusted Means 33-0 33.1 Adjusted Means 33-0 33.1 Percent Abnormality 13-9 1*1.8 SGPT RH COM 20.3 20.5 20.3 20.5 7.8 8.6 GGPT RH 40.2 40.1 10.8 COM 39.3 39.3 10.3 Alk. Phos. RH COM 7.68 7.53 7.69 7.52 T. Bill RH COM 0.57 0.58 0.57 0.58 D. Bili RH COM 0.23 0.24 0.23 0.24 29.0 29.7 LDH RH COM 142.1 141.7 142.1 141.7 1.7 2.1 RH 212.2 212.2 26.0 COM 216.6 216.6 27.7 RH 121.8 121.9 34.7 COM 124.3 124.1 36.1 CHOL TRIG *COM denotes original fully compliant comparisons. XIV-3 17.3 16.9 1 .8 2.0 �Table XIV-3 SUMMARY OF RESULTS UNMATCHED ANALYSES OF N I N E BIOCHEMICAL V A R I A B L E S REFLECTING P V a l u e s for Models w i t h Interaction P V a l u e s for models without Interaction Gp X ALC Gp X anti HB^Ag Gp X 1C Gp X DC .032 - - - .805 <.001 .867 <.001 - anti HBsAg .052 - .663 <.001 .662 .003 - VAR ANAL SCOT CC CD DD .127 <.001 .278 <.001 .578 <.001 -* - - - CC CD DD .736 <.001 .309 .005 .592 - - CC CD DD .731 <.001 .050 <.001 .782 <.001 - - .6 06 - ALK PHOS CC CD DD .405 .142 .734 - _- _- .0 09 _ .010 _ - _ _ TOT BILI CC CD DD .113 .0 66 .800 _ 0 .014 _. _6 3 _ .001 .100 _ - _ - _ - - CC CD DD .494 .371 .869 .0 04 .091 - - - CC CD DD .063 .024 .526 .090 SGPT GGPT DIR BILI LDH CHOL TRIG J3p_ ALC 1C _ .001 - DC .032 - .027 .6 09 - - - - Gp .483 <.001 .421 <.001 _ DC anti HB«;Ag — — « _ _ _ - _ .078 - .0 09 .011 .001 .071 .066 .423 .400 .0 09 - .011 <. 001. 095 .9 09 .770 .755 .0 03 .025 - - _ .037 - .836 .711 - - .022 <.001 .031 .020 - .601 .616 _ .016 - .023 - - _ - _ .011 _ - - - - .8 06 - CC CD DD .062 <.001 .079 .216 .014 .466 .053 - - - - CC CD D D .911 .284 .589 - - -- - - 1C .140 .115 - _ - - ALC _ _ - - L I V E R FUNCTION - _ - .061 - - - - * - denotes P > 0.050 for main effects, P > 0.100 for Interation effects In Tables XIV-2 and XIV-3, there is a very slight indication of overall group differences in the GGPT with the Ranch Hand mean greater than the comparison mean and a P value of 0.050 in the CD analysis with interaction terms. However, when interaction terms are not considered, P = 0.^21. This may indicate some interaction effects even though they were not detected as statistically significant. Additionally, no difference is detected in the CC or DD XIV-4 �analyses. A stronger indication of overall group difference is seen with LDH; however, it is interesting to note that while the Ranch Hand mean LDH is greater than the comparison mean, the Ranch Hand percent abnormal LDH is less than that of the comparison group. The Ranch Hand cholesterol mean is lower than that of ,the comparison group and the result appears unlikely to have occurred by chance (P value of 0.062 in the full model CC analysis; P values of 0.022 and 0.031 in the CC and CD analyses respectively not using interaction terms). These group differences in GGPT, LDH and CHOL are all small. Further group specific differences are noted in interaction effects with covariables. Ranch Hand SCOT levels are correlated more highly with alcohol ingestion than are comparison SCOT levels. The Ranch Hand SCOT - alcohol regression slope is 0.0178 logarithmic units per drink per day, while the comparison SCOT - alcohol slope is 0.0113 logarithmic units per drink per day. This difference in slopes is statistically significant with P = 0.032, and could represent differing hepatic sensitivities to alcohol. A borderline group by industrial chemical exposure is noted in the DD analysis of SGPT levels. This interaction is shown in Table XIV-4. Table XIV- 4 INDUSTRIAL CHEMICAL EXPOSURE AND % ABNORMAL SGPT IN RANCH HAND AND COMPARISON GROUPS Ranch Hand Exposure 8.84$ (38 of 430) No Exposure 7.19$ (42 of 584) Comparison 6.71$ (23 of 3^3) 10.1$ (42 of 416) Ranch Hand personnel exposed to industrial chemicals have a higher proportion of abnormal SGPT values than do Ranch Hand personnel who are not exposed to industrial chemicals. The situation is reversed in the comparison group. The relative risk for abnormal SCOT in the Ranch Hand group associated with industrial chemical exposure is 1.23, while the comparison relative risk is 0.66, and this difference carries a P value of 0.052. Two group-by-covariate interactions are noted in the LDH data. In the comparison group neither alcohol ingestion nor exposure to degreasing chemicals was associated with change in LDH levels, while in the Ranch Hand group, increased levels were noted to occur in association with both exposures. Specifically, in the comparison group the LDH--alcohol slope is -0.0008 logarithmic units per drink per day which is not statistically significantly different from XIV-5 �zero (P = 0.577). Also, the comparison LDH-degreasing chemical slope is -0.08 x 10"5 units per exposure day (P = 0.735 against the null hypothesis of zero slope). On the other .hand, the Ranch Hand LDH-alcohol slope is 0.0041 units per drink per day (P < 0.001 against hypothesis of zero slope) and the LDH-degreasing slope is 0.51 X 10~5 units per exposure day (P = 0.003 against zero slope hypothesis). c. Exposure Analyses Analyses within the Ranch Hand cohort are presented contrasting the hepatic clinical variables against the herbicide exposure index. For this exposure index work, separate analyses were run for each of 3 occupational groups: officers, enlisted flying and enlisted ground. The 9 hepatic variables were analyzed as continuous dependent variables after logarithmic transformation. As with the Ranch Hand^-comparison group analyses, alcohol use, industrial chemical exposure, degreasing chemical exposure and antibody to Hepatitis B surface antigen were used as adjusting covariates, and individuals with body temperature greater than or equal to 100°F were omitted from the analysis as were individuals with hepatitis B surface antigen. For this exposure index effort, alcohol use, industrial chemical exposure and degreasing chemical exposure were used as continuous variables. Table XIV-5 is a display of exposure means adjusting for covariates without invoking interaction. Table XIV-6 provides a summary of P values for the testing. Analyses of covariance or generalized linear models with and without interaction were employed. An overall or main exposure effect on GGPT levels is indicated among officers and enlisted ground personnel. However, clear-cut dose-response patterns are not-noted, rather, in the officer cohort the medium exposure subgroup has the highest mean GGPT while in the enlisted ground cohort the subgroup with low exposure has the highest GGPT. Six exposure group-by-covariate interactions were found at P _<_ 0.050. These interactions are written out in Table XIV>-7. In this table, the slope of the dependent variable with respect to the covariate of interest is provided for each of the 3 exposure levels. XIV-6 �An exposure-by-degreasing chemical interaction was noted in SCOT in officers. Low herbicide exposure is associated with a possible depression of SCOT levels with increasing degreasirig chemical exposure, while individuals in the high herbicide exposure group show increasing SCOT levels with increasing degreasing chemical exposure. Table XIV-5 ADJUSTED BIOCHEMICAL MEANS BY EXPOSURE AND OCCUPATIONAL CATEGORY, WITH TYPICAL SAMPLE SIZES Variable Occupational Category Low Exposure Medium Exposure High Exposure SCOT Officer Enl. F. Enl. G. 33.3 31 .8 33.6 32.2 33-5 32.7 33.0 31.7 34.1 SGPT Officer Enl. F. Enl. G. 20.2 18.5 21.3 19.9 20.8 21.1 19.4 18.4 20.6 GGPT Officer Enl. F. Enl. G. 37.1 39.5 45.9 40.2 37.5 37.8 40.5 41.4 43.0 Officer Enl. F. Enl. G. 6.91 8.13 7.93 7.24 7.88 7.85 7.47 7.98 8.04 T. Bili. Officer Enl. F. Enl. G. 0.56 0.53 0.58 0.55 0.56 0.58 0.57 0.54 0.60 D. Bili. Officer Enl. F. Enl. G. 0.22 0.18 0.25 0.23 0.23 0.24 0.23 0.21 0.26 LDH Officer Enl. F. Enl. G. 111.3 143.1 142.9 139.4 141 .0 140.8 139.3 149.3 144.9 Choi. Officer Enl. F. Enl. G. 214.6 214.0 208.7 213.0 212.6 210.4 209.4 222.5 211.4 Trig. Officer Enl. F. Enl. G. 111 .9 129.8 118.6 127.4 126.4 114.5 129.0 128.4 121.1 Typical Sample Sizes Officer Enl. F. Enl. G. 107 58 143 122 58 170 120 63 146 Alk. Phos. XIV-7 �Table XIV-6 SUMMARY OF P VALUES FOR EXPOSURE INDEX ANALYSIS OF NINE HEPATIC VARIABLES P Values for Models with Interaction VAR SCOT SGPT GGPT ALK PHOS TOT BID DIR BID LDH CHOL TRIG OCC CAT EXP CAT ALC 1C OFF .563 <.001 ENL.F. .885 <.001 ENL.G. .698 <.001 aHb EXP X ALC - .037 - - _ EXP X 1C -* OFF .463 •c.001 ENL.F. .909 ENL.G. .467 DC - - OFF .192 ENL.F. .685 ENL.G. .629 OFF. .643 ENL.F. .449 ENL.G. .606 OFF .992 ENL.F. .399 ENL.G. .823 OFF .516 ENL.F. .656 ENL.G. .300 OFF .290 ENL.F. .310 ENL.G. .096 .394 OFF. ENL.F. .468 ENL.G. .890 - — - - _ .029 _ .010 _ _ _ _ _ .031 .045 - _ - _ _ _ _ _ _ _ _ _ - .060 _ _ _ .050 _ _ _ _ _ _ _ _ - .086 _ — .089 .049 _ _ .018 - - - <.001 _ _ .009 EXP X antl HBsAg - _ - EXP X DC .081 .010 OFF .052 <.001 ENL.F. .427 <.001 ENL.G. .093 <.001 P Values for Models With No Interaction — — .044 _ ' _ .026 — ~ .058 - *. .006 Exp Cat .512 .538 .409 <,001 <.001 <.001 .812 .411 .862 <.001 .696 .224 .574 <.001 <.001 <.001 .280 .855 .710 .885 .560 .642 .856 .310 .697 .758 - .174 .049 .360 - .602 .343 .841 .244 .980 .768 * - Indicates P > 0.050 for main effects P > 0.100 for Interactions. XIV-8 ALC - _ .011 1C _ DC anti HBsAg .047 .035 _ _ _ _ .040 - _ .020 _ _ _ _ _ _ _ _ - .023 .008 - _ _ _ _ - - - - - - - _ .019 .034 .036 - - _ .037 _ - _ _ — — _ - - • - _ — _ _ _ _ �Table XIV-7 EXPOSURE - COVARIATE INTERACTION EFFECTS FOR NINE HEPATIC VARIABLES Exposure Level P Va 1 ue on Test of Slope Against Nul 1 Hypothesis of Zero Slope Van Occ Cat Interact Level of Interact SCOT Officers Exp x DC .009 Low Mod High -.201 x 10~4 units/day .021 x 10"4 units/day .674 x 10~4 units/day GGPT En 1 1 sted Flying Exp x ALC .4 09 Low M»d High . 8 8 unlts/drk/day 02 .056*1 units/drk/day . 2 8 unlts/drk/day 08 <.001 .002 .037 ALK PHOS Officers Exp x ALC <.001 Low Med High -.0442 unlts/drk/day .0131 unlts/drk/day -.0015 unlts/drk/day <.001 .254 .6 84 DIR Bill En 1 1 sted Antl Exp x HBsAg .0 06 Low Med High .3713 mqm/dl -.2246 mgm/dl .1752 mgm/ml .013 .071 .134 LDH Enlisted Antl Exp x HbsAg .4 09 Low Med High .0329 units -.0407 units -.0330 units .159 .085 .128 Exp x ALC .026 Low Med High . 0 9 mgm/dl/drk/day 03 -.0065 mgm/dl/drk/day .0054 mgm/dl/drk/day .284 .4 03 .147 Flying Ground CHOL En 1 1 sted Ground Slope XIV-9 .286 .924 .0 02 �Alcohol use is associated with increasing GGPT levels among enlisted flying personnel, but the increase in GGPT falls smoothly with increasing exposure levels. On the other hand, alcohol use is associated with decreasing alkaline phosphatase levels among Ranch Hand officers in the low exposure group. There are 2 interactions between exposure group and antibody to Hepatitis B antigen. Direct bilirubin levels are higher in enlisted flying personnel who' are antibody positive and are in the low or high exposure groups. Direct bilirubin levels are lower in individuals who are antibody positive but in the medium exposure group. LDH is higher among enlisted ground Ranch Handers who are antibody positive and are in the low herbicide exposure group while LDH levels are lower among antibody positive individuals in the medium and high exposure groups. An exposure-by-alcohol use interaction effect on cholesterol levels shows positive slopes in the low and high exposure categories but a negative slope in the medium exposure category. Thus, of the 6 statistically significant interactions noted in this exposure index analysis only 1, the SGOT-degreasing chemical interaction, supports an interpretation of. herbicide effect. But this interpretation is markedly weakened by the presence of the 5 uninterpretable patterns. 3. Urlnalysis Determinations Related to Porphyrin Metabolism Three components associated with porphyrin metabolism were determined and are analyzed here: uroporphyrin, coproporphyrin and d-aminolevulinic acid. Data addressing these 3 variables were analyzed looking for differences between the Ranch Hand and comparison groups and looking for associations with indexed herbicide exposure within the Ranch Hand group. In examining the uroporphyrin, coproporphyrin and d-aminolevulinic acid data for Ranch Hand - comparison group differences, adjustments were accomplished for the following 6 variables: current alcohol use in drinks per day (ALC), blood urinary nitrogen (BUN), creatinine clearance (CCL), days of exposure to industrial chemicals (1C), days of exposure to degreasing chemicals (DC) and presence or absence of antibody to hepatitis B antigen. Adjustments were accomplished treating the dependent variable and all independent variables except antibody to hepatitis B antigen as continuous variables in a generalized linear model analysis. Since the compounds uroporphyrin, coproporphyrin and d-aminolevulinic acid are all measured in 2M-hour urine collections, only data from subjects who complied with the full collection of urine are used in the analysis (620 Ranch Handers and ^39 comparisons). Also, febrile participants and individuals with HB3Ag have been removed. In the adjusted analyses the dependent variable was normalized by using a logarithmic (base 10) transformation. XIV-10 �Table XIV-8 provides uroporphyrin, coproporphyrin and d-aminolevulinic acid unadjusted means, adjusted means and percent abnormality. For uroporphyrin, values greater than 60 were considered abnormal, for coproporphyrin, values greater than 235 and for d-aminolevulinic acid, values greater than tfQOQ were counted . as abnormal. .*t \ . : ; Table XIV-8 UNADJUSTED MEANS, ADJUSTED MEANS AND PERCENT ABNORMALITY FOR THREE COMPOUNDS RELATED TO PORPHYRIN METABOLISM Unadjusted Means Adjusted Means % AbnormaI Uroporphyrin RH COM 30.5 30.8 * * 6.5? 6.8% Coproporphyrin RH COM 31.2 30.8 * * 0.2% 0.0? d - a m l n o l e v u l i n i c acid RH COM 2328.9 2383.2 2337.1 2371.4 0.0? 0.0? * adjusted means not represented due to interaction Table XIV-9 SUMMARY OF RESULTS UNMATCHED ANALYSES OF THREE COMPOUNDS RELATED TO PORPHYRIN METABOLISM P-VALUES FOR MODELS WITH INTERACTION VAR Gp ALC URO .227 COPRO .490 ALA .145 - BUN <.001 <.001 - <.001 <.001 <.001 1C DC Gp x ALC .049 CCL AntI HBsAg .014 .045 - _ — - Gp x BUN .077 .097 _ Gp x CCL _ _ Gp x Gp x Anti DC HBsAg Gp x _IC _ _ Table XIV-9 displays the detailed analyses. No overall group differences are observed. With uroporphyrin a borderline significant group-by-BUN interaction (P = 0.077) was observed. In the Ranch Hand group, the uroporphyrin-BUN slope was -0.010 uroporphyrin logarithm units per BUN unit, while the comparison slope was steeper (-0.017). A borderline group-by-BUN interaction was also noted in the coproporphyrin data. In the Ranch Hand group, the coproporphyrinBUN slope was -0.01*1 coproporphyrin logarithmic units per BUN unit, while the comparison slope was again steeper (-0.023). Lastly, a group-by-alcohol interaction was detected in the coproporphyrin data (P = 0.045). The Ranch Hand slope was positive (+0.013) while the comparison slope was negative (-0.008). XIV-11 �Table X I V - 1 0 SUMMARY OF P VALUES FOR EXPOSURE INDEX ANALYSES OF THREE COMPOUNDS RELATED TO PORPHYRIN METABOLISM EXP VAR URO COPRO ALA OCC CAT EXP CAT BUN OFF .207 ENL F. .670 ENL 6. .882 .010 OFF .630 ENL F. .498 ENL G. .699 <.001 .016 OFF .279 ENL F. .135 ENL G. .312 CCL <.001 ^C_ JDC _ a_Hb EXP x ALC EXP x x ALC BJJN CCL EXP x 1C - EXP x DC .033 Exp x Ant I HBcAg - .050 .022 .035 - .015 <.001 <.001 <.001 .016 .040 .020 - .028 - .042 - Table XIV-11 TABLE OF UNADJUSTED MEANS FOR THREE COMPOUNDS RELATED TO PORPHYRIN METABOLISM Variable Occupational Category N__ Low Exposure Medium Exposure High Exposure Uroporphyrin Officers Enlisted Fly. Enlisted Gnd. 212 106 282 28.9 38.7 31.1 26.9 27.8 32.4 31.3 31.6 29.8 Coproporphyrin Officers Enlisted Fly. Enlisted Gnd. 212 106 282 32.4 36.4 31.6 26.7 31.1 30.9 29.9 32.5 32.8 d-amino levulinic Acid Officers Enlisted Fly. Enlisted Gnd. 212 106 282 XIV-12 2221 2460 2290 2312 2510 2441 2211 2381 2271 �Table XIV-12 EXPQSURE-COVARIATE INTERACTIONS FOR THREE COMPOUNDS RELATED TO PORPHYRIN METABOLISM Variable Occupational Category P Value for Exposure Interaction Interaction Level Uropophyrin Officer Exp x DC .033 Low Med High -.000043 .000074 .000190 Coproporphyrin Enlisted Ground Exp x 1C .016 Low Med High .301 X TO'11 -.540 X 10"11 .176 X 10"11 d-amino levulinic acid Enlisted Flying Exp x ALC .028 Low Med High .00045 -.02922 .01445 d-amino levulinic acid Enlisted Ground Exp x 1C .040 Low Med High -.1450 X 10'14 -.2944 X 10"11 .0315 X 10-^ d-amino levulinic acid Enlisted Ground Exp x DC .042 Low Med High -.0538 X 10-1* .0398 X 10-4 .0394 X 10-4 The literature indicates elevated porphyrin compound excretion resulting from sufficient dioxin exposure. The pattern found here is one of higher Ranch Hand uroporphyrin or coproporphyrin levels relative to comparisons when there are concomitantly higher BUN levels, or, in the case of coproprophyrin, when there is higher alcohol ingestion. No overall group differences are observed. Tables XIV-10, XIV-11 and XIV-12 display the results of exposure index analyses within the Ranch Hand group. Starting with Table XIV-10, no statistically significant overall group differences are seen and 5 statistically significant(P £ 0.050) group-covariate interactions are noted. Table XIV-11 displays unadjusted group means for the porphyrin metabolism related variables and, as indicated by the statistical testing of overall group differences, no trends with exposure index are observed. XIV-13 �The 5 exposure-by-covariate interactions are listed in Table XIV-12; however, only the exposure index by degreasing chemical interactions follow a classical dose-response pattern. Specifically, Ranch Hand officers with greater herbicide exposure, as measured by the exposure index, have greater increases in uroporphyrin output in response to degreasing chemical exposures than do Ranch Hand officers with less herbicide exposure. The same pattern is seen in the enlisted ground d-aminolevulinic acid data. 4. Clinical Variables Sixteen of 1027 Ranch Handers (1.56%) were diagnosed as having hepatomegaly while 6 of 769 comparisons (0.78$) had that finding (P = 0.138) with an approximate 70% power. In the Ranch Hand group, the cases of hepatomegaly appear to be randomly distributed within the 3 exposure categories; however, due to the small number of cases statistical testing is not powerful. These data on hepatomegaly are shown in Table XIV-13 (febrile participants and individuals with HBsAg have been removed). Table XIV-13 CASES OF HEPATOMEGALY IN THE RANCH HAND COHORT BY OCCUPATION AND EXPOSURE CATEGORY Exposure Index Medium Low Occupational Category Officers Enlisted Flying Enlisted Ground Cases N Cases N 2 1 0 110 59 148 2 2 3 124 58 176 High Cases 2 2 1 _N 123 63 147 Eighteen of 1027 Ranch Handers (1.75?) reported an enlarged liver during response to questionnaire inquiry while 13 of 760 comparisons (1.71?) reported the same. XIV-1 �The study questionnaire also inquired about a medical history of hepatitis, jaundice, cirrhosis, and a general category called other liver conditions. Ranch Hand and comparison responses to these questions are shown in Table XIV-14, Ranch Hand respondents differ from comparisons only in the other liver category. Thirteen of the 16 Ranch Handers reporting other liver conditions have had their report verified by medical record. One comparison has had his condition verified. A display of the verified findings is shown in Table XIV-15 (febrile individuals and HBsAg positive individuals were left in the analysis). Table XIV-14 SPECIFIC LIVER DISORDERS REPORTED ON QUESTIONNAIRE Reported Event Ranch Hand Comparison P Value Yes No Yes No Hepatitis 40 1005 32 741 >0.50 Jaundice 44 1001 35 738 >0.50 Cirrhosis 4 1041 3 770 >0.50 16 1029 2 771 0.004 Other , Table XIV-15 OTHER LIVER CONDITIONS REPORTED BY STUDY PARTICIPANTS AND VERIFIED BY MEDICAL RECORDS Ranch Hand: ICD Code Code Meaning 2724 570 5739 7904 Comparison: Hyperlipidemia Liver necrosis Unspecified Enzyme elevation 5719 Chronic unspecified XIV-15 Number 1 1 10 1 1 �Table XIV-16 REPORTED SKIN PATCHES, BRUISES OR SENSITIVITY IN RANCH HAND PARTICIPANTS BY OCCUPATION AND EXPOSURE CATEGORY Exposure Index Occupational Category Cases Officers Enlisted Flying Enlisted Ground Medium Low 36 27 74 % 32.4 45.8 49.0 High N Cases % N Cases 111 59 151 48 28 82 37.5 47.5 45.8 128 59 179 44 37 76 % 35.2 56.1 51.4 N 125 66 148 Seeking historical evidence of porphyric symptoms, questions concerning skin changes that could have been associated with porphyria cutanea tarda were asked (specifically, skin patches, bruisibility or sensitivity). Of 1045 Ranch Hand respondents, 462 or 44.2$ reported these skin symptoms while 278 of 773 comparisons or 36.0$ reported these conditions. These reported cases indicate a statistically significant group difference (P <0.001); however, no regression with exposure index was noted (data given in Table XIV-16). The historical and hepatomegaly data support an interpretation of some group difference. However, no positive association with herbicide exposure has been noted. 5. Summary and Conclusion Ranch Handers have slightly greater GGPT and LDH levels than the comparisons while having lower cholesterol levels. Also, Ranch Hand SGOT, SGPT and LDH levels are more highly correlated to (and therefore may be more influenced by) materials with an hepatic effect, namely, alcohol, degreasing compounds and industrial chemicals. No group differences were noted in alkaline phosphatase or bilirubin levels. Borderline statistically significant group differences have been detected in uroporphyrin and coproporphyrin levels in association with BUN, and in coproporphyrin levels in association with alcohol ingestion. No overall group differences were detected in these compounds or delta aminolevulinic acid values. Twice as many Ranch Handers as comparisons had enlarged livers on physical examination, but this difference was not, statistically significant. Statistically significant group differences were noted in the occurrence of miscellaneous liver disorders exclusive of hepatitis, jaundice and cirrhosis, verified by XIV-16 �medical record review. Ranch Handers self reported 23% more skin changes of the type associated with porphyria cutanea tarda than did the comparison participants, and the group difference was statistically significant. Clinically apparent porphyria was not evident at physical examination. The observed group differences in liver-related biochemical variables found in the blood, and in porphyrin metabolism compounds found in the urine are most likely of minor or negligible medical importance at the present time. The verified reports of liver morbidity are of greater clinical interest. The exposure index analyses do not support an interpretation of herbicide effect with respect to any of the group differences summarized. XIV-17 �Chapter XV DERMATOLOGIC EVALUATION A thorough dermatologic assessment was deemed essential because chloracne is the only recognized definitive clinical end point following exposure to chlorophenols and dioxin. Over one-half of all veteran complaints recorded in the Veterans Administration Herbicide Registry cited dermatologic symptoms. These facts, coupled with the knowledge that chloracne is transient following a single point exposure (Homberger, 1979), suggested that there is a significant potential to misclassify adolescent acne and chloracne. While the issue of correct diagnosis could be resolved by biopsies and histopathologic characterizations in all participants, this approach was rejected on ethical grounds, as well as concern for the adverse impact of biopsy procedures on future study participation. Consequently, the dermatologic assessment was carefully planned to collect historical and distributional dermatologic data by questionnaire, followed by a detailed corroborative physical examination, supplemented by voluntary biopsies when indicated. Most data reported in this chapter are from the 10*15 Ranch Handers and the 773 originally selected comparison individuals enrolled in the study. Minor fluctuations from these denominators reflect missing dependent variable or covariate data. Relative risks and confidence intervals are shown for all dependent variables in Appendix XVIII. 1 • Questionnaire Data The in-home study questionnaire collected detailed medical histories on the occurrence of acne. These data are displayed in Table XV-1 and show that the Ranch Handers reported slightly more acne than their comparisons. Table XV-1 REPORTED OCCURRENCE OF ACNE BY GROUP Group No Acne Number Percent Reported Acne Number Percent Total Number Percent Ranch Hand 659 63.3 382 36.7 1041 100 Comparison 498 64.8 271 35.2 769 100 Reported acne group contrast: P = 0.52 XV-1 �Beginning and end dates of up to three sustained periods of acne activity re recorded for each individual on the questionnaire. Since only acne after 61 could be possibly induced by herbicide exposure, cases of post-1961 acne re placed in time reference to each individual's RVN tour(s). This temporal stribution was not statistically different with respect to group membership, ese data are reflected in Table XV-2. Table XV-2 REPORTED POST-1961 ACNE BY TIME OF THE SOUTHEAST ASIA [SEA] TOUR(S) BY GROUP roup Pre-SEA Only Number Percent Post-SEA Only Number Percent Pre- and Post-SEA* Number Percent nch Hand = 179 62 3^.6 31 17.3 86 48.0 mparison = 116 51 44.0 17 14.7 48 41.4 iported acne by group by pre/post SEA: P = 0.27 ^ported acne (Post SEA) relative risk: 1.18, 95/S Conf. int. (.67, 2.18) *Such acne could have been separate cases or the same case starting before his RVN tour and ending afterwards. Durations of the cumulative acne episodes were distributed by 5-year itervals and contrasted by group and SEA category. These data are shown in tble XV-3. XV-2 �Table XV-3 DURATION OF ACNE IN 5-YEAR CATEGORIES BY SEA TOUR AND GROUP MEMBERSHIP Pre-SEA ONLY J5_ Duration in Years 5 <Yr <10 10 <Yr <1 5 15 <Yr ^20 Total Ranch Hand 44 15 2 1 62 Comparison 38 12 0 1 51 P = 0.63 Post-SEA ONLY Ranch Hander Comparison 15 9 4 11 1 31 2 4 2 1 7 P = 0.61 Thus, these SEA tour categories suggested that there were no group differences for the pre-SEA or post-SEA acne. Questionnaire information on whether the participant consulted a physician for his acne was used as an indirect measure of the clinical severity of the acne. Of 70 Ranch Handers with acne post-1961 who were asked this question, 29 (41.4/0 responded as having visited a physician as contrasted to 15 of the 45 (33.3/0 comparisons (P = 0.38), suggesting that there was not a statistically significant difference in the clinical severity of their acne. Since chloracne, following mild to moderate exposures, is classically found in skin areas on the temples, eyes/eyelids, and ears (eyeglass distribution), questions on rash locations and combinations of locations were presented to each participant reporting acne. Qt the 117 post-SEA plus pre- and post-SEA cases of acne in Ranch Handers after SEA duty, 75 (6455) reported no acne at any of these locations, while 36 (55%) of the 65 post-SEA plus pre- and postSEA comparisons reported none. These proportions are not significantly different (P = 0.25), and the occurrence of skin disease which could potentially be chloracne does not differ in the two groups. There were only four individuals, two in each group, with acne confined exclusively to the classical chloracne areas. As further corroboration of these anatomically categorized data, a Venn diagram was constructed for post-1961 acne lesions on the temples, ears, and eyes for the Ranch Hand group and the entire comparison group. These data are shown in Figure XV-I and display remarkable visual concordance. XV-3 �Figure XV-1 VENN DIAGRAM OF POST-1961 TEMPLE, EAR, AND EYE ACNE BY GROUP ENTIRE COMPARISON GROUP RANCH HAND GROUP (POST 1961) (POST 1961) ACNE, OTHER SITES: 157 N = 202 (ACNE REPORTED) 2. ACNE. OTHER SITES: 156 N = 198 (ACNE REPORTED) Physical Examination Data All physical examination data were described using a diagnostic checklist, and abnormalities were annotated on a full body diagram. Color photographs were obtained at the dermatologist's discretion, and 14 lesions were biopsied. Of the 14 biopsies collected from 11 patients, none were suggestive of chloracne. No cases of chloracne were diagnosed. Histologic descriptions of these biopsies are presented in Table XV-4. XV- �Table XV-4 BIOPSY RESULTS Number Histologlc Description Active degeneration Inclusion cysts Epidermal cysts Basal cell carcinoma Intradermal melanosis Seborrheic keratosis Pigmented nevus Psoriasiform dermatitis Chronic inflammation Insect bite 3 2 2 1 1 1 1 1 1 1 The five most common diagnoses and the P value for group differences are shown in Table XV-5. Abnormal skin findings were prevalent but almost identical in both groups (i.e., 45.0/& in Ranch Handers, and 44.9% in the comparisons; P = 0.97). Only for the miscellaneous diagnoses of "Other Abnormalities" (which included 15 diagnostic categories) was there a statistically significant group difference, with the comparisons having more disease than the Ranch Handers. Table XV-5 PREVALENCE OF DERMATOLOGIC DIAGNOSES IN PERCENT Diagnoses Comedones Acneiform lesions Acneiform scars Cysts Hyperpigmentation Other abnormalities Any abnormality Ranch Hand N = 1045 21. 18. 11, 11. 8. 12.6 Comparison N = 773 20.7 17.5 10.4 10.5 7.1 16.3 44.9 P Value 0.60 0.66 0.57 0.46 0.35 0.03 0.97 Relative Risk 1.05 1.05 1.08 1.10 1.17 .77 1 .00 95$ Conf int (.87,1.26) (.85,1.29) (.82,1.43) (.84,1.46) (.84,1.65) (.81, .98) (.90,1.11) Based upon the four most prevalent diagnoses in Table XV-5 (comedones, acneiform lesions, acneiform scars, and dermal cysts), all of which should encompass the diagnostic possibility of chloracne, a dermatologic index was constructed for each study participant. A score of zero was given if none of the four lesions were noted, and a score of 1 was assigned if one lesion was diagnosed, etc. These data are displayed in Table XV-6. XV-5 �Table XV-6 DERMATOLOGIC INDEX SCORK BY GROUP Grout 0 Number % 1 Number % Ranch Hand (N = 1045) 633 60.6 23*1 22.4 Comparison (N = 773) 487 63.0 157 Scores 2 Number % 20.3 3 Number % 4 Number % 124 11.9 42 4.0 12 1. 12.3 27 3.5 7 0.9 95 P = 0.74 The distributions of these scores did not differ significantly, suggesting a similar crude clinical severity between the groups. 3. Questionnaire - Examination Correlations The dermatologic index was contrasted to the historical occurrence of acne by group. These data are shown in Table XV-7. Table XV-7 DERMATOLOGIC INDEX IN PERCENT BY QUESTIONNAIRE HISTORY OF ACNE BY GROUP Score 2_ 3. _4_ 21.4 18.1 9.4 9.6 2.4 2.6 0.5 0.6 • 0.72 55.3 55.1 25.1 21.8 13.4 17.7 4.5 4.1 1.7 1.4 0.84 47.3 48.4 23.2 26.6 17.7 16.9 8.9 6.4 3.0 1.6 0.82 History Group No Acne Ranch Hand Comparison 66.3 69.1 Acne S1961 Ranch Hand Comparison Acne >1961 Ranch Hand Comparison 0 1 XV-6 P Value �These data show that the dermatologic index does not differ significantly by group for any historical subset. And, as can be observed in Table XV-7, there is a positive association between the history (and time) of acne and the dermatologic index, regardless of group membership. An additional analysis of the dermatologic index for each individual who reported acne after his SEA tour (post-SEA only) did not reveal significant Ranch Hand-comparison differences (P = 0.50). 4. Exposure Index Analyses Several comparisons were made using the exposure index and both historical and examination findings in the Ranch Hand group. Two historical parameters (incidence of acne and severity of acne) and the dermatologic examination findings were contrasted to the exposure index after stratifying for occupational categories by log-linear models. The historical-exposure analyses were essentially negative. Major dermatologic lesions from the examination were contrasted to the exposure index by occupational category. This analysis is presented in Table XV-8. XV-7 �Table XV-8 PERCENTAGE OF SPECIFIC SKIN LESIONS IN RANCH HANDERS BY EXPOSURE LEVEL BY OCCUPATIONAL CATEGORY (POST 1961 DATA ONLY) Exposure Level Low Medium High Condition Qcoupational Group P Value All skin abnormalities Officers Enlisted Flying Enlisted Ground 57.1 22.2 14.3 16.7 39.5 35.8 21.4 60.0 25.0 0.20 0.17 0.40 Comedones Officers Enlisted Flying Enlisted Ground 14.3 22.2 57.1 50.0 18.6 24.5 21 .4 20.0 31.2 0.91 0.42 0.45 Acneiform Lesions Officers Enlisted Flying Enlisted Ground 0 33.3 57.1 16.7 37.2 22.6 50.0 20.0 37.5 0.08 0.23 0.21 Acneiform Scars Officers Enlisted Flying Enlisted Ground 28.6 1 1 . 1 71.4 50.0 10.9 28.3 21 .4 40.0 31 .2 0.68 0.53 0.57 Inclusion Cysts Officers Enlisted Flying Enlisted Ground 14.3 0 14.3 50.0 18.6 18.6 14.3 20.0 27.1 0.49 0.32 0.53 Hyperpigmentation Officers Enlisted Flying Enlisted Ground 0 11.1 14.3 16.7 9.3 15.1 7.1 0 3.1 0.72 0.64 0.20 Thus, of the 18 exposure analyses, none were statistically significant (although based upon small sample sizes). Similarly, the relationship between the dermatologic index and exposure index was explored. For all three occupational categories, the dermatologic index showed no significant correlation to the exposure index, as reflected in Table XV-9. XV-8 �Table XV-9 RANCH HAND DERMATOLOGIC INDEX IN ALL OCCUPATIONAL CATEGORIES BY THE EXPOSURE INDEX (POST 1961 DATA ONLY) Exposure Level Dermatologic Index > 1 0 Number Percent Number Percent Low 26 45.6 31 54.4 Medium 28 41.2 40 58.8 High 20 39.2 31 60.8 P 5. - 0.78 Summary A comprehensive dermatologic assessment was conducted by questionnaire and physical examination. The questionnaire data revealed that the incidence of past acne, its time of occurrence relative to the individual's SEA tour(s), its severity and duration, and its anatomic location did not significantly differ between the Ranch Hand and comparison groups. No cases of chloracne were diagnosed at physical examination or by biopsy. No group differences were noted for the five most prevalent dermatologic diagnoses. The category, other abnormalities (containing 15 dermatologic conditions), was significantly larger for the comparison group than for the Ranch Hand group. However, when all skin abnormalities were considered, the group rates were essentially identical. A dermatologic index was constructed to account for the number of skin abnormalities per individual (severity index) that might encompass a diagnosis of chloracne. The index was not associated with group membership but showed some correlation with a total history of past acne in both groups. There were no associations between historical or dermatological examination findings and exposure level in any occupational category of the Ranch Hand group. XV-9 �Chapter XVI-1 CARDIOVASCULAR EVALUATION 1. Introduction The effects of Herbicide Orange and its dioxin contaminant on the cardiovascular system are not well defined. Both bradycardia and tachycardia have been suggested in acute heavy exposures to the 2,4-D and 2,^,5-T components, but the cardiovascular effects following chronic low dose exposure are essentially unknown. The thrust of this cardiovascular evaluation has been to collect important data by questionnaire, physical examination, and laboratory testing, that would identify Ranch Hand-comparison group differences after accounting for the effects of confounding variables. Of the well-established risk factors for cardiovascular disease, smoking, cholesterol level or cholesterol to high density lipoprotein (HDL) ratio, and age were selected as covariates in most analyses (Brand et al, 1976). The covariates were categorized as follows: age, S^O, ^0 years 1 month - 59 years 11 months (abbreviated HO < > 60), and 60 years or more; smoking, 0 pack-years, 1-10 pack-years, and 11 or more pack-years; cholesterol, S180 mg/dl, 181-279 mg/dl, and S280 mg/dl; and cholesterol-HDL ratio, <5.3» £5.3. In complex analyses with sparse data, trichotomous covariates were reduced to dichotomous ones. The outpoint for cholesterol-HDL ratio was derived from data on rated Air Force personnel referred for cardiovascular diagnostic examination; it is an unweighted average of means of flyers verified at cardiac catheterization as having or not having occlusive coronary atherosclerosis. A more optimal approach, based upon a median HDL value of the comparison group, will be used in subsequent reports. Statistically significant interactions between these covariates were not explored in detail when there was no effect on group membership and when the interactions were consonant with the classical epidemiology of cardiovascular disease. Analyses of weak risk factors in the data will be presented in subsequent reports. Because of the low proportion of Black participants in both groups, covariate adjustment by race was not possible. Consequently, a variety of dependent variable analyses by race, unadjusted for age, smoking, and cholesterol, are discussed throughout this chapter. In addition, where adjusted group differences were found to be statistically significant, other covariates (e.g., percent body fat, current smoking, history of intermittent claudication, testosterone level, differential cortisol level, etc.) have been used to reanalyze all data in an attempt to clarify the clinical significance of the finding. Most analyses herein are based upon Ranch Hand contrasts to the "originals" of the comparison group. Where group associations are statistically significant or of general interest, other comparison group denominators have been used (e.g., matched originals only and the entire comparison group). Further, for specific analyses, participants with diabetes and pedal edema have been deleted. Small denominator fluctuations are also inherent in these analyses because of missing covariate or dependent variable information. Thus, tabular data may not be directly comparable between analyses because of the type of XVI-1-1 �covariate adjustment, or the denominator of the comparison group, or the deletion of certain medical conditions thought to confound a specific clinical diagnosis. In general, covariates having a nonsignificant association with the dependent variable were removed from the analysis. The statistical analyses are based on log-linear models (BMDP-4F), logistic regression (BMDP-LR), and generalized linear models, chi-square, t tests, and matched covariate analyses (Breslow, 1982). Relative risks and confidence intervals, computed using the hypergeometric distribution (Thomas, 1971) and the normal approximation (Fleiss, 1981), are shown for all dependent variables in Appendix XVIII. 2 * Central Cardiovascular System a. Sy s t o1ic Blood Pressure Abnormal systolic blood pressure was defined as pressure in excess of 140 mmHg by standard observer auscultation. All blood pressures were obtained in a sitting position. Second or third readings were recorded on those individuals who manifested an initial elevation. There was no significant difference in systolic blood pressure (P = 0.248) between the non-Black Ranch Hand and the non-Black original comparison group after adjusting for age, smoking, and cholesterol level. These data are reflected in Tables XVI-1-1 and XVI-1-2. Diabetics (2-hour postprandial glucose £120 mg/dl) were removed from the analyses. Table XVI-1-1 SYSTOLIC BLOOD PRESSURE RANCH HANDERS AND THE ORIGINAL COMPARISONS VERSUS AGE (NON-BLACKS ONLY) Ranch Hand Age Abnormal % Abnormal Normal Original Comparisons Abnormal % Abnormal Normal Total Both Groups Abnormal % Abnormal Normal <40 36 10.4 309 32 14.3 192 68 11.9 501 £40 113 23.1 377 94 24.6 288 207 23.7 665 Systolic pressure between groups: P = 0.248 Age versus systolic Relative risk under 40: .73,9555 Conf int ( . 4 6 , 1 . 1 8 ) pressure (unadjusted Relative risk over 40: .94, 95/5 Conf int (.73,1.20) for smoking and cholesterol): P <0.0001 The unadjusted systolic blood pressure by smoking history association, presented in Table XVI-1-2, is not significant (P-0.179) in these data. XVI-1-2 �Table XVI-1-2 SYSTOLIC BLOOD PRESSURE PARTICIPANTS BY SMOKING HISTORY (NON-BLACKS ONLY) Smoking History in Pack-Years Abnormal 0 1-10 >10 % Abnormal 17.8 16.1 20.8 70 44 161 Normal 324 230 612 0.179 Ranch Handers and original comparisons reflected in these tables were also compared on systolic blood pressure as a continuous variable with adjustment for age, smoking history, HDL ratio, and body fat, via a general linear model. There was no significant difference between the groups on systolic blood pressure (P » 0.976). The Ranch Hand and original comparison adjusted means were 133.12 and 133.15, respectively. The covariates of age and body fat were both significantly associated with systolic blood pressure (P = 0.0001). Additional categorical analyses comparing Non-Black Ranch Handers with the total non-Black comparison group adjusted for age, smoking, and cholesterol showed comparable nonsignificant intergroup differences (P = 0.366) for systolic blood pressure. The effects of age and smoking were statistically significant, P <0.0001 and P = 0.04, respectively. In addition, a chi-square analysis of Black Ranch Handers and Black individuals from the entire comparison group (diabetics removed) showed no group difference (P = 0.265) in systolic pressure. b. Diastolic Blood Pressure Diastolic blood pressure in excess of 90 mmHg was categorized as abnormal. No significant intergroup difference was noted after adjustment for age, smoking, and cholesterol level. These data are based upon non-Black, nondiabetic denominators and are presented in Table XVI-1-3. XVI-1-3 �Table XVI-1-3 DIASTOLIC BLOOD PRESSURE IN RANCH HANDERS AND THE ORIGINAL COMPARISONS VERSUS AGE (NON-BLACKS ONLY) Ranch Hand Age Abnormal %_ Abnormal Normal <40 18 >40 57 5.2 Original Comparisons Abnormal % Abnormal Normal Total Both Groups Abnormal % Abnormal Normal 327 12 5.4 212 30 5.3 539 11.6 433 53 13.9 329 110 12.6 762 Diastolic blood pressure P = 0.351 between groups: Relative risk under 40: .97,95? Conf. int. (.45,2.18) Relative risk over 40: .84,95? Conf. int. (.58,1.21) Age versus diastolic pressure (unadjusted for smoking and cholesterol): P <0.0001 The Ranch Handers and original comparisons (as represented in Table XVI-1-3) diastolic blood pressure was also compared as a continuous variable with adjustment for age, smoking history, HDL ratio, and body fat, via a general linear model. There was a borderline significant diastolic blood pressure by group by age interaction (P = 0.0585), indicating a change in the blood pressure by group association with level of age (<40, £40). However, separate analyses at each level of age revealed no significant group differences. In the under-40 age group, the diastolic blood pressure by group association was not significant (P = 0.435); the adjusted group means were 78.2 and 77.02 for Ranch Handers and comparisons, respectively. In the 40-and-over age group, the diastolic blood pressure by group association was not significant (P = 0.904); the Ranch Hand and comparison adjusted means were 80.7 and 81.7, respectively. An intergroup log linear analysis of diastolic blood pressure for Blacks and non-Blacks using original comparisons showed comparable nonsignificant results (P = 0.573). Age was a significant covariate (P <0.0001) while the history of past smoking was not. An unadjusted contrast of Black Ranch Handers and Black individuals from the entire comparison group also showed similar nonsignificant group differences (P = 0.533). c. Electrocardiograms (ECG's) ECG's were obtained on all participants, following a minimum fast of 4 hours and abstinence from tobacco for 4 hours. The vast majority of ECG's were obtained by 1 or 2 technicians on dedicated and calibrated machines. The tracings were read by a contract clinic' cardiologist and categorized into normal and abnormal groups, the latter consisting of right bundle branch block, left bundle branch block, nonspecific T wave changes, bradycardia, tachycardia, and XVI-1-4 �other diagnoses. Grave findings were immediately discussed with the participant's family physician and appropriate follow-up was arranged. As shown in Table XVl-1-4, abnormal EGG findings were not associated with group membership (P = 0.987). For both the non-Black Ranch Hand and original comparison groups, there was a highly statistically significant (P <0.0006) association between abnormal ECG's and increased age. Table XVI-1-4 EGG FINDINGS IN RANCH HANDERS AND THE ORIGINAL COMPARISONS BY AGE, ADJUSTED FOR SMOKING HISTORY AND HDL RATIO (NON-BLACKS ONLY) Ranch Hand Age Abnormal % Abnormal Normal Original Comparisons Abnormal % Abnormal Normal Total Both Groups Abnormal % Abnormal Normal <40 69 20.1 274 51 23.1 170 120 21.3 444 >40 148 30.2 342 107 28.4 269 255 29.4 611 Abnormal ECG findings between groups: P = 0.987 EGG findings in both Relative risk under 40: .87,95% Conf. int. (.62,1.23) groups by age (unRelative risk over 40: 1.06,95$ Conf. int. (.86,1.32) adjusted for smoking and HDL ratio): P = < 0.0006 When the ECG data in Table XVI-1-4 were redistributed into the categories of tachycardia, bradycardia, other abnormalities, and normal, an unadjusted analysis showed no significant differences between the Ranch Hand and original comparison group (P = 0.881). An additional cardiac assessment was made on all past or present flying personnel in both groups. Participants' names and social security numbers were computer matched to the USAF ECG Repository, the world's largest ECG repository on flying personnel (Lancaster and Ord, 1972; Hiss and Lamb, 1962). Three hundred and fifty-four Ranch Handers and 282 original comparisons had between one and 10 previous tracings on file which had been diagnostically coded by stringent criteria. Accordingly, USAF cardiologists reviewed all 636 physical examination ECG's (without knowledge of group membership) and coded them by the standardized USAF criteria. The physical examination ECG was contrasted to the past ECG's and categorized as no change or degraded (no ECG's were improved in either group). These data analyzed by group membership and age are shown in Table XVI-1-5. Blacks and diabetics were removed from the analysis. This analysis is not adjusted for elapsed time between ECG readings. XVI-1-5 �Table XVI-1-5 CLINICAL COMPARISON OF CURRENT EGG'S TO PAST EGG'S IN FLYING PERSONNEL BY GROUP MEMBERSHIP AND AGE (NON-BLACKS ONLY) Age Comparison Ranch Hand Total Degraded No Change No Change Degraded No Change Degraded Number Number Percent Number Number Percent Number Number Percent <40 45 2 4.2 29 2 6. 4 74 >40 226 271 20. 22 8. 1 182 211 17 19 8.5 408 482 4 ' 5.1 37 41 8.3 Because of sparse data in the under-40 age group, an analysis adjusted for both age and smoking was not possible; the unadjusted ECG change by group association was not significant (P = 0.652). In the 40-and-over age group, the ECG change by group association was not significant (P = 0.939), adjusted for smoking history. The smoking history covariate was borderline significant, P = 0.0852. In both the Ranch Hand and comparison groups combined, the age by ECG association (P = 0.412) was not significant. The unadjusted ECG change by smoking history association was significant (P = 0.018). An overall analysis of systolic/diastolic blood pressures and ECG abnormalities was performed by group membership and adjusted for smoking (0, 1-10, >10 pack-years), cholesterol-HDL ratio «5.3, S5.3), age «40, £40) and differential cortisol level (continuous); Blacks and diabetics were omitted. The differential cortisol level is defined as the 7:30 AM cortisol measurement minus the 9:30 AM cortisol measurement. A logistic regression analysis showed similar nonsignificant results (as in Sections a-c above) that are presented in Table XVI-1-6. Table XVI-1-6 RANCH HAND AND ORIGINAL COMPARISON GROUP CONTRAST FOLLOWING ADJUSTMENT FOR AGE, SMOKING, CHOLESTEROL-HDL RATIO, AND DIFFERENTIAL CORTISOL RESULTS (NON-BLACKS ONLY) Dependent Variable P Value Systolic Blood Pressure Diastolic Blood Pressure ECG Abnormality 0.195 0.351 0.999 XVI-1-6 �d. Heart Sounds All valvular sound abnormalities were recorded following detailed auscultation. Fourth heart sounds were considered abnormal. If the participant indicated that the heart sound abnormality was a new finding, the diagnostician confirmed the abnormality. A review of the heart sound abnormalities in the non-Black Ranch Handers and original comparisons revealed that the data were too sparse for a fully adjusted analysis. An unadjusted group comparison was nonsignificant (P = 0.414), as was the unadjusted effect of age (P = 0.375). Similarly, an unadjusted analysis of Black Ranch Handers and comparison individuals did not demonstrate statistical significance (P = 0.799). A combined race and fully adjusted (age, smoking, cholesterol level) analysis of Ranch Handers and the entire comparison group is presented in Table XVI-1-7. These data also show no group differences (P «= 0.592) but do reflect a significant association of heart sound abnormalities and increasing age (P <0.002). Table XVI-1-7 HEART SOUND ABNORMALITIES IN BLACK AND NON-BLACK RANCH HANDERS AND ALL COMPARISONS BY AGE Ranch Hand Age Abnormal % Abnormal Normal <40 5 40060 11 >60 2 1.3 2.3 11.1 367 476 16 Comparison Abnormal % Abnormal Normal 8 15 2 1.9 2.7 8.3 Abnormal heart sounds between groups: 3. 417 542 22 Total Both Groups Abnormal % Abnormal Normal 13 26 4 1.6 2.5 9.5 784 1018 38 P = 0.592 Heart sound abnormalities in both groups by age: P < 0.002 Peripheral Cardiovascular System The status of the peripheral cardiovascular system was evaluated by ophthalmoscopic examination of the eyegrounds for arterial-venous nicking and hemorrhages, auscultation of the carotid arteries, and bilateral palpation for the presence and quality of 5 peripheral pulses. The finding of a bilateral abnormality (e.g., bruits in both carotid arteries) was scored as 1 abnormality. Diminished or absent peripheral pulses were both designated as abnormal. While there is clearly recognized misclassification of the specific causes for the examination findings, it is judged to be of a minor nature; thus, the examination findings are deemed to be generally indicative of the presence XVI-1-7 �or absence of severe arteriosclerosis. Abdominal x-rays to confirm the severity of the peripheral vessel arteriosclerosis were not obtained because of the possible impact of detected asymptomatic or clinically irrelevant kidney stones upon the flying status of active pilots. a. Eyegrounds Abnormal funduscopic findings were not associated with group membership (P = 0.965), but were highly correlated with increased age (P < 0.0001), as reflected in Table XVI-1-8. The additional covariates of smoking history and cholesterol-HDL ratio were nonsignificant in the analysis. Table XVI-1-8 FUNDUSCOPIC ABNORMALITIES IN RANCH HANDERS AND ORIGINAL COMPARISONS BY AGE (NON-BLACKS ONLY) Ranch Hand_ Age Abnormal % Abnormal Normal _ Comparison_ Abnormal % Abnormal Normal Total _ Both Groups _ Abnormal % Abnormal Normal 8 2.3 333 6 2.7 214 14 2.5 547 42 <40 8.7 441 31 8.4 339 73 8.6 780 Funduscopic abnormalities between groups: P = 0.965 Funduscopic abnormalRelative risk under 40: .86,95? Conf. int. (.26,2.97) ities in both groups Relative risk over 40: 1.04,95? Conf. int. (.65,1.67) by age (unadjusted for smoking and cholesterol-HDL ratio) P <0.0001 An unadjusted contrast of Black Ranch Handers and Black individuals from the entire comparison group showed similar nonsignificant results (P = 0.860). b. Carotid Bruits The prevalence of carotid bruits in both groups combined was 1.47/5. Because of sparse data, an unadjusted analysis comparing non-Black Ranch. Handers with non-Black original comparisons was performed; the group by carotid bruits association was nonsignificant (P = 0.269), as was the unadjusted age by carotid bruits association (P = 0.353). However, the larger analysis of both Black and non-Black Ranch Handers with the entire comparison group showed a group membership association of interest (P = 0.183) and a significant relationship between bruits and increasing age (P = 0.03). XVI-1-8 �c. Peripheral Pulses The absence or diminished quality of 5 peripheral pulses was determined by detailed clinical palpation. One or more abnormal pulses were found in 12.8%(106/829) of the non-Black Ranch Handers as contrasted to 9.4$ (56/596) in the non-Black original comparisons (P = 0.05) giving an unadjusted relative risk of 1.36 with a 95% confidence interval (.99, 1.88). The reader is referred to Appendix XVIII for complete relative risks and confidence intervals. Data on specific pulses are presented in Table XVI-1-9. The covariates of cholesterol-HDL ratio and percent body fat (<25%, S25/0 were noncontributory in all of the analyses. Thus, the pulse variables were adjusted for age «40, S40) and smoking (0, 1-10, >10 pack-years). Blacks, diabetics, and individuals with peripheral pitting edema were omitted from the analysis. Since most abnormalities were concentrated in the over 40 and > 10 pack-year group, these data were re-analyzed on that subset with the results shown in column three of Table XVI-1-9. Table XVI-1-9 SUMMARY OF PERIPHERAL PULSE QUALITY: RANCH HANDERS AND ORIGINAL COMPARISONS (NON-BLACKS ONLY) Pulse Examined,- Unadjusted P Value Number of and Direction of Participants Group Abnormalities Unadjusted P Value for Age ^40 Years and >10 Pack-Years Unadjusted P Value Age Versus Pulse (Groups Combined) Radial N = 1414 0.147 (RH > C) Sparse Data 0.668 Femoral N = 1414 0.147 (RH > C) 0.117 (RH > C) 0.157 Popliteal N = 1414 0.0255 (RH > C) 0.0159 (RH > C) 0.0065 0.0375 0.0003 Dorsalis Pedis N = 1413 Posterior Tibial N • 1413 0.0644 (0.0406)* (RH > C) 0.312 (0.250)* (RH > C) (RH > C) 0.123 (RH > C) *Adjusted for age and smoking XVI-1-9 0.0022 �Although only two pulses reached statistical significance (P 50.05) in Table XVI-1-9, the consistent directional findings in all peripheral pulses were sufficient to merit additional clarifying analyses. Further, these directional findings were present after accounting for diabetes and the clinically confounding physical effects of peripheral pitting edema and obesity. Accordingly, various aggregates of pulses were constructed to determine more precisely the anatomic patterns of the abnormalities. This approach, adjusted by age and smoking history, is displayed in Table XVI-1-10. Table XVI-1-10 SUMMARY OF PERIPHERAL PULSE ABNORMALITY COMBINATIONS: RANCH HANDERS AND ORIGINAL COMPARISONS ADJUSTED BY AGE AND SMOKING HISTORY (NON-BLACKS ONLY) Pulse Abnormalities Combination Adjusted P Value and Direction of Group Abnormalities Unadjusted P Value Age Versus Pulse Combination Leg Pulses* (Femoral, Popliteal, Dorsalis Pedis, Posterior Tibial) 0.03.02 (RH > C) 0.0001 All Pulses (Carotid, Femoral, Radial, Popliteal, Dorsalis Pedis, Posterior Tibial) 0.0257 (RH > C) 0.0005 Peripheral Pulses (Radial, Femoral, Popliteal, Dorsalis Pedis, Posterior Tibial) 0.0235 (RH > C) 0.0002 *In nondiabetic, non-Black, Ranch Handers and the original comparisons, leg pulses were associated with a history of intermittent claudication (P = 0.0113), and this association was the same in both groups (P = 0.962). The data in Table XVI-1-10 did not point to specific anatomic groupings but rather suggested a generalized phenomenon. As a result of this finding, the pulse data were reanalyzed using testosterone and differential cortisol results as new covariates. No substantial change in the significance of the pulse findings was observed. In order to provide a complete approach to the peripheral pulse findings, 2 supplemental contrasts using other denominators were performed: 1) an analysis of both Black and non-Black Ranch Handers versus Black and non-Black comparisons from the entire comparison group, adjusted for XVI-1-10 �age, smoking history in pack-years, and cholesterol level; and 2) an unadjusted analysis of Black Ranch Handers versus Black comparisons from the entire comparison set. The data from these analyses are presented in Table XVI-1-11. Table XVI-1-11 SUMMARY OF PERIPHERAL PULSE QUALITY: ALL RANCH HANDERS VERSUS ALL COMPARISONS*, ASSOCIATION OF AGE, UNADJUSTED CONTRAST OF BLACK RANCH HANDERS AND BLACK COMPARISONS Blacks and Non-Blacks P Value and Direction of Pulse Examined, Group Abnormalities Number of Participants Blacks Only P Value of Age Association Both Groups** Unadjusted P Value Radial N = 1884 0.047 (RH > C) 0.012 0.890 Femoral N = 1882 0.134 (RH > C) 0.007 0.219 Popliteal N = 1883 0.0174 (RH > C) <0.001 0.219, Dorsalis Pedis N = 1881 0.006 (RH > C) <0.001 0.789 Posterior Tibial N = 1882 0.067 (RH > C) <0.001 0.557 *Adjusted for age, smoking, and cholesterol level **Unadjusted for smoking and cholesterol The data in Table XVI-1-11 are thus corroborative of diminished pulse quality in the Ranch Hand group. These data also weakly suggest that the Ranch Hand - comparison pulse differences may be aggregated in the non-Black population (or may be spurious due to small sample size). A matched pair analysis (matching variables: age, job, race) of data sets for 3 pulses (see Table XVI-1-9), adjusting for percent body fat and smoking history, are shown in Table XVI-1-12. Due to sparse data, only main effects were included in these analyses. XVI-1-11 �Table XVI-1-12 MATCHED PAIR ANALYSIS FOR THREE PERIPHERAL PULSES: RANCH HANDERS VERSUS ORIGINAL COMPARISONS (NON-BLACKS ONLY) Pulse Variables P Value and Direction of Group Abnormalities Popliteal Pulse 0.053 (RH > C) Dorsalis Pedis 0.050 (RH > C) Posterior Tibial 0.081 (RH > C) Thus, the data in Table XVI-1-12 reaffirm the overall finding of significant peripheral pulse deficits in the Ranch Hand group. *(. Risk Factors in Central and Peripheral Cardiovascular Disease This section emphasizes cardiovascular disease relationships that are highlighted by significant risk factors or combinations of risk factors identified in the preceding sections or in the general literature. a. Cholesterol and HDL Cholesterol Nondiabetic non-Black Ranch Handers and the non-Black original comparisons were contrasted for continuous cholesterol and HDL levels via a general linear model, adjusting for age (<40, S^O), smoking history (0, 1-10, >10 packyears), and body fat (<25%, £25%). Although no group membership differences were found for cholesterol and HDL, several of the covariates were of profound influence. These data are shown in Table XVI-1-13. XVI-1-12 �Table XVI-1-13 CHOLESTEROL AND HDL IN RANCH HANDERS AND ORIGINAL COMPARISONS (NON-BLACKS ONLY) Adjusted Ranch Hand - Comparison P Value Age Cholesterol 0.355 0.038 0.002 0.919 HDL 0.178 0.788 0.028 0.0001 Dependent Variable Covariate P Values Smoking Body Fat Similar results were found in the contrast of nondiabetic Blacks. Because of small sample size, covariate adjustment was not possible. The contrasts were made by t tests and the results are shown in Table XVI-1-14. Table XVI-1-14 CHOLESTEROL AND HDL RESULTS IN RANCH HANDERS AND ORIGINAL COMPARISONS (BLACKS ONLY) IN Ranch Hand Standard Mean Deviation N Comparison Standard Mean Deviation Cholesterol 49 214.3 34.6 37 209.8 41 .3 0.595 HDL 49 55.5 17.2 37 52.4 14.6 0.375 P Value b. Age, Past Smoking, Current Smoking Risk Factors Several analyses have shown the substantial effects of age and smoking on the cardiovascular system. Because of the unknown influence of antismoking campaigns in recent years on Air Force personnel, the covariate of smoking history (0, 1-10, >10 pack-years) may not be fully appropriate, particularly if smoking ceased several years before the examination. Consequently, all dependent variables were reanalyzed for group differences restricting to older (>40), heavy past smokers (>10 pack-years), adjusted for current smoking (yes, no). These contrasts are presented in Table XVI-1-15. Blacks and diabetics were removed for the analysis. XVI-1-13 �Table XVI-1-15 RANCH HANDERS AND ORIGINAL COMPARISONS ADJUSTED FOR CURRENT SMOKING (NON-BLACKS, > 10 YEARS, > 10 PACK YEARS ONLY) Dependent Variable(s) P Value and Direction of Group Significance V Systolic Blood Pressure Diastolic Blood Pressure ECG Abnormalities Heart Sound Abnormalities Eyegrounds Carotid Bruits Radial Pulse Femoral Pulse Popliteal Pulse Dorsalis Pedis Pulse Posterior Tlbial Pulse All Pulses Leg Pulses 0.571 0.350 0.322 0.833 0.628 0.026 0.258 0.033 0.001 0.002 0.051 0.002 0.003 RH > C RH RH RH RH RH RH > > > > > > C C C C C C These specific data, when compared to the broader previous analyses in Table XVI-1-9, show decreasing P values. In addition, there is a suggestion that the peripheral pulse deficits are targeted in the older heavy smokers who are currently still smoking. c. Reported and Verified Heart Disease All participants were asked 2 questions during the in-home interview that were intended to capture a history of heart disease. The questions were: "Did you ever have a heart condition?" and "Did you ever have any other major health condition?" All affirmative responses were medically coded by the International Classification of Diseases, 9th Edition, Clinically Modified (ICD CM). Twenty-seven distinct cardiac classifications were identified for the Ranch Hand group and 19 were found in the comparison group. Medical records were sought on all of these individuals in order to verify the reported conditions. Table XVI-1-16 summarizes the verification results for the specific question on past heart disease. XVI-1-14 �'t ,,,a' Table XVI-1-16 MEDICAL RECORD VERIFICATION OF REPORTED HEART DISEASE Ranch Hand Group Original Comparison Group Number of reported cardiac conditions 139 Medical Records Reviewed 117 ?96 '•''fe.i -81 Medical Records Pending -22 ^tf- % Cardiac Conditions Verified 82.9 85.2 % Cardiac Conditions Unsupported 17.1 !l^«8 Overall, these data show a high confirmation proportion of reported cardiac conditions. Since Table XVI-1-16 does not include :results from the second overlapping question (Other major conditions?) and siri&%:individuals may have multiple heart disease responses, the following analyses have different numerators and denominators. All Ranch Handers (diabetics, Blacks, edemics included;) were contrasted to the original comparisons for reported heart disease and reported heart attacks. This analysis was supplemented by an analysis on verified heart disease and heart attacks; all these data are summarized in Table XVI-1-17. The unadjusted relative risk and 95% confidence interval for verified heart disease are 1.00 and (.79, 1.27). Table XVI-1-17 f : f."f RANCH HAND AND ORIGINAL COMPARISON GROUPS VERSUS REPORTED AND VERIFIED HEART DISEASE AND HEART:ATTACKS Heart Disease Parameter Ranch Hand Yes ^Jo Comparison Yes |> k Reported Reported Verified Verified 181 10 1 47 7 136 4 109 3 Heart Heart Heart Heart Disease Attack Disease Attack 864 1035 898 1038 XVI-1-15 ' 637 769 664 770 ; P Value 0.878 0.296 0.982 0.432 �While the lack of group differences in Table XVI-1-17 is of interest, and the good agreement between subjective responses and medically verified responses is notable, additional covariate analyses were conducted to rule out any hidden effect of a risk factor interaction that might be associated with group membership. Thus, Ranch Handers and their comparisons were again contrasted for reported heart disease and verified heart disease, adjusting for the covariates of age, smoking, body fat or HDL. As age was confounding for both reported and verified disease, the analyses are age specific. Further, there are significant interactions between smoking, group membership, and disease; these findings are shown in Table XVI-1H8. Table XVI-1-18 RANCH HAND AND ORIGINAL COMPARISON GROUP: COVARIATE ANALYSES OF REPORTED AND VERIFIED HEART DISEASE Adjusted Intergroup P Value and Direction of Association Parameter and Covariates Reported Heart Disease:* Body Fat, smoking <40 **£40, less than 10 pack-years >40, greater than 10 pack-years 0.530 0.0038 0.139 (RH = C) (RH < C) (RH = C) 0.506 0.008 0.0712 (RH = C) (RH < C) (RH > C) Verified Heart Disease* HDL, smoking <40 , less than 10 pack-years , greater than 10 pack-years *Age confounding variable **Group "- heart disease - smoking interaction: P = 0.0054 ***Group - heart disease - smoking interaction: P = 0.004? These data, in contrast to Table XVI-1-17, demonstrate associations of significance. Young Ranch Handers are equivalent to their young comparisons for both reported and verified heart disease; whereas, the older Ranch Handers smoking more than 10 pack-years are manifesting more verified heart disease than their counterparts. Conversely, older Ranch Handers smoking less than 10 pack-years are faring significantly better than their comparisons for both reported and verified heart disease. These associations, in light of essentially negative blood pressure and ECG findings at the physical examination, XVI-1-16 �could be speculatively attributed to a wide array of post hoc explanations: e.g., a true disease process that will evolve more clearly in the future, an enigmatic finding akin to the peripheral pulse deficits, chance, etc. d. Cardiovascular Examlnation Findings and Verifled Hlstorrloalr. Heart Disease The cardiovascular examination findings were contrasted to the history of cardiovascular disease as verified by detailed medical record review. The purposes of this analysis were to determine the degree of positive correlation between the examination and the past medical history, and to determine if peripheral pulse abnormalities were associated with known cardiovascular disease. These data are presented in Table XVI-1-19. Table XVI-1-19 ASSOCIATION OF CENTRAL AND PERIPHERAL CARDIOVASCULAR ABNORMALITIES WITH VERIFIED HEART DISEASE BY AGE: RANCH HANDERS VERSUS ORIGINAL COMPARISONS* (NON-BLACKS ONLY) Dependent^Variable P Value (Unadjusted) Dependent Variable l^^H^J^^^Ili^lJiE^^JPA?.®3^. Systolic Blood Pressure Diastolic Blood Pressure Electrocardiogram Heart Sounds Carotid Bruits Radial Pulse Femoral Pulse (£40) Posterior Tibial Pulse (>40) Popliteal Pulse (£40) Dorsalis Pedis Pulse (£40) All Pulses «40) (£40) Peripheral Pulses «40) (£40) Leg Pulses «40) (£40) <0.00001 <0.00001 <0.00001 0.292 0.084 0.023 0.14? 0.103 0.074 0.002 0.0004 ._ --..„ °-°007 n nno° °2 ' 03 *Pitting edema omitted for pulse analyses XVI-1-17 P Value (Adjusted for Age) Ranch Hand VersusComparison 0.229 0.391 0.875 0.316 0.223 0.152 0.104 0.082 0.022 0.094 °'205 0.0691 0.261 0.048 0.369 0.044 �Systolic, diastolic blood pressure and EGG abnormalities at physical examination showed exceptionally significant (P = 0) associations with medical record histories of cardiac disease, regardless of group membership or age. While moderately positive associations are to be expected, the unusual strength of the associations suggests that very few new cases of hypertension or EGG abnormalities were diagnosed at examination, reflecting perhaps, up-to-date medical records due to the overall medical sophistication and free access to medical care by most members of both groups. The association of carotid bruits and previously diagnosed cardiovascular disease was marginally positive but based upon small numbers. Table XVI-1-19 was most revealing for the peripheral pulse abnormalities. For the radial pulse, the data were too sparse for age adjustment but for all other pulse abnormalities, age was confounding, primarily due to a relative lack of abnormalities in the under-40 age group. A remarkably consistent observation in the 40-and-older age group was that significant or borderline significant Ranch Hand - comparison differences were found almost exclusively in those individuals without a history of cardiovascular disease. This uniform pattern is best exemplified by the popliteal pulse data, as shown in Table XVI-1-20. Table XVI-1-20 ASSOCIATION OF POPLITEAL PULSE ABNORMALITIES WITH VERIFIED HISTORY OF CARDIOVASCULAR DISEASE BY AGE AND GROUP MEMBERSHIP* History of Cardiovascular D i sease Group Membership Popliteal Pulse Findings in £40 Age Group Abnormal Normal Yes (Verified by record review) Ranch Hand Comparison 2 2 68 59 No Ranch Hand Comparison 11 0 404 313 Popliteal pulse by disease history: P = 0.074 Popliteal pulse by disease by group interaction: P = 0.022 *No pulse abnormalities in <40 group Interpretation of this intriguing finding at the baseline physical examination is not clear. The fact that the abnormal pulses, regardless of group membership, are associated with increased age, heavy past smoking, current smoking (and possibly race), and verified past heart disease and are largely substantiated by the use of 3 related denominators suggest that the finding is real rather than spurious. While there was most likely a tendency to diagnose additional abnormal pulses, given the first abnormal pulse, this XVI-1-18 �possible examination bias would not likely aggregate in the Ranch Hand group (because of the blind examination) nor in individuals without a history of prior cardiovascular disease. The speculative interpretation of concern is that the finding of substantial "subclinioal" peripheral pulse abnormalities (i.e., without a history of past cardiovascular disease) in the Ranch Handers may be a precursor to either clinically manifest arterial disease or central cardiovascular abnormalities. This possibility will receive detailed attention at the first follow-up examination because an analysis of onset times for verified heart disease (adjusted for race, occupation, and age) did not show a significant difference between the Ranch Hand and comparison group (P = 0.395). This finding suggests that if the observed pulse abnormalities are a precursor to central cardiovascular disease, this pathogenesis is not manifested by premature heart disease at this time. 5. Exposure Index Analyses All of the dependent variables within the Ranch Hand group were compared to the exposure index. Systolic and diastolic blood pressure elevations, and EGG, heart sound, and eyeground abnormalities were adjusted for age (<40, £40). The peripheral pulse analyses were not age adjusted because of sparse data; subjects with peripheral pitting edema were omitted from these comparisons. The exposure index was stratified into 3 categories: low, medium, and high. All analyses were performed on each of 3 occupational categories: officer, flying enlisted, and ground enlisted. This analysis is presented in Table XVI-1-21. Separate age analyses were performed when age was found to be a confounding variable. When some data were too small for valid analysis, the word sparse is written instead of a P value. Table XVI-1-21 SUMMARY OF EXPOSURE INDEX ANALYSES WITHIN THE , RANCH HAND GROUP* P Value Dependent Variable** Systolic Blood Pressure Occupation Adjusted for Age Age (***°Unadjusted for Age) <40 £40 Officer Flying Enlisted Ground Enlisted 0.731 Officer Flying Enlisted Ground Enlisted 0.313 EGG Officer Flying Enlisted Ground Enlisted 0.858 0.209 0.450 Heart Sounds Officer Flying Enlisted Ground Enlisted 0.397*** 0.395*** Diastolic Blood Pressure XVI-1-19 0.560 0.746 0.499 0.701 Sparse 0.739 0.567 0.214 0.255 0.638 �Table XVI-1-21 (Cont'd) SUMMARY OF EXPOSURE INDEX ANALYSES WITHIN THE RANCH HAND GROUP1 Dependent Variable** Eyegrounds OcGupatlon P Value Adjusted for Age Age (***=UnadJusted for Age) <40 £40 Officer Flying Enlisted Ground Enlisted 0.513 0.395*** Carotid Bruits Officer Flying Enlisted Ground Enlisted 0.616 0.992 0.094 Popliteal Pulse Officer Flying Enlisted Ground Enlisted Sparse 0.814 Dorsalis Pedis Pulse Officer Flying Enlisted Ground Enlisted 0.288 0.719 0.531 Posterior Tibial Pulse Officer Flying Enlisted Ground Enlisted 0.643 Sparse 0.654 All Pulses Officer Flying Enlisted Ground Enlisted 0.305 0.624 0.624 Peripheral Pulses Officer Flying Enlisted Ground Enlisted 0.338 0.784 0.746 Leg Pulses Officer Flying Enlisted Ground Enlisted 0.350 0.784 0.882 0.255 0.638 Sparse *Peripheral edema omitted for peripheral pulse analyses **Radial and femoral pulses omitted; data too sparse ***Unadjusted for age. The data in Table XVI-1-21 clearly indicate that there is no detectable association between the herbicide exposure index adjusted by occupational category and any of the cardiovascular variables. XVI-1-20 �6. Summary Central cardiovascular system abnormalities, as manifested by elevated systolic or diastolic blood pressure, abnormal ECG's, and abnormal heart sounds, showed no statistically significant Ranch Hand - comparison group differences, but did reflect a strong correlation to increased age and, to a lesser degree, heavy past smoking. The 3 risk factors of age, smoking, and cholesterol were strongly associated with each other. Unadjusted analyses of Blacks were essentially negative. The prevalence of funduscopic abnormalities and carotid bruits was not associated with group membership but was significantly dependent upon age. Abnormal peripheral pulses were associated with the Ranch Hand group. A series of detailed covariate analyses showed that pulse abnormalities, regardless of group membership, were associated with increased age (S^O years), heavy past smoking, current smoking, and a verified history of past cardiovascular disease. Substantial Ranch Hand pulse abnormalities were also found in members without prior cardiovascular disease. All significant or borderline significant pulse findings in the Ranch Handers were largely sustained regardless of the comparison group used (originals, matched originals, or all comparisons). Both the femoral and carotid pulses revealed substantial, but statistically nonsignificant, abnormalities in the Ranch Hand group. More biologic credence is assigned to the large artery observations in light of the small artery findings. Peripheral pulse abnormalities will merit extensive clinical inquiry at the first follow-up examination. The history of cardiovascular disease obtained during the in-home interview was verified by a review of medical records. Both reported and verified past heart disease and heart attacks were adjusted by age, smoking, and body fat or HDL. This analysis revealed that the older (£*IO years) smoking Ranch Handers manifested significantly more verified heart disease than their equivalent comparisons. Alternatively, the older less smoking Ranch Handers have substantially less reported and verified cardiovascular disease than their comparisons. Detailed herbicide exposure analyses showed no associations to any of the central or peripheral cardiovascular findings. Future reports will explore a theoretical synergism between cigarette smoking and herbicide exposure. XVI-1-21 �Chapter XVI-2 IMMUNOLOGY 1. Introduction Recent experimental data in animals have suggested that TCDD has deleterious effects on the immune system (Dean et al, 1984). As a result, the Science Panel Committee recommended that the immunotoxic potential of TCDD be evaluated during the physical examination portion of this study. Parameters selected for assessment included: (1) the enumeration of T-lymphocytes, T-lymphocyte subsets and B-lymphocytes using monoclonal surface marker analysis and (2) functional ability of lymphocyte to respond to selected antigen or mitogen stimuli in the lymphocyte transformation assay. Five hundred ninety-two participants were randomly selected for this examination using the terminal digit of the participant's case number. This selection occurred during the time period March 1982 through September 1982. Of the 592 participants, 297 were Ranch Handers and 295 were comparisons. Of the 295 comparisons, 180 were original comparisons. The statistical testing presented in this chapter is all based on this basic set of 297 Ranch Handers and 180 original comparisons. However for each test performed, differing data deletions occurred. Specifically, data from professed homosexuals were removed from all analyses. Also, data were removed from all analyses if covariate information (age, smoking, alcohol use) was missing. Finally, data were removed from certain analyses (T-j-j, T^, Tij, TQ, Tij/Tg, BI counts and percentages) if: (1) differential counts were unavailable, (2) if samples exhibited greater than 30% background fluorescence, or (3) if samples had a T^ or T-) 1 proportion of less than 10$. Surface marker analysis and lymphocyte function studies were performed on purified mononuclear cells obtained from heparinized whole blood drawn at Kelsey-Seybold Clinic early on the second day of the examination period. Peripheral blood mononuclear leukocytes (PEL) were separated from erythrocytes and polymorphonuclear leukocytes using a density gradient centrifugation technique. Unfortunately, blood specimens were collected and processed in glass tubes with resultant variable loss of adherent PEL. White cell differential counts were not obtained on purified PEL so that the number of lymphocytes actually placed into functional assays could not be ascertained. Due to these laboratory difficulties, coupled with relatively small sample sizes, exposure index analyses are not provided in this chapter. 2• Analysis of I_mmunological^ Cell Count_JData Mouse monoclonal antibodies directed against various lymphocyte surface antigens were incubated with PEL. Following washing, fluorescent anti-mouse antibodies were added. After the cells had been stored for a variable period in paraformaldehyde, the presence or absence of fluorescent antibody on each PBL was determined and counted using a cytofluorograph. The percentage of cells positive for each surface marker is reported as the number of fluorescent XVI-2-1 �cells divided by the total number of lymphocytes in a given specimen. Since differential counts were not obtained on the purified PEL, a 250 cell differential count was performed at the recommendation of the Peer Review Committee on paraformaldehyde-fixed cells. These cells had been stored for 6 to 12 months. Although cell morphology was not optimal, determination of the percentage of lymphocytes in each specimen was possible. The number of surface marker positive cells per mm3 was calculated by multiplying the percent marker positive cells by the total lymphocyte count. The cells counted and analyzed for this report are classified as having TJI , Tg, Ti|, TQ, or BI cell surface markers. The T|-| surface marker identifies thymus dependent lymphocytes which form rosettes with sheep erythrocytes (also called E+ cells). The Tj surface marker is found on nearly 100$ of circulating T-lymphocytes cells (Reinherz and Schlossman, 1980). Cells with Tj| cell surface markers proliferate in response to soluble antigens and have an inducer or helper function in T-T, T-B and T-macrophage interactions (Reinherz and Schlossman, 1980). JQ cells have cytotoxic and suppressor functions (Reinherz and Schlossman, 1980). B1 cells, or bursa equivalent cells, are producers of immunoglobulins (David, 1979). The number of TII, Tj., Tij, TQ, and B-| positive cells per mm3 are provided below by group, along with the TII/TS ratio and total lymphocyte count. Additionally, percentages of T-||, T^, TH, TS, and B-| positive cells are reported by group. The data were analyzed for statistically significant group differences using the Kolmogorov-Smirnov Two Sample Test. Also, crude group (Ranch Hand versus comparison) means were contrasted, and then the groups were contrasted while adjusting for age, smoking history in pack-years and alcohol intake measured as drink-years. The literature does not yet provide clear guidance to the selection of covariates for analysis as attempted here. Age, smoking and alcohol were chosen based on the observation that these variables frequently correlate with general measures of health and impact upon hematologic parameters. Group interactions with age, smoking or alcohol indicate group differences associated with these covariables. When group-covariate interaction is observed, group and associated covariate main effects are not reported, rather the interaction is detailed. The probability level used to indicate an interaction of interest is P = 0.100. In the absence of interaction, group and covariate main effects are reported in the usual manner. When P > 0.100 for all interaptions, P values for the reduced model, consisting of main effects only, are provided. Table XVI-2-1 provides the results of Kolmogorov-Smirnov testing of the number of surface marker positive cells per mm3. A borderline statistical difference is seen in the B-| count with Ranch Handers having lower values. However, BI cells are an adherent set of cells. The purification process resulted in a variable loss of adherent cells, therefore, this data must be interpreted with extreme caution. Table XVI-2-2 provides the KolmogorovSmirnov testing of cell percentages and no statistically significant differences are observed. Table XVI-2-3 provides unadjusted means for the number of surface marker positive cells per mm3. No statistically significant group mean XVI-2-2 �differences are observed. Table XVI-2-4 provides unadjusted means for the cell percentages, and again no statistically significant group mean differences are observed. Both counts and percentages are provided to aid with interpretation. Table XVI-2-1 KOLMOGOROV-SMIRNOV TESTING OF NUMBER OF SURFACE MARKER POSITIVE CELLS (THOUSANDS/mm3) Percentiles Variable Group N W% 50% 9055 Tn COMP RH 144 235 0.77 0.70 1.23 1.25 2.02 1.96 0.74 T3 COMP RH 0.73 0.70 1.28 1.27 2.13 1.96 0.39 233 147 231 0.48 0.398 0.78 0.794 1.42 1.251 0.81 147 235 0.277 0.296 0.604 0.569 1.168 0.985 RH 147 231 0.64 0.64 1.38 1.41 2.62 2.70 0.78 B-j COMP RH 147 235 0.022 0.023 0.071 0.071 0.247 0.188 0.097 TLC COMP RH 177 290 1.35 1.34 1.91 1.92 2.74 2.54 0.63 TH COMP RH TS COMP RH TH/TS COMP COMP = comparison group RH = Ranch Hand group XVI-2-3 �Table XVI-2-2. KOLMOGOROV-SMIRNOV TESTING OF PERCENTAGE. QF SURFACE MARKER POSITIVE CELLS (THOUSANDS/DP3) Percentiles Variable N ' 1 0% 50$ 90$ 66,0 68,0 87.5 88.4 0.90 P Value Tn COMP RH 144 235 42,0 41.6 T 3 COMP RH 144 233 48,5 48.4 66,5 66.0 83.5 83.6 0.79 T4 COMP RH 147 231 26.8 23.0 42,0 44.0 58.0 61,0 0.45 T 8 COMP RH 147 235 17.8 16.6 31,0 29,0 47.0 49.0 0,82 B, COMP RH 147 235 1.0 1.0 3.0 4.0 13.2 10.4 0.48 . COMP = comparison, group RH = Ranch Hand group XVI-2-4 �Table XVI-2-3 UNADJUSTED MEANS FOR NUMBER OF SURFACE MARKER POSITIVE CELLS (THOUSANDS/mm) AND P VALUES FOR TESTS BETWEEN GROUPS MEANS Group N COMP 139 228 1.33 1 .29 0.050 0.034 0.47 139 226 1.36 1 .29 0.052 0.031 0.21 142 224 0.877 0.846 0.038 0.027 0.49 142 228 0.660 0.606 0.029 0.020 0.11 RH 142 224 1.54 1.65 0.075 0.075 0.34 B1 COMP RH 142 228 0.117 0.102 0.011 0.008 0.26 TLC COMP RH 171 280 2.00 1.92 0.046 0.028 0.14 Variable TH RH T3 COMP RH Ti| COMP RH Tg COMP RH TH/TS COMP RH TLC SEM COMP = = = = Unadjusted Means comparison group Ranch Hand group total lymphocyte count standard error of the means XVI-2-5 SEM P Values �Table XVI-2-4 UNADJUSTED MEANS FOR PERCENTAGE OF SURFACE MARKER POSITIVE CELLS AND P VALUES FOR TESTS BETWEEN GROUPS MEANS Variable Tn Group N Unadjusted Means SEM Ti, Tg 139 65.0 1.44 228 65.7 1.20 COMP 139 65.6 1.22 RH T3 COMP RH 226 65.1 0.97 COMP RH 142 224 42.1 43.1 1.13 1.07 COMP 142 32.0 1 .02 RH B1 P^ Values 228 30.8 ' Qf71 0 75 0.53 0.80 COMP 142 5.80 0.50 RH 228 5.35 0 36 0.41 0>i)8 COMP = comparison group RH = Ranch Hand group SEM = standard error of the means Table XVI-2-5 provides the adjusted surface marker positive cell count means, along with P values for main (group, age, smoking and alcohol) and interaction (group by age, group by smoking, and group by alcohol) effects. No main or interaction effect associated with group is noted to be statistically significant. The number of lymphocytes and Tg positive cells per mm3 decreased with increasing age in both the Ranch Hand and comparison groups. The effect was -0.0043 thousand cells per mm3 per year of life for Tg and was -0.0110 thousand cells per mm3 per year of life for the lymphocyte count. Smoking was observed to be associated with increased cell counts on all variables except for B-) positive cells. Specifically, the slope was 0.0036 thousand cells per mm3 per pack-year for T-\ 1 ; 0.0076 thousand cells per mm3 per pack-year for T3; 0.0070 thousand cells per mm3 per pack-year for T4; 0.0022 thousand cells per mm3 per pack-year for Tg; and 0.0083 thousand cells per mm3 per pack-year for total lymphocyte count. XVI-2-6 �Table XVI-2-5 ADJUSTED MEANS, PLUS MAIN AND INTERACTION P VALUES FOR THE NUMBER OF MARKER POSITIVE CELLS (THOUSANDS/mm3) P Values for P Value A for Gp x Gp x Gp x | Alco Age Smkng Alco Adj ' d Adj ' d Age Smkng Mean Means Effect Effect Effect Effect Effect Effect Variable Group (Gp) T11 COMP 139 RH 228 1.29 T 3 COMP RH 139 1.35 0.38 226 1.30 <0.001 Til COMP RH 142 0.864 0.82 224 0.854 <0.001 T8 COMP RH 142 228 T4/T8 COMP RH 142 1.52 224 1.66 COMP RH 142 0.117 228 0.102 COMP RH 171 1.99 280 1.92 TLC N 1.33 0.52 0.660 0.12 0.606 0.057 0.029 0.025 0.22 0.27 0.20 <0.001 <0.001 COMP = comparison group RH = Ranch Hand group = P > 0.050 for main effects or P > 0.100 for interactions. When P > 0.100 for all interactions, P values for the reduced model, consisting of main effects only, are provided. TLC = total lymphocyte count Table XVI-2-6 shows adjusted means for percentage of surface marker positive cells. No statistically significant overall group differences are observed. The T^ and TH percentages are influenced by smoking, but this effect is essentially the same in both study groups. The effect of smoking on the Tg percentage is 0.124 percentage points per pack-year, while the effect of smoking on the TH percentage is 0.171 percentage points per year. A weak indication of a group specific alcohol intake effect was noted on the T-|-| percentage. The association of alcohol use with the percentage of T-)-| positive cells was 0.0980? per drink-year in the comparison group and -0.0042$ per drink-year in XVI-2-7 �the Ranch Hand group. This pattern could reflect a diminished Ranch Hand immunological response to drinking in reference to the comparisons; the biological relevance of this borderline finding is uncertain at this time. Table XVI-2-6 ADJUSTED MEANS AND OTHER MAIN AND INTERACTION EFFECTS FOR PERCENTAGE OF SURFACE MARKER POSITIVE CELLS Variable Group (Gp) T11 COMP P Value for Gp x Gp x Gp x Adj'd Adj'd Age Smkng Alco Age Smkng Alco Mean Means Effect Effect Effect Effect Effect Effect * RH 139 226 * - COMP RH 139 226 65.2 65.4 0.92 - 0.005 COMP RH 41.6 43.4 0.27 - <0.001 224 COMP RH 142 32.0 0.34 228 30.7 COMP RH 142 228 0.087 5.79 0.52 5.36 COMP = comparison group = Ranch Hand group RH * = that a group interaction effect was noted rendering overall group mean differences and the associated main effect not meaningful. = P > 0.050 or P > 0.100 per footnote in Table XVI-2-3. In summary, the lymphocyte surface marker analyses reported in Tables XVI-2-5 and XVI-2-6 show no detectable differences between the Ranch Hand arid comparison groups on these measures, except possibly for the borderline group difference in the T-|-| percentage by alcohol use association. 3. T and B Cell Functional Studies T and B lymphocyte function was determined by measuring the ability of these cells to transform in response to antigen or mitogen stimuli. Briefly, this assay is performed by culturing PBL in the presence of mitogens (plant lecthins which stimulate the cells to divide) or antigen. After a certain length of incubation time, the rate of DNA synthesis is estimated by adding tritiated thymidine (a radioactive DNA precursor). Thus, the counts per minute XVI-2-8 �of thymidinc incorporated into the cell culture is a measure of the ability of those lymphocyte to proliferate in response to the added stimulus. Mitogona stimulate lymphocytes non-specifically. Phytohemagglutin (PHA) and concanavallin A (conA) stimulate T-lymphocytes to divide, while pokeweed mitogen (PW) stimulates B-lymphocytes through a T-lymphocyte. On the other hand, antigen require that lymphocytes recognize specifically antigen as a substance to which the host has been exposed. Tetanus toxoid (TT) is a T-lymphocyte dependent B-lymphocyte recall antigen. Kolmogorov-Smirnov testing of the 4 stimulation and 2 control measurements are shown in Table XVI-2-7. No statistically significant group differences are noted. Unadjusted group mean net counts per minute for the stimulation studies and control measurements are shown in Table XVI-2-8. No statistically significant group differences are noted except in Control #1 where the Ranch Hand group was found to have a lower unstimulated proliferation rate. A comparable differential is also noted in Control #2, but is not statistically significant. The group differences noted are of unknown biological significance. Table XVI-2-7 KOLMOGOROV-SMIRNOV TESTING OF T AND B CELL FUNCTIONAL STUDIES 1035 Percentiles 50% 90% P Value 168 279 138 140 448 374 1483 1320 0.20 COMP RH 168 279 13596 17741 58394 54190 99104 91724 0.38 After PHA COMP RH 168 279 30143 33027 84339 79342 135684 130064 0.51 Control #2 COMP RH 168 142 132 404 388 1079 917 0.85 After PW COMP RH 168 274 12232 12700 27916 29623 53662 58288 0.64 After TT COMP RH 168 274 1001 866 3719 3726 16058 13979 0.81 Group N Control #1 COMP RH After conA Variable COMP = comparison group RH = Ranch Hand group XVI-2-9 �Table XVI-2-8 UNADJUSTED MEANS FOR T AND B CELL FUNCTIONAL STUDIES BY GROUP, AND P VALUES FOR TESTS BETWEEN GROUP MEANS Unadjusted Means (nCPM) P Value for Unadj' d Means Variable Group Control #1 COMP RH 163 269 652 535 After conA COMP RH 163 269 57454 54637 2248 1658 0.31 After PHA COMP RH 163 269 83808 80433 3048 2244 0.37 Control #2 COMP RH 163 264 523 480 After PW COMP RH 163 264 32092 33710 1337 1151 0.37 After TT COMP RH 163 264 6848 7051 650 787 0.86 COMP RH. nCPM SEM = = = = SEM 49.2 29.4 37.1 23.9 0.031 0.31 comparison group Ranch Hand group net counts per minute (stimulated CPM - control CPM). standard error of the mean Table XVI-2-9 shows adjusted net CPM means. A statistically significant group difference is noted in Control #1. Other group effects are noted as interactions with smoking and alcohol. Specifically, smoking was associated with a decreased proliferation rate to concanavallin A stimulation,(-113 nCPM per pack-year) in the comparison group, while smoking was associated with an increased proliferation rate in the Ranch Hand cohort (+169 CPM per pack-year). Two comparable group differences were observed as interactions of concanavallin A and phytohemagglutinin stimulation with alcohol use. Alcohol use was associated with an increased proliferation after concanavallin A stimulation in the comparison group (+212 CPM per drink-year), while an increase of 12 CPM per drink-year was found in the Ranch Hand cohort. Alcohol use in the comparison group increased proliferation after phytohemagglutinin by 167 CPM per drinkyear, while alcohol use in the Ranch Hand group decreased proliferation by 76 CPM per drink-year. This alcohol effect has no known biologic explanation. The finding is of questionable significance and will need to be examined further in subsequent immunologic analyses. XVI-2-10 �In addition to these group specific effects, some effects not associated with group were observed. Age and smoking were covariates which were found to be highly statistically significant. Lymphoproliferative responses to phytohemagglutinin and concanavallin A decreased monotonically in both Ranch Hand and comparison groups with advancing age. Lymphocyte response to pokeweed mitogen increased with increasing pack-years in both Ranch Hand and comparison groups. Table XVI-2-9 ADJUSTED MEANS, PLUS MAIN AND INTERACTION P VALUES FOR T AND B CELL FUNCTIONAL STUDIES BY GROUP P Value for Gp x Gp x Gp x Adj'd Adj'd Age Smkng Alco Age Smkng Alco Mean Means Effect Effect Effect Effect Effect Effect Variable Group (Gp) Control #1 COMP RH 163 269 After conA COMP RH 163 269 After PHA COMP RH 163 269 Control #2 COMP RH 163 264 After PW COMP •RH 163 264 31982 0.32 33778 After TT COMP RH 163 264 6929 0.95 7001 657 0.023 532 * # <0.001 0.089 0.025 * * 518 484 » x <0.001 0.041 0.41 0.01 COMP = comparison group RH = Ranch Hand group = P > 0.050 or P > 0.100 per footnote in Table XVI-2-3. 4. Summary The analysis of these data has provided a valuable insight into the rapidly changing area of clinical immunology. Analysis has revealed no statistically significant differences in mean TII, T^, Tij, TQ, TIJ/TS ratio or B-| 1 counts between the Ranch Hand and comparison groups. Similarly, there were no statistically significant overall mean differences in PHA, conA, PW, or TT stimulation responses between the groups. There were significant differences in XVI-2-11 �unstimulated (control) thymidine incorporation (P = 0.023) with less activity in the Ranch Hand group. In both groups, lymphoproliferative responsiveness to PHA and conA decreased significantly with increasing age, and total lymphocyte counts were correlated with age and smoking. The subsets of T-lymphocytes Ti), TQ, and T-| 1) also were correlated with smoking. From the clinical vantage point, the immunological findings do not present a picture indicative of immunological alteration in the herbicide-exposed group. However these data are of such quality that concern must be taken for a possibility of both false positive and false negative statements. Due to previously defined difficulties in surface marker analyses and lymphocyte stimulation assays, these data cannot be reliably referenced to other published data. Nonetheless, no gross adverse immunological effects were noted between the herbicide-exposed group and the comparison group. XVI-2-12 �Chapter XVI-3 HEMATOLOGICAL VARIABLES In this section, 8 hematological variables are reported. These 8 variables are listed in Table XVI-3-1 along with abbreviations used, units of measure, and normal ranges employed in the analyses. Ranch Hand-comparison group differences have been analyzed using general linear models with all variables except the group indicator treated as continuous variables. Group differences have also been evaluated using log-linear models with all variables treated as categorical. In both the general linear and log'-linear model analy^ ses, the hematological variables were adjusted for smoking history available from the questionnaire as pack-years of cigarette use (Wintrobe, 1974). In the general linear models analyses, pack-years were used directly as a continuous variable. In the log-linear models, smoking history was treated as a tricotomous variable by grouping together: (1) nonsmokers, (2) smokers with 10 packyears or less contact, and (3) smokers with greater than 10 pack~years cigarette smoking. Also, in the log-linear models analyses, the dependent (hematologic) variable was dichotomized as normal (within range) or abnormal (out of range). Analyses using the exposure index were also accomplished using the Ranch Hand participant data. These within-group analyses were performed in much the same manner as the Ranch Hand-comparison group contrasts, except that in the within-group analyses, exposure category took the place of the cohort indicator. Data on all Ranch Hand and original comparison participants are presented in this section. Table XVI-3-1 HEMATOLOGICAL VARIABLES STUDIED Variable Name Abbreviation Red Blood Cell Count RBC White Blood Cell Count WBC Hemoglobin Hgb Hematocrit Mean Corpuscular Volume Mean Corpuscular Hemoglobin Mean Corpuscular Hemoglobin Concentration Platelet Count Units Of Measure Normal Range Hot MCV Million per Cubic mm Thousand per Cubic mm Grams per 100 ml ml/100 ml Cubic Micra 42.0 - 52.0 80.0 - 101 .0 MCH Micromicrogram 27.0 - 31 .0 MCHC Percent 32.0 - 36.0 PLT Thousands Per Cubic mm 150 - 450 XVI-3-1 4.6 *- 6.2 4.8 - 10.8 14.0 - 18.0 �Table XVI~3~2 provides the results of the Ranch Hand - comparison group contrasts. The abbreviation CC is used to denote linear model analyses on continuously distributed data, DD denotes categorical log linear analyses. Two group differences are seen in Table XVI-3-2. The Ranch Hand group has a statistically significantly larger red blood cell corpuscular volume than does the comparison group (P = 0.05 in the CC analysis) and, perhaps paralleling this finding on corpuscular volume, the Ranch Hand group has a larger mean corpuscular hemoglobin (P = 0.04 in the CC analysis). In performing these analyses of group differences, smoking history was an important variable in essentially all instances. All of the hematological variables except RBC and MCHC increase with cigarette use. A summary of P values and slopes is provided in Table XVI-3~3In Table XVI-3-1* analyses are provided within the Ranch Hand group, examining for differences between exposure categories. Sample sizes in these analyses are provided in Table XVI~3i-5. Table XVI-3-6 provides variable means and percents by occupation and exposure group. Table XVI-3-2 P VALUES FOR RANCH HAND-COMPARISON GROUP DIFFERENCES, ADJUSTED MEANS, AND ABNORMAL PERCENTAGES Var Anal Group Pack-yr Group x Pack-yr RH Adj'd Mean Comp. Adj'd Mean RH ABN % Com ABN % RBC CC DD 0.62 0.36 0.08 0.65 0.71 5.20 NA 5.21 NA NA 7.43 NA 6.28 WBC CC DD 0.14 0.62 <0.001 0.48 0.83 7.51 NA 7.38 NA NA 12.45 NA 11.65 HGB CC DD 0.15 0.97 <0.001 0.77 0.65 16.04 NA 15.97 NA NA 3.28 NA 3.27 HCT CC DD 0.23 0.62 <0.001 0.25 0.32 46.16 NA 46.01 NA NA 8.30 NA 7.59 MCV CC DD 0.05 0.70 <0.001 0.58 0.71 89.04 NA 88.60 NA NA 3.76 NA 3.40 MCH CC DD <0.001 0.04 0.005 0.73 0.64 30.83 NA 30.66 NA NA 46.24 NA 39.66 MCHC CC DD 0.63 0.47 0.005 0.15 0.84 34.68 NA 34.66 NA NA 94 .6 NA 10.47 PLT CC DD 0.06 0.16 <0.001 0.76 0.33 276.74 NA 271.48 NA NA 1.16 NA 1.97 XVI-3-2 �Table XVI-3-3 SMOKING EFFECTS ON HEMATOLOGIC VARIABLES AS SEEN BY CONTINUOUS VARIABLE LINEAR MODELS P Value for Smoking Effect Dependent Variable Smoking Slope (Units/Pack-yr) RBC 0.08 -0.00089 WBC <0.001 0.0389 HGB <0.001 0.00743 HCT <0.001 0.0266 MCV <0.001 0.0675 MCH <0.001 0.0200 MCHC 0.005 "0.00376 PLT <0.001 0.322 Variable XVI-3-3 �Table XVI-3-4 P-VALUES FOR RANCH HAND OCCUPATION AND EXPOSURE GROUP ANALYSES Var Occ Cat Generalized Linear Model Exp Pack-yr Exp X Pack-yr Effect Effect Log Linear Model Exp Pack-yr Exp X Effect Effect Pack-yr 0.83 0.35 * 0.66 0.50 0.09 0.59 0.22 <0.001 0.26 0.09 0.52 0.51 0.85 0.47 0.06 0.69 0.62 0.75 0.74 <0.001 0.07 0.03 0.68 0.40 0.38 0.27 0.19 0.46 0.32 0.13 0.23 0.56 0.48 0.26 0.77 0.50 0.19 <0.001 0.001 0.008 0.37 0.22 0.23 * * 0.19 * * 0.28 0.01 0.06 0.49 OFF ENL F. ENL G. 0.38 0.84 0.45 <0.001 <0.001 <0.001 0.58 0.18 0.19 0.98 0.83 0.39 0.93 0.84 0.45 0.18 0.61 0.49 MCH OFF ENL F. ENL G. 0.38 # 0.84 <0.001 # <0.001 0.84 0.08 0.51 0.05 0.47 0.99 0.04 0.01 0.05 0.43 0.38 0.47 MCHC OFF ENL F. ENL G. 0.24 0.77 0.65 0.01 0.003 0.73 0.32 0.59 0.55 0.03 0.88 0.39 0.08 0.08 0.60 0.97 0.63 0.17 PLT OFF ENL F. ENL G. 0.66 0.26 0.97 0.02 <0.001 0.91 0.56 0.17 0.71 0.30 0.24 0.32 0.95 0.93 0.88 0.99 0.95 0.58 OFF ENL F. ENL G. 0.69 * 0.06 OFF ENL F. ENL G. * 0.61 * HGB OFF ENL F. ENL G. HCT RBC 0.83 # 0.13 0.53 0.03 0.26 <0.001 * 0.37 0.59 0.08 OFF ENL F. ENL G. MCV WBC * * *P values not relevant due to Exposure by Pack-year interaction term. XVI-3-4 �Table XVI-3-5 SAMPLE SIZES FOR RANCH HAND OCCUPATION AND EXPOSURE GROUP ANALYSES Occupational Category Exposure Category Officer Enlisted FlyIng Enlisted Ground Low 111 56 150 Medium 128 58 178 High 125 65 146 In Table XVI-S-iJ, 2 statistically significant (P < 0.05) overall exposure group effects are seen and 7 exposure-smoking interaction effects (P £ 0.10) are also present. First, the overall exposure group effects will be described. The 2 overall exposure group effects occur in the Ranch Hand officer cohort and involve the variables MCH and MCHC. An increasing dose-response relationship is clear in the MCH data, and the high exposure group also has the highest rate of mean corpuscular hemoglobin concentration (MCHC) abnormalities. These findings are suggestive of a herbicide effect, however, similar trends are not noted in the other 2 occupational categories thus decreasing the likelihood of a bonafide herbicide effect by raising the possibility that an unknown confounding variable is operative. XVI-3-5 �Table XVI-3-6 HEMATOLOGICAL VARIABLE MEAN AND PERCENTS FOR RANCH HAND OCCUPATION-EXPOSURE GROUP ANALYSES Var Adjusted Variable Means ENL ENL Ground Exp Level OFF Flying Percent Abnormal** ENL ENL Officers Flying Ground 5 .23 5 .34 5 .27 8.11 9• 38 8.80 8.93 8.62 6.45 4.00 6.74 8.22 7.63* 7.66* 7.81* 11 .71 7.03 10 .53 16 .07 13 .79 21 .54 12.00 14.04 12.33 16 .04 16 .26 16 .09 3 .60 2.34 0 .80 3 .57 1 .72 9.23 4.67 2.25 4.11 45.36 46 .22 45. 40 46 .42 45.59 46 .84 46 .36 46 .82 46 .39 11 .71 10 .94 11 .20 8.93 8.62 10 .77 8.00 3-37 6.16 MCV Low 89.02 90 .09 Medium 89.84 89 .53 89.56 89 .70 High 88 .75 88 .10 88 .46 3 .60 3 .91 4.00 3 .57 1 .72 3.08 2.67 5.06 3.42 MCH Low Medium High 30 .61 30 .50 30 .56 41 .44 52 .34 58 .40 46 .43 44 .83 53 .85 40.67 41.01 42.47 MCHC Low 34.80 34.56 Medium 34.73 34.65 High 34.94 34.52 34.54 34.66 34.61 9.91 6.25 16.00 8.93 6.90 7.69 10.67 6.74 9.59 PLT Low 262.13 294.48 Medium 268.20 290.97 High 264.09 277.78 280.94 282.09 282.53 0.00 1.56 0.00 3.57 0.00 0.00 2.67 0.56 1.37 RBC Low Medium High 5.11 5.07 5.11 WBC Low Medium High 7.03* 8.25 6.93* 7 .91 7.15* 7.89 HGB Low Medium High 15.82 15.80 15.95 15 .99 16 .11 16 .19 HCT Low Medium High 5 .15* 5 .20* 5 .24* 30.94 31 .08* 31.14 30 .97* 31.22 30 .91* *Unadjusted means interaction. given due to smoking (pack-years) by dependent variable **A11 percents given are unadjusted. XVI-3-6 �The general linear model analysis of the red blood cell count shows an interesting interaction with smoking in Ranch Hand enlisted flying personnel. In the low exposure set of enlisted flying Ranch Handers, smoking cigarettes is associated with increasing RBC values (slope = 0.00562), but the medium exposed and high exposed individuals show decreasing RBC values with smoking (slopes -0.00124 and -0.00457 respectively). This gradient of slopes with exposure is suggestive of a true herbicide effect. Log-linear analysis of the red blood cell count shows a smoking-exposure interaction among Ranch Hand officers. The data for these officers is given in Table XIV-3-7. Table XVI-3-7 SMOKING-EXPOSURE INTERACTIONS ON RBC IN RANCH HAND OFFICERS % ABNORMAL RBC Exposure Zero Pk-Yrs 1HO Pk-Yrs >10 Pk-Yrs Low Med High 0.00 8.51 9.52 16.67 10.53 5.88 13.16 9.68 9.09 . This interaction is compatible with an herbicide effect, and reinforces the finding in the enlisted flying personnel. The WBC count in Ranch Hand officers shows a smoking-exposure interaction (P <0.001). In the low exposure officer set, cigarette use is associated with an increased WBC value (slope = 0.0691), but this association is less in the higher exposure categories (slope in medium exposure category = 0.0251, and slope in the high exposure category is 0.0307). These data suggest that the correlation of leucocyte count to cigarette smoking might be affected by herbicide exposure in Ranch Hand officers. This pattern of decreasing association of leucocyte counts to cigarette smoking with increasing exposure is also suggested by the data for Ranch Hand enlisted ground personnel. In the low exposure set, cigarrette use is also associated with increased WBC values (slope = 0.0466) but this association is least in the high exposed group (slope = 0.0192). An exposure - pack-year interaction in the HCT data was noted in the officer cohort (P = 0.01) and an interaction was also seen in the enlisted flying group. The data describing these interactions is shown in Table XIV~3~8. Relatively smooth dose-response trends are seen in each officer smoking category, but the same regularity is not apparent in the enlisted flying group. It is of interest that the HCT pattern seen in the officer data of Table XIV-S-S appears to parallel the RBC pattern in the officer data of Table XIV-3-7. XVI-3-7 �Table XIV-3-8 SMOKING-EXPOSURE INTERACTIONS ON HOT IN RANCH HAND OFFICERS AND ENLISTED FLYING PERSONNEL Occupation Exposure Zero Pk-yr Abnormal HCT 1-10 Pk-yr >10 Pk-yr Officers Low Med High 6.12 8.51 23.81 16.67 15.79 11.76 15.79 11.29 3.03 Enlisted Flying Low 37.50 0.00 5.00 Med 0.00 33-33 5.13 High 18.18 16.67 7.U Lastly, a smoking-exposure interaction is seen in the MCH data in the flying enlisted group. In the low exposure group the MCH - pack-year slope is -0.00^78, while this slope is positive in the medium and high exposure sets (0.0207 and 0.03083 respectively). Summary and Conclusions The ranch hand group has a higher mean corpuscular volume and mean corpuscular hemoglobin than does the comparison group. Also, a dose-response pattern of increasing mean corpuscular hemoglobin and mean corpuscular hemoglobin concentration was found in the Ranch Hand officer cohort. Seven hematologic variable by cigarette use by exposure level interactions were also found. Five of these interactions involved decreasing associations of hematologic measures with smoking with increasing exposure levels. One interaction (for MCH) showed increasing associations with smoking at increased exposure levels, and one interaction was uninterpretable. These statistical findings display some degree of consistency. However, the statistical differences do not appear to be significant in terms of current medical morbidity. XVI-3-8 �Chapter XVI-4 PULMONARY FUNCTION AND DISEASE Bronchitis, cough, dyspnea and acute respiratory irritation and distress have been reported as acute effects following exposure to phenoxy herbicides and dioxin (Berwick, 1970; Bauer et al, 1961; Bashirov, 1969). Little is known about the presence or absence of chronic pulmonary disease following herbicide exposure. These acute effects and the high likelihood of inhalation exposure to herbicide among operation Ranch Hand personnel in Vietnam prompted the evaluation of the pulmonary status of the study participants. In-home questionnaire responses concerning history of pulmonary disease were reviewed to determine the history of reported pulmonary disease in the Ranch Hand and comparison groups. The analysis of past pulmonary disease included data from the total comparison group. All other analyses in this subchapter were performed on all Ranch Hand individuals (1045) and the subset of original comparisons (773) who participated in the physical examination, except for a few individuals omitted due to missing pulmonary function data. Table XVI-4-1 presents the distribution of reported pulmonary disease in the Ranch Hand group, the entire comparison group, and in the subset of original comparisons. Table XVI-4-1 DISTRIBUTION OF REPORTED PULMONARY DISEASE IN THE RANCH HAND AND COMPARISON GROUPS Group Original Comparison Ranch Hand Total Comparison Tuberculosis and fungal infection (010-018; 114-116) 9 11 10 Pneumonia and Acute infections (480.-487; 460-466) 10 6 11 1 3 2 Diagnosis (ICD-9 Code) Neoplasia (160-165; 212) Chronic sinusitis and other upper respiratory disease (470-478; 480-519) 426 689 \ \ / / P=0.20 687 \ \ / / P=0.63 The distribution of reported disease is not significantly different between the Ranch Hand group and either the original comparisons or the entire comparison group. XVI-4-1 �Two measures of pulmonary function obtained during the physical examination and a third variable, derived from the other two, were analyzed. The forced expiratory volume in one second (FEV-| ) and the forced vital capacity (FVC) were determined. Prior to being analyzed, these two quantities were expressed as a percent of the predicted values for healthy, nonsmoking males (Morris et al, 1971). The third variable analyzed was the derived ratio of FEV-\ to FVC. Group differences were tested using both an unadjusted one-way analysis of variance and an analysis of covariance adjusting for age and smoking habits. The results of the analysis of the unadjusted mean values for the FVC, FEV-) and the FEV^FVC ratio are presented in Table XVI-4-2. Table XVI-4-2 ANALYSIS OF THE UNADJUSTED MEANS OF PULMONARY FUNCTION PARAMETER Mean Parameter Std. Dev. P Value FVC (^Predicted) Ranch Hand Comparison 1033 761 98.87% 98.84$ 13.15 12.98 0.97 FEV-] (% Predicted) Ranch Hand Comparison 1033 761 105.58$ 105.8755 15.65 15.36 0.69 Ranch Hand Comparison 1035 764 0.0663 0.0670 0.87 /FVC 0.8031$ 0.8026$ There are no significant unadjusted group differences between the Ranch Hand and comparison group. However, there were statistically significant interactions between age, group and pulmonary function in the analysis of both FVC and FEV-| (p = 0.04 and 0.01 respectively). Similarly, smoking .habits interacted significantly with the FEV/FVC ratio (P = 0.03). As a result, fully adjusted testing was considered to be inappropriate. However, comparison of the regression planes using the mean values of the covariables revealed P values of 0.86, 0.79, and 0.85 respectively for the FVC, FEV-] and FEV-|/FVC ratio. These values are observed to be quite similar to those seen in the unadjusted analyses. An analysis of variance of the unadjusted means for low, medium, and high exposure among the Ranch Hand group was conducted in each occupational category. These analyses revealed no consistent association between exposure level and pulmonary function. The results are presented in Table XVI-4-3. The only significant findings were in the FEV-|/FVC ratio in the enlisted categories. However, these findings were inconsistent, with the lowest exposed individuals in the enlisted flying category having the lowest mean ratio (percent performance) and higher exposed individuals doing better. In the enlisted ground personnel the mean ratio was lowest in the most heavily exposed group. Thus, while statistically significant, these findings do not conform to classic dose-response relationships. XVI-4-2 �Table XVI-4-3 HERBICIDE EXPOSURE ANALYSIS OF PULMONARY FUNCTION PARAMETERS, UNADJUSTED FOR COVARIATES OF AGE AND SMOKING I FVC (% Predicted) Low Medium High 110 128 125 Low Medium High 110 128 125 Low Medium High 110 128 125 Low Medium High 56 57 65 99.84 95.78 96.68 14,.19 11..88 14..12 Low Medium High Officer Exposure Level (% Predicted) Occupational Category 56 57 65 102.75 104.13 103.80 17,.36 14.,52 Low Medium High 56 58 65 Low Medium High 150 178 145 98.22 98.44 97.70 12.17 13-88 11.97 0.87 Low Medium High 150 178 145 105.60 105.00 102.47 14, 54 15, 42 14, 85 0.16 Low Medium High 150 178 146 0.817 0.819 Parameter /FVC Enlisted Flying FVC /FVC Enlisted Ground FVC FEVi FEV-j /FVC Std Dev P Value 100.81 99.61 101.40 12.80 13.53 12.96 0.55 108.17 107.2? 108.94 15.46 16.37 14.46 0.69 Mean 0.799 0.792 0.798 0.067 0.062 0.056 0.64 0^24 0.90 16.89 0.768 0.070 0.819 0.106 0.803 0.063 0.794 0 .056 0 .058 0.068 0.003 0.0005 Analyses of covariance adjusting for age and smoking were possible in some of the occupational categories, and the results of these analyses are presented in Table XVI-4-4. XVI-4-3 �Table XVI-4-4 ANALYSES OF PULMONARY FUNCTION AND HERBICIDE EXPOSURE, ADJUSTED FOR SMOKING AND AGE Occupational Category Parameter P Value for the Exposure _ Analysis ___ Officer FVC 0.26 0.28 0.68 Enlisted Flying FVC 0.13* FEV-j Enlisted Ground 0.90* FEV^FVC O.OOM FVC FEV-, 0.62 0.47 0.03* *= Significant covariable interaction These adjusted analyses identified significant associations in the FEV-j / FVC ratio in both enlisted categories. However, there was significant interaction between exposure level, the FEV-|/FVC ratio, smoking habits and age in the enlisted ground category. As noted in Table XVI'-4-i|, there was also interaction in the enlisted flying category for both FVC and FEV-] . When the regression planes were compared using the mean values of the age and smoking covariables, the resultant P values were as follows: Enlisted flying, FVC P = 0.10; Enlisted flying FEV-) p = 0.98; Enlisted ground FEVi/FVC P = 0.02. These P values are essentially the same as those observed in the interactions. They are also similar to those seen in the unadjusted analyses. As noted in the unadjusted analysis in Table XVI-il-3 the pattern did not suggest a consistent dose response. Summary In a few instances the results of the statistical analyses revealed significant (P % 0.05) or suggestive (P = 0.10 to 0.20) differences in pulmonary function. There were no differences detected between the Ranch Hand and comparison groups. Where significant differences were noted in the exposure index analyses, they were isolated and inconsistent in character. There were differences in the age by smoking by exposure interaction in the two groups, but it is not possible to characterize these further at this time. It may be possible to clarify these differences during follow-up phases of the study. In summary, there is no indication in the baseline physical examination that exposure to herbicide in Vietnam adversely affected pulmonary function as measured 10 to 20 years after the exposure. �Chapter XVI-5 RENAL DISEASE AND FUNCTION 1 . Introduction Overt kidney disease is not an acknowledged clinical end point following chronic exposure to low doses of Herbicide Orange or dioxin. However, since both 2,4-0 and 2,4,5-T are excreted by the kidney as unmetabolized compounds, it is understandable that acute ,renal dysfunction, as measured by a variety of laboratory tests, has been reported following acute, high dose exposure to phenoxy herbicides and dioxin. Consequently, in this study, renal function and disease were determined by general laboratory testing and history obtained by a review-of-systems questionnaire administered at the examination site. The laboratory tests emphasized measures of glomerular function rather than those of tubular function. Age of the subject (£1(0, >40 years) and 2-hour postprandial glucose levels «120, £120 mg/dl) were used as dichotomous covariates in all log-linear analyses, but were used as continuous variables in the analyses of covariance. Because of the small numbers of Black participants, the analyses are not race specific. The Ranch Hand denominator consists of all fully compliant individuals (1045) minus those few for whom covariate or dependent variable data were missing. The comparison group denominator is formed by the 773 original comparisons (i.e., shifted and replaced comparisons omitted) minus those few with missing data. Relative risks and confidence intervals are shown for all dependent variables in Appendix XVIII. 2. Laboratory Test Results The presence of occult urinary blood and protein was measured by standard reagent strips for urinalysis. The results are shown in Table XVI-5-1. After these data were placed into normal-abnormal categories, log-linear models were fitted using the covariates of age and 2-hour postprandial glucose results. These covariates were not confounding or involved in higher order interactions. Therefore, unadjusted probability values from the likelihood-ratio chi-square test statistics are used. XVI-5-1 �Table XVI-5-1 URINARY OCCULT BLOOD AND PROTEIN RESULTS BY GROUP MEMBERSHIP Occult 0 Number (50 Blood Protein >0 Number (50 0 Number (50 >0 Number (%) Comparison (N = 773) 763 (98.7) 10 (1.3) 753 (97.4) 20 (2.6) Ranch Hand (N = 1045) 030 (98.7) 14 (1.3) 030 (98.7) 14 (1.4) Occult blood group contrast P = 0.94 Relative risk: 1.037, 95% Conf. Int. (.46, 1.18) Protein group contrast P = 0.0545 Relative risk: .50, 95$ Conf. Int. (.24, 1.07) The data in Table 'XVI-5-1 show that there is no statistically significant difference in the prevalence of urinary occult blood between the Ranch Hand and comparison groups. However, the prevalence of proteinuria is borderline significant (P = 0.0545), comparisons greater than Ranch Handers. For blood urea nitrogen (BUN), urine specific gravities, and the finding of white blood cells (WBC's) in the urine, abnormalities were too sparse for log-linear analysis. Distributional data of these 3 variables were tested by an analysis of covariance, again using age and 2-hour postprandial glucose levels as continuous covariates. These data analyses and the interaction of the covariates are displayed in Table XVI~5~2. XVI-5-2 �Table XVI-5-2 MEAN BUN, URINE SPECIFIC GRAVITY AND WHITE CELL RESULTS BY GROUP MEMBERSHIP: ANALYSIS BY COVARIANCE BUN (mg/dl) P Value WBC/HPF 14.65 13-99 Comparison Ranch Hand (Adjusted Means) Specific Gravity 1.02103 1.02099 1.204 1.192 0.91 0.83 0.18 Dependent Variable Covariate Relationship P-Values Age: < 0.001 Glucose: 0.36 < 0.001 < 0.001 0.53 0.59 The data in Table XVI-5-2 show that there are no statistically significant differences in the mean BUN, specific gravity, or urinary white cells between the Ranch Hand and comparison groups, although the directional difference in the mean BUN (P = 0.18), comparison greater than Ranch Hand, is of interest. As expected, the age covariate was significantly related to BUN and specific gravity, while the glucose covariate was associated only with the specific gravity. The pattern of such classical covariate effects lends credence to the lack of group differences for these 3 dependent variables. Urine creatinine clearance levels were determined by the formula: Concentration of urine creatinine X urine volume Concentration of plasma creatinine Plasma creatinine was determined from blood samples obtained at the start of the 24-hour urine collection. Noncompliance to the full 24-hour urine collection was determined by direct questioning at the end of the sample collection and was noted to occur slightly more frequently in the comparison group (P =0.18), and significantly more (P <0.001) in older members of both groups. Air Force monitors at the examination facility frequently noted that the study participants were not fully conscientious about collecting a complete specimen, thereby casting some doubt on the overall accuracy of the creatinine clearance data. The data were not adjusted for cases of mild congestive heart failure or for high dose aspirin usage because of the rarity of these conditions in a young ambulatory population. Notwithstanding, the creatinine clearance results were tested by a log-linear model with age and glucose levels as covariates, after removing the known noncompliants. The abnormality outpoint of <110 ml/min was based upon data from the USAFSAM clinical data base, but this application produced unduly high abnormality proportions of 39.3$ and 37.4? for the Ranch XVI-5-3 �Handera and comparisons, respectively (P = 0.52). Therefore, continuous creatinine clearance values were subjected to an analysis of variance. These data are presented in Table XVI-5-3. Table XVI-5-3 MEAN VALUES OF CREATININE CLEARANCE BY GROUP, UNADJUSTED FOR COVARIATES Group Number Mean (ml/min) Standard Deviation Comparison 439 119.43 30.70 Ranch Hand 628 116.60 31.26 0.142 The concordance between group percents under 110 ml/min and the group means shown in Table XVI-5-3 is due to the left tail skew of the Ranch Hand creatinine clearance distribution as compared with that of the original comparisons. These are shown in Figure XVI-5-1. Figure XVI-5-1 CREATININE CLEARANCE FREQUENCY DISTRIBUTION BY GROUP 30-69 70-109 110-149 150-189 CLEARANCE (ML/MIN) XVI-5-4 190-230 HI RANCH HAND El COMPARISON �An analysis of covariance using age and glucose values was also performed. The glucose slopes were nonhomogeneous (P = 0.075), indicating that the group creatinine clearance difference varies with the level of glucose. 3. Questionnaire Versus Laboratory Results Log-linear models were fitted to data obtained at the time of physical examination from the question, "Have you ever had kidney disease?" with age and the 2-hour postprandial glucose level as covariates. This analysis is presented in Table XVI-5-4. These data show that the Ranch Hand group reported significantly more past kidney disease than the comparison group. Age and glucose values were not statistically significant as adjusting variables. Table XVI-5-4 HISTORY OF KIDNEY DISEASE BY GROUP Group History of Kidney Disease No (%) Yes (%) Comparison 745 (96.5) 27 (3.5) Ranch Hand 985 (94.4) 58 (5.6) 1043 Report disease group contrast: P = 0.039 (unadjusted) Relative risk: 1.6, 95% Conf. Int. (1.00, 2.59) Although analyses of 6 clinical variables had been negative with respect to group membership, it was theoretically possible that cumulative numbers of abnormalities might corroborate the historical findings. To test this notion, abnormalities were scored for 5 of the variables which exceeded normal range, i.e., BUN >26 mg/dl, creatinine clearance <110 ml/min, presence of occult blood, urine WBC ^5/HPF, and the presence of urine protein. These data were analyzed by a log-linear model, using age and glucose values as covariates. The results are presented in Table XVI-5-5. XVI-5-5 �Table XVI-5-5 ABNORMALITIES FROM FIVE RENAL FUNCTION TESTS BY HISTORY OF KIDNEY DISEASE AND GROUP MEMBERSHIP Group Abnormalities No^History;_Jjf)_ History ($) Total Comparison 0 £ 1 406 (96.4) 339 (96.6) 15 (3.6) 12 (3.4) 421 351 Ranch Hand 0 2: 1 524 (94.4) 462 (94.5) 31 (5.6) 27 (5.5) 555 489 P = 0.94 (History by abnormality interaction) These data show that the reporting of kidney disease is associated only with group membership and not with abnormal findings on the physical examination. 4. Herbicide Exposure Analyses Each Ranch Hand member was placed into an occupational stratum of flying officer, flying enlisted, or ground enlisted, which was further categorized into low, medium, or high exposure to herbicide (see Chapter VIII). Nonflying officers were assigned to the "low" exposure category of the flying officer group because of their nonherbicide administrative duties. Log-linear models were constructed for the variable of history of kidney disease, creatinine clearance, occult blood, and urinary protein; analyses of covariance were performed on the variables of BUN and urinary WBC's. Both tests used covariate adjustments based on age and 2-hour postprandial glucose results. Of the 18 exposure analyses, only 1 was borderline significant; these data are presented in Table XVI-5-6. Table XVI-5-6 HISTORY OF KIDNEY DISEASE IN RANCH HAND FLYING ENLISTED PERSONNEL BY EXPOSURE CATEGORY Ranch Hand Occupational Category Flying Enlisted Exposure Low Med High History of Kidney Disease No (%) Yes (%) 58 (98.3) 52 (88.1) 64 (97.0) P = 0.0504 XVI-5-6 1 (1.7) 7 (11 .9) 2 (3.0) Total 59 59 66 �While these exposure data are borderline significant, the association is nonlinear from low to high and is based upon very low numbers of positive histories. 5. Summary Six clinical measures of renal function and data from a review-of-systems questionnaire were tested for group membership differences by log-linear models or analysis of covariance with age and 2-hour postprandial glucose results as covariates when appropriate. A two^fold increase in proteinuria (P = 0.05*15) was found in the comparison group. Ranch Hand versus comparison group creatinine clearance differences were difficult to assess due to manifest compliance problems to the 24-hour urine collection process. While the Ranch Handers reported a significantly higher history (P = 0.0389) of past kidney disease, these historical differences were not correlated to cumulative abnormalities of 5 clinical variables. Herbicide exposure analyses in the Ranch Hand group were essentially negative. XVI-5-7 �Chapter XVI-6 ENDOCRINE FUNCTION 1. Introduction TCDD is known to produce a broad spectrum of metabolic phenomena in animal experimental subjects treated with sufficiently large doses. The pattern of effects is quite complex. Hypothyroxinemia has been produced in rats (Potter et al, 1983), and this may be associated with increased biliary elimination of thyroxine (Bastomsky, 1977). Hypoglycemia has been produced in rats (Gasiewicz et al, 1980, Potter et al, 1983) at the same time that serum and pancreatic insulin levels fell (Potter et al, 1983). TCDD has been observed to reduce hepatic catabolism of testosterone in the rat (Nienstedt et al, 1979). Based on animal data, the physical examination in this study obtained data for thyroid function (T3 uptake, serum T4 and the free thyroxine index or FTI), glucose metabolism (blood glucose level taken 2 hours after a standard carbohydrate load) and serum testosterone level. These 5 variables are listed in Table XVI-6-1 together with a description of normal and abnormal levels provided by the Kelsey-Seybold contract effort. Table XVI-6-1 FIVE ENDOCRINOLOGICAL VARIABLES AND THEIR NORMAL AND ABNORMAL LEVELS Variable Name Variable Abbreviation Abnormal (Low) Normal Range Abnormal (High) 27%-37% T3 Uptake T3 Serum T4 T4 Free Thyr6xine Index FTI 2 Hour Postprandial Glucose GLU 2 HR NA <120 mg/dl Serum Testosterone TEST <400 ng/dl 400-1200 ng/dl <4.7 yg/dl 4.7-12.5 ug/dl >12.5 1.3-^.6 >_120 mg/dl >1200 ng/dl Each study subject was asked to follow a standardized diet prior to arrival at the examination site. Not all participants complied with the diet. Table XVI-6-2 shows dietary compliance by group. XVI-6-1 �Table XVI-6-2 DIETARY COMPLIANCE BY GROUP Group Complied with Diet Did Not Comply With Diet Dietary Compliance Unknown Ranch Hand 896 (86?) 96 53 Comparison 676 (87%) 70 27 The groups are not different as regards dietary compliance (P = 0.262). Also dietary compliance was not found to be associated with the likelihood of being in the high abnormal GLU 2 HR category. Thus, in Tables XVI-6-3 and XVI-6-5 participants were used irrespective of dietary compliance status. 2. Data Analysis Table XVI-6-3 shows unadjusted percentages of the 5 endocrinological variables by variable level and group. (For this table and all other analyses in this chapter, all Ranch Hand participants (N = 10*15) and all original controls (N = 773) were used as the basic data set). In the analysis of thyroid hormones, data from individuals with thyroidectomies were removed (7 Ranch Handers and 3 original comparisons), and in the analysis of testosterone, data from individuals with orchiectomies (5 Ranch Handers and 1 original comparison) were removed. Other denominator variations occurred due to missing covariates. A group difference in T3 uptake is noted in Table XVI-6-3. The Ranch Hand group has fewer individuals in the low category and more individuals in the high category than does the comparison group. The same directionality is noted with the T4 and FTI variables. No group differences are found in GLU 2 HR or TEST. Since hormone levels can be correlated with age and physical habitus, an analysis of the 5 endocrinological variables was attempted adjusting for age in years (dichotomized as less than or equal to 40 years and greater than 40 years) and for percent body fat (trichotomized as less than 10?, 10-25?, greater than 25?). There are too few abnormalities for a full analysis of any of the 5 endocrinological variables. However, for T3 and TEST, analyses could be performed on those individuals with 10? body fat or greater and having low abnormal or normal dependent variable values. Similarly, an analysis of GLU 2 HR values was possible on those individuals with 10? body fat or greater. The data for these 3 adjusted analyses are presented in Tables XVI-6-4, XVI-6-5 and XVI-6-6. Log-linear models were used in these analyses. XVI-6-2 �Table XVI-6-3 UNADJUSTED PERCENTAGES FOR FIVE ENDOCRINOLOGICAL VARIABLES BY VARIABLE LEVEL AND GROUP Variable Level Variable Low Normal High P Value For Group Difference RH COM 1032 767 5.72$ 8.47$ 93.41$ 91.26$ 0.87$ 0.26$ 0.020 RH COM 1033 767 0.10$ 0.39$ 99.13$ 99.22$ 0.77$ 0.39$ 0.250 FT I RH COM 1033 767 0.00$ 0.26$ 99.71$ 99.74$ 0.29$ 0.00$ 0.085 GLU 2 HR RH COM 1040 770 NA NA 84.81$ 82.73$ 15.19$ 17.27$ 0.234 TEST RH COM 1034 769 4.93$ 6.37$ 94.58$ 93.11$ 0.48$ 0.52$ 0.414 T3 Table XVl<-6-4 shows a group difference in T3 uptake which is age specific (P = 0.005). There are more low T3 values in the comparison group than in the Ranch Hand group in the 40 and under-40 age group, but the groups are similar above 40 years Of age. A highly statistically significant association of T3 hypothyroxinemia with body fat is noted within the groups (P = 0.004). Table XVI-6-5 shows no group difference in the observed proportions of hyperglycemia (>_ 120 mg/dl). Age and body fat are seen to influence these proportions (P < 0.001 in both instances), and the effect is about the same in both groups. Table XVI-6-6 shows no group difference in the observed proportions of low testosterone. Age and body fat both influence these proportions (P = 0.022 for age and P < 0.001 for body fat), and the effect is approximately the same in both groups. Using the categories for normal and abnormal levels shown in Table XVI-6-1, it was not possible to meaningfully carry out an exposure index analysis of the 5 endocrinological variables, due to sample size limitations. XVI-6-3 �Table XVI-6-4 PERCENT OF ABNORMALLY LOW Tj VALUES BY GROUP, AGE AND BODY FAT CATEGORY* % T3 Low Abnormal in 10-25? Body Fat Subgroup % T3 Low Abnormal in > 25% Body Fat Subgroup Age <40 RH 2.59 (9/347) 6.58 (5/76) _<40 COM 7.89 (18/228) 19.15 (9/47) >40 RH 6.49 (30/462) 10.94 (14/128 >40 * Group COM 7.43 (28/377) 9.26 (10/108) Abnormally high individuals and lean individuals (less than ^0% body fat) were removed from the analysis due to sample size limitations. Table XVI-6-5 PERCENT ABNORMAL GLU 2 HR VALUES BY GROUP, AGE AND BODY FAT CATEGORY* % GLU 2 HR in Abnormal Category in 10-2555 Body Fat Subgroup % GLU 2 HR in Abnormal Category in >25% Body Fat Subgroup Age £40 RH 6.25 (22/352) 17.11 (13/76) <_40 COM 6.55 (15/229) 17.02 (8/47) >40 RH 18.01 (85/472) 28.46 (37/130) >40 * Group COM 18.25 (69/378) 36.36 (40/110) Lean individuals (less than *\Q% body fat) were removed from the analysis due to sample size limitations. XVI-6-4 �Table XVI-6-6 PERCENT ABNORMAL LOW TESTOSTERONE VALUES BY GROUP, AGE AND BODY FAT CATEGORY* Age Group % Testosterone Low Abnormal in 10-25? Body Fat Subgroup % Testosterone Low Abnormal in > 25% Body Fat Subgroup 2.00 7.89 (6/76) 3.52 (8/227) 10.64 (5A7) 3.46 (16/463) 16.15 (21/130 it.00 * (7/350) (15/375) 19.09 (21/110) Abnormally high individuals and lean individuals (less than 10% body fat) were removed from the analysis due to sample size limitations. Analysis of covariance is less vulnerable to the data limitations of sparse or empty cells than are log-linear models. Thus, the Ranch Hand group was contrasted with the comparison group in terms of the 5 endocrinological variables using analysis of covariance adjusting for age and percent body fat. In these analyses, all variables except group indicators were used as continuous variables. In the analysis of thyroid hormones, data from individuals with thyroidectomies were removed, and in the analysis of testosterone levels, individuals with orchiectomies were removed. In the analysis of glucose levels, all participant data were used irrespective of dietary compliance as compliance was not found to influence glucose levels. Table XVI-6-7 provides unadjusted and adjusted means. When a group-by-age or group-by-body fat interaction was observed with P < 0.10, adjusted means, and age and body fat main effects are not reported. One overall group difference is noted in Table XVI-6-7. Specifically, the Ranch Handers show a higher testosterone level than do comparison participants (P = 0.02 unadjusted, 0.06 adjusted). Both increasing age and increasing body fat were found to be associated with decreasing testosterone level with slopes being -3.8 ng/dl per year of life and -12.6 ng/dl per % body fat. XVI-6-5 �Table XVI-6-7 RANCH HAND - COMPARISON GROUP MEANS OF ENDOCRINE VARIABLES P Variable Group N Unadj'd Mean T3 Com 770 30.14 Uptake RH 1037 30.28 T, Com 770 8.39 (ug/dl) RH 1038 8.46 FTI Com 770 2.51 RH 1038 2.54 GLU 2HR Com 773 102 (mg/dl) RH 1045 104 TEST Com 772 634 (ng/dl) RH 1039 P Value for Value for Remarks about Unadj'd Adj'd Adj'd Adjusting Means Mean Means Covariates 654 0.21 Group-by-age interaction (P = 0.026) 0.31 8.39 0.38 None significant at P<.05 0.13 Age (P<.001) 8.45 0.07 2.51 2.54 0.37 * % Body fat (P<.001) * Group-by-age interaction (P-.006) 0.02 637 652 XVI-6-6 0.06 Age (P<.001) % Body fat (P<.001) �Two other group differences are noted in Table XVI-6-7; however, these are associated with group-by-age interactions. In both the Ranch Hand and comparison groups, decreasing T3 uptake is observed associated with advancing age, but the slope was found to be -0*0068? per year in the comparison group while it is -O.Q4955& per year in the Ranch Hand group. Glucose levels, measured 2 hours into the glucose tolerance test, were observed to increase with age in both the comparison and Ranch Hand group; however, the rate of increase is 0.77 mg/dl per year. in the comparison group and 1.53 mg/dl per year in the Ranch Hand group. Dose-response data within the Ranch Hand group are provided in Tables XVI-6-8, XVI-6-9 and XVI-6-10. No overall statistically significant doseresponse relationships were detected; however, 5 exposure group by covariate interactions were noted. These interactions are summarized in Table XVI-6-11. No interactions are seen with respect to the variables T3 or T4. XVI-6-7 �Table XVI-6-8 RANCH HAND OFFICERS ENDOCRINE DOSE-RESPONSE DATA P P Value for Unadj'd Adj'd Mean Mean 30.6 125 30.6 30.6 L 110 8.21 126 8.15 8.15 125 8.22 8.22 L 110 2.51 126 2.4? 125 2.49 L 111 106.7 M 128 104.2 H 125 106.8 L 111 614.8 M 127 614.2 H 123 604.5 Value for Remarks about Adj'd Adjusting Mean Covariates 30.7 H TEST 126 M GLU 2 HR 30.9 H FTI 110 M T4 L H T3 N M Variable Unadj'd Mean Group 0.39 0.12 30.8 8.23 0.88 Age (P-0.033) % Body fat (P-0.039) 0.89 None 0.59 * * Age-exposure interaction (P=0.042) 0.90 * * % Body fatexposure interaction (P-0.041) 0.85 * * % Body fatexposure interaction (P-0.011) XVI-6-8 �Table XVI-6-9 RANCH HAND - FLYING ENLISTED PERSONNEL ENDOCRINE DOSE-RESPONSE DATA Variable Group N P Value for Unad j ' d Unadj'd Adj'd Mean Mean Mean 30.0 64 30.0 30.1 L 59 8.85 59 8.48 8.49 64 8.48 8.50 L 59 2.60 59 2.51 H 64 2.60 L 59 102.3 M 59 105.9 108.0 H 66 105.6 103.8 L 59 663.5 M 58 657.8 653.5 H 66 658.5 666.7 Remarks about Adjusting Covariates 30.0 M TEST 59 H GLU 2 HR 29.6 M FTI 59 H T4 L M T3 P Value for Adj'd Mean 0.57 0.32 0.45 0.88 0.98 XVI-6-9 29.6 8.85 * 102.3 659.8 0.59 None 0.32 None * % Body fatexposure interaction (P=0.03) 0.78 Age (P=0.01) 0.90 % Body fat (P<0.001 ) �Table XVI-6-10 RANCH HAND - GROUND ENLISTED PERSONNEL ENDOCRINE DOSE-RESPONSE DATA Variable Group N P Value for Unad j ' d Unad j ' d Mean Mean P Value for Remarks about Ad j ' d Adj'd Adjusting Mean Mean Covariates 30.2 148 30.3. 151 8.58 177 8.67 8.67 148 8.59 8.58 L 151 2.55 177 2.58 2.56 H 148 2.60 2.61 L 151 99.9 M 179 104.8 H 148 103.7 L 151 686.4 M 179 680.5 678.2 H 146 683.0 684.4 Age (P<0.001) 30.4 L 0.18 30.1 M TEST 176 H GLU 2 HR 29.8 M FTI 151 H T4 L M T3 0.30 0.89 0.69 0.60 0.97 XVI-6-10 29.9 8.59 2.55 * 685.6 % Body fat (P<0.003) 0.89 None 0.53 Age (P-0.01) % Body fat (P-0.03) * % Body fatexposure interaction (P-0.09) 0.93 Age (P-0.02) % Body fat (P<0.001) �Table XVI-6-11 ENDOCRINE DOSE - COVARIATE INTERACTIONS T3 Ranch Hand No Officers interactions FTI T4 No interactions GLU 2 Hr Age-exposure % Body fatinteraction exposure interaction (P-0.042) (P=0.041) TEST % Body fatexposure interaction (P-0.011) Ranch Hand No Flying interactions Enlisted No interactions % Body fatexposure interaction (P-0.03) No interaction Ranch Hand No Ground interactions Enlisted No interactions No interactions % Body fat- No interactions exposure interaction (P-0.09) No interactions The FTI shows an age*-exposure interaction among the officers and a % body fatexposure interaction in the flying enlisted Ranch Hand group. Among the officers, FTI increased by 0.00*11 per year of life in the low exposure group but decreased by 0.0127 and 0.0079 per year in the medium and high exposure groups respectively. No effect of body fat was suggested by the officer data. Among the flying enlisted, FTI did not appear affected by age, but increased with increasing % body fat in the low and medium exposure groups (0.00295 and 0.00378 per % body fat respectively) while it decreased with body fat (-0.0241 per % body fat) in the high exposure group. These FTI effects are interesting; however, the lack of consistency between occupational and exposure categories leads to doubt that an actual herbicide effect exists. Both Ranch Hand officers and ground enlisted personnel show comparable body fat-exposure interactions affecting glucose levels. The glucose level-body fat slopes are given in Table XVI-6-12. In both the officer and ground enlisted categories, the low exposed individuals show a decreasing blood glucose with increasing % body fat, but this relationship changes to a positive correlation in the medium and high exposure categories. XVI-6-11 �Table XVI-6-12 CHANGE IN GLUCOSE LEVEL PER % BODY FAT (mg/dl PER % BODY FAT) BY HERBICIDE EXPOSURE LEVEL IN TWO RANCH HAND GROUPS Exposure Category Ranch Hand Officers Ranch Hand Ground Enlisted Low M.18 -0.30 Medium +2.94 +1.75 High +1.26 +1.36 A % body fat by exposure interaction is also observed to affect testosterone levels in Ranch Hand officers with a very low probability that the effect could be due to chance (P = 0.011). Low exposed officers show a decrease in serum testosterone levels of 4.5 ng/dl per % body fat while medium and high exposed officers show decreases of 16.6 ng/dl and 15.3 ng/dl per % body fat respectively. 3. Summary The Ranch Hand group was found to differ from the comparison group with respect to proportions of individuals in normal and abnormal thyroid hormone categories. The difference is a tendency toward hyperthyroxinemia which is directionally opposite to what would be expected on the basis of subacute animal studies. On the other hand, decreasing T3 uptakes are associated with advancing age in both groups with the slope being much steeper in the Ranch Hand group. Finally, no meaningful association of thyroid hormone levels with the exposure index were found. Thus, in sum, no definite herbicide effect on thyroid function can be considered demonstrated; however, it also cannot be confidently asserted that a herbicide effect on thyroid function has not occurred. As a group, Ranch Hand personnel have higher testosterone levels than comparison individuals and Ranch Hand officers evidence a decrease in testosterone level with increasing body fat that is related to herbicide exposure category (higher exposures are associated with greater decreases in testosterone with body fat). Since subacute animal studies have shown decreased catabolism of testosterone, higher serum levels could be expected. Thus, this finding in the present study may reflect an herbicide effect, whose long-term impact will require further clinical evaluation. Overall, Ranch Hand blood glucose levels are not statistically significantly different from those of comparison individuals. However, positive associations of glucose levels with age are greater in the Ranch Hand group than in the comparison group, and in both the Ranch Hand officer and ground enlisted groups significant exposure - body fat interactions exist on glucose levels. Thus, a subtle toxicological effect of herbicide on glucose metabolism may have been detected. It will be important and interesting to follow these groups in time with respect to the incidence of diabetes. XVI-6-12 �Chapter XVII INDIVIDUAL HEALTH ASSESSMENT 1. Personal Habi ts and Character1stics The personal characteristics of the Ranch Hand and comparison individuals were obtained from the in-home questionnaire. The areas of tobacco, alcohol, and marijuana use, personal and family income, education, religion, active duty, retired/separated status, and risk-taking behavior received particular attention. The number of Ranch Hand and comparison group individuals reporting a listing of past traumatic injuries, poisonings, and/or toxic effects (ICD9-CM Codes 960-999) were also determined. The smoking and alcohol use habits of the study subjects are displayed in Table XVII-1. XVII-1 �Table XVI1-1 HISTORY OF TOBACCO AND ALCOHOL USE AMONG THE STUDY PARTICIPANTS Group Original Comparisons Habit Current Use of Cigarettes Past History of Cigarettes Past History of Cigar Use Yes ($) All Comparisons Ranch Hand No Yes ($) No Past History of Marijuana Use 478 (45.7?) 567 / \ / \ 484 (39.6$) / / P = 0.003 739 552 (72.3/0 212 \ \ P = 0.67 758 (73.2$) 278 / \ / \ 861 (71.150 / / P = 0.28 350 92 (11.9$) \ \ 680 99 (9.5$) / / 942 \ \ 157 (20.4$) 613 \ \ P = 0.62 22 (2.8$) \ \ 750 Past History of Alcohol Use 141 (11.5$) 1081 / / P = 0.12 200 (19.4$) 829 / \ / \ 53 (5.1$) 992 / \ / \ P = 0.02 Current Use of Alcohol No 313 (40.550 459 \ \ P = 0.03 P = 0.10 Past History of Pipe Use Yes ($) 246 (20.2$) / / P = 0.64 970 62 (5.1$) 1160 / / P = 1.00 447 (58.6$) 316 \ \ P,= 0.89 J 609 (58.9$) 425 694 (57.3$) / \ / / \ / P = 0.43 518 478 (63.0$) 281 \ \ P = 0.74 635 (62.2$) 386 / \ / \ 421 XVII-2 773 (64.7$) / / P = 0.21 �The mean number of cigarettes currently smoked and the mean number of alcohols-containing drinks consumed per day by those currently reporting use of these substances were determined. Similarly, the mean pack-years, cigar-years, pipe-years, drink-years and marijuana joint-years were determined for the groups in the study. These data are presented in Table XVII-2. Table XVII-2 MEAN USE OF TOBACCO PRODUCTS AND ALCOHOL IN THOSE REPORTING USE OF THESE SUBSTANCES Mean Usage Level Substance Original Comparisons Mean (Median) Ranch Hand Mean (Median) All Comparisons Mean (Median) Cigarettes per day (current use) 28.28 (30) 27.21 (25) 27.72 (30) Cigarette pack-years (cumulative) 23. (20.12) 23.89 (20.91) 22.92 (19-58) Cigar-years (cumulative) 21.26 (8.11) 19.12 (9.38) 20.80 (7.33) Pipe-years (cumulative) 26.96 (6) 26.32 (7.23) 26.26 (5.71) Marijuana Joint-years (cumulative) 7.60 (2.52) 7.12 (3.54) 8.26 (2.88) Alcohol drinks per day (current use) 2.33 (2) 2.35 (2) 2.38 (2) Drink-years (cumulative) 36.48 (26.31) 40.48 XVII-3 (24.23) 34.87 (25.08) �In most of the cumulative measurements (e.g., pack-years) the median level of use was lower than the mean level, indicating that the heavy users of these substances skewed the distributions. However, in the measurements of current use, there was little evidence for this effect. The median income levels of the Ranch Handers and the original comparison were the same with personal income ranging from $20,000 - $24,999 and total family income ranging from $30,000 - $34,999. The median personal income of the entire comparison group was also in the $20,000 - $24,999 range, but the median family income remained in this same category. The educational backgrounds of the groups were not significantly different. Religious preferences of the groups were also similar. These data are shown in Tables XVII-3 and XVII-4. Table XVII-3 EDUCATIONAL BACKGROUND BY GROUP Group Educational Level Original Comparisons Number (%) High School/GED Associate Degree BA/BS Degree Graduate Degree Unknown 430 53 152 132 6 Ranch Hand Rai Number (%} (55.63) 580 (6.86) 6? (19.66) 197 (17.07) 187 (0.78) 14 \ / \ / P = 0.78 XVII-4 (55.50) (6.41) (18.85) (17.89) (1.34) \ \ All Comparisons Number (%) 661 96 249 206 12 P = 0.48 (54.01) (7.84) (20.34) (16.83) (0.98) / / �Table XVII-4 RELIGIOUS PREFERENCE BY GROUP Group Religion Original Comparisons Number (50 Protestant Catholic Jewish Other None 699 218 9 34 85 Ranch Hand Number (50 (66.89) (68.69) 531 (20.86) 162 (20.96) 12 (0.86) (1.55) 20 (3.25) (2.59) (6.21) (8.13) \ /' \ \ \ P = 0.29 All Comparisons Number (%) 816 263 16 49 80 (66.68) (21.49) (1.3D (4.00) (6.54) 0.50 The current military status of each individual was determined as either active duty, retired, separated, reserve status, or deceased, and there were no statistically significant differences between the Ranch Handers and the subset of original comparisons (P = 0.23); however, there was a significant difference (P = 0.01) between the Ranch Handers and the total comparison group. These data are presented in Table XVII-5. Table XVII-5 MILITARY STATUS BY GROUP Group Military Status Original Comparisons Number (%) Active Duty Retired Separated Reserve Forces Deceased* 113 420 196 39 4 Ranch Hand Number (%) (14.64) (14.66) 153 (54.40) 515 (49.33) (29.21) (25.39) 305 64 (5.05) (6.13) (0.52) 7 (0.67) \ / \ /' \ \ P = 0.23 All Comparisons Number (50 184 593 247 69 6 (16.74) (53.96) (22.47) (6.28) (0.55) P = 0.01 *Deceased subsequent to the physical examination. Risk-taking behavior patterns were assessed by a series of questions (i.e., "Have you participated three or more times in activity?") that emphasized participation in potentially dangerous recreational activities. These data are tabulated in Table XVII^-6. XVII-5 �Table XVII-6 RISK-TAKING BEHAVIOR BY GROUP Group Activity Scuba Diving Auto, Boat or Motorcycle Racing Original Comparisons Yes (%} No 88 (11.40) 684 103 (9.87) \ / \ / P = 0.29 77 (9.97) \ \ 695 All Comparisons Yes (%} No_ Ranch Hand Yes (%) No 941 \ \ 155 (12.68) 106? / / P = 0.04 132 (12.64) 912 / \ / \ 140 (11.46) 1082 / / P = 0.39 P = 0.08 Acrobatic Flying 25 (3.24) \ \ 747 29 (2.78) / / 1015 \ \ 39 (3-19) 1183 / / P = 0.57 14 (1 .3*0 / / 1030 \ \ 29 (2.37) 1193 / / P = 0.07 6 (0.57) / / 1038 \ \ P = 0.57 Sky Diving 12 (1.55) \ \ 760 P = 0.71 Hang Gliding 4 (0.52) \ \ 768 P = 0.87 Mountain Climbing 35 (4.53) \ \ 737 172 (22.3) \ \ 61 (5.84) / / 601 1209 / / P = 0.20 P = 0.22 One or More Risk-taking activities 13 (1.06) 253 (24.2) / / P = 0.33 XVII-6 983 63 (5.16) 1159 \ / \ / . P = 0.47 792 308 (25.2) \ \ P - 0.60 916 / / �Only in motor vehicle racing (automobile, boats and motorcycles) was there a borderline suggestion of a difference in risk-taking behavior between the Ranch Handers and the original comparison subset. In contrast, there was a statistically significant difference between the Ranch Handers and the entire comparison group in scuba diving (P = 0.0*1) and a borderline difference (P = 0.07) in sky diving. In both of these instances, the comparisons had higher rates of participation. In combining all activities, there was no significant difference in risk-taking behavior between the Ranch Handers and the original or entire comparison group. Table XVII-7 contains the distribution of reported past injuries and poisonings by ICD code for each group. Conditional unadjusted chi-square testing reveals no significant group differences in these distributions. Table XVII-7 DISTRIBUTION OF REPORTED INJURIES AND POISONINGS BY GROUP Group Injury (ICD Code) Fractures, Dislocations, Sprains (800-848) Original Comparisons All Compa^risons Ranch Hand 11 11 Intracranial, chest; abdominal and pelvic injuries; open wounds; nerve and spinal cord injuries (850-897; 925-929; 950-957) 3 4 Late effects; superficial injuries and contusions; burns (905-924; 940-949) 5 2 6 Traumatic complications (958-959) 5 9 8 Poisonings, toxic effects; other specified causes (960-989) 3 \ \ 0 / \ / \ P =0.23 XVTI-7 17 ' • 8 4 / / P = 0.31 �2. Health Abnormalities Detected at Physical Examination Throughout previous chapters, health of the participants has been assessed in a variety of interrelated ways. Normal-abnormal categorizations, or continuously distributed clinical variables have been defined organ system by isolated organ system, categorized into physical, mental, reproductive, biochemical, and machine-results parameters, all of which were qualified by overall historic and diagnostic impressions. This research approach has not been suitable to assess total individual health. Since such a task would involve complete listings of all past abnormalities and current normalitiesabnormalities by individual, these citations would exceed the scope of this report. This chapter section attempts to assess the overall health of individuals in three ways: the summation of abnormalities of major components of each of the 12 organ systems; the summation of a weighted score of the same abnormalities; and a summary count of medical codes for historical disease and disease suspected/detected at the physical examination. a. Summation of Individual Abnormalities In 8 of the 12 clinical areas, virtually all individuals were found to have complete examination data, and all of the selected parameters of individual health could be evaluated. Table XVII-8 provides the number of Ranch Hand and original comparison group individuals with incomplete data who were not included in the tabulation for each organ system. Table XVII-8 DISTRIBUTION OF INDIVIDUALS WITH INCOMPLETE DATA OMITTED FROM ANALYSIS OF INDIVIDUAL HEALTH Organ System General Health Malignancy Reproductive Neurological Psychological Hepatic Dermatology Cardiovascular Hematologic Pulmonary Renal Endocrine Ranch Hand 8 0 473 31 4 0 " 0 4 0 5 0 9 XVII-8 Comparison 6 0 352 19 0 0 0 3 0 3 0 3 �The assessment of the reproductive system is based solely on the sperm count. Those individuals noncompliant for the collection of semen or those having had vasectomies or orchiectomies were excluded from this analysis. In the psychologic, hepatic and neurologic clinical areas, there were sufficient numbers of individuals with missing data to warrant separate analyses of individuals with complete data and individuals with partial data. The data and results of the analysis of abnormalities by organ system are presented in Table XVII-9. As noted for the psychologic, neurologic and hepatic data, subset analyses were accomplished. XVII-9 �Table XVI1-9 COUNT DATA NUMBER OF HEALTH ABNORMALITIES BY ORGAN SYSTEM AND GROUP (UNADJUSTED FOR MATCHING VARIABLES OR RISK FACTORS) Organ System Group 0 Number of Abnormalities 1 2 3 4 General Health RH C 791 573 228 186 18 8 Ma 1 1 gnancy RH C 997 755 48 17 0 1 RH C 374 263 198 158 0 Reproductive Neurological - _ Unadjusted P Values 5-6 _ 0.27 0.01 - - 1 2 3 0.34 4-9 (Full Data Subset) RH C 113 112 268 179 238 186 126 92 84 57 0.17 (Subset with 1 Missing Parameter) RH C 59 40 64 46 36 27 20 9 6 6 0.79 0 1 2 3 4 RH C 341 243 301 234 121 75 10 3 - RH C 143 129 114 83 11 6 - - RH C 184 134 206 134 143 94 68 54 26 18 3 7 0.45 (Subset with 3 Missing Parameters) RH C 114 74 134 115 90 77 44 42 29 24 4 0 0.27 Dermato logic RH C 470 347 575 426 - - Cardiovascular RH C 491 365 324 232 151 117 53 42 16 12 Hemato logic RH C 428 341 432 311 147 98 35 20 3 3 0.59 Pulmonary RH C 655 463 289 232 52 56 32 15 12 4 0.05 Renal RH C 1002 740 42 31 1 2 - Endocrine RH C 787 551 207 182 36 33 5-6 Psychological (Full Data Subset) (Subset with 1 Missing Parameter) 0.29 0.38 Hepatic (Ful 1 Data Subset) XVII-10 6 4 0.97 - - 6 2 0.92 0.70 0.20 �These data demonstrate statistically significant group differences only for malignancy (a result of the identified increase in skin cancer in the Ranch Hand Group) and in pulmonary function (due to more abnormalities in the comparison group). All other analyses were not statistically significant. The reader is cautioned that the data in Table XVII-9 are crude counts, unadjusted for the matching variables or risk factors known to affect organ system parameters. The number of abnormalities per organ system may be considered a crude index of severity. All individuals and their abnormality counts were summed, regardless of the degree of completeness of their data. The frequency distribution of these abnormalities is shown in Figure XVII-1. Figure XVII-1 EXAMINATION ABNORMALITIES 20 10 0-2 3-4 5-6 7-8 9-10 1M3 NUMBER OF ABNORMALITIES I RANCH HAND COMPARISON There was a maximum of 61 abnormalities in this analysis. The median number of abnormalities in both the Ranch Hand and comparison groups was seven. There were 0.96$ of the Ranch Handers and 1.55% comparison individuals who had no abnormalities, and 2.58$ and 2.07%, respectively, with 16 or more abnormalities. Log linear analysis of these distributions revealed no differences between the groups for numbers of abnormalities or degree of completeness of data (P values of 0.26 and 0.59, respectively). XVI1-11 �b. Weighted Score of Individual Abnormalities The count of abnormalities (Table XVII-9) was subjected to a weighting scale of 1 to 10 depending on the clinical seriousness of each abnormality. While such weighting is arbitrary, the resulting data serve as a complementary analytic technique to the basic count of abnormalities in which, for example, acne is considered to be equivalent to systemic cancer or a major EGG abnormality. The assignment of a weight to each abnormality was made before organ system results were known. Appendix VII contains a listing of all parameters and their relative weight scores for each organ system. The weighted score histogram is depicted in Figure XVII-2. Figure XVII-2 ABNORMALITY WEIGHTED SCORE 0-9 10-19 20-29 30-39 SCORE 4049 ="49 • RANCH HAND H COMPARISON Scores between zero and nine were achieved by 9.09/5 of the Ranch Handers and 7.24$ of the comparisons, with 8.805? of the Ranch Handers and 8.02$ of the comparisons scoring above 50 (out of a maximum possible score of 236). The median score was in the 20 to 24 range for both groups. The weighted score analysis showed statistical significance for cancer, again due to the aggregation of skin cancer in the Ranch Hand group. Statistical differences of interest were noted for renal disease (P = 0.09), general health (P = 0.114), and hepatic disease (P = 0.11). The relevance of these P values is minimal in view XVII-12 �of the predominantly negative analyses observed in the clinical chapters. All weighted scores were combined across clinical areas and no statistically significant differences were noted (P = 0.20). From these analyses on crude and weighted abnormalities, it is clear that there were not significantly more ill or more severely ill individuals in the Ranch Hand group than in the comparison group. c. Physical Examination Diagnostic Codes The diseases or conditions listed by the diagnostician in the diagnostic summary of the review of systems, the medical history, and the physical examination were coded according to the 9th ICD^-CM manual. These diseases were coded as being reported by history, or suspected or actually diagnosed conditions. One individual could account for more than one diagnosed disease or condition. The diagnostician listed 219 suspected diseases among the 10^5 Ranch Handers and 160 suspected conditions in the 773 original comparisons (P = 0.91). In both groups, there were 0.21 suspected diagnoses per individual. Similarly, 19^9 definitive diagnoses were made in the Ranch Handers and 1^37 in the original comparisons yielding an average of 1.87 diagnoses per Ranch Hander and 1.86 per comparison individual (P = 0.96). While the mean numbers of suspected and definitive diagnoses were essentially the same in both groups, the mean number of diseases and conditions reported by the participants were different in the two groups. There were 113 diseases reported by history in the Ranch Handers, but only 57 in the comparisons (mean number of conditions of 0.11 per person and 0.07 per person (P = 0.02), respectively). The similarity in diagnosed and suspected conditions in the two groups parallels the findings in the analysis of examination abnormalities. The difference in reported conditions may reflect differential reporting, or actual difference in past health. However, if past illness was different in the two groups, these experiences have apparently not resulted in long-term sequelae detected at the examination. 3- Summary The anecdotal comments of the examining physicians and psychologists suggested that the study participants were remarkably healthy both physically and mentally for a group of mid-aged men. These comments were made about the entire group of participants based on the medical experience of each examiner, without knowledge of which individuals were Ranch Handers and which were comparisons. The statistical analyses discussed in this chapter support the clinical impressions of the examiners. Both the Ranch Handers and the original comparisons had somewhat similar health habits, although significantly more Ranch Handers are current cigarette smokers and more had reported smoking marijuana in the past. The two groups were also similar in risk-taking activities, religion, education, income, and military status. XVII-13 �The distribution of identified health abnormalities by individual, and weighted scores of these abnormalities were not significantly different in Ranch Hand and comparison groups. Similarly, the mean number of diagnoses individual at the conclusion of the examination was not different in the groups. the the per two Overall, the health of individuals in the two groups appears to be quite comparable. As individuals, they seem to be in quite good health for men of their age. These findings and observations are most likely a result of the healthy worker effect, previously noted in the baseline mortality study. XVII-14 �Chapter XVIII FUTURE COMMITMENTS The large volume and complexity of the data collected during this baseline phase of the Ranch Hand II study have made it difficult to completely fulfill all aspects of the analytic plan envisioned in the study protocol. While most of the major anticipated analyses have been completed and included in this report, other important tasks remain to be done. The results cited in this report logically lead toward a commitment by the USAF and the study principal investigators to pursue further evaluations of these data, and follow the study participants over time. There are 5 key areas requiring additional effort: (1) database refinement, (2) definition of requirements and examination refinements for the follow-up phase of the study, (3) refinement and expansion of exposure indices (4) additional statistical analyses and (5) collaborative activities with other organizations involved in herbicide/dioxin research. 1. Database Refinement The database derived from the questionnaire and from the physical examination was very extensive in size and scope, and a quality control program was initiated to identify coding, keypunching, and editing errors in the database provided by the contractors. This data validation has been an on-going task, and is not yet complete in some areas. After the remaining questionnaire and physical examination data have been validated by comparison with the source documents, epidemiologic and statistical analyses of these data will be completed. Additionally, validation of illnesses and conditions reported on the in-home questionnaire will continue to be accomplished as medical records and birth certificates are received. Methods of validating smoking histories, and a reassessment of flying status and its impact on compliance will be pursued. The completion of this process will provide a verified database for subsequent analyses. This process will also allow an assessment of the degree of differential reporting present in the study. 2. Follow-up^ Examination Requirements One of the purposes of the baseline phase was to identify clinical areas requiring in-depth evaluation in the follow-up portions of the study. Focused questionnaire and physical examination formats will be developed for use during the reexamination scheduled for 1985. At that time detailed evaluation of skin cancer, and known risk factors affecting its occurrence will be obtained. Additional data on fertility and reproductive history will be gathered and updates of conceptions and live births occurring since the baseline questionnaire will be obtained. The cardiovascular status of the participants will also be closely examined, using doppler measurements of peripheral pulses and electrocardiographic monitoring during stress testing. New, fully validated psychological scales will be used to assess additional psychological parameters such as sleep patterns. Further immunologic evaluations with strict laboratory quality control will also be accomplished. Steps will also be taken to insure that all participants comply with dietary and 24-hour urine collection requirements. At XVIII-1 �the time of the follow-up physical examination, all participants will be requested to authorize an autopsy at the time of their deaths and have copies of those reports and tissue specimens provided to the Air Force. The participants will also be asked to forward copies of hospitalization summaries and other significant medical events to the Air Force for inclusion in their records at Brooks AFB. 3. Exposure Index Refinement The index of exposure to phenoxy herbicide and dioxin used in this report is not as complete or refined as planned in the study protocol. As it is currently calculated, each of the major occupational categories (Officers, Enlisted Flying, and Enlisted Ground) must be analyzed separately since the index is not necessarily equivalent in each category. A series of flights in a C-123 aircraft is planned. The aircraft will be configured and flown to simulate the Vietnam spray missions and a herbicide simulant will be released. Industrial hygiene sampling techniques will be used to measure differential exposure for aircrewmembers, ground support personnel, and administrative staff members. These data will then be used to calculate a weighting factor for use in the exposure index. In this way, a common index can be applied to all 3 occupational categories. The individual records of flying time ("Form 5's") will be used wherever possible to more clearly define the opportunity for in-flight exposures. Adjustment of the exposure analyses for confounding factors such as age and time spent in Southeast Asia will also be conducted to refine the index and make it more specifically a measure of herbicide exposure. This exposure index will also be modified to assess the degree of exposure to other chemicals such as arsenical herbicides (Herbicide Blue) and malathion. **• Additional Statistical Analyses Expanded statistical analyses and procedures are planned on the baseline data of this study. More detailed statistical power estimates will be developed for the analyses contained in this report, and an overall assessment of the ability of this study to detect adverse health effects in the populations studied will be made. Specifically, the analyses of reported and verified birth defects will be reaccomplished with the nature of the anomalies categorized as severe, moderate, and of minor medical consequence. The defects will also be classified as being congenital or teratogenic in origin. The results of the semen analyses and the father's occupation will also be considered. Efforts will be made to more fully define and correct sources of potential bias in the subsets of the comparison group so that all analyses can be conducted using the entire group of comparison individuals. This will maximize study power, and allow the use of the replacement strategy outlined in the protocol. Additional matched pair analyses will also be conducted in each clinical area, thus taking full advantage of the most powerful statistical techniques. The full spectrum of clinical end points and covariables will be analyzed as well. Case by case reviews of individuals with testicular, bladder, oropharyngeal, and skin cancer and those with pulse abnormalities will be conducted. This review may highlight additional risk factors and may suggest alternative epidemiologic and statistical methodologies for subsequent reanalysis (e.g., case-control studies). XVIII-2 �Other techniques will be used to address correlations between clinical areas in the data. An organ system does not operate independently, and interactions between systems will be evaluated in subsequent reports. The effects of differential reporting are potentially significant in this study, and analyses aimed at differences in reporting between groups, and between study participants and their spouses will be evaluated. Questionnaire data was collected from the next-of-kin of deceased individuals and from totally noncompliant individuals, and time constraints have not permitted an analysis of these data. However, these are potentially valuable sources of information and appropriate evaluation will be conducted as time permits. Additional testing using more multi-variate techniques, expanded model-fitting, and goodness-of'-fit testing will also be carried out via contract. 5. Collaborative Activities Over the past 5 years, the principal investigators have worked closely with other organization and scientists involved in the herbicide/dioxin issue, and these collaborative activities will be strengthened and expanded. The common problems encountered by this study and the studies of Vietnam veterans being conducted by the Centers for Disease Control and the Veterans Administration can be more effectively resolved through the sharing of approaches and solutions. Collaboration has benefited all of these studies in the past, and should continue to be of benefit in the future. In addition to U.S. governmental agencies, the principal investigators have interacted with the. epidemiologic staffs at DOW Chemical Company, Monsanto Company and with researchers in Australia, New Zealand and Europe. The value of these interactions cannot be overstated, and these contacts will be maintained as the study progresses. More importantly, a closer working relationship will be developed between the principal investigators and the Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicide and Contaminants. Continued coordination with this panel will be invaluable as the complex findings of this study emerge over time. XVIII-3 �Chapter XIX INTERPRETATION OF STUDY RESULTS AND CONCLUSIONS 1. Introduction This section presents a cautionary note to both scientific and lay readers who may wish to assert that this study, in whole or in part, is supportive or nonsupportive of a causal relationship between exposure, to Herbicide Orange (and its dioxin contaminant) and adverse health. It is important to recognize that this observational study cannot prove the "negative," nor can it be construed as "definitive" science. The process of determining causality is complex and must entail a methodical consideration of many factors (Lilienfeld and Lilienfeld, 1980). 2. Causality Factors j In general, the following factors are very important in making an inference of causality: strength of association; dose-response; biologic plausibility; consistency; time relationships; specificity; and coherence. In an epidemiologic study, not all these factors are required to be present in order to make a correct inference, but clearly, substantial conflict between one or more factors casts doubt 6n an inference of causality. In this study, numerous group differences (associations) were detected and expressed in terms of probability (P) values. In any given analysis, statistically significant P values «0.05) represent the strength of the association, but in and of themselves, do not imply an herbicide causation. As expected under the null hypothesis, most a_ priori hypothesis tests were negative (P >0.05), but the validity of these findings must be assessed by the power of the given test. As expected, many positive associations were found in the clarifying analyses, or' as expressions of the influence of specific risk factors (e.g., age, smoking, etc.). Highly significant associations must also be viewed in the context of relative risk. A very significant association with a relative risk of less than two is generally of minor interest from the traditional epidemiologic perspective. In this study, only four objectively determined group differences of P <0.05 had a relative risk of two or greater. Moreover, statistically significant differences in the group means of a laboratory parameter were often detected, but the overall distributions were similar, the values were within normal range, and the clinical relevance of these shifts was not readily apparent (e.g., LDH, testosterone, T3, etc.). A positive linear dose response relationship is a substantial feature in establishing a cause and effect association. A careful counting of the 388 exposure index analyses cited in this report shows that only 11? are statistically significant, and only 2.855 are increasing from low to high XIX-1 �exposure. While these proportions are suggestive of chance associations, this possibility should be modified by the fact that positive exposure analyses, although not totally consistent throughout all occupational categories, tend to aggregate in only several of the organ systems. Additionally, it is recognized that the exposure index has not been fitted to the most specific format, as further experimental studies are still in progress. Thus, the exposure index used herein is a very indirect measure of exposure, making these analyses less certain than the observed group differences. Numerous other subcategorical exposure analyses (also predominantly negative) were accomplished, but were not included at the discretion of the author. Descriptive opinions of the positive exposure associations were often the sole choice of the responsible principal investigator within each chapter. The time interval from herbicide exposure to onset of subclinical or clinical manifestations is an important concept for proper interpretation of these study findings. The observational period for the detection of possible latent health effects ranges from 10-20 years for all Ranch Handers. While 10-20 years may be insufficient time for the induction of many systemic cancers, and possibly skin cancer, clearly it is of sufficient length to have already "caused" transient biochemical aberrations, birth outcome abnormalities, fertility problems, chloracne, porphyria cutanea tarda, neurologic sequelae, psychological deficits, etc. Thus, if the above acute/subacute conditions are found attributable in these data, it must be acknowledged that the end result of many of the disease processes is being observed. That notion must be reconciled with essentially identical mortality rates in both groups to date, as many of the proposed diseases would most likely have exerted a subtle mortality influence. Alternatively, the suggestion that the release of dioxin from fat may result in slow systemic poisoning, if true, may account for a delay of clinical manifestations beyond classically accepted latent periods. Another influential time-onset relationship is that of "crossover," i.e., a sequential time-disease association based upon a linkage to a pulsed exposure. While many pre/post-SEA analyses have been performed in this study, reapplication of exposure to herbicides (to complete the crossover) via non-SEA vocations or avocations has not, as yet, been exploited. Other causal factors merit comment. The finding of no cases of soft tissue sarcoma, porphyria cutanea tarda or chloracne in the Ranch Hand group may reflect a lack of specificity and/or a weak toxicity of the received dose of the putative agent (dioxin), or may reflect the low statistical power to detect group differences for these diseases in this study. The absence of these three diseases may also suggest that a synergism with a yet-to-be-discovered factor is required to induce disease. Findings of this study are, as yet, not fully consistent with other human dioxin studies performed in industrial populations. However, this inconsistency may be attributable to different exposure levels. In terms of biological plausibility, there is no discernible syndrome or symptom cluster that has emerged from this study that makes sense, has an XIX-2 �identifiable pathogenesis, or has an analogous animal model. A systemic poisoning theory carries with it the expectation of finding more biochemical abnormalities than were detected in this study. 3. Other Factors Chloracne has been proposed as a prerequisite to systemic disease. This premise is not wholly consistent with spectrum of illness concepts or other studies which have suggested attributable soft tissue sarcoma in predominantly nonchloracne populations. However, if the premise is true to the extent that the induction of chloracne represents moderate to high exposures to dioxin, then overall, it may be inferred that the Ranch Hand group (with no chloracne) has received relatively low exposure vis-a-vis industrial populations. Assuming a dose-response hierarchy, this inference may be extended to the contemplated studies of U.S. military ground personnel, for if the Ranch Hand study is deemed "negative," so probably will be the other studies of comparable size. The question of the validity of this study is paramount. Overall, the processes of data collection have been quite good. To the extent possible, biases have been minimized in both the data collection and data analytic phases. Notwithstanding, a general predominance of adverse findings can be noted in the Ranch Hand group. A closer inspection of this aggregation suggests that most statistically significant findings are found in the subjective data sets, as contrasted to the objective measures. Many of these subjective findings in the Ranch Hand group are in various stages of medical record verification at this time. Unfortunately, some areas, e.g., psychological testing by questionnaire, can never be totally verified. Throughout this study, there is a suggestion of differential reporting (MMPI K and Hypochondriasis scales), albeit unanalyzed, that must temper the interpretation of the subjective results. For the objective data, there is good evidence that the laboratory measurements and the clinical assessments were reasonably accurate. This study has duplicated the classical effects of numerous risk factors (age, smoking, alcohol, etc.) on the clinical measurements throughout all organ systems. The detected effects of age and smoking in the functional and count immunologic tests are new observations, to the best of our knowledge. Thus, the effects of these risk factors have been taken into account throughout the study and lend strong credence to the accuracy of the overall group associations, whether statistically significant or not. It is our belief that this physical examination has reflected the true health status of all participants and groups to the maximum extent possible. XIX-3 �4. Conclusions a. Preface This section places into context the thousands of statistical tests which have been accomplished on the enormous data bases generated by the population ascertainment efforts, and the administration of the in-home questionnaire and the physical examination. The total baseline study, including all preparatory tasks and the Baseline Mortality Report, has spanned more than 5 years, has required approximately 100 man-years of in-house work, and has cost about $11M in direct and indirect costs. The Ranch Hand study has been characterized by solid resource support and stringent timetables throughout all levels of government, intense media interest, and outstanding participation of the study subjects. As part of the mosaic of all dioxin research, the Ranch Hand study has been directed to the herbicide-health effect issue in veterans, and specifically, to heavily exposed Air Force personnel. b. Study Performance Aspects Of all live Ranch Hand and comparison individuals who were selected for this study, almost all (99.5%) were contacted, eliminating a major element of bias concern. Participation in the in-home questionnaire was 97% and 93% for the Ranch Handers and comparisons, respectively; and similarly 87% and 76% for physical examination. Differential compliance to the examination may have introduced a participation bias, a bias that is potentially related to the true health status of the participant. Age, race, participation in flying, and military status were also significant factors in determining attendance at the examination, but the relative contribution of each factor has not as yet been determined. Traditionally, individuals in either military or civilian commercial flying occupations do not readily volunteer for physical examinations that might disclose even minor ailments that jeopardize their flying careers. Early in the study, it was discovered that 18% of the comparison group was ineligible for the study because of inappropriate selection due to a computer programming error. Some selected USAF organizational units containing cargo-hauling aircraft were found not to be engaged in RVN duties (a study requirement). Thus, the direction of the error was for overselection and not for underselection of the comparison group. Ineligible individuals were removed from the randomly ordered comparison set. The replacements for the ineligibles were the next-in-line proper comparisons. For both these "shifted" comparisons and the next-in-line comparisons who were also used as substitutes for noncompliant individuals, later statistical analyses suggested that they differed from the original eligible comparisons in a variety of subtle and often opposite ways. Because of the possible bias suggested in their use, and because of the time constraints of this report, a conservative management decision was made to base the bulk of statistical tests upon a contrast of the original comparisons to the Ranch Hand group. Several analyses, using the entire XIX-4 �comparison group, were also performed and found not to differ consistently from the analyses based upon the original comparison group. For those analyses which showed differences between the original versus the total comparison group contrast, it is unclear whether these differences are primarily due to true subset variances or to a sample size effect. A full clarification of the complex biases (selection, compliance, overreporting, etc.) must be conducted before the first follow-up phase of the study. Most of the stringent quality control aspects of the study were monitored and maintained throughout the data collection phases. As a USAF contract requirement, all contractors were required to maintain "blindness" with respect to the exposure status of each individual, thereby reducing examiner bias to an absolute minimum. In addition, by contract all data are the property of the USAF. Study codes were not provided to the contractors. c. Clinical Aspects In terms of overall health, the Ranch Handers perceive their state of health to be poorer than that of the comparisons. This finding parallels the examiner's independent assessment. Percent body fat is similar in both groups as are the hematocrit determinations. A higher proportion of abnormal red cell sedimentation rates is found in comparisons under 40 years of age. The proportions are the same in both groups older than 40. The sedimentation rate, hematocrit, percent body fat, self-perception of health, and age are associated pairwise irrespective of group; these relationships are expected as all variables are traditional indicators of nonspecific illness. There are no significant group differences for malignant or benign systemic tumors. One case of soft tissue sarcoma is noted in a member of the comparison group. A slight nonsignificant aggregation of genitourinary cancers is identified in the Ranch Hand group, and an aggregation of digestive system cancers is observed in the comparison group. Two Ranch Hand bladder cancers are noted at earlier-than-expected ages. A borderline association between systemic cancer and s,mpking is observed in both groups. Significantly more nonmelanotic skin cancer is observed and verified by medical record review in the Ranch Handers. The predominant cancer, basal cell carcinoma, is the most common skin cancer in the U.S. White male population, and a proper excision is curative. While this finding is of interest, it is emphasized that these data are not adjusted for sunlight exposure, the recognized primary cause of these cancers. This analysis must await more complete data to be collected at the first follow-up examination. Overall there is no consistent data to show that the Ranch Handers are developing uncommon systemic cancers, or cancer in unusual sites, or at a younger age. Both systemic and skin cancers in the Ranch Hand group do not correlate consistently with the herbicide exposure index. XIX-5 �The fertility and reproductive analyses show mixed findings. As these results are largely based upon subjective self-reports, and must be verified by complete medical record and birth certificate reviews, the findings are judged preliminary at this time. A semen analysis on those participants willing and able to provide a specimen shows essentially identical sperm counts and percent abnormal forms between groups. The finding of an increase in sperm count by age is discounted as physiologically significant because of concomitant noncompliance by increasing age. Four measures of fertility show no difference between the Ranch Hand and comparison groups: number of childless marriages; couples with the desired number of children; the fertility index; and the infertility index. There are no significant findings in conception outcomes for miscarriages; stillbirths, induced abortions, or live births. With respect to live birth outcomes, no group differences are observed for prematurity, learning disability, or infant deaths. Birth defects, as cited by parental history, show no group differences for severe or moderate classifications; however, for minor birth defects (simple birth marks, birth rashes, port wine stains, etc.) Ranch Hand offspring show a significant excess. Reported neonatal deaths and physical handicaps significantly predominate in the Ranch Hand group when contrasted to the full comparison group. All analyses are adjusted for as many of the relevant risk factors as possible, e.g., maternal age, maternal smoking, maternal use of alcohol, paternal age, pre/post-RVN service, etc. Herbicide exposure analyses show several findings of statistical significance but the patterns of association are not fully consistent across all occupational categories. A thorough neurological assessment of the cranial nerves, peripheral nerves, and central nervous system functioning does not disclose any substantive Ranch Hand-comparison group differences. Past history of neurological disorders is similar for both groups. An increased proportion of abnormal Babinski reflexes are noted in the Ranch Handers but this finding is not statistically significant. Detailed nerve conduction velocities are not associated with group membership but are profoundly influenced by alcohol and diabetes. Similarly, abnormalities in sensation to light touch, vibration, and two reflexes are related to abnormal postprandial glucose levels. Exposure index analyses are predominantly negative. Detailed psychologic evaluations from the in-home questionnaire and physical examination show consistent findings. Educational level of the participant profoundly influences most all of the subjective test results. Due to the inherently high correlation between military rank and educational level, these variables are considered interchangeable. It is emphasized that the majority of psychologic data are based upon highly subjective self-reporting, most of which can never be fully verified by medical record reviews. There are no group differences for reported past emotional or psychological illnesses. However, the high school educated (mostly enlisted) Ranch Handers demonstrate significant findings or deficits in the following categories: fatigue, anger, anxiety, erosion, fear, startle, psychosomatic behavior, hypochondria, XIX-6 �Iriity, and manla/hypomanla. It is noted that the high school educated comparisons exhibit a higher degree of denial in most of these categories. These findings are not observed in the college educated Ranch Handers (mostly officers). The Ranch Hand group demonstrates significant hypochondria, depression, hysteria and schizophrenia vis-a-vis the comparison group, after adjustment for education. In sharp contrast, there are no substantial group differences for the more "objective functional and performance psychologic tests (e.g., Halstead-Reitan battery, IQ testing). Almost all exposure index analyses are negative. In full context, differential reporting is strongly suggested, albeit unproven. The roles of an overreporting bias and the Post-Vietnam Stress Syndrome will be clarified in subsequent follow-up psychological evaluations. The hepatic status is assessed by 9 biochemical tests and a variety of questionnaire and medical record data. The results are mixed. Ranch Hand GGPT and LDH levels are slightly higher while cholesterol levels are lower than the comparisons. Alcohol history is associated with most enzymatic elevations in both groups. Ranch Handers report significantly more skin changes compatible with a historical diagnosis of porphyria cutanea tarda (PCT). However, laboratory determinations for delta-aminolevulinic acid, uropo'rphyrin and coproporphyrin are similar between groups and no cases of PCT were diagnosed at the physcial examination. Reported miscellaneous liver disorders, verified by medical record reviews, are found significantly more in the Ranch Handers. The exposure index analyses are generally inconsistent. A comprehensive dermatologic evaluation reveals no substantial findings in the Ranch Hand group. No cases of chloracne are diagnosed clinically or by biopsy of suspicious lesions. Questionnaire data show that the incidence, severity, duration, and anatomic locations of past acne do not portray a pattern consistent with significant historical chloracne in the Ranch Handers. The classical "eyeglass" distribution of acne (suggesting chloracne) is the same in both groups. Historical acne correlates with the total cumulative acne found at physical examination. All exposure index analyses are negative. Examination of the central cardiovascular system reveals no remarkable differences between the groups for systolic blood pressure, diastolic blood pressure, abnormal electrocardiograms, past versus present electrocardiograms, or abnormal heart sounds. As expected, abnormalities in most of these parameters are significantly associated with age, smoking, and a past history of heart disease. The three risk factors: age, smoking, and cholesterol level are strongly associated with each other, and HDL cholesterol is significantly influenced by percent body fat and smoking. An analysis of questionnaire data shows that the Ranch Handers are not having premature heart attacks or generalized heart disease, although subset analyses show differing age and smoking effects. As an unexpected finding, two peripheral pulses are significantly diminished or absent in the Ranch Handers, and several other pulses show weak group differences. Clarifying statistical analyses show that XIX-7 �the the aggregate of Ranch Hand peripheral pulses, predominantly leg pulses, are significantly associated with age, past smoking, current smoking, and verified past heart disease. The weak but similar directional findings in the Ranch Hand carotid and femoral pulses are assigned more significance in view of the peripheral pulse observations. State-of-the-art measurement techniques and a specific medical questionnaire will be used to determine the relevance of these pulse deficits at the first follow-up examination. Detailed herbicide exposure analyses show no associations to any of the central or peripheral cardiovascular findings. Detailed immunological tests, via B and T lymphocyte enumeration and lymphocyte function studies on a randomized subset of all participants, do not demonstrate significant group differences. Because of the high variability of the quality control data, an independent peer review panel evaluated testing methodology and established criteria for analysis. The numbers of T-|-|, Tg, Tij, Tg, B-), positive cells and total lymphocyte counts are similar in both groups. Smoking history is observed to significantly affect the TII, T3» Ti|, Tg, marker counts and the total lymphocyte count. Age is seen to affect the Tg count and the total lymphocyte count. No group differences are observed for the functional studies using phytohemagglutinin, concanavallln A, pokeweed mitogen, and tetanus toxoid. Although the baseline proliferation rate (Control #1) was significantly lower in the Ranch Handers, the biologic relevance of this finding is unclear, particularly in the absence of group differences for concanavallin A and phytohemagglutinin stimulation studies. Age is observed to profoundly affect concanavallin A and phytohemagglutinin results while smoking history is seen to significantly influence pokeweed mitogen results. Because of the overall variability of quality control data, interpretation of a specific individual's immunocompetence is not attempted. Of 8 measured blood elements and parameters, the mean corpuscular volume and the mean corpuscular hemoglobin level are statistically significantly elevated in the Ranch Hand group, but the relative differences are exceptionally minor and are not of clinical relevance or understanding at this time. Seven of the 8 blood measurements are significantly affected by smoking history. Several exposure index analyses demonstrate positive correlations but a consistent pattern by occupational strata is not observed. There is no group difference in the distribution of reported past pulmonary disease. Forced expiratory volume for one second and forced vital capacity measurements obtained at the physical examination do not reveal group differences that are consistent in character. There are age/smoking/exposure interactions but it is not possible to further delineate these findings at this time. Several statistically significant herbicide exposure index analyses do not conform to classic dose-response relationships. XIX-8 �Ranch Handers report significantly more kidney disease than the comparisons but this history is not corroborated by 6 laboratory measurements obtained at the physical exam. Proteinuria is of borderline significance in the comparison group. Creatinine clearance may be considered of borderline significance in the Ranch Handers, depending on the laboratory value chosen to determine the abnormal category. Because of the substantial problem of compliance to the 2*1 hour urine collection, little credence is assigned to the creatinine clearance results. Age is observed to influence the blood urea nitrogen and urine specific gravity results while diabetes affected only the specific gravity results. Herbicide exposure analyses are essentially unrevealing. A comprehensive assessment of thyroid function and insulin and testosterone production show mixed results. Distributional shifts are noted in thyroid function between the Ranch Hand and comparison groups but the test results are generally within the limits of normal values. There are no group differences for diabetes as determined by abnormal 2 hour 'postprandial glucose levels. Age and percent body fat determinations are associated with abnormalities in T^ uptake, 2 hour postprandial glucose- levels, and testosterone levels. Herbicide exposure analyses show a variety of positive correlations but many are inconsistent across occupational strata. Evaluations of personal habits and individual health show that Ranch Handers currently smoke cigarettes more than the comparisons, equally participate in high risk sports activities, and have a similar background of traumatic injuries. An unrefined assessment of the total number of abnormalities found at the physical examination show no Ranch Hand aggregations in the high range nor do arbitrary clinically weighted scores. Overall, both groups are comparable in most health respects, and are probably faring better than similarly aged men in the general population. d. Final Conclusion This study has disclosed numerous medical findings, mostly of a minor or undetermined nature, that require detailed follow-up. In full context, the baseline study results should be viewed as reassuring to the Ranch Handers and to their families at this time, because this study has not identified statistical group differences . for illnesses commonly .attributed to dioxin exposure. The data herein suggest that group differences exist which tend to favor the comparisons, but the cause and clinical relevance of these differences is unclear. This baseline report concludes that there is insufficient evidence to support a cause and effect relationship between herbicide exposure and adverse health in the Ranch Hand group at this time. XIX-9 �REFERENCES Abramson JH, Tereapolsky L, Brook JG, Kark SL. Cornell Medipal Index as a Health Measure in Epidemiblogical Studies. A Test of the Validity of a Health Questionnaire. Brit J Prev Soc Med 19: 103-110, 1965. An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides - Baseline Mortality Study Results. Epidemiology Division, Data Sciences Division, USAF School of Aerospace Medicine, Brooks Air Force Base, Texas, 61 p, 1983. Available from NTIS, Springfield VA. Anderson S, Auquier A, Hauck WW, Oakes D, Vandaele W, Weisberg HI. Statistical Methods for Comparative Studies. John Wiley and Sons, New York, 1980. Arnold BC. Hypothesis Testing with a Preliminary Test of Significance. Stat Assoc 65: 1590-1596, 1970. Axelson 0, Sundell L. 2888, 1977. Phenoxy Acids and Cancer. J Am Lakartidningen 7*1(35): 2887- Bancroft TA. Analysis and Inference for Incompletely Specified Models Involving the Use of Preliminary Test(s) of Significance. Biometrics 20: 427-442, 1964. Bashirov AA. Health Conditions of Workers Producing Herbicides of Amine Salt and Butyl Ester of 2,4-D Acid. Vrach Delo 10: 92-95, 1969. Bastomsky CH. Enhanced Thyroxine Metabolism and High Uptake Goiters in Rats After a Single Dose of 2,3,7,8-Tetrachlorodibenzo*-p-dioxin. Endocrinol 101: 292-296, 1977. Bauer H, Schulz KH, Spiegelberg U. Occupational Intoxication in the Manufacture of Chlorophenol Compounds. Arch Gewerbepathol Gewerbehyg 18: 538-555, 1961. Berkley MC, Magee KR. Neuropathy following exposure to a dimethylamine salt of 2,4,-D. Arch Int Med 1 1 1 : 133-13^, 1963. Bernstein RA, Giefer EE, Rimm AA. Gallbladder Disease - I. Assessment of Validity and Reliability of Data Derived from a Questionnaire. A Study of 62,739 Weight-Conscious Women. J Chronic Dis 29: 51-58, 1976. Berwick P. 2,4-Dichlorophenoxyacetic Acid Poisoning in Man. Some Interesting Clinical and Laboratory Findings. J Am Med Assoc 214(6): 1114-1117, 1970. Bishop YMM, Feinberg SE, Holland PW. Discrete Multivariate Analysis: and Practice. The MIT Press, Cambridge MA, 1975. Theory �Boeri R, Bordo B, Crenna P, et al. Preliminary Results of a Neurological Investigation of the Population Exposed to TCDD in the Seveso Region. Riv Pat Nerv Ment 99: 111-128, 1978. Boyle JM, Stackpole A, Siegrist S. Final Report on the Air Force Health Survey: Operation Ranch Hand II Volume 1: Study Procedures and Results. Louis Harris and Associates, Inc, 1982. Bozivich IH, Bancroft TA, Hartley HO. Power of Analysis of Variance Test Procedures for Certain Incompletely Specified Models. Ann Math Stat 27: 1017~ 1013, 1956. Brand RJ, Rosenman RH, Sholtz RI, Friedman M. Multivariate Prediction of Coronary Heart Disease in the Western Collaborative Group Study Compared to the Findings of the Framingham Study. Circulation 53:348-355, 1976. Breslow N. Covariance Adjustment of Relative-Risk Estimates in Matched Studies. Biometrics 38:661-672, 1982. .<* .. Buckingham WA Jr. Operation Ranch Hand The Air Force and Herbicides in Southeast Asia 1961-1971. Office of Air Force History, United States Air Force, Washington DC, 1982: 9-69, 199-201. Case-Control Study of Congenital Anomalies and Vietnam Service (Birth Defects Study). Report to the Minister for Veterans' Affairs January 1983. 127 p. Australian Government Publishing Service, Canberra 1983. Christiansen RE, Vandenberg BJ, Oscsli FW. Incidence of Congenital Anomalies Among White and Black Live Births with Long-term Follow-up. AM J Pub Hlth 71:1333-1341, 1981 Cohen A. .To Pool or Not to Pool in Hypothesis Testing. 721-725, 1974. Colombotos J. Personal Versus Telephone Interviews: Public Health Rep 84: 773~782, 1969. J Am Stat Assoc 69: Effects on Responses. Cook TD, Campbell DT. Quasi-Experimentation. Design & Analysis Issues for Field Settings. Rand McNally Publishing Company, Chicago, 1979. Courtney KD, Gaylor DW, Hogan MD, et al. Science 168: 864-866, 1970. Crow KD. Chloracne. Teratogenic Evaluation of 2,4,5-T. Trans St John Hosp Derm Soc 56: 79-99, 1970. Dalstrom WG, Welsh GS, Dalstrom LE. An MMPI Handbook, Volume 2: Research plications. University of Minnesota Press, 1960. Ap- David J. The Organs and Cells of the Immune Systems. Rubenstein and Federman, eds. Scientific American Medicine, 1979. REF-2 �Dean JH, Luster MI, Doorman GA, et al. Assessment of Immunotoxicity Induced by the Environmental Chemicals 2,3,7,8~Tetrachlorodibenzo-p"dioxln, Diethylstilbestrol and Benzo(a)pyrene. Advances in Immunopharmacology. Hadden J, ed, Pergaraon Press, New York, 1981: 37^50. Dean JH, Murray MJ, Ward EC. Toxic Modifications of the Immune System. Doull J, Klaassen CD, Andur MO. eds. Casarett and Douell's Toxicology: The Basic Science of Poisons, 3rd ed. MacMillan, New York 1984 (in press). Dudley AW, Thapar NT. Fatal Human Ingestion of 2,H~D a Common Herbicide. Arch Path 9*»: 270-275, 1972. Fleiss JL. Statistical Methods for Rates and Proportions. Sons, New York, 1981. John Wiley and Fry HG, McNair S. Data Gathering by Long Distance Telephone. Public Health Rep 73: 831-835, 1958. Gibbons JD. Nonparametric Statistical Inference. New York, 1971 : 127. McGraw-Hill Book Company, Goldstein JA, Hickman P, Bergman H, Vos JG. Hepatic Porphyria Induced by 2,3,7,8~Tetrachlorodibenzo'~p-dioxin. Res Com Chem Path Pharmacol 6: 919--928, 1973. Goldstein NP, Jones PH, Brown JR. Peripheral Neuropathy After Exposure to an Ester of Dichlorophenoxyacetic Acid. J Am Med Assoc 171(10): 1306-1309, 1959. Hardell L, Sandstrom A. Case-control study: Soft'-Tissue Sarcomas and Exposure to Phenoxyacetic Acids or Chlorophenols. Br J Cancer 39: 711 <-717, 1979. Hay A. The Chemical Scythe. Plenum Press, New York, 1982: Herman JB. Mixed*-mode Data Collection: Applied Psychology 62: 399-^04, 1977. 89-146. Telephone and Personal Interviewing. J Hiss RG, Lamb LE. Electrocardlographic Findings in 122,0*13 Individuals. Circulation 25:9il7196l, 1962. Hodgdon JA, Marcinik EJ. Unpublished Data. Health Research Center, San Diego, 1983. Department of the Navy, Naval Hoffman RA, Kung PC, Hansen WP, Goldstein G. Simple and Rapid Measurement of Human T Lymphocytes and their Subclasses in Peripheral Blood. Proc Natl Acad Sci USA 77: 4914-4917, 1980. Homberger E, Reggiani G, Sambeth J, Wipf HK* The Seveso Accident: Its Nature, Extent and Consequences. Ann Occup Hyg 22: 327^370, 1979. REF-3 �Honchar PA, Halperin WE. 2,4,5,~T, Trichlorophenol, and Soft Tissue Sarcoma. Lancet, Jan 31: 268-269, 1981. Hyman H, et al, Interviewing in Social Research, University of Chicago Press, Chicago IL, 1954. Jirasek J, Kalensky J, Kubec K. Acne Chlorina and Porphyria Cutanea Tarda During the Manufacture of Herbicides, Part I. Czech Dermatol 48 (5): 306^-31 7, 1973. Jirasek L, Kalensky J, Kubec K, et al. Acne Chlorina, Porphyria Cutanea Tarda and Other Manifestations of General Intoxication During the Manufacture of Herbicides, Part 2, Czech Dermatol 49(3): 145-157, 1974. Johnson NL, Katz S. Distributions in Statistics. tributions-2. John Wiley and Sons, New York, 1970 Continuous Univariate Dis- Kale BK, Bancroft TA. Inference for Some Incompletely Specified Models Involving Normal Approximations to Discrete Data. Biometrics 23: 335-348, 1967. Kimbrough RD, Halogenated Biphenyls, Terphenyls, Napthalenes, Dibenzodioxins and Related Products. Amsterdam, The Netherlands Elsevier/North-Holland Biomedical Press, 1980. Kleinbaum DG, Kupper LL, Morgenstern H: Epidemiologic Research. Principles and Quantitative Methods. Lifetime Learning Publications, California, 1982. Kociba RJ, et al. Long-Term Toxicologic Studies of 2,3.7,8-Tetrachlorodibenzop-dioxin (TCDD) in Laboratory Animals. Ann NY Acad Sci 320: 397-404, 1979. Kociba RJ, Keeler PA, Park CN, Gehring PJ. 2,3,7,8~Tetrachlorodibenzo-p-dioxin (TCDD): Results of a 13~Week Oral Toxicity Study in Rats. Toxicol Appl Pharmacol 35: 553-574, 1976. Kociba RJ, Keyes DG, Beyer JE, et al. Results of two-year Chronic Toxicity and Oncogenicity Study of 2, 3, 7, 8'-Tetrachlorodibenzo~p-dioxin in Rats. Toxicol Appl Pharmacol 46:279-303, 1978. Kouri RE, Rude TH, Joglekar, et al. 2,3,7,8-Tetrachlordibenzo-p-dioxin as a Cocarcinogen Causing 3-Methylcholanthrene-Initiated Subcutaneous Tumors in Mice Genetically "Nonresponsive" at Ah Locus. Cancer Res 38(9): 2777~2783, 1978. Kramer CG. Health of Employees Exposed to 2,4,5<-T. Findings of the DOW Chemical Company, Corporate Medical Department, 2030 DOW Center, Midland, Michigan, 19 p, 1974. Lachar D. The MMPI: Clinical Assessment and Automated Interpretation. Western Psychological Services, 1974. REF-4 �Lamb JC, Moore JA, and Marks TA. Evaluation of 2,4-dichlorophenoxyacetic acid (2,4<-D), 2,4,5-trichlorophenoxyacetic acid (2,3,5-T), and 2,3,7,8-tetrachlorobidenzo-p-dioxin (TCDD) toxicity in C57BL/6,mice: Reproduction and fertility in trated male mice and evaluation of congenital malformations in their offspring. National Toxicology Program, Research Triangle Institute, Research Triangle Park NC. Report No. NTP-80>-44. 57 pp. Lancaster MC, Ord JW. The USAF Central Eleetrocardiographic Library. Medical Service Digest 23-.8MO, 1972. Last JM. Maxcy-Rosenau Public Health and Preventive Medicine, 11th ed. New York: Appleton-Century-Crofts. 1980. Lathrop GD, Moynahan PM, Albanese RA, Wolfe WH. Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Baseline Questionnaires. USAF School of Aerospace Medicine Technical Report 82-42, 276 p, 1982. Available from NTIS, Springfield VA. Lathrop GD, Wolfe WH, Albanese RA, Moynahan PM. Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Study Protocol. USAF School of Aerospace Medicine Technical Report 82-44, 172 p, 1982. Available from NTIS, Springfield VA. Lilienfeld A. Reviews in Cancer. Epidemiology, Vol.1, Elsevier North Holland, New York, 1980. Lilienfeld AM, Lilienfeld DE. Foundations of Epidemiology. University Press, New York, 1980. 2nd ed. Oxford Louis Harris and Associates, and Inc. Myths and Realities: A Study of Attitudes Towards Vietnam Era Veterans. Veterans Administration, Washington DC, 1980. McCormick MC, Shapiro S., Starfield B. Factors Associated with Maternal Opinion of Infant Development—Clues to the Vulnerable Child? Pediatrics 69:537-543, 1982 McNulty WP. Toxicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin for Rhesus Monkeys: Brief Report. Bull Environ Contam Toxicol 18: 108-109, 1977. Monarca G, di Vito G. Acute poisoning from weed killer (2,4-dichlorophenoxyacetic acid). Folia Med 44: 480-485, 1961. Monson R. R. 1980. Occupational Epidemiology. CRC Press Inc, Boca Raton, Florida, Mulcahy MT. Chromosome Aberrations and "Agent Orange." Med J Aust 2: 573"574, 1980. REF-5 �Nienstadt W, Parkki M, Uotila P, Aitio A. Effects of 2,3,7,7-Tetrachlorodibenzo-p-dioxin on the Hepatic Metabolism of Testosterone in the Rat. Toxicol 13: 233-236, 1979. Oliver RM. Toxic Effect of 2,3,7,8-Tetrachloro-dibenzoH,4-dioxin in Laboratory Workers. Br J Ind Med 32: 46-53, 1975. Paggiaro PL, Martino E, Mariotti S. A case of 2,4-Dichlorophenoxyacetic Acid (2,4-D) Intoxication. Med Lav 65(3-4): 128M35, 1974. Poland A, Greenlee WF, Kende AS. Studies on the Mechanism of Action of the Chlorinated Dibenzo-pr-dioxins and Related Compounds Ann NY Acad Sci 320: 214-230, 1979. Poland AP, Smith D, Metter G, Possick, P. A Health Survey of Workers in a 2,4-D and 2,4,5"T Plant. With Special Attention to Chloracne, Porphyria Cutanea Tards, and Psychologic Parameters. Arch Environ Health 22(3): 316^327, 1971. Potter CL, Sipes IG, Russell DH. Hypqthyroxinemia and Hypothermia in Rats in Response to 2,3,7,8-Tetrachlorodibenzo-p~dioxin Administration. Toxicol Appl Pharmacol 69: 89-95, 1983. Reggiani G. Acute human exposure to TCDD in Seveso, Italy. J Toxicol Environ Health 6: 27-43, 1980. Reinherz EL, Schlossman SF. The Differentiation and Function of Human T Lymphocytes. Cell 19: 821-827, 1980. Robins LN, Helzer JE, Ratcliff KS, Seyfried W. Validity of the Diagnostic Interview Schedule, Version II: DSM-III Diagnoses. Psychol Med 12: 1855*1870, 1982. Rubin P. 1974. Current Concepts in Cancer. Chicago, American Medical Assocation, Schantz SL, Barsott DA, Allen JR. Toxicological Effects Produced in Nonhuman Primates Chronically Exposed to Fifty Parts per Trillion 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD). Toxicol Appl Pharmacol 48: A180, 1979. Schottenfield D, Fraumeni J. Cancer Epidemiology and Prevention. phia, W.B. Saunders, 1982: 926, 950-951. Philadel- Scotto J, Fears TR, Fraumeni JF. Cancer Epidemiology and Prevention. W. B. Saunders, Philadelphia, 1982: 254-276. Simon RJ, Fleiss JI, et al. Two Methods of Psychiatric Interviewing: phone and Face to Face. J Psychol 88: 141-146, 1974. Tele- Stevens KM. Agent Orange Toxicity: A Quantitative Perspective. Human Toxicol 1: 31-39, 1981. REF-6 �Thigpen JE, Faith RE, McConnell EE, Moore JA. Increased Susceptibililty to Bacterial Infection as a Sequela of Exposure to 2,3,7,8-Tetrachlorodibenzo-pdioxin. Infection and Immunity 12: 1319-1324, 1975. Thigpen JE, McConnell EE, Moore JA, Faith RE. Effects of 2,3,7,8-Tetrachlorodibenzo-p~dioxin (TCDD) on Host Resistance to Infectious Agents. Environ Health Perspec 20: 245, 1977. Thomas DG. Exact Confidence Limits for the Odds Ration in a 2 x 2 Table. Applied Statistics Appl Stat 20:105-110, 1971. Toth K, Somfai-Relle S, Sugar J, Bence, J. Carcinogenic!ty Testing of Herbicide 2,4,5<-Trichlorophenoxyethanol Containing Dioxin and Pure Dioxin in Swiss Mice. Nature 278: 538-549, 1979. Townsend JC, Bodner KM, Ven Peenen PFD, Olson RL, Cook RR. Survey of Reproductive Events of Wives of Employees Exposed to Chlorinated Dioxins. Am J of Epidemiology 115(5): 695-713, 1982. Wallis WE, Van Poznak A, Plum F. Generalized Muscular Stiffness, Fascicula'tions, and Myokymia of Peripheral Nerve Origin. Arch Neurol 22: 430-439. 1970. Westing AH. Ecological Consequences of the Second Indochina War. Stockholm International Peace Research Institute. Almgrist and Wiksel Internation, Stockholm, Sweden, 119 p, 1976. Wintrobe MM, Lee GR, Boggs DR, Bithell TC, Athens JW, Foerster J. Hematology, Lea & Febiger, Philadelphia, 1974. Clinical Wolfe WH, Lathrop GD. A Medical Surveillance Program for Scientists Exposed to Dioxins and Furans. Tucker RE, Young AL, Gray AP, eds. Human and Environmental Risks of Chlorinated Dioxins and Related Compounds. Plenum Press, New York, 1983: 707-716. Young AL, Calcagni JA, Thalken CE, Tremblay JW. The Toxicology, Environmental Fate, and Human Risk of Herbicide Orange and Its Associated Dioxin, Technical Report OEHL-TR-78-92, USAF Occupational and Environmental Health Laboratory, Brooks AF&, Texas, 247 p, 1978. Zack J, Suskind R. The Mortality Experience of Workers Exposed to Tetrachlorodibenzodioxin as a Trichlorephenol Process Accident. J Occup Med 22: 11-14, 1980. REF-7 �Appendix I PRINCIPAL INVESTIGATORS AND KEY PERSONNEL A. Principal Investigators George D. Lathrop, MD, MPH, PhD, FACPM Colonel, USAF, MC Chief, Epidemiology Division William H. Wolfe, MD, MPH< -SLColonel, USAF, MC Chief, Epidemiology Services Branch Epidemiology Division Richard A. Albanese, MD, GM-15 Chief, Biomathematical Modeling Branch Data Sciences Division Patricia M. Moynahan, BSN, MS Colonel, USAF, NC Chief, Occupational Epidemiology Section Epidemiology Division B. Coinyesti gators Joel E. Michalek, PhD, GS-13 Mathematical Statistician Data Sciences Division Alton J. Rahe, MS, GS-13 Mathematical Statistician Data Sciences Division John R. Herbold, DVM, MPH, PhD, Dipl ACVPM Lt Colonel, USAF, BSC Chief, Disease Surveillance Section Epidemiology Division Richard C. McNee, MS, GM-13 Chief, Advanced Analysis Branch Data Sciences Division AI-1 �C. Contributors F. Page Armstrong, BSN Lt Colonel, USAF, NC Infection Control Consultant Epidemiology Division William J. Besich, BS, GS-12 Computer Systems Analyst Data Sciences Division Vincent V. Elequin, BS, RRA, GS-11 Medical Record Librarian Epidemiology Division William G, Jackson Jr., MS, 03-12 Mathematical Statistician Data Sciences Division Thomas V. Murphy, MBA, GS-9 Statistical Assistant Data Sciences Division William E. Nixon, BS, GM-13 Computer Systems Analyst Data Sciences Division Carolyn J. Oakley, BA, GS-11 Mathematical Statistician Data Sciences Division Kenneth P. Pankratz MSgt, USAF NCOIC, Project Ranch Hand II Epidemiology Division Melody Prihoda, MS, GS-9 Mathematical Statistician Data Sciences Division Vianney V. F. Simonnet, BS 1st Lieutenant, USAF Chief, Data Control Epidemiology Division Clarence F. Watson, Jr., MD, MPH, FACPM Life Extension Institute 540 North Michigan Avenue Chicago IL 60611 AI-2 �Thomas J. White, MA Senior Subject Matter Specialist Data Sciences Division James A. Wright, MD, MPH, FACPM Major, USAF, MC Chief, Preventive Medicine Headquarters Pacific Air Force Hickam AFB HI 96853 D. Medical Specialty Consultants Jerry J. Tomasovic, MD Colonel, USAF, MC Department of Neurology Wilford Hall USAF Medical Center Lackland AFB TX John S. Silva, MD Major, USAF, MC Department of Surgery USAF Medical Center Keesler Keesler AFB MS Richard N. Boswell, MD Lt Colonel USAF, MC Department of Medicine Wilford Hall USAF Medical Center Lackland AFB TX John W. White, MD Colonel, USAF, MC Department of Dermatology Wilford Hall USAF Medical Center Lackland AFB TX Gary L. Mueller, MD Lt Colonel, USAF, MC Department of Medicine Wilford Hall USAF Medical Center Lackland AFB TX Leonard Gardner, PhD Captain, USAF, BSC Department of Neuropsychology Clinical Sciences Division AI-3 �E< Ifflrcuriology, Review Panel J. George Bekesi, MD Dept 6f Neoplastic Disease and the Environmental Sciences Laboratory Mount Sinai School of Medicine New York NY 10029 Richard N. Boswell, MD Lt Col, USAF, MC Department of Medicine Wilford Hall USAF Medical Center Lackland AFB tX 78236 Jack H. Dean, MD Immunotoxicology Division Chemical Industries Institute of Technology Research Triangle Park NC John S. Silva, MD (Chair) Major, USAF, MC Department of Surgery USAF Medical Center Keesler Keesler AFB MS 39534 F. Management Personnel Project Director: Royce Moser, Jr., MD, MPH, FACPM Colonel, USAF, MC Commander, USAF School of Aerospace Medicine Project Resource Manager: G. Ray Sims, BS, MA, MSEE Major, USAF Chief, Engineering Division Director for Systems Acquisition Aerospace Medical Division Air Force Systems Command Coordinator: John H. Wolcott, PhD Colonel, USAF, BSC Deputy Chief of Staff for Medical and Life Sciences HQ, Air Force Systems Command �Air Force Surgeon General Coordinator: Robert A. Capell, BS, MS Lt Colonel, USAF, BSC Assistant for Bioenvironmental Engineering Office of the Surgeon General G. On-Site Physical Examination Monitors George D. Lathrop, MD, MPH, PhD, FACPM Colonel, USAF, MC Chief, Epidemiology Division William H. Wolfe, MD, MPH, FACPM Lt Colonel, USAF, MC Chief, Epidemiology Services Branch Epidemiology Division James A. Wright, MD, MPH, FACPM Major, USAF, MC Chief, Preventive Medicine Headquarters, Pacific Air Force Hickam AFB HI 96853 Clarence F. Watson, Jr., MD, MPH, FACPM Colonel, USAF, MC (Ret) Life Extension Institute 540 North Michigan Avenue Chicago IL 60611 Roy L. DeHart, MD, MPH, MS, FACPM Colonel, USAF, MC (Ret) Director, Times of Oklahoma, Inc. 1511 North Sheridan Rd Tulsa OK 71115 James A. Boydstun, MD Colonel, USAF, MC Deputy Surgeon AF Systems Command Andrews AFB MD 20334 Paul H. Grundy, MD, MPH, FACPM Captain, USAF, MC US Military Training Mission Dhahran, Saudi Arabia AI-5 �Appendix II ADVISORY COMMITTEE ON SPECIAL STUDIES RELATING TO THE POSSIBLE LONG-TERM HEALTH EFFECTS OF PHENOXY HERBICIDES AND CONTAMINANTS George W. Corastock, MD, MPH, DrPH Professor" of Epidemiology Johns Hopkins University Johns Hopkins Research Center Box 2067 Hagerstown MD 21740 John Doull, PhD, MD Professor of Pharmacology and Toxicology Department of Pharmacology University of Kansas Medical Center Kansas City KA 66103 Robert W. Miller, MD, MPH, DrPH (Chair) Chief, Epidemiology Branch National Cancer Institute Bethesda MD 20205 Richard R. Monson, MD, ScD Professor of Epidemiology Harvard School of Public Health 677 Huntington Avenue Boston MA 02115 John A. Moore DVM, MS (Former Chair) Director, Toxicology Research and Testing Program National Institute of Environmental Health Sciences P.O. Box 12233 Research Triangle Park NC 27709 Norton Nelson, PhD Professor of Environmental Medicine Institute of Environmental Medicine New York University Medical Center 550 First Ave New York NY 10016 Alan P. Poland, MS, MD Associate Professor of Oncology McCardle Laboratory for Cancer Research University of Wisconsin-Madison Madison WI 53706 Irving J. Selikoff, MD Director, Environmental Sciences Laboratory Mt Sinai School of Medicine New York NY 10029 All �Appendix III CONTRACT MANAGEMENT The Aerospace Medical Division, Air Force Systems Command, Brooks AFB TX, was designated as the primary management agency responsible for the Air Force Health Study. The program is managed by the Commander, USAF School of Aerospace Medicine, with scientific, technical, and business management support from the Epidemiology Division and the Data Sciences Division of the USAF School of Aerospace Medicine and business support from the Director for Systems Acquisition, Aerospace Medical Division, respectively. The Commander, USAF School of Aerospace Medicine, coordinates business and technical inputs from the interfacing organizations and consolidates program status and direction. He is responsible for informing higher headquarters of management or technical situations which could impact the success of the program. The 3303rd Contracting Squadron, Air Training Command, Randolph AFB TX, provides all procurement support to the Ranch Hand II Program. Contracted efforts, to date, have included software development, statement of work preparation, questionnaire development, questionnaire administration and the conduct of physical examination. To the maximum extent practical, fixed price contracts with cost reimbursement for travel, lodging and stipend expenses were used. The contractor(s) provided data as required to the contracting agency and the Program Manager. Reports were provided on technical progress, expenditure of funds, and overall program progress against the contractual schedules. Data were used to assess program progress and to initiate corrective actions where required. A contract to assist in the development of a statement of work for the questionnaires was let to Research Statistics, Inc., Houston TX, at a cost of $11 ,900. The study questionnaire was developed by the National Opinion Research Center, New York NY, a nationally recognized survey research firm. A sole source contract was awarded on 26 September 1980 and was concluded on 31 July 1981 at a total cost of $3^8,000. Louis Harris and Associates, Inc., New York NY, was competitively selected to administer the questionnaire and awarded a contract on 18 September 1981. The original effort was scheduled for completion in April 1982, but due to data collection as well as questionnaire/physical examination contractor interface requirements, the contract was extended to November 1982. The final cost for the questionnaire administration effort was $1.076 million. AIII-1 �A formal source selection process was also used to select the Kelsey-Seybold Clinic, P.A., Houston TX, as the single site for conducting the physical examinations. The Initial contract period was scheduled for 10 months (23 November 1981 - 30 September 1982) but was extended to 15 December 1982. The total contract cost was $6.161 million, Which included the physical examinations, travel expenses, lodging, meals and stipend allowances. An Air Force on-site physician monitor in-briefed all study participants and conducted quality control checks on all medical aspects of the physical examination. Additional medical and contracting specialists periodically visited the examination site to ensure adherence to all aspects of the contract. All three contracting efforts were characterized by this type of close interaction and control. AIII-2 �Appendix IV KELSEY-SEYBOLD NORMAL VALUE REPORT BLOOD CHEMISTRY AGE-ADJUSTED NORMALS AGES PARAMETERS 10 - 29 Years BUN (mg/dl): 10-26 Great (mg/dl): 0.7-1.4 Glue (mg/dl): 70-115 Choi (mg/dl): 106-210 Trig (mg/dl): 30-110 HDL (mg/dl): 32-72 T Bil (mg/dl): 0^2-1 .'2 D Bil (mg/dl): 0-0.36 Alk Phos (U/dl): 2.5-9.7 SCOT (U/L): 0-41 SGPT (U/L): 0-45 GGTP (U/L): 15~85 LDH (U/L): 0-200 CPK (U/L): 35-232 Alcohol (mg/dl): None 30 - 39 Years BUN (mg/dl): 10-26 Great (mg/dl): 0.7-1.4 Glue (mg/dl): 70H15 Choi (mg/dl): 119-240 Trig (mg/dl): 30-150 HDL (mg/dl): 32~72 T Bil (mg/dl): 0.2-1.2 D Bil (mg/dl): 0-0.36 Alk Phos (U/dl): 2.5~9.7 SCOT (U/L): 0-41 SGPT (U/L): 0-45 GGTP (U/L): 15-85 LDH (U/L): 0-200 CPK (U/L): 35-232 Alcohol (mg/dl): None 40 - 49 Years BUN (mg/dl): 10-26 Great (mg/dl): 0.7-1.4 Glue (mg/dl): 70-115 Choi (mg/dl): 131-265 Trig (mg/dl): 30-160 HDL (mg/dl): 32~72 T Bil (mg/dl): 0.2-1.2 AIV-1 �D Bil (mg/dl): 0-0.36 Alk Phos (U/dl): 2.5~9.7 SCOT (U/L): 0-41 SGPT (U/L): 0-45 GGTP (U/L): 15~85 LDH (U/L): 0-200 CPK (U/L): 35-232 Alcohol (mg/dl): None 50 - years and older BUN (mg/dl): 10-26 Great (mg/dl): 0.7~1.4 Glue (mg/dl): 80-125 Choi (mg/dl): 144-265 Trig (mg/dl): 30-190 HDL (mg/dl): 32-72 T Bil (mg/dl): 0.2-1.2 D Bil (mg/dl): 0-0.36 Alk Phos (U/dl): 2.5-9.7 SCOT (U/L): 0-41 SGPT (U/L): 0-45 GGTP (U/L): 15-85 LDH (U/L): 0-200 CPK (U/L): 35-232 Alcohol (mg/dl): None Unknown BUN (mg/dl): 10*-26 Great (mg/dl): 0.7-1.4 Glue (mg/dl): 70-125 Choi (mg/dl): 106-265 Trig (mg/dl): 30-190 HDL (mg/dl): 32-72 T Bil (mg/dl): 0.2H.2 D Bil (mg/dl): 0-0.36 Alk Phos (U/dl): 2.5~9.7 SGOT (U/L): 0-41 SGPT (U/L): 0-45 GGTP (U/L): 15~85 LDH (U/L): 0-200 CPK (U/L): 35-232 Alcohol (mg/dl): None AIV-2 �Appendix V DEFINITION OF BIRTH DEFECTS, LEARNING DISABILITIES AND PHYSICAL, MENTAL OR MOTOR IMPAIRMENTS Birth Defects ICD-9 Code 7*10 741 742 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 216 228 239.2 363.2 426.7 524.0 550 550.1 550.9 553.1 553.29 658.8 685.1 778.6 Condition Anencephalus and similar anomalies Spina Bifida Other nervous system anomalies Anomalies of eye Anomalies of ear, face, and neck Bulbus cordis/cardiac septal closure anomalies Other anomalies heart (valves) Other anomalies of circulatory system Other anomalies of respiratory system Cleft palate and cleft lip Other anomalies of upper alimentary tract Other anomalies of digestive system Anomalies of genital organs Anomalies of urinary system Certain congenital musculoskeletal deformities Other anomalies of limbs Other musculoskeletal anomalies Anomalies of the integument Chromosomal anomalies Other and unspecified anomalies Benign neoplasm of skin Hemangioma and Lymphangioma, any site Neoplasms of unspecified nature of bone, skin, connective tissue Chorioretinitis Wolff-Parkinson-White syndrome Major anomalies of jaw size Inguinal hernia gangrene Inguinal hernia with obstruction, no mention of gangrene Inguinal hernia, no mention of obstruction or gangrene Umbilical hernia Epigastric hernia Amniotic bands (constricting bands) Pilonidal Sinus or dimple Hydrocele AV-1 �Learning Disabilities (Developmental Delays) 313 314 315 317 318 319 Disturbance of emotions specific to childhood and adolescence Hyperkinetic syndrome of childhood Specific delays in development Mild mental retardation Other specified mental retardation Unspecified mental retardation Physical, Mental, Motor Impairments 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 270 271 272 273 274 275 276 Fetus or newborn affected by maternal conditions which may be unrelated to present pregnancy Fetus or newborn affected by maternal complications of pregnancy Fetus or newborn affected by complications of placenta, cord and membrane Fetus or newborn affected by other complications of labor and delivery Slow fetal growth and fetal malnutrition Disorders relating to short gestation and unspecified low birthweight Disorders relating to long gestation and high birthweight Birth trauma Intrauterine hypoxia and birth asphyxia Respiratory distress syndrome Other respiratory conditions of fetus and newborn Infections specific to the perinatal period Fetal and neonatal hemorrhage Hemolytic disease of fetus or newborn, due to isoimmunization Other perinatal jaundice Endocrine and metabolic disturbances specific to the fetus and newborn Hematological disorders of fetus and newborn Perinatal disorders of digestive system Conditions involving the integument and temperature regulation of fetus and newborn Disorders of amino-acid transport and metabolism Disorders of carbohydrate transport and metabolism Disorders of lipoid metabolism Disorders of plasma protein metabolism Gout Disorders of mineral metabolism Disorders of fluid, electrolyte, and acid-~base balance AV~2 �277 278 279 340 3^1 3^3 359 250 Other and unspecified disorders of metabolism Obesity and other hyperalimentation Disorders involving the immune mechanism Multiple sclerosis Other demyelinating diseases of central nervous system Infantile cerebral palsy Other paralytic syndromes Epilepsy Muscular dystrophies and other myopathies Diabetes mellitus AV-3 �Appendix VI PHYSICAL EXAMINATION FORMS Patient History and Health Questionnaire Conduct of the Examination (Internal Medicine) Neurological Examination Specialty Examination-Dermatology Pulmonary Function Diagnostic Summary AVI-1 �Pg. 1 of 7 KS# NAME DATE: PATIENT'S HISTORY AND HEALTH QUESTIONNAIRE FAMILY HISTORY; HAVE ANY MEMBERS OF YOUR FAMILY EVER HAD THE FOLLOWING? IF SO, PLEASE CHECK BELOW AND NOTE WHICH FAMILY MEMBER. Mother Father Sister Child Brother Diabetes Epilepsy Cancer High Blood Pressure Heart Disease Stroke Allergy Stomacn Trouble Nervous Trouble Blood Disease Deformities Arthritis Other familial diseases: Please list: FATHER: Living-Age Dead-Age _ MOTHER: Living-Age Dead-Age NUMBER BROTHERS: Living Condition of Health? Cause of Death? Condition of Health? Cause of Death? Ages: Dead Causes: NUMBER SISTERS: Ages: Dead Causes: Living ARE YOU MARRIED? WIFE'S AGE: NO. OF YEARS HUSBAND'S AGE Health of Husband or Wife? If spouse dead, give age, year, and cause of death:_ Previous Marriages?_ NUMBER OF CHILDREN: Boys: _ All Healthy? If yes, explain Gives dates: Ages: Girls: Any dead?_ Any birth defects?_ PLEASE LIST HERE ANY PHYSICAL OR NERVOUS COMPLAINTS WHICH YOU HAVE: KS-AF-1 AVI-2 Ages: �Pg. 2 of 7 PERSONAL HISTORY Allergy or severe reaction to medicines, foods, plants, chemicals, etc.: Please list: List Average Do you take regular exercise? What is your usual weight? What is the most you ever weighed? At what age or year? Have you lost or gained weight? If so, how much? Hours worked per day: Hours sleep per night: Days worked per week: Days vacation per year: Number cigarettes per day: Other tobacco per day: Cups coffee per day: Alcoholic drinks per day: PUT A CIRCLE AROUND ANY OF THE FOLLOWING CONDITIONS WHICH YOU NOW HAVE OR HAVE HAD IN THE PAST: Cataracts Tonsillitis Sinusitis Goiter Hay fever Asthma Bronchitis Pleurisy Pneumonia Tuberculosis Breast Trouble Heart Trouble Stomach Trouble Gallstones Ulcer Jaundice Liver trouble Hepatitis Worms Dysentery Colitis Loss of sensation Loss of sex drive Polio Mumps Measles Hemorrhoids Skin Trouble Acne Excess hair growth Change of skin color Other Malaria Arthritis Gout Kidney Trouble Kidney Stones Bladder Trouble Rheumatic fever Anemia Prostate Trouble Syphyilis Diabetes Gonorrhea Hernia (rupture) Varicose Veins Phlebitis Fainting Fits or convulsions Nervous Breakdown Depression Paralysis Muscle Pain Muscle weakness Numbness AVI-3 Cancer or Tumor Rheumatoid Arthritis Severe A r t h r i t i s Systemic Lupus E r y t h e r n a t o s u ; Scleroderma �Pg. 3 of 7 ' PAST HISTORY: Please list previous operations, injuries, serious illnesses, etc. and year; including those checked off above. 1. 2. 3. 4. ; 5. . 6. When was your last physical examination? Any abnormality found? Are you under any medical treatment now? List any medications you take now or occasionally: PERSONAL PHYSICIAN: Name Street Address City, State § Zip Code IF YOU HAVE HAD REPEATED CASES OF ANY OF THE FOLLOWING IN THE PAST YEAR, PLEASE CIRCLE Pneumonia Kidney Infections Skin Boils Other Infections (specify) Ks-AF-1 rev. AVI-4 �Pg. 4 of 7 IF YOU HAVE ANY OF THE FOLLOWING COMPLAINTS, PLEASE CIRCLE YES, IF NOT, CIRCLE NO. THE DOCTOR WILL ASK ABOUT DETAILS LATER. ANSWER ALL QUESTIONS. IF IN DOUBT, GUESS YES OR NO. Severe headaches or head pains .............. Yes Do you have: Any disturbance in vision ... ............ Yes Pain or discomfort in eyes ............... Yes Wear glasses .......... . ............ Yes Constant noise in ears . . . . . . ............ Yes Hard of hearing ........ . ............ Yes Ear ache with colds ( ) plane flights ( ) ....... Yes Chronic running ear ...... .... ......... Yes Chronic stuffy or runny nose ............... Yes Need to use nose drops frequently ............ Yes Bad nose bleeds at times ............ ^ . . . . Yes Frequent severe colds or sore throat ........... Yes Any known dental problems ................ Yes Soreness or bleeding of gums ............... Yes More than a year since teeth checked ... ........ Yes Sore mouth or tongue ................... Yes Goiter or thryoid trouble ................ Yes Thyroid test -- too high ( ) too low ( ) ........ Yes Feeling of lump in the throat .............. Yes Need to take thyroid medicine .............. Yes Hoarseness at times ................ ... Yes Recent or chronic cough ................. Yes Chronic coughing up of sputum ........... ... Yes Ever coughed up blood .................. Yes Ache all over ...................... Yes Having chills or fever .................. Yes Severe soaking night sweats ............... Yes Lived with anyone having T.B ............... Yes Worried about your heart . * ............... Yes Blood pressure -- too high ( ) too low ( ) ...... Yes Pains in heart or chest ................. Yes Pounding or skipping of heart .............. Yes Heart starts racing suddenly ............... Yes Shortness of breath or wheezing ............. Yes KS-AF-1 No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No �Pg. 5 of 7 Trouble getting a deep breath Swelling ankles Leg cramps in bed or sitting still Leg cramps while walking Pain or trouble with swallowing Poor appetite -- recently ( ) always ( ) Nausea or vomiting Vomiting of blood Belching, bloating or indigestion Yellow skin or eyes (jaundice) Burning or hunger pains in stomach Use antacids for stomach burning Soreness or pain in stomach, abdomen Suspect ulcers or stomach trouble Cramps in stomach or low down Loose bowels or diarrhea Black or tarry stools (bowel movement) Fresh or bright blood with stools Miicus (slime or plegm) in stools Constipation Use laxatives ( ) or enemas ( ) frequently v Recent change in bowel habits Rectal trouble or pain List any foods which always disagree: Pain in the kidney region Get up nights to urinate (Number of times Blood or pus in urine Albumin in urine Sugar in urine Spells of frequent urination Severe burning or pain on urination Pains over bladder or low down Trouble starting urine Urinary stream has become weak Hard to empty bladder completely Lose control of passing urine Painful or sore genitals (privates) KS-AF-1 AVI-6 Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes ) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes '. Yes Yes No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No �Pg. 6 of 7 Swollen or painful joints . . . . . ..... Stiffness of muscles or joints . Severe pains in arms or legs . Painful feet Backache Pains in neck , . . . * Easy to sunburn Itch or rash (where?) ____ , _ Subject to acne . . * Subject to boils or infections Subject ot athlete's foot, skin fungus Subject to hives or skin reactions Easy bleeding or bruising Mole or sore which is not healing Swelling, lump, or soreness anywhere on body (where?)_ Severe dizziness Numbness or tingling (where?) , ; _ Twitching muscles (where?) Generalized weakness Muscle weakness Nail biting Sleep walking Bed wetting after age 12 Chronically tired or overworked Irregular living habits Can't go to sleep Or stay asleep Nearly always in poor health Prom sickly or nervous family Considered to be a nervous person Tremble and sweat easily Have trouble making up your mind Easily mixed up or confused Clumsy or have frequent accidents Feel sad, lonely or depressed Cry often Wish you were dead Worry continually .* _, , ._ » . KSAF1 AVT-7 Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No No No No No No No No No No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No No No No No No No No No No No No No No No No No No No No �Page 7 of 7 Upset by little things Yes No A perfectionist Yes No Sensitive or feelings easily hurt Yes No . Yes No Often act on sudden impulse Yes No Easily angered or have violent rages Yes No Frequently keyed up and jittery Yes No Easily scared by sudden noise Yes No Yes No Yes i No . . . . Yes No Often misunderstood Have bad dreams or thoughts " Suspect a serious disease or cancer Having trouble getting along with someone at home or work Have you ever been exposed to any of the following substances or types of radiation? Exposure is defined as skin or respiratory contact more than one day's duration. 1. Coal tar Yes No 2. creosote Yes No 3. anthrocene Yes No 4. benezene Yes No 5. benzidine Yes No 6. naphthylamine . . . . . Yes No 7. aminodiphenyl Yes No 8. mustard gas Yes No 9. vinyl chloride Yes No 10. chloromethyl ether Yes No 11. arsenic Yes No 12. chromstes Yes No 13. asbestos Yes No 14. cutting oils Yes No 15. trichloroethylene Yes No AVl-8 �Page 7a <^, 16. U l t r a - v i o l e t l i g h t ( o t h e r than sun) Yes No 17. x-rays Yes No 18. ionizing radiation1-1' Yes No COMMENTS: 1. (other than r o u t i n e ) ' For each " Y E S " exposure in the preceding l i s t , u l e a s e fill out the f o l l o w i n g : Type of Exposure (coal t a r , e t c ) . __ ^ A. Was exposure received on the job? B. If y e s , job t i t l e _ C. If no, how exposure received D. Circle frequency of exposure that best fits your experience: Daily E. 2. Yes Ho t __ ^ Weekly u Monthly In what year(s) were you exposed? .^^ ^.^ u ___ _ Yearly _>_• Type of Exposure (coal tar, etc) A; Was exposure received on the job? B. If yes, job title C. If no, how exposure received D. Circle frequency of exposure that best fits your experience: Daily E. 3. Weekly Yes Monthly In what year(s) were you exposed?___ Type of Exposure (coal tar, etc) NO Yearly ____ : .___ _ A. Was exposure received on the job? B. If yes, job title C. If no, how exposure received D. Circle frequency of exposure that best fits your experience: Daily E. Weekly Yes ,, Monthly In what year(s) were you exposed? __ ... _ AVI-9 No Yearly �CONDUCT OF THE EXAMINATION NAME DATE DATE OF BIRTH (DOB) CASE NO. PHYSICAL EXAMINATION 1. General Appearance a. b. c. d. Appearance/Stated Age: ( ) Younger Than Well -nourished ( ) ( ) Obese Appearance of illness or distress Hair distribution: ( ) Normal Specify: ( ( ( ( ) Older Than ( ) Same As ) Under -nourished ) Yes ( ) No ) Abnormal Weight (Undressed) kg Sitting Blood Pressure Right Arm at Heart Level Systolic Diastolic Temp. Oral F. 2. Height (cm) 3. •Pulse rate Regular: ( ) Yes ( ) No * a. Irregular ( ) b. Irregularly irregular ( ) c. VPBs per minute Describe any irregularities: •4. Eye Grounds: ( ) Normal ( )" Abnormal Describe any vascular lesions, hemorrhages, exudates, or papilledema: ( ( ( ( ) A-V nicking ) t light reflex ) Arteriolar spasm ) Papilledema ( ) Hemorrhage s ( ) Exudates ( ) Disk Pallo ( ) t Cupping 5. Arcus Senilis: 6. ENT: ( ) Normal ( ) .Abnormal Right Tympanic Membranes intact ( ) fes ' Left Tympanic Membranes intact ( ) fes ' Nasal Ulcerations . . ) fes ( ' 7. ( ) Present ( ) Absent 5a. Abnormal Ocular Pigmentation ( ) Yes ( ) No Neck (Especially thyroid gland): Thyroid gland 'palpable Enlarged Nodules Tenderness J ( ( ( ) ) ) ) Describe any abnormality: ( ) No ( J No ( ) No ' (. ) Normal AVI-10 . ( ) Abnormal Parotid gland enlargement Carotid pulse absent Carotid bruit r . Right ( Right ( Right ( Left ( ) Left ( ) Left ( ) �;8. Thorax and Lungs: ( ( ( ( ( ( ) Normal C 3 Abnormal ) Asymmetrical ^expansion ) Hyperresonance 3 Dullness . ) WheeZes ) Rales Circumference at nipple level: Expiration 9. Describe any abnormality especially basilar rales: Heart: ( ) Normal cm Inspiration cm ( ) Abnormal Displacement of apical ijnpulse ( ). Yes C 3No Heart sounds normal ( ) Yes ( 3 Ho ( ) S1 Precordial thrust ( ) Yes C 3No ( ) S2 ( ) S3 ( )S Heart and Other Observations Murmur Ao P.u.. Apex Mitral (It. Tat) Sys f~3 (} Dia ( ) (3 (3 Describe •any enlargement , irregularity of rate , murmurs, or thrills: , . „ „ , . 1 . , , 10. ( ) No ( ) Yes ,-„-,,,! | .. , -- , J...J..,.|__.±. ,,..., -„, J , , , .. „,-,-. , . ,,,,, , „ , .- .„ J -.„.„„._, . Abdomen: ( ) Normal ( ) Abnormal Waist Measurement ( ) Heptomegaly Describe any abnormality with special cm. Liver span attention to the spleen and liver: ( ) Spleenomegaly Tenderness Liver ( ) Tenderness Other () OtVif^T mn*t*? * ( 11. Extremities: \ —..-.. ( ) Normal. ( ) Abnormal . ( ) Absence, specify: ( ) Edema Pitting ( 3 Non-pitting ( ) ( ) Clubbing of nails ( ) Varicosities • ( } Loss of hair on toes right ( ) Loss of hair on toes left .KS-AF-6 ._,«_,....... AVI-11 Describe any edema or signs of vascular insufficiency: �12. Peripheral Pulses Radial Femoral Popliteal Dorsalis Pedis Posterior Tibial Nonrlal Dimin. Absent 1 1 : Comments: ! ! 1 1 i * 13. Musculoskeletal : ( ) Normal ( ) Abnormal MUSCLE SPINE Weakness ( ) Scoliosis Kyphosis Tenderness ( )' Abnormal Consistency ( ) Tenderness ' Atrophy ( ) • Tenderness level Decreased range of motion Comments : ... , . ,. . _,. Pf'Hrir f i l t -,^-.r Sninr SIR RT Spine SLR LFT « C C C C C ) ) ) ) ) () () C ) Normal ( ) Abnormal ( ) Varicocele ( )Hemorrhoids ( ) Epididymis () Prostatic enlargement ( ) Scrotal mass Enlarged Atrophic emdia. ( ) Rectal mass ( ) } ( Comments : ( ) ( ) 14. Genitourinary /Rectal/Hernia ( ) Inguinal hernia Rt. ( ) Inguinal hernia Lft. ; Absent Testes;Rt. ( ) Testes Lft. ( ) 15. Lymph Nodes (Check all areas) Cervical ' Occipital Supraclavicular Axillary Epitrachlear Inguinal . Femoral Enlarged ( ) () () () () C ) C ) ( ) Tender C C C C C C C Normal Hard ) ) ) ) ) ) ) ( C ( C C ( ( ) ) ) ) ) ) ) Fixed ( ) ( ) C D C ) C ) C ) C ) ( ) Abnormal -Specify: Confluent C) C ) () () C ) C) C) 16. Nervous System - Separate Examination Page 3 of 4 KS-AF-6 AVI-12 �17. Other Tests Ordered: ( Tests ordered (Specify) ) Yes ( ) No Signed: Examining Physician Examining facility: Printed Name of Examining Physician Form 531 incl. KS-AF-6 AVI-13 Page 4 of 4 �or (.) A N A T O M I C A L FIGURE NAME Date: (Sioned - Examiner) Standard Form 53 AVI-1.4 �Name: Date: No. JSianed - Examiner) EXAMI MAT I ON - EXAMINER AVI-15 - (1 Of 2) �NEUROLOGICAL EXAMINATION DATE NAME CASE NO. Yes HEAD S NECK Normal to Palpations/Inspections Specify: ( ) Scar No ( ) ( ) C ) Assemraetry ( ) Depression Neck Range of Motion - Normal Decreased ( ) Left C ) Right ( ) Forward ( ) Back C ) MOTOR SYSTEM Handedness ) Right ) Left Gait - Normal or ( ( ( ( ) Broad Based ) Ataxic ) Small Stepped ) Other Comments MUSCLE STATUS {Strength. Tone, Volume, Tenderness, 1 Increased Right Left Normal Rt. Arm Swing Left Arm Swing Muscle Bulk Tone - upper ext. Lower ext. Strength - Distal Wrist Extensors Ankle/Toe Dors/. Flexors Proximal Deltoids Hip Flexors Abnormal ( ) ( ) ( ) C ) ( ) ( ) ( ) ( ) • C C Decreased Right Left C ) ( ) ) C ) ( ) ( ) C ) AVI-16 ) ) ( ) ( ) ) () ) () C C C ) ( ( ^ KS-AF-4 rev. ) ( ) ) C ) Fibrillations) C ) ( ) C ) C ) ( }- • �c Neurological ExaminaL. Page 2 Yes No ABNORMAL MOVEMENTS (Tremors, Tics, choreas, etc) fMABiMiiSM " If yes (1-4*) Tenderness If yes, (1-4+) Tremor (if yes, specify below) Upjier Upper Lbwer Lower Extremities Extremities Extremities Extremities Rt. Lt. Rt. Lt. ( ) Resting ( 3 ( 3 b. Intention Normal Equilibfatory - Eyes Open Right Foot Left Foot Equilibratory - Eyes Closed Right Foot Lefit Foot ( 3 Non-Equilibratory Normal Finger to Nose Finger t6 Finger Abnormal Finger to Nose to Finger Heel'-Knee-Shin Succession Movements (Including cheek, rebound posture-holding) Rapidly Alternative Movements Handwriting (if indicated) Speech (articulation, aphasia, agnosia) Grossly . if abnormal specify ( ) Dysarthia ' ( ) Aphasia KS-AF-4 ( ( AVI-17 ». ) 3 [ ) ( 3 [ 3 ( 3 C ( 3 ) (3 ( 3 i 3 3 ( 3 Normal SKILLED ACTS- PRAXIS Both } ; ( ) ( ) j Lett (3 ( Heel to Knee a. b. Other Abnormal Kt. c. 3 ( 3 COORDINATION a. Essential C (3 {. ( 3 ( 3 Abnormal �Neurological Examination Page 3 Reflexes (Code Oabsent, l=sluggish, 2 = a c t i v e , 3«verv a c t i v e , 4«trans1ent clonus, 5=sustained clonus, 6=other SPECIFY UNDER COMMENTS BELOW) Right Biceps Triceps PatellarAchilles Cremasteric Abdominal ( ( ( ( ( ( Left ) ) ) ) ) ) ( ( ( ( ( ( ) ) ) ) ) ) Abnormal No, Yes Babinski Present? ( ) ( ) COMMENTS MENINGEAL IMITATION Normal Abnormal Rt. Straight leg raising ( Both Left ( ) ) SENSORY SYSTEM (tactile, pain vibration, position . If positive sensory signs are present summarize below and indicate details on Anatomical - Standard Form 531) Normal L i g h t Touch Pin Prick V i b r a t i o n (6 ankle, 128hz Tuning Fork) Position (Great Toe) CRANIAL NERVES Right-smell L e f t - sme'll C ) (. ) ( ) Present C ) AVI-18 KS-AF-4 Abnormal Rt.left Absent ( ) ( ) Both �Neurological Examination Page 4 Normal Abnorma1 . _„_ TT~ FUNDUS - Right: lT~abnormal: ( ) bisk Pollar Atrophy ( ) Exudate ( ) Papilledema ( ) Hemorrhage Fundus - Left: If abnormal: ( ) Disk Pollar Atrophy ( ) Exudate ( ) Papilledema ( ) Hemorrhage FIELDS - Right Fields - Left C ) ( ) to confrontation PUPILS fnm Size Shape, position Light, reaction Position of Eyeballs C ) C ) ( ) equal ( ) unequal difference ( ) round ( ) other rt. ( ) other left ( ) normal ( ) abnormal rt.( ). abnormal left ( C ) ) ) ) normal deviation medial rt. ( ) deviation lateral rt deviation medial It. ( ) deviation lateral It deviation medial both ( ) deviation lateral C ) Movements if abnormal describe NYSTAGMUS ( ) rotary ( ) horizontal ( ) vertical ( ( C Draw position: ( PTOSIS ) None ) right MOTOR Clench. Jaw, rt. Clench" Jaw, left SENSORY Symme tr i c C ) ( ) Normal v ( ) ( ) Sensory right Sensory left VC.AC-d »-rt»r l AVI-19 ) left Deviated Left RigHt ( ) C ) ( ) C ) Abnormal \r__ 1 \ T -) ) ( ) None �Neu.ro logical Examination Page S ( -') present r i g h t ( ) present l e f t CORNEAL REFLEX ( ( ) Absent Right ) Absent left MOTOR RIGHT Palpebral Fissure ( PALATE AND UVULA Movement No Yes ( ) C ) Normal Smile Rt. Normal Smile Left ) C Normal ( ) ) Abnormal Normal Deviation C ) Palatal Reflex Rt. Palatal Reflex Left ( ( ) normal ) normal Tongue-Protruded Atrophy ( C ) Central ) No Rt. ( ) ( ( ) abnormal ) abnormal ( ( ) Left )Yes )Right MENTAL STATUS (Alert, clear, cooperative, etc) Gross abnormalities If yes, specify: ( ) yes C ) no ADDITIONAL COMMENTS: Signed Examining Physician KS-AF-4 Left ( ) AVI-20 �Ol C L J . M L . I I CLINICAL RECORD Neurology LA; ."Illin I .(2 of 2) ' A N A T O M I C A L FIGURE NAME Date: (Sirtned - E x a m i n e r ) iCrvT'S IO£H7 i r i C ^ / of - r ;'/ I »n .n [ > / . » t-": /V/i cr.»— y, ,; I HEKI5TC ftUtii.*ih--C For Standard Form 53] . KS-AF-4 AVI-21 �Name: Date: No. jSigned - Examiner) SPECIALTY EXAMINATION - NEUROLOGY - (1 of 2) Standard Form 531 KS-AF-4 AVI-22 �UALTY EXAMINATION - OERMATOIO( NAME_ Oate_ No. Skin: Normal Abnormal ( ( ) - Indicate type and location of lesions on the anatomical figure - attached ( ) Comedones ( ) Palmer Keratosis ( ) Acneifom .lesions ( .) Petechiae ( ) Acnelfprjn Scars ( ) Ecchymoses .'•' . V*; '' - ( ) Deplgjnenjtation ( ) Conjunctiva (- ) InclusliWKGysts ( ) Oral Mucosa ( ) Cutls Rhpmpptdalis ( ) Finger Nails ( } Hyperpigmentation ( ) Toe^Nails ( ) Jaundice ( ) Soles of Feet (" ) ^eniiatpgriaph^ia ( •-')•••' ( : J^ .****** •Photographs taken? If so indicate area> photographed: (ONLY: SUSPECT. LESIONS) ( ) Face (right) ( ) Neck ( ) Chest ( ) Face (left) ( ) Shoulders ( ) Stomach ( ) Face (Full) BIOPSY ( Yes No •) Skin Biopsy Performed (Check if yes) Consent Form obtained? ' ( ) ( ) ^ COMMENTS: Biopsy location .•' Signed AVI-23 �SPECIALTY EXAMINATION - DERMATOLOGY CLINICAL RECORD (2 of 2) ANATOMICAL FIGURE NAME Date: (Sinned - Examiner)_ T ( f ^ T 'S J D I K 1 i r 1C AT ION (F oi ifft-dc-r ~ r it i *n o n f i it t i iff : /V«r*»«"— V* j.', / i / > ! . H K C S I S T C I NO. AUvA.tK.-C Furra an Standard Form 531 , KS-AF-2 AVI-24 �Name: Date: No. _(_Sjqned - Examiner) _SPECIALTY_ EXAMINATION - DERMATOLOGY - ANATOMICAL FIGURE ( 1 of 2) Standard Form 531 KS-AF-2 AVI-25 �P U L M O N A R Y F U N C T I O N Test Date NAME_ Age Case No. Actual Predicted FVC FEV-1 FEV-1/FVC 'Comments regarding test performance: Testing Technician *complete only if performance is questioned - i.e., cold, bronchitis, etc. Equipment Used: Breon Spirometer COMMENTS OF REVIEWING PHYSICIAN Signature_ Reviewing Physician KS-AF-3 rev. AVI-26 �NUMBER: NERVE -:!_,.? J ' | RiST W ! GAIN i i 5K C I1 RECORD SITE 5K BELOW ELBOW ! ABOVE ELBOW I 1 ANKLE FIBULAS >: ' R A • ! ; H[A[^ : I .o t CM 1 j ; CM j ! i ! CM | ji CM 1 14 CM ! i. j i i ! i i I i .10 { i i 1 2K i DIFF. (STMr-CURR. | C.V. | LAT. 2K i i ?£SGS£AL DISTANCE .5K f ] TEMPERATURE: AGE: ! M • j !j ' | l i I ! r ! i ! 1 ! 1 i f I i 1 1 I i . ! �-DIAGNOSTIC SUMMARY *. SYNOPSIS OF POSITIVE FINDINGS Medical History: 1. 2. • . 3. y ; ' '" ~ ' " " " " " " ' * """"" ' ~ I J ™"' — *" '-> •""• "~" •«"""'» ,,.— 4. 5. iWSICAL EXAMINATION 1. General 1. (Complete below and continue on additional page - reference no.) Dermatologic Neurological Including Nerve Conduction Studies Psychological (Bianary Provided) KS-AF-7 • AVI ~28 �DIAGNOSTIC SUMMA *v SYNOPSIS OF POSITIVE FINDINGS Medical History: 1. 2. . • ..,_,...,.,.,, . ____ 3. 4. 5. PHYSICALEXAMINATION (Complete below and continue on additional,page - reference no.) 1. General 2. Dermatologic 3. Neurological Including Nerve Conduction Studies Psychological (Bianary Provided) AVI-29 of. �Appendix VII EXAMINATION PARAMETERS AND ABNORMALITY WEIGHTS USED IN ASSESSING INDIVIDUAL HEALTH Relative Weight Assigned to an Abnormality Organ System Parameter Hematologic RBC WBC Hemoglob'in/Hematocrit RBC Indices (MCV/MCH/MCHC) Platelets Cancer Skin Cancer Systemic Cancer Endocrine T3 Tij FTI Glucose 2-hour Postprandial Testosterone 2 2 3 2 it 3 10 3 6 3 it i» it Pulmonary FEV 1 FVC FEV 1/FVC Ratio X-ray Hepatic Enzymes (SGOT, SGPT, GGTP, Alkaline Phosphatase) Total Bilirubin Direct Bilirubin LDH Cholesterol HDL Triglycerides Uroporphyrins Coproporphyrins ALA Hepatomegaly Reproductive Sperm Count Psychological MMPI (10 Major Scales) Halstead-Reitan IQ Scores (VRQ, PRQ, FLQ) 3 3 3 it 5 4 it 3 3 6 AVII-1 it 5 �General Health Examiner's Assessment Percent Body Fat Sedimentation Rate Systolic Blood Pressure Diastolic Blood Pressure ECG Heart Sounds Eye Grounds Proximal Pulses (Carotid/Femoral) Distal Pulses (Popliteal/Dorsal is Pedis/Posterior Tibial) 3 3 2 7 8 9 7 6 5 BUN Occult Blood WBC in Urine Protein in Urine Specific Gravity 5 2 2 Dermatologic Normal/Abnormal 1 Neurological Smell (Bilateral) Visual Fields (Bilateral) Pupils (Reaction and Movement) Sensation/Corneal Reflex/Jaw Clench (Bilateral) Smile/Palpebral Fissure Palate Movement and Reflex/Neck Range of Motion Speech/Tongue Protrusion Pinprick/Light Touch/Vibration Sense Muscle Status Central Function (Finger-to-Nose/ Romberg/Tremor/Gait) Babinski Tendon Reflexes (Patellar/ Archilles/Biceps) Ulnar Velocities (Above and Below) Peroneal Velocities 1 3 3 3 Cardiovascular Renal 5 AVII-2 3 3 3 3 3 3 4 5 3 3 �Appendix V I I I TOTAL MORTALITY AND MORBIDITY STUDY SITE SPECIFIC MALIGNANT NEOPLASMS Site ICD Code (9th Ed) Mortality Comparison Ranch Hand (First cohort only) Lip, oral cavity, Pharynx (140-149) 1 Morbidity Comparison Ranch Hand 0 S R t 2 0 0 Digestive organs, peritoneum (150-159) Respiratory, intrathoracic (160-165) 0 1 2* 1 0 0 Bone, connective tissue, skin, breast (170-175) 0 0 0 1 Genitourinary organs (179-189) 1 1 6 2 Brain (191-192) 0 1 1 0 0 0 Thyroid (193) 0 0 0 1 0 0 Lymphatic and hematopoietic tissue (200-208) 0 1 0 0 0 No site specification (199) J_ J_ 0_ 0_ £ 0_ TOTAL 4 10 13 0 = Original S = Shifted R = Replaced *Includes 1 Ranch Hand and 1 comparison who expired following interview. AVI II 10 0 1 1 2 �Appendix IX GENERAL HEALTH ANALYSES USING DATA FROM ALL COMPARISONS SELF-PERCEPTION OF HEALTH BY GROUP All Comparisons Number(?) , Ranch Hand Number (jt) Perception Excellent Good Fair Poor 180(10) 523(11) 113(12) 16( 1) 1192 p=0.05 392(38) 159(15) 53( 5) 1039 SELF-PERCEPTION OF HEALTH BY GROUP MEMBERSHIP AND OCCUPATIONAL CATEGORY Occupational Group Excellent Perception of Health Good Fair/Poor p value Officer, flying Ranch Hand Comparison 198 225 121 115 12 10 0.66 Enlisted, flying Ranch Hand Comparison 59 65 83 89 12 13 0.97 Enlisted, ground Ranch Hand Comparison 126 176 225 280 127 103 0.005 DISTRIBUTION OF BODY FAT (PERCENT) Lean <10!0 Obese Q25?) 13 12 Ranch Hand Comparison Normal (10-25?) 821 961 208 217 P = 0.83 PERCENTILE DISTRIBUTION OF SEDIMENTATION RATE RESULTS 5% Ranch Hand Comparison 0 0 25% 50% 75% 95% 1 1 2 2 1 1 12 12 AIX �Appendix X FERTILITY AND REPRODUCTIVE ANALYSES; RANCH HANDERS VERSUS ALL COMPARISONS ANALYSES OF CONCEPTION OUTCOMES, UNADJUSTED FOR MATERNAL COVARIABLES (COMPLETE AND PARTIAL DATA SUBSETS); RANCH HANDERS VERSUS ALL COMPARISONS Yes Pre-SEA (%) Yes No Post/-SEA (?) No Miscarriage Ranch Hand Comparison 295 (14.4) 282 (11.9) 1754 2089 190 (16.0) 233 (14.0) P = 0.01 1001 1430 P •= 0.15 Stillbirth Ranch Hand Comparison 13 (0.6) 21 (0.9) 2036 2350 16 (1.3) 12 (0.7) P = 0.34 1175 1651 P = 0.10 Induced Abortion Ranch Hand Comparison 13 (0.6) 18 (0.8) 2036 2353 62 65 P = 0.62 (5.2) (6.0) 1129 1563 P = 0.36 Live Birth Ranch Hand Comparison 1723 (84.1) 2042 (86.1) 326 329 917 (77.0) 1309 (78.7) P = 0.06 274 354 P = 0.27 AX-1 �CONCEPTION OUTCOMES (COMPLETE DATA SUBSET) BY GROUP MEMBERSHIP AND TIME; RANCH HANDERS VERSUS ALL COMPARISONS Yes Pre-SEA (*) No Post-SEA (*) No Yes Miscarriage Ranch Hand Comparison 239 233 (13.7) (11.6) 1505 1776 156 (15.0) 883 188 (13-2) 1238 P = 0.20 P = 0.05 Stillbirth Ranch Hand Comparison 9 13 (0.5) (0.6) 1735 1996 12 12 (1.2) 1027 (0.8) 1414 P = 0.43 = 0.60 Induced Abortion (0.5) (0.4) Ranch Hand Comparison 1736 2001 37 53 (3.6) (3.7) 1002 1373 0.84 = 0.76 Live Birth Ranch Hand Comparison 1487 1752 (85.3) (87.2) 257 257 833 (80.2) 1170 (82.0) 206 256 0.24 0.08 RESULTS OF THE ANALYSIS OF CONCEPTION OUTCOMES RANCH HANDERS VERSUS ALL COMPARISONS Relationship P value Miscarriage by Group by Pre/Post-SEA 0.70 Stillbirth by Group by Pre/Post-SEA 1.00 Induced Abortion by Group by Pre/Post-SEA 1 .00 Live Birth by Group by Pre/Post-SEA 0.78 AX-2 �ANALYSES OF LIVE BIRTH OUTCOMES, UNADJUSTED FOR MATERNAL COVARIABLES (COMPLETE AND PARTIAL DATA SUBSETS); RANCH HANDERS VERSUS ALL COMPARISONS Yes Pre-SEA (%) No Yes Post-SEA (%) No Learning D1sabi1ity Ranch Hand Comparison 61 81 (3-5) (8.0) 1662 1961 77 81 P = 0.49 (8.4) (6.2) 840 1228 P = 0.05 Physical Handicaps Ranch Hand Comparison 144 176 (8.4) (8.6) 1579 1866 132 130 P = 0.77 (14.4) 785 (9.9) 1179 P = <0.01 Infant Death Ranch Hand Comparison 8 4 (0.5) (0.2) 1715 2038 4 3 P = 0.15 (0.4) 913 (0.2) 1306 P = 0.39 Birth Defects Ranch Hand Comparison 90 123 (5.2) (6.0) 1633 1919 80 84 P = 0.29 (8.7) 837 (6.4) 1225 P = 0.04 Neonatal Death Ranch Hand Comparison 25 28 (1.5) (1.4) 1698 2014 P = 0.84 AX-3 14 3 (1.5) 903 (0.4) 1305 P = <0.01 �LIVE BIRTH OUTCOMES (COMPLETE DATA SUBSET); RANCH HANDERS VERSUS ALL COMPARISONS No Yes Post-SEA (%) No 1430 1680 75 74 (9.0) 758 (6.3) 1096 (9.0) (9.1) 1353 1592 126 118 (15.1) 707 (10.1) 1052 7 (0.5) 3 (0.2) 1480 1749 3 1 78 (5.2) 113 (6.4) 1409 1639 76 77 (9.1) 757 (6.6) 1093 20 (1.3) 28 (1.6) 1467 1724 14 4 (1.7) 819 (0.3) 1166 Yes Pre-SEA (%) Learning Disability Ranch Hand Comparison 57 (3.8) 72 (4.1) Physical Handicap Ranch Hand Comparison 134 160 Infant Death Ranch Hand Comparison (0.4) (0.1) 830 1169 Birth Defects Ranch Hand Comparison Neonatal Death Ranch Hand Comparison RESULTS OF THE ANALYSIS OF LIVE BIRTH OUTCOMES; RANCH HANDERS VERSUS ALL COMPARISON Relationship P Value Learning Disability by Group by Pre-Post SEA 0.12 Physical Handicap by Group by Pre-Post SEA 0.02 Infant Death by Group by Pre-Post SEA 1.0 Birth Defects by Group by Pre-Post SEA 0.02 Neonatal Death by Group by Pre-Post SEA 0.03 AX-4 �Appendix XI INTRODUCTORY LETTERS Secretary of Air Force USAF Surgeon General with Fact Sheet AXI-1 �D E P A R T M E N T OF THE AIR FORCE WASHINGTON. D C 2O33O OFFICE OF T H E S E C R E T A R Y James W. Doe 1215 Middle Grove Norfork, MD 23456 Dear Mr Doe The Air Force will soon begin conducting a very comprehensive health assessment of certain Air Force members who served our Nation in the Vietnam conflict. This health assessment is part of a medical study designed to help determine if you or your fellow Vietnam veterans may have had any compromise to your health as a result of exposure to the complex environment of Southeast Asia. Scientists at the USAF School of Aerospace Medicine have been given the responsibility for conducting this important project. The Air Force Surgeon General will contact you soon with more details and ask for your voluntary participation. A major focus of the President's program for veterans is the resolution of health issues raised by them. The Air Force and I are committed to doing our part in resolving these issues. I ask that you help us and all Vietnam veterans by voluntarily participating in this major study. Sincerely, Verne Orr Secretary of the Air Force AXI-2 �DEPARTMENT OF THE AIR FORCE HEADQUARTERS UNITED STATES AIR FORCE BOLUNG AFB DC 2O332 James W. Doe 1215 Middle Grove Norfork, MD 23456 Dear Mr Doe The Air Force is conducting a very comprehensive health assessment of certain Air Force members who served our Nation in the Vietnam conflict. The USAF School of Aerospace Medicine has been given the responsibility for conducting this study. The purpose of the study is to determine whether there may be any causal relationship between health problems and exposure to the complex and unique environment of the war in Southeast Asia. Simply stated, we do not know if such health effects exist. You are being asked to voluntarily participate in this study because of your unique Southeast Asia experience. Your participation is critical to the success of this study. However, you should not view this invitation to participate as a cause for alarm nor as an implication that you are at risk for any known disease. To insure the scientific validity of the study, both an in-depth interview and a detailed physical examination will be conducted. The administration of the interview will begin soon under the direction of a nationally recognized health survey organization. You will be contacted by phone or letter to arrange a convenient time for an in-home interview which will take from two to three hours. Shortly after the interview you will again be contacted to schedule a physical examination at a nationally recognized civilian medical facility. The physical examination will take approximately four days. Every effort will be made to minimize disruption of your normal activities and to facilitate your participation in the study. Travel and per diem will be paid by the Air Force. For those not precluded by law, a stipend of $100 per day will be paid as a partial compensation for your time. Our intent is to maintain all individual health data in strictest confidence. In case outside parties attempt to gain access to the data, the Air Force and the Department of Justice are committed to protect this individual confidentiality. Only in the event of an adverse final court decision, or in the highly unlikely instance where serious medical deficiencies must be shared with appropriate medical authorities to protect public health and safety, will any personal health data be revealed. You are referred to the Fact Sheet for further information regarding this matter. AXI-3 �This is perhaps one of the most important health studies undertaken by the Air Force. Your voluntary participation is critical to its success. Although you may feel healthy, numerous Vietnam veterans believe that they have illnesses which may be attributable to service in Southeast?Asia. The only way we can get clarification of these difficult questions is through your cooperation and participation. Sincerely PAUL W. MYERS 1 Atch Lieutenant General, USAF, MC Surgeon General Fact Sheet AXI-4 �FACT SHEET INTRODUCTION - The USAF School of Aerospace Medicine, Brooks AFB, Texas, is conducting the study. You are being invited to participate in this study because of your specific duties and period of assignment in Southeast Asia. PURPOSE - To determine whether there is a causal relationship between adverse health effects and exposure to the complex environment of Southeast Asia. METHODS - An in-depth health questionnaire will be administered to you by a member of a health evaluation team from Louis Harris and Associates, Inc. - A complete profile of your current health will be obtained by a physical examination which will be conducted by a nationally recognized outpatient clinic* - Follow-up abbreviated health questionnaires and physical examinations will be conducted at years 3, 5, 10, 15, and 20 of the study. - Travel expenses (including board and lodging) for the physical examination will be paid by the Air Force. - Stipend of $100 per day will be paid to study participants who are not on active duty, Government employed or otherwise precluded by law from receiving such a stipend. - Confidentiality is to be maintained except in two cases: - A judicial order to release personal medical data following an Air Force and Justice Department defended lawsuit. Serious medical findings which impact public health and safety. Two examples of situations in which public health and safety would raise the questions of disclosure are: a participatnt has typhoid fever, a participant who directly impacts the safety of others either in his profession, or as a volunteer, is found to have a serious nerve, heart or mental disorder. In this instance a committee composed of a physician (whose specialty is the area of the identified problem), a physician of your choice, a flight surgeon, a judge advocate (lawyer) and a representative from your field of expertise will be convened to review the medical findings. Before any disclosure Is made to medical authorities, the committee must determine that the findings jeopardize the public health and safety. AXI-5 �BENEFITS TO YOU - You will receive a complete health review and physical examination of top level executive calibre at no cost to yourself. - You will be completely informed of all examination results. - The Information from this study will be provided to a physician of your choice if you so request. - Questions concerning the study may be referred to the USAF School of Aerospace Medicine, Epidemiology Division, Brooks Air Force Base, Texas 78235, or by calling collect AC 512 536-3309. - If you have recently changed your address or have an unlisted phone number, please advise the USAF School of Aerospace Medicine at the above address and phone number so that your records may be properly updated. AXI-6 �Appendix XII OCCUPATIONAL CATEGORY AND RACE OF THE FULLY COMPLIANT POPULATION IN PERCENT AND COUNTS Occupation Code Ranch Hand % Counts Comparisons Original Shifted Replaced % Counts % Counts % Counts Won- Black 1 2 3 Officerr-Pilot 82 Officer-Navigator 96 Officers-Other 83 278 76 20 71 81 77 218 58 10 78 100 67 32 6 8 59 71 100 94 12 7 Officer Subtotal 374 73 286 78 46 61 113 Enlisted-Fit Eng Enlisted-Other 93 86 172 436 84 75 141 301 94 75 17 91 70 75 26 133 Enlisted Subtotal 88 608 77 442 77 108 74 159 8? 982 76 728 77 154 68 272 Officer-Pilot 67 Officer-Navigator 100 Officer-Other 0 4 2 0 80 100 100 4 2 1 f 0* 0* 0* - * * * Officer Subtotal 67 6 88 7 ~ 0* 1 * 93 90 13 44 67 76 10 28 69 0 9 83 55 5 11 Enlisted Subtotal 90 4 5 85 57 73 38 69 9 62 16 88 63 75 45 69 9 62 16 87 1045 76 773 77 163 68 288 Total Non-Black Black 1 2 3 4 5 Enlisted-Fit Eng Enlisted-Other Total Black Entire Population AXII �Appendix XIII SELF-REPORTED REASONS FOR NONCOMPLIANCE TO QUESTIONNAIRE Reason Fear of Physical Job Commitment Dissatisfaction with the Military No Time - No Interest No Travel, Distance, Family Confidentiality/ Active Duty Health Reasons Passive Refusals* TOTAL Ranch Hand - Original Shifted Replaced Total 12 (242) 2 (22) 14 (112) 2 (3%) 2 (3%) - 3 (92) 9 (13) 1 (82) 1 (22) 11 (82) 23 (68?) 36 (5150 9 (752) 15 (292) 60 (452) 2 (5%) 0 6 (18$) 10 (14*) 2 (162) 12 (242) 1 (22) 10 (202) 22 (172) 1 (12) 23 (172) 34 11. (16« 70 12 ^Unresponsive to scheduling attempts. AXIII 51 133 �Appendix XIV SELF-REPORTED REASONS FOR NONCOMPLIANCE TO PHYSICAL EXAMINATION Reason Ranch Hand Fear of Physical 6 (5%) Job Commitment 29 (2450 Dissatisfaction With the Military 5 (450 No Time - No Interest 53 (13*) No Travel-Distance 4 (450 Family ConfidentialityActive Duty 11 (951) Health Reasons 5 (450 Passive Refusals* 9 (7$) Original 3 (2$) 51 (29%) r Shifted Replaced Total 3 (4$) 6 (2?) 10 (21%) 20 (24$) 81 (27$) - T". 0 94 (53%) 17 (46$) 10 (5?) 4 (11?) 7 (950 21 (7$) 8 (450 3 (250 10 (550 2 (555) 1 (3$) 3 (8$) 6 (750 1 (150 2 (355) 16 (5%) 5 (2$) 15 (555) 122 179 *Unresponsive to scheduling attempts. AXIV 37 43 (52JO 154 (5250 82 298 �Appendix XV COEFFICIENT OF VARIATION FOR TRI-LEVEL CONTROLS Control values were analyzed on 15 different laboratory tests for the period from January 14 thru December 13, 1982. Triplicate values were collected on each laboratory test at each of three different ranges (I, II, and III) except for triglyceride and alcohol which each had only ranges II & III. These control data were received from 91 groups of study participants reporting for physical examination (usually 2 groups per week). A total of 91 sets of control values were received for II & III and a total of 78 for I. A one-way analysis of variance procedure was used, on each trilevel laboratory test to determine whether or not the data varied significantly among the 91 (or 78) groups. The error term used was the pooled variance (o£) from the {* triplicate values recorded for each group. The group means differed signifi" cantly at the 0.01 level on nearly all of the analyses (40 out of 42). Hence, the variability among the groups was significantly more than can be explained by the variability among the triplicate readings. A variance component for the group-to-group variability (o^) was estimated from the one-way analysis of variance and the standard deviation of a single measurement/group was estimated as: 7 /XO °e « + ° Each coefficient of variation given in the table below was computed as: o x 100 x CV5& where the x is the mean of the control values for each trilevel/laboratory test. Ninety-five percent confidence limits were computed as follows: / N(N-1)y2 <.. N(N-1)y2 V where v2 is the square of the observed CV, N = 91 or 78 (depending on the trilevel of interest) and o and u are the population parameters associated with o and k respectively. The interval for the CV$'s marked with an asterisk in the table below did not contain the USAFSAM required CV%, implying that the estimated CV$ differed significantly from the required at the 5% level. The estimate exceeded the required on 12 of the 40 trilevel sets. The average CV$ was not tested. AXV-1 �SAMPLE MEAN, STANDARD DEVIATION AM) COEFFICIENT OF VARIATION FOR TRI-LEVEL CONTROLS USED FOR 15 BIOCHEMICAL ASSAYS Test x 6.6 0.296 Creatinine 0.602 - 16.6 0.415 45.9 1.6970.024 0.702 _n III CV% 0* BUN USAFSAM Require- I III II CVJt CV% Average 1.53 2.84 2.00 1.40 0.93 1.16 2.50 1.41 0.69 1.19 3.50 1.49 1.89 1.50 4.50* 2.50* 5.6370.053 nent Glucose 49.4 0.719 100.2 1.408 212.6 1.457 1.46 Cholesterol 104.2 2.357 151.7 2.257 2.15* 2.04* 72.39 1.869 177.4 2.464 2.58* 1.39 1.98 2.10 31.6 0.786 1.535 5.42* 2.48 4.06* 3.99 3.50 2.236 115.8 Triglycerides HDL 20.5 1.111 37.8 Total Bilirubin 0.930 0.040 1.4370.045 5.4700.133 Conjugated Bilirubin 0.400 0.043 0.8110.043 2.3830.110 Alk Phos 5.2740.203 9.8550.273 28.37 .438 4.34* 3.12* 2.42* 3.29 10.74* 5.33 1.50 4.60 6.89 6.00 3.85* 2.77 1.54 2.72 2.70 SCOT 38.32 1.18 56.73 1.41 171.2 2.18 3.08 2.48 1.27 2.28 4.00 SGPT 28.16 2.697 26.65.0.999 101.6 1.133 2.70 3.75 1.12 2.52 5.00 GGPT 31.97 0.985 43.68 1.033 186.79 2.20 3.08 2.37 1.18 2.21 5.00 2.612 441.7 1.35 1.57 0.93 1.28 2.20 2.08 4.00 2.57 2.88 5.00 1.54 1.53 1.54 LDH CPK Alcohol 147.9 65.5 1.997 165.8 1.362 139.1 5.559 48.5 0.749 4.104 440.9 11.34 99.2 1.518 *P<0.05, reject the hypothesis that the sample CV% came from the population with required AXV-2 �Appendix XVI SPECIFIC RULES FOR ENTRY INTO THE MORBIDITY STUDY RULES CIRCUMSTANCES Ranch Hander (RH) Dies Following Initial Data Collection Control Followed Throughout and Replaced as Necessary RH Dies of Combat Cause Medical Records Reviewed; No Control Set Formed RH Dies of Noncombat Cause Prior to Initial Data Collection 1st Order Surrogate Interview Accomplished; Control Selected and Followed Throughout; as Necessary RH Noncompliant for Baseline Questionnaire and Physical Control Followed Throughout the Study; Replaced as Necessary RH Compliant for Questionnaire Noncompliant for Baseline Physical Examination Control Followed Throughout the Study; Replaced as Necessary RH Noncompliant During Follow-up Control Followed Throughout the Study; Replaced as Necessary Control Dies Following Initial Data Collection Not Replaced in the Prospective Study of Morbidity Control Dies of Combat Cause Medical Records Reviewed; Excluded from Further Study Control Dies of Noncombat Cause Prior to Initial Data Collection Included in Mortality and Retrospective Morbidity Studies; Surrogate Interview Accomplished. Not Included in Prospective Morbidity Study and Replaced by a Living Compliant Control. Control Noncompliant for Baseline Physical Examination Control Followed Throughout Study Replace as Necessary Control Noncompliant During Follow-up Control Followed Throughout Study Replace as Necessary Noncompliant Control Returns to Study Both Primary and Replacement Controls will be Continued in Study AXVI �Appendix XVII PERCENT COMPLIANCE BY FLYING CODE AND MILITARY STATUS OF THE RANCH HAND AND COMPARISON POPULATION NON-BLACK OFFICERS Military Status** and Flying Code* Fully Compliant Participation Tart Tally NonCompliant Compliant Total Ranch Hand 19 .4 10 .3 36 ,5 3 .8 4 .8 11 .1 AF RF SVF AN RN SVN TOTAL 84.7 3.2 3.5 11.6 0.0 1.6 1.9 100 100 100 100 100 100 12.1 3.2 100 8.9 0.0 39,3 10.0 2.3 8.5 100 100 100 100 100 100 9.3 100 12.5 0.0 0.0 0.0 4.0 0.0 100 100 100 100 100 100 Comparison Original AF RF SVF 58.9 86.0 39.3 75.0 86.6 62.9 AN RN SVN TOTAL Comparison 72.9 17.8 Shifted 87.5 100.0 37.5 75.0 96.0 61.5 AF RF SVF AN RN SVN TOTAL Comparison 78.0 18.6 Replaced 7.9 0.0 43,.3 1 1 ,,1 10.4 17.4 57.9 83.3 32.4 88.9 77.1 63.0 AF RF SVF AN RN SVN TOTAL 61.4 20.7 Flying Nonflying Active Retired **SV - Separated/Reserve *F *N **A **R 100 = = = = AXVII 100 100 100 100 100 100 17.9 100 �Appendix XVIII RELATIVE RISKS FOR SELECTED CLINICAL END POINTS CLINICAL PARAMETERS Self Perception of Poorer Health S40 yrs Self Perception of Poorer Health >40 yrs Older Than Stated Age Lean by Body Fat Obese by Body Fat Sed Rate S40 Sed Rate >40 Skin Cancer Systemic Cancer Percent* RH C Relative Risk Confidential Interval Exact Normal Approx 19.3 10.6 1.82 (1.18,2.10) (1.17,2.87) 21.4 0.8 1.2 19.8 0.5 5.8 15.8 0.1 0.9 20.3 4.2 5.4 1.35 5.92 1.37 0.97 0.13 1.07 (1.05,1.76) (.80,262.37) (.51,1.043) (.80,1.18) (1.05,1.75) (.76,126.11) (.51,3.78) (.80,1.18) (.66,1.78) (.66,1.77) 2.35 1.20 (1.18,5.11) (.46,3.33) (1.16,4.90) (.47,3.15) 3-35 1.24 1.42 1.03 Childless Marriages Not Having Desired Children Abnormal Sperm Miscarriage Stillbirth Induced Abortion Non-live Birth Learning Disability Physical Handicaps Infant Death Birth Defects Neonatal Death 20.9 19.5 1.07 (.93,1.23) (.93,1.23) 18.3 4.6 15.9 1.3 5.2 23.0 8.4 13.8 -0.5 8.7 1.5 19.9 4.6 13.6 0.8 6.8 22.1 6.9 11.4 0.4 6.5 0.4 0.92 0.99 1.17 1 .60 0.76 1.04 1.22 1.21 1.35 1.35 3-78 (.76,1.10) (.54,1.86) (.95,1.45) (.65,4.30) (.54,1.087) (.89,1.22) (.86,1.75) (.93,1.58) (.26,8.67) (.94,1.95) (1.06,20.45) (.76,1.10) (.54,1.83) (.95,1.45) (.65,4.06) (.54,1.09) (.88,1.22) (.86,1.75) (.93,1.58) (.28,7.09) (.95,1.94) (1.03,16.50) Reported Neuro Disease Smell, Left Smell, Right Visual Fields, Left Visual Fields, Right Light Reaction Ocular Movement Sensation, Left Sensation, Right Corneal Reflex Jaw Clench Smile Palpebral Fissure Balance Gag Reflex Speech Tongue in Midline Palate and Uvula Neck Motion Pin Prick 5.18 4.59 .1.82 1.56 1.43 1.63 .26 0.29 .39 0.19 .52 0.77 34.8 35.3 .52 .67 .38 .39 .19 .13 .096 0 .26 .38 5.65 5.43 19.9 19.5 1.44 1.68 .28 o.30 .45 .29 .13 3^92 3.23 9.41 9.56 .89 1;17 1.14 1.12 .49 1.48 .99 1.29 .99 1.48 (.58,1.37) (.54,2.63) (.51,2.68) (.13,13.28) (.041,4.31) (.40,6.68) (.86,1;12) (.33,6.03) (.17,6.74) (.077,87.25) (.58,1.37) (.55,2;55) (.51,2.59) (.15,9.46) (.06,3.60) (.41,5.80) (.86,1.13) (.34,5.25) (.19,5.54) (.11,41.17) 1.48 1.04 1.04 .86 (.21,16.35) (.70,1.57) (.86,1.27) (.38,1.94) (.24,11.59) (.70,1.56) (.86,1.26) (.39,1.89) 1.50 2.22 1.21 .98 (.22,16.60) (.18,116.45) (.73,2.06) (.73,1.33) (.24,11.78) (.21,55.29) (.73,2.04) (.73,1.33) *Categorical values displayed as % abnormal with relative risk. AXVIII-1 �Appendix XVIII (Cont) RELATIVE RISKS FOR SELECTED CLINICAL END POINTS CLINICAL PARAMETERS RH Percent* C Relative Risk 95/5 Confidential Interval Exact Normal Approx Light Touch Muscle Status Vibration Patellar Reflex Achilles Reflex Biceps Reflex Babinski Tremor Coordination Romberg Gait 7.08 3.56 7.56 0.385 3.77 0.771 0.871 5.29 4.62 19.9 2.31 7.46 .95 3.62 .98 8.76 .86 .59 0.649 1.12 3.37 0.519 1.49 0.259 3.36 4.01 1.32 3.88 1.89 1.04 19.2 1.27 1.83 (.67,1.35) (.59,1.65) (.62,1.20) (.12,2.75) (.67,1.90) (.40,6.72) (.70,31.96) (.84,2.10) (.75,1.93) (.86,1.26) (.64,2.65) (.67,1.35) (.59,1.64) (.62,1.20) (.14,2.53) (.67,1.88) (.41,5.84) (.69,22.50) (.84,2.08) (.74,1.91) (.-86,1.26) (.64,2.58) Psychological Illness Isolation U14) Halstead-Reitan SCOT SGPT GGPT Alk Phos T Bill D Bili 3-45 4.62 33.5 13.9 7.8 10.8 17.3 1.8 29.0 1.7 26.0 34.7 2.07 2.34 33.5 14.8 8.6 10.3 16.9 2.0 29.7 2.1 27.7 36.1 1.67 1.97 1.00 .93 .91 1.053 1.020 .90 .98 .80 .94 .96 (.91,3.20) (1.14,3.58) (.85,1.17) (.74,1.18) (.66,1.26) (.79,1.40) (.83,1.26) (.43,1.90) (.84,1.13) (.38,1.67) (.80,1.10) (.85,1.10) (.90,3.12) (1.13,3.50) (.86,1.17) (.-74,1.18) (.66,1.26) (.79,1.40) (.83,1 .26) (.44,1.87) (.84,1.13) (.39,1.65) (.80,1.097) (.85,1.097) Uroporphyrins Coproporphyrins d-Aminolevulinic Acid 6.5 0.2 0.0 6.8 0.0 0.0 .94 (.58,155) (.58,1.54) Verified Hepatitis Jaundice Cirrhosis Other Hepatic Verified Reported Hepatomegaley Observed Hepatomegaley Skin Patches, etc., Reported Reported Acne (Post SEA) Reported Acne Severity Reported Chloracne Comedones Acneiform Lesions Acneiform Scars Cysts Hyperpigmentation Other Abnorms Any Abnormality 3.83 4.14 4.53 .39 .39 1.71 0.78 .93 ;93 .99 3.93 1.02 2.00 (.57,1.51) (.59,1.48) (.17,6.72) (1.13,21.07) (.48,2.26) (.75,6.21) (.57,1.50) (.59,1.47) (.19,5.52) (1.09,16.99) (.48,2.20) (.74,5.69) 1.23 1.18 1.24 .80 1.050 1.047 1.082 1.11 1.17 .78 1.00 (1.09,1 .40) (.67,2.18) (.74,2.21) (.55,1.21) (.87,1.26) (.85,1.29) (.82,1.43) (.84,1.46) (.84,1.65) (.61, .98) ( . 90 , 1 . 1 1 ) (1.09,1.39) (.67,2.15) (.74,2.20) (.55,1.21) (.87,1.26) (.85,1 .29) (.82,1.43) (.84,1.46) (.84,1.64) (.61, .98) (.90,1.11) LDH Choi Trig 4.21 .38 1.53 1.75 1.56 44.2 17.3 41.4 36 21.7 18.3 11.2 11 .6 8.3 12.6 45.0 36.0 14.7 33.3 45 20.7 17.5 10.4 10.5 7.1 16.3 44.9 AXVIII-2 �Appendix XVIII (Cont) RELATIVE RISKS FOR SELECTED CLINICAL END POINTS CLINICAL PARAMETERS Percent* C RH Relative Risk 95% Confidential Interval Normal Approx Exact Systolic Blood Pressure 10.4 14.3 .73 (.46,1.18) (.46,1.17) 23.1 24.6 .94 (.73,1.20) (.73,1.20) 5.2 5.4 .97 (.45,1.28) (.46,2.12) .84 13-9 23.1 .87 28.4 1.061 6.4 .66 8.5 .95 .86 2.7 8.4 1.038 9.5 1.35 .98 17.6 .52 1.85 .96 14.06 14.10 1.00 .670 .390 1.73 (.58,1.21) (.62,1.23) (.86,1.32) (.05,8.73) (.49,1.88) (.27,2.97) (.65,1.67) (.99,1.88) (.80,1.21) (.54,8.05) (.79,1.27) (.40,10.32) (.58,1.21) (.62,1.22) (.86,1.32) (.068,6.43) (.49,1.86) (.28,2.76) (.65,1.66) (.99,1.88) (.80,1.21) (.54,6.97) (.79,1.27) (.41,8.39) RBC WBC HGB HCT MCU MCH MCHC PLT 7.43 6.28 12.45 11.65 3.28 •3.'27 8.30 7.59 3.76 3.40 46.24 39.66 9.46 10.47 1.16 1.97 (.82,1.71) (.82,1.71) (.82,1.40) (.82,1.39) (.59,1.74) (.59,1.72) (.78,1.53) (.78,1.53) (.66,1.90) (.66,1.86) (1.043,1.30) (1.043,1.31) (.68,1.21) (.68,1.21) (.25,1.34) (.26,1.32) Occult Blood in Urine Protein in Urine Reported Kidney Disease T3+ T3 + 1.341 1.293 1.037 2.6 .50 1.3 1.60 5.6 3.5 .26 3.34 .87 5.72 8.47 .68 .77 • 39 1.98 .10 .39 .25 .29 0 0 .26 .88 15.19 17.27 .48 .52 .93 .77 4.93 6.37 <40 yrs Systolic Blood Pressure £40 yrs Diastolic Blood Pressure <40 yrs Diastolic Blood Pressure yrs EGG Findings <40 yrs EGG Findings <:40 yrs EGG A <40 yrs EGG A £40 yrs Eye gnds <40 yrs Eye gnds £40 yrs Peripheral Pulses Reported Heart Disease Reported Heart Attack Verified Heart Disease Verified Heart Attack T4 + FTIt FTI+ GLU TEST* TEST* 11.6 20.1 30.2 4.2 8.1 2.3 8.7 12.8 17.3 AXVIII-3 1.18 1.069 1.003 1.094 1.11 1.17 .90 .59 (.43,2.60) (.24,1.07) (1.00,2.59) (.69,31.77) (.47, .96) (.48,11.55) (.005,3.08) (.44,2.50) (.25,1.067) (1.00,2.56) (.68,22.37) (.47, .96) (.48,9.38) (1.13,2.64) (.71,1.09) (.20,4.67) (.52,1.16) (.71,1.094) (.22,4.098) (.52,1.15) �Appendix XVIII (Cont) MEAN SHIFTS FOR SELECTED CLINICAL END POINTS CLINICAL PARAMETERS R MEAN VALUE H C Conceptions per Participants 2.80 Mean Number of Marriages 1.24 Ulnar Nerve Cond (Above) 55.89 Ulnar Nerve Cond (Below) 60.52 Peroneal 48.23 Fatigue Score (HS ed) 15.33 Anger Score (HS ed) 11.27 Erosion (HS ed) 22.34 Anxiety (HS ed) 24.62 Depression (HS ed) 5.79 Fatigue (Coll ed) 12.79 Anger (Coll ed) 9.55 Erosion (Coll ed) 20.19 Anxiety (Coll ed) 21.23 Depression (Coll ed) 5.22 Cornell Index (HS ed) 9.21 Cornell Index (Coll ed) 3.66 MMPI Validity Scale (HS ed) 1.85 MMPI Defensiveness Scale (HS ed) 51.99 MMPI Consistency (HS ed) 51.95 MMPI Denial (HS ed) 53.95 MMPI Hypochondria (HS ed) 59.74 MMPI Depression (HS ed) 60.47 MMPI Hysteria (HS ed) 60.12 MMPI Psychopathic (HS ed) 56.38 MMPI Masc/Fem (HS ed) 55.94 MMPI Paranoia (HS ed) 51.72 MMPI Anxiety (HS ed) 57.27 MMPI Schizo (HS ed) 57.53 MMPI Mania (HS ed) 56.03 MMPI Social (HS ed) 52.31 MMPI Validity (Coll ed) 1.48 MMPI Defensiveness (Coll ed) 50.26 MMPI Consistency (Coll ed) 48.74 MMPI Denial (Coll ed) 58.46 MMPI Hypochondria (Coll ed) 55.42 MMPI Depression (Coll ed) 55.34 MMPI Hysteria (Coll ed) 59.75 MMPI Psychopathic (Coll ed) 55.21 MMPI Masc/Fem (Coll ed) 59.15 MMPI Paranoia (Coll ed) 53.62 MMPI Anxiety (Coll ed) 53.62 MMPI Schizo (Coll ed) 54.70 MMPI Mania (Coll ed) 55.22 MMPI Social Introversion (Coll ed) 46.83 Verbal IQ (HS ed) 110.61 Verbal IQ (Coll ed) 117.00 Perf IQ (HS ed) 102.40 Perf IQ (Coll ed) 113.70 AXVIII-4 MEAN SHIFT 2.79 1.22 56.12 60.71 48.93 13.64 9.99 20.00 21.91 5.30 12.83 9.46 19.90 20.51 4.46 6.44 3.44 1.73 52.03 50.65 55.63 57.22 58.39 58.90 55.89 54.85 50.68 55.59 55;97 54.49 50.80 1.95 50.33 48.44 58.41 54.65 54.57 59.32 55.66 57.87 53.26 54.18 54.89 54.05 .0036 .0164 -0.004 -0.003 -0.014 .1239 .1281 .1170 .1237 .0925 -.0031 .0095 .0146 .0351 .1704 .4301 .0640 .0694 -.0008 .0257 -.0302 .0440 .0356 .0207 .0088 .0199 .0205 .0302 .0279 .0283 .0297 -.241 -.0014 .0062 .0009 .0141 .0141 .0072 -.0081 .0221 .0068 -.0103 -.0035 .0216 47.50 101.73 116.84 104.14 112.37 -.0141 .0873 .0014 -.0167 .0118 �Appendix XVIII (Cont) MEAN SHIFTS FOR SELECTED CLINICAL END POINTS CLINICAL PARAMETERS R MEAN VALUE H C MEAN SHIFT Full Scale IQ (HS ed) Full Scale IQ (Coll ed) 101.18 117.30 102.74 116.59 -.0152 .0061 SCOT SGPT GGPT Alk Phos T Bill D Bill LDH Choi 33.0 20.3 10.1 7.69 .57 .23 142.1 212.2 33.1 20.5 39.3 7.52 .58 .24 141.7 216.6 -.0030 -.0098 .0204 .0226 -.0172 -.0417 .0028 -.0203 Trig Uroporphyrins Coproporphyrins d-Aminolevulinic Acid 121.9 30.2 30.8 2337.1 AXVIII-5 124.1 30.8 30.8 2371.4 -.0177 -0.0195 0.0 -0.0145 �Appendix XIX SPOUSE AND PARTICIPANT REPORTED BIRTH DEFECTS NOT MEETING STUDY CRITERIA Original Total Ranch Hand Comparison Comparison I CD NAME 140-239 NeopI asms Malignant melanoma-skin Uncertain behavior of skin Unspecified nature, ovarian 240-279 EndocrIne-MetaboIi c-Nutr i 11onaI-1mmune Gout Cystic flbrosis HypogammaglobulInemla Alblnlsm(ocular) 280-289 Blood & Blood-Forming Chronic lymphadenitis 0 1 2 290-319 Mental 8 2 8 16 23 Hyperkinetlc syndrome Dyslexia Learning dlsabiIIty Mental retardation 320-389 Nervous System & Sense Organs Epllepsy Meningitis Unspecified brain damage Polyneuropathy Visual disturbance LagophthaImos Esotropla Cerebral palsy Congenital deafness Endophthalmltls AmbIyopIa Acoustic nerve disorder Hearing loss Chronic otltls media 390-459 Heart Disease Unspecified 0 2 460-519 Respiratory 4 6 21 AIlergy Asthma Pulmonary congestion & hypostasis Unspecified disease of respiratory system 520-579 Digestive Tooth disorders Esophagitis Unspecified hernia Ruptured rectum 580-629 Genitourinary Kidney disorders 2 680-709 Skin and Subcutaneous Tissue Eczema Unspecified skin disorders 2 AXIX-1 �Appendix XIX (continued) SPOUSE AND PARTICIPANT REPORTED BIRTH DEFECTS NOT MEETING STUDY CRITERIA Or IgInaI TotaI Ranch Hand Comparison Comparison ICO NAME 710-739 Musculoskeletal & Connective Tissue Arthralgla Juvenile osteochondrosls of spine Sco11os i s Arthrogryposis Foot Deformity 760-779 Conditions Originating In the Perinatal Period Premature Hyaline membrane disease Birth trauma Atelectasls Perinatal Infection RH ISO Immunization Neonatal jaundice Transient neonatal electrolyte disturbance Unspecified hematological disorder Complications of labor & delivery ABO ISO Immunization FetaI hemorrhage 32 31 42 780-799 Symptoms, Signs, and ill-Defined Conditions Sudden death syndrome Functional & undlagnosed cardiac murmurs Enlarged lymph glands Others Jaundice, not of newborn Rash Other umbilical hernia SwelI ing or lump Lack of physiological development Billuria 26 15 18 120 81 122 TOTAL AXIX-2 �Appendix XX OBSERVED CANCER VERSUS SEER* DATA EXPECTED IN 1174 RANCH HANDERS (RH) AND 956 ORIGINAL COMPARISONS (COM) (QUESTIONNAIRE COMPLIANT) Probability Observed of Probability of Olbserved or Larger Cancer Type Expected Observed RH RH RH RH RH Testicle Bladder Digestive Lip and Oral Genitourinary 1 .09656 1.05838 4.00809 1.31739 3.59195 2 2 0 it 6 .2009 .1945 .0180 .0336 .0822 ;•; .1545 COM COM COM COM COM Testicle Bladder Digestive Lip and Oral Genitourinary .912751 .927593 3.52238 1.15221 3.11509 0 1 4 2 2 .4012 .3671 .1898 .2099 .2154 1.0000 .6047 .4684 .3201 .8179 *Surveillance, Epidemiology, and End Results (SEER) ^Statistically significant deficit ***Statistically significant excess AXX .2997 .2857 1.0000** .0448*** � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 058 Folder The folder containing the original item. 1573 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. Patricia M. Moynahan Richard A. Albanese William H. Wolfe Description An account of the resource <strong>Corporate Author: </strong>United States Air Force School of Aerospace Medicine, Aerospace Medical Division (AFSC), Brooks AFB, Texas Date A point or period of time associated with an event in the lifecycle of the resource February 24 1984 Title A name given to the resource An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Baseline Morbidity Study Results Subject The topic of the resource Air Force Health Study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/4ca43099304adab60cb910c56dd30288.pdf 11fc206d9fa1651135565ef9cbb253a7 PDF Text Text Item ID Number 01570 Aether Lathrop, George D. Corporate Author United States Air Force School of Aerospace Medicine, Report/Article Title Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Study Protocol Journal/Book Title Yeer MODth/Day December Color D Number of Images 192 Descrlpton Notes Wednesday, May 23, 2001 Page 1571 of 1608 �ALV1N L. YOUNG, Major, USAF Consultant, Environmental Sciences Report SAM-TR- 82-44 EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES: STUDY PROTOCOL George William Richard Patricia D. H. A. M. Lathrop, Colonel, USAF, MC Wolfe, Lieutenant Colonel, USAF, MC Albanese, M.D. Mpynahan, Colonel, USAF, NC December 1982 Initial Report for Period October 1978 - December 1982 Approved for public release; distribution unlimited. Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 USAF SCHOOL OF AEROSPACE MEDICINE Aerospace Medical Division (AFSC) Brooks Air Force Base; Texas 78235 �NOTICES This initial report was submitted by personnel of the Epidemiology Division and the Data Sciences Division, USAF School of Aerospace Medicine, Aerospace Medical Division, AFSC, Brooks Air Force Base, Texas, under job order 2767-00-01. When Government drawings, specifications, or other data are used for any purpose other than in connection with a definitely Government-related procurement, the United States Government incurs no responsibility or any obligation whatsoever. The fact that the Government may have formulated or in any way supplied the said drawings, specifications, or other data, is not to be regarded by implication, or otherwise in any manner construed, as licensing the holder, or any other person or corporation; or as conveying any rights or permission to manufacture, use, or sell any patented invention that may in any way be related thereto. The Office of Public Affairs has reviewed this report, and it is releasable to the National Technical Information Service, where it will be available to the general public, including foreign nationals. This report has been reviewed and is approved for publication. WILLIAM H. W0fcfE, Lt Col, USAF, MC Project Scientist ROY L. DEHART Colonel, USAF, MC Commander G E O R G E . LATHROP, Colonel, USAF, MC Supervisor �UNCLASSIFIED SECURITY CLASSIFICATION OF THIS P A G E (When Data Entered) READ INSTRUCTIONS BEFORE COMPLETING FORM REPORT DOCUMENTATION PAGE 1. REPORT NUMBER 2. GOVT ACCESSION NO SAM-TR-82-44 4. TITLE (and Subtitle) 3. RECIPIENT'S C A T A L O G NUMBER 5. TYPE. OE REPORT 4 PERIOD COVERED Initial Report October 1978 - December 1982 EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES: STUDY PROTOCOL 6. P E R F O R M I N G O^G. R E P O R T N U M B E R 7. AUTHOR(s) 8. C O N T R A C T OR G R A N T N U M B E R f s ; George D. Lathrop, Colonel, USAF, MC William H. Wolfe, Lieutenant Colonel, USAF, MC Richard A. Albanese, M.D. Pat.Hrla M., Mnynahan, Colonel, USAF. NC 9. P E R F O R M I N G ORGANIZATION N A M E AND ADDRESS 10. P R O G R A M ELEMENT. PROJECT, TASK AREA a WORK UNIT NUMBERS USAF School of Aerospace Medicine (EK) Aerospace Medical Division (AFSC) Brooks...Air Force Base, Texas 78235 65306F 2767-00-01 12. REPORT DATE 11. CONTROLLING OFFICE NAME AND ADDRESS December 1982 USAF School of Aerospace Medicine (EK) Aerospace Medical Division (AFSC) <3. NUMBER OF PAGES 185 4. MONITORING AGENCY NAME ft ADDRESSfM different from Controlling Office) 15. SECURITY CLASS, (of this report) Unclassified o. DECLASSIFI CATION/DOWN GRADING" SCHEDULE 6. DISTRIBUTION STATEMENT (of this Report) Approved for public release; distribution unlimited. 7. DISTRIBUTION STATEMENT (of the abstract entered In Block 20, It dltlerent from Report) 8. SUPPLEMENTARY NOTES 9. KEY WORDS (Continue on reverse aide If necessary and Identify by block number) Epidemiologic investigation Study design protocol Air Force Health Study RANCH HAND Matched cohort design 20. Nonconcurrent prospective design Defoliants Phenoxy herbicides Herbicide Orange Agent Orange 2,4-D 2,4,5-T Dioxin TCDD A B S T R A C T (Continue on reverse side If necessary and identity by block number) In 1979 the United States Air Force (USAF) made the commitment to Congress and to the White House to conduct an epidemiologic study of the possible health effects from chemical exposure in Air Force personnel who conducted aerial herbicide dissemination missions in Vietnam (Operation RANCH HAND). The urpose of this epidemiologic investigation is to determine whether long-term i FORM J A N 73 EDITION OF i MOV 65 is onsoi. !:TE UNCLASSIFIED S E C U R I T Y C L A S S I F I C A T I O N OF THIS P A G E ("IVhen Data Entered) �UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE(HT>en Data Entered; 20. ABSTRACT (Continued) health effects exist and can be attributed to occupational exposure to herbicides. This study uses a matched cohort design in a nonconcurrent prospective setting incorporating mortality, morbidity, and followup studies. Detailed computer searches of Air Force personnel records, with several crossreferencing techniques, have ensured total ascertainment of the RANCH HAND population. The unique circumstances of exposure in this population of 1264 individuals will permit a semiquantitative estimate of exposure. A comparison group will be formed from a population of 23,978 flight crew members and support personnel who were assigned to duty in Southeast Asia (SEA), but were not occupationally exposed to herbicides. These individuals will be matched to RANCH HAND personnel for the variables of age, type of job, and race. Since both the exposed subjects and their selected controls performed similar combat or combat-related jobs, many of the physical and psychophysiologic effects of combat stress and the SEA environment will also be equivalent in the two groups. In the analysis of mortality, each exposed subject and five randomly selected controls will be followed yearly for at least 20 years, constituting a 1:5 mortality design. The first of the mortality controls will be selected and entered into the questionnaire and physical examination phases of the study, producing a 1:1 morbidity design. The initial questionnaire will look backwards in time and will reconstruct occupational, social, and medical data to quantitate morbidity endpoints and confounding factors. All RANCH HAND personnel and their primary controls will be asked to participate in a comprehensive physical examination, with special emphasis being placed on dermatologic, neuropsychiatn'c, hepatic, immunologic, reproductive, and neoplastic conditions. The questionnaire will be developed and administered by a civilian opinion research organization of national stature under contract to the U.S. Air Force. In-home, face-to-face interviews will be conducted to maximize data quality. Medical and occupational data will be obtained from the study subjects. Fertility data will be obtained from the subject's spouse and/or former spouses whenever possible, preferably by face-to-face interview. In addition, next-of-kin interviews will be obtained for all study subjects who have died of noncombat-related causes between the time of their assignment to SEA and the initiation of this study. The physical examination will be conducted under Air Force contract at a single center by a civilian medical organization of national stature. Blind assessment protocols and strict quality control measures will be used to avoid bias and limit data variability,. Adaptive physical examinations and questionnaires will be developed for use in years 3, 5, 10, 15, and 20 of the followup study. Expected biases arid study difficulties include risk-taking behavior bias in the predominantly volunteer RANCH HAND group, response bias, interviewer bias, loss to study bias, and variability of procedures performed. UNCLASSIFIED S E C U R I T Y CLASSIFICATION OF THIS PAGE (Wien Dalit Entnrfd) �PREFACE In 1979 the United States Air Force (USAF) made the commitment to the Congress and the White House to conduct an epidemiologic study of possible health effects resulting from chemical exposure to Air Force personnel who conducted aerial herbicide dissemination missions in Vietnam (Operation RANCH HAND). The purpose of this epidemiologic investigation is to determine whether long-term health effects exist and, if so, whether they can be attributed to occupational exposure to herbicides or their contaminants. The study protocol for this effort incorporates a matched cohort design in a nonconcurrent prospective setting. The scientific protocol of the Air Force Health Study is presented here and is the result of a maturation process which began in October 1978. At that time, an epidemiologic strategy was developed. After approval of the basic approach was obtained from the USAF Surgeon General in early 1979, fullscale protocol development began in preparation for a series of peer reviews by a variety of expert panels. Throughout this review process, the advice and recommendations of each panel were used to enhance the protocol where appropriate. The following discussion summarizes key recommendations made by each review panel. These reviews were independent of one another, and the approval of one version of the protocol does not imply that those reviewers have approved the protocol in its final form. Although several members of the panels reviewing early protocol versions have received periodic courtesy progress reports, they have not had the opportunity to formally review the final product. The University of Texas School of Public Health, Houston, Texas, conducted the first review on 8 June 1979. The reviewers stressed the need to insure that the population groups selected for the study were fully ascertained, and that sources of potential bias should be carefully addressed. The advantages of face-to-face interview technique over telephone techniques were discussed as well. On 6 and 7 August 1979, a panel appointed by the USAF Scientific Advisory Board recommended that face-to-face interviews should be used and that the mortality phase of the study be expanded from a 1:1 to a 1:3 design to increase statistical power. Toxicologic aspects of the study and their impact on the scope of the physical examination were extensively discussed. A subcommittee of the Armed Forces Epidemiologic Board conducted a review on 30 and 31 August 1979. The committee members recommended the appointment of an independent monitoring panel to oversee the conduct of the study on a periodic basis. They felt that it was necessary to expand the mortality study to a 1:5 design, with subjects randomly drawn from a 1:10 cohort matrix. Quality control concerns and the advisability of using a single examination center were also recommended. The National Academy of Sciences (NAS) reviewed the protocol on 18 December 1979. The NAS recommendations stressed the need to place increased emphasis on reproductive endpoints, and to expand statistical power calculations, methods of population ascertainment, location, and long-term followup. They reiterated the value of ongoing peer review by a monitoring group. They also strongly encouraged the Air Force to conduct the • • N.idy by contract to an independent agency to avoid the appearance of conflict (•> interest. Following the NAS review, additional reviews by the Science I ,:!iel of the Agent Orange Working Group and the Advisory Committee on Special �Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants were obtained. A subcommittee of this Advisory Group, chaired by Dr. John Moore, Director of Toxicology and Testing Programs, National Institute of Environmental Health Sciences, was appointed to monitor the study. Reviews by this subcommittee continue on a regular basis. The edition of the protocol presented in this technical report is the protocol in effect at the time the physical examination phase of the study began in January 1982. Subsequently, circumstances beyond the control of the principal investigators led to some modifications in portions of the design. These modifications are discussed in annexes to the basic protocol (Chapters XVII, XVIII, XIX of this report) and are summarized. The principal investigators' increasing knowledge of the operational environment of the Vietnam War and the herbicide dissemination programs, and a more complete knowledge of the advantages and limitations of available records, contributed to the refinement of this document. Initially, an individual-specific exposure index or estimate was planned, but these highly specific estimates of exposure were not feasible. Objective data sources were not available to permit development of the index on the individual level, and therefore the use of a more generalized index is required. The initial ascertainment of the control population was conducted by a computer search of the Air Force personnel records system coupled with a manual search of noncomputerized records. This process resulted in the inadvertent overselection of some comparison individuals who were subsequently found not to meet the criteria for inclusion in the study. These ineligible individuals were removed from the study cohorts, and appropriate subjects were substituted for them. Analysis of the problem revealed that there was true overselection of subjects, and that no eligible subjects had been overlooked. Thus, the statistical and scientific validity of the study has been preserved. As a result of this event, the comparison cohort matrix was reduced from 1:10 to 1:8. This reduction will have minimal consequences, since the 1:5 mortality analytic design and the 1:1 morbidity design are maintained. The primary focus of this study is the potential effects of herbicide/ dioxin exposure on health outcomes. However, the flexibility of the statistical methodology, the comprehensive nature of the data being collected, and the high rates of participation in the questionnaire and examination process will permit the analysis of other factors. This final protocol represents a synthesis of the comments of all of the peer reviews, coupled with the increasing sophistication of knowledge concerning record sources and operational features of the war. The evolution of this document has occurred over a four-year span of time. This evolutionary process is outlined in the following table. Refinements of concepts and procedures were the only changes made to the study design since November 1979. There have been no substantive changes in study design methods or procedures since that time. Analytic techniques may be further refined to represent state-of-the-art statistical methodology. ii �PROTOCOL EVOLUTION Protocol Version 1 2 Major Areas of Change Date 6 June 1979 1979 - Expanded discussion of epidemiclogic design - Expanded statistical analytic strategy - Consideration of bias sources 30 July 1979 - Discussion of exposure index - Development of survival analysis techniques - Expanded discussion of physical examination procedures 30 August - Expanded discussion of exposure concepts - Expansion of mortality study to a 1:3 design - Discussion of compliance factors - Further expansion of physical examination procedures 10 July 1979 31 October 1979 - Expansion of mortality cohorts to - Single center examinations - Discussion of the replacement concept for bias correction 28 November 1979 » Expanded exposure index discussion - More detailed discussion of statistical analytic strategy 8 October 1980 - Increased emphasis on fertility and reproductive endpoints - Enlarged discussion of the mortality analysis - Enlarged discussion of statistical power - Discussion of Quality Control methods 26 November 1980 - Presentation of refined data on study population demographic characteristics in �9 15 June 1981 - Discussion of matching procedures - Consideration of time-in-study effects 10 September 1981 - Expanded discussion of matching procedures and results 11 28 January 1982 - Refinement of the exposure index - Presentation of modified performance schedules IV �ACKNOWLEDGMENT The services of many staff members and consultants of the United States Air Force School of Aerospace Medicine are acknowledged. Special acknowledgment is made to the following co-investigators: Clarence F. Watson, Jr., M.D., M.P.H.; Alvin L. Young, B.S., M.S., Ph.D.; Joel E. Michalek, Ph.D.; Phelps P. Crump, Ph.D.; Richard C. McNee, M.S.; Alton J. Rahe, M.S.; Michael A, Sairi, M.D., M.P.H., and T.M.; Richie S. Dryden, M.D., M.P.H.; James A. Wright, M.D., M.P.H., who provided consultation on study design and physical examination development. In addition, special acknowledgment is made to the United States Air Force School of Aerospace Medicine, Management and Air Training Command Procurement Personnel: Hugh F. Mulligan, Colonel, USAF, BSC, Chief, Program Acquisition Division; Donald F. Norville, Air Training Command Contracting Officer, Randolph Air Force Base, Texas, who coordinated the requirements for the physical examination implementation contract. �PROJECT RANCH HAND II EXECUTIVE SUMMARY OF THE PROTOCOL The Air Force has made the commitment to Congress and to the White House to conduct an epidemiologic study of possible health effects in the Air Force personnel (RANCH HAND) who conducted aerial herbicide missions in Vietnam. The purpose of this investigation is to determine whether long-term health effects exist and can be attributed to occupational exposure to Herbicide Orange. The extensive use of herbicides in Vietnam between 1962 and 1971 was terminated when it became known that TCDD, a contaminant present in 2,4,5-Tcontaining herbicides, caused congenital abnormalities when administered to pregnant rodents. Subsequent extensive research into the toxicity of TCDD in animals remains equivocal from the point of view of human population risks. Presently, the potential for teratogenicity and carcinogen!city of TCDD seems to be significant, but species specific. The scientific literature on the toxicity of the components of Herbicide Orange reveals that the two main ingredients, 2,4-D and 2,4,5-T, have extremely low toxicity, and are distinctly different in nature than TCDD. TCDD has been shown to be embryotoxic at markedly lower doses in animals. Only recently have comprehensive prospective studies in humans been undertaken. Most previous epidemiologic studies dealing with TCDD exposure in humans have suffered from weakness in design and statistical power. These studies have only validated a link between TCDD exposure and the subsequent development of chloracne. However, the public's perception of the toxicity of Herbicide Orange/TCDD is generally different from that of the scientific community. A review of veteran inquiries submitted to the Veterans Aiiniini strati on reveals an awesome spectrum of alleged symptoms and diseases. This study uses a matched cohort design in a nonconcurrent prospective setting incorporating mortality, morbidity, and followup studies. Detailed computer searches of Air Force personnel records, with several cross-referencing techniques, have ensured total ascertainment of the RANCH HAND population. The unique circumstances of exposure in this population of 1264 individuals will permit a semi-quantitative estimate of exposure. Specifically, since there was a documented higher concentration of TCDD contamination prior to 1965, this factor will be incorporated in the development of an exposure index. A control group will be formed from a population of 23,978 C-130 crewmembers and support personnel who were assigned to duty in Southeast Asia (SEA), but were not occupationally exposed to herbicides. Control Individuals will be matched to RANCH HAND personnel for the variables of age, type of job, and race. Since both the exposed subjects and their selected controls performed similar combat or combat-related jobs, many of the physical and psycho-physiologic effects of combat stress and the SEA environment will also be equivalent in the two groups. Ten statistically equivalent matches for each exposed subject will form the control set for each exposed subject. In the analysis of mortality, each exposed subject and a 50% random selection from each control set will be followed yearly for at least 20 years, constituting a 1:5 mortality design. The first of the randomized mortality controls w1Tl be selected and entered into the questionnaire and physical examination �phases of the study, producing a 1:1 morbidity design. The initial questionnaire will look backwards in time and will reconstruct occupational, social, and medical data to quantitate morbidity endpoints and confounding factors. Subsequent questionnaires and physical examinations will constitute a followup morbidity study of living exposed subjects and suitable living controls. In this followup phase, primary controls who are noncompliant will be replaced by another suitable control from the control set so that both statistical power and loss to study bias in the followup study may be improved. Controls dying after the initiation of the followup will not be replaced. All RANCH HAND personnel and their primary controls will be asked to complete a questionnaire and participate in a comprehensive physical examination, with special emphasis being placed on dermatologic, neuropsychiatric, hepatic, immunologic, reproductive, and neoplastic conditions. The questionnaire will be developed and administered by a civilian opinion research organization of national stature under contract to the U.S. Air Force. In-home, face-to-face interviews will b,e conducted to maximize data quality; however, noncompliant individuals will be requested to participate in a shortened telephone interview. Medical and occupational data will be obtained from the study subjects. Fertility data will be obtained from the subject's spouse and/or former spouses whenever possible, preferably by faceto-face interview. In addition, next-of-kin interviews will be obtained for all study subjects who have died of non-combat-related causes between the time of their assignment to SEA and the initiation of this study. The physical examinations will be conducted under Air Force contract at a single center by a civilian medical organization of national stature. Blind assessment protocols and strict quality control measures will be used to avoid bias and limit data variability. A $100 per day stipend will be paid to all eligible subjects to maximize participation in the study. Adaptive physical examinations and questionnaires will be developed for use in years 3, 5, 10, 15, and 20 of the followup study. Expected biases and study difficulties include risktaking behavior bias in the predominantly volunteer RANCH HAND group, response bias, interviewer bias, loss to study bias, and variability of procedures performed. Since this study is dealing with nonspecific clinical endpoints, identification or elucidation of a disease state or syndrome by statistical methodology is a prime thrust of the investigation. Inferences about a disease state will be developed by identifying symptom complexes or physical findings which in themselves may represent disease. By comparison of symptoms, signs, and laboratory tests within and between groups, a logical decision-making scheme can be utilized to calculate relative risks from baseline data. If appropriate, these results will be used to sharpen adaptive approaches in the followup study. By the use of combinational and correlational analysis, statements about the probability of a disease state, a subclinical state, and/or over-reporting bias will be attempted. In addition, the application of regression techniques to a normalized exposure index among exposed individuals exhibiting symptoms and/or signs will also assist in the clarification of a disease state or syndrome. Mortality data will be analyzed using several different approaches, including age and age-disease specific rates, standardized mortality rates, and modified life table approaches, as well as more sophisticated logistic and multiplicative models. Analysis of questionnaire and vii �physical examination data will utilize log-linear models for dichotomous or polytomous data to verify the appropriateness of the standard statistical methodologies (e.g., McNemar's test for dichotomous rates). Continuous variables will undergo covariance analysis to remove noncontrolled effects, followed by the use of a paired difference statistic. Some data will naturally fall into groups or batteries (e.g., fertility/reproduction, liver function tests); in which case, group scoring techniques will be used as appropriate. viil �CONTENTS I. II. III. IV. Purpose of the Investigation „ Synopsis of Background A. The USAF Commitment B. The Peer Review Process C. The Military Use of Herbicides 2 2 2 3 Goals of the Investigation 6 Synopsis and Discussion of Literature A. Overview B. Pharmacokinetics of 2,4-D, 2,4,5-T and TCDD.. C. Proposed Cellular Mechanisms of Action 7 7 7 for TCDD D. E. F. 6. V. 9 Animal Studies Case Reports Veteran Concerns Epidemiologic Studies 10 11 12 13 Epidemiologic Study Design... A. Design Considerations , B. Selection and Ascertainment of the Populations for Study C. Overview of Statistical Methodology D. Mortality Study ... E. Morbidity Study 17 17 18 28 31 37 F. Followup Study 43 G. Determination of "Disease" H. Exposure Indices VI. 1 , 48 50 Special Statistical Considerations A. False Reporting/Misrepresentation 55 55 B. Adequacy of Sample Sizes 56 C. The Replacement Concept D. Statistical Analysis of Large Data Sets E. Time-In-Study Effects 62 74 74 IX �CONTENTS (Continued) Page VII. VIII. IX. X. XI. XII. XIII. XIV. XV. XVI. XVII. XVIII. XIX. Data Repository , Recognized Study Difficulties and Corrective Measures A. Medical Precedence B. Group Accountability Bias C. "Risk Taking" Behavior Bias D. Response Bias E. Interviewer Bias.... F. Changes to the Protocol.. G. Loss to Study—Statistical and Bias Consi derat i ons H. Statistical Power Limitations I. Variability of Procedures J. Confounding Exposure Factors . 77 .. 79 79 79 80 81 82 82 83 85 86 86 Quality Assurance and Management Considerations A. Quality Control. B. Management Structure 88 88 90 Reporting Procedures 93 Questionnai re 94 Physical Examination Design A. General Instructions B. Conduct of the Examination C. Special Procedures 96 96 97 103 Principal Investigators, Co-Investigators and Contributors........ 112 Selected Bibliography. 115 Appendi x 125 Examiner's Handbook A. General Instructions B. Conduct of the Examination C. Examination Format D. Special Procedures.... E. Forms 138 138 139 140 146 160 Annex 1 - Exposure Index Construction 166 Annex 2 - Comparison Group Ineligibility ., 168 Annex 3 - Supplemental Analysis 170 Glossary of Abbreviations 171 �List of Figures F_igure ^No^ Page 1 Mortality Analysis Cohorts 2 Mortality Matrix.. 25 3 Selection Procedure for the Questionnaire, Physical Examination, and Followup Study 28 4 Design Schematic.... 29 5 Effect of Non-Random Loss to Study in The Control Population 45 6 The Replacement Strategy 46 7 Control Replacement for Morbidity and Followup Studies 47 8 Interpretation of Horizontal Comparisons 49 9 False Reporting/Misrepresentation 55 10 Misrepresentation in Ranch Hand II 56 11 Apparent Relative Risk Versus Specificity 57 12 Power Versus Relative Risk, 1:5 Mortality Study by Three Disease Incidences. 67 Power Curves of the Marine Study Considering Relative Exposure and Misclassification of the Study Population; 5,900 Marines 68 Power Curves of the Marine Study Considering Relative Exposure and Misclassification of the Study Population; 21,900 Marines 69 15 Rationale of Replacement 84 16 Management Structure, 91 13 14 A-l A-2 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) .. 25 135 Estimated Identification/Participation of RANCH HAND Population A-3 136 Study Design Format 137 �List of Tables Table No. 1 Page Estimated Quantities of Herbicides and TCDD Sprayed in RVN, Jan 1963 - Feb 1972 2 4 Summary of Descriptive Characteristics of Patients in the VA Herbicide Registry, as of 10 February 1981... 13 3 VA Herbicide Registry Symptom Reporting 13 4 Feasibility of Identifying Aircraft Maintenance Personnel (Total Population) Exposed to Herbicide Orange 19 Comparisons of the Study Group to Possible Control Groups by Known and Estimated Factors. 23 6 Results of the Matching Process (1:10) 26 7 Stratified Format of Age-Specific Death Rates 32 8 Format of McNemar's Test... 35 9 Format of Categorical Representation of Retinal Changes 42 10 Format of Pairing for Grades of Retinal Findings 42 11 Power Calculations.,... 12 Power Calculations for the Dichotomous Variable Case as a Function of Efficacy of Paired Designs 59 13 Power Calculations as a Function of Herbicide Effect.... 61 14 Mortality Analysis. Power Comparison of the RANCH HAND Study to the Marine Population; 5,900 Marines, Rare Disease 63 Mortality Analysis. Power Comparison of the RANCH HAND Study to the Marine Population; 5,900 Marines, Common Disease 64 Mortality Analysis. Power Comparison of the RANCH HAND Study to the Marine Population; 21,900 Marines, Rare Disease 65 5 15 16 ,. 58 �List of Tables (Continued) Table No. 17 18 Page Mortality Analysis. Power Comparison of the RANCH HAND Study to the Marine Population; 21,900 Marines, Common Disease... 66 Control Distributions by Examination, Matching 1000 RANCH HAND Personnel... 71 19 Factors Affecting Compliance 73 20 Titne-In-Study Effects 76 A-l Summary of 2,4-D, 2,4,5-T and TCDD Animal Studies 126 A-2 "Symptom Complex" Derived from Literature Review of Case Studies Exposed to 2,4-D, 2,4,5-T and/or TCDD 127 Detailed Listing of Symptoms/Signs by Major Category from Literature Review of Case Studies Exposed to 2,4-D, 2,4,5-T and/or TCDD 128 Age Comparison of Exposed Subjects and their Matched Controls 130 A-5 Statistical Description of the Matching Process 131 A-6 Specific Rules For Entry Into the Morbidity Study 132 A-7 Schedule and Mode of Contacts with Study Subjects 133 A-8 Monte Carlo Simulation.... 134 A-3 A-4 xiii �PROJECT RANCH HAND II EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDE ORANGE MATCHED COHORT DESIGN I. Purpose of the Investigation The purpose of this epidemiologic investigation is to determine whether long-term health effects exist and can be attributed to occupational exposure to Herbicide Orange. �II. Synopsis of Background A. The USAF Commitment Since 19713 news media presentations have focused attention on possible adverse health effects in former military personnel, allegedly due to Herbicide Orange [a mixture of 2,4-dichlorophenoxyacetic acid (2,4-D) and 2,4,5-trichlorophenoxy-acetic acid (2,4,5-T)] which was used as a defoliant during the Vietnam Conflict. Other herbicides containing 2,4,5-T were also used extensively, and as commonly used by the news media, the term "Herbicide Orange" refers to all of these 2,4,5-T products. These herbicides were contaminated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (Figure A-l, Section X V ) , and the presence of this toxin is the basis for much of the concern over exposure to these defoliants. Claims for compensation have been filed against the Veterans Administration (VA), by more than 3,000 veterans. In response to Congress, the General Accounting Office (6AO) investigated the issue and subsequently recommended that the Department of Defense (DOD) conduct a longterm epidemiologic study of the problem. The Department of the Air Force has made a formal commitment to the Congress and the White House to conduct such a study. On 16 September 1980, the White House directed the DOD to initiate the RANCH HAND study with reasonable speed and high quality. This decision was subsequently reaffirmed by the new administration. B. The Peer Review Process This protocol has received rigorous peer review. From the outset, the Air Force principal investigators have acknowledged the scientific complexities of the effort and voluntarily sought outside peer review and consultative guidance. The following reviews have been conducted: Reviewing Agency Date University of Texas, School of Public Health June 1979 Air Force Scientific Advisory Board August 1979 Armed Forces Epidemiologic Board August 1979 National Research Council, National Academy of Sciences December 1979 Members of each independent review agency were provided copies of the protocol and key references in advance of the review. An extensive briefing of the protocol was presented to three of the four agencies. Each review group provided a report of their opinions and recommendations. The Air Force principal investigators responded to reports from the first three peer reviews and indicated concurrence or nonconcurrence with each of the recommendations. Most of �the peer group recommendations were gratefully accepted and incorporated appropriately within the protocol. Because the National Research Council's report cited "major deficiencies in design" and emphasized public credibility issues, the protocol was referred to the Interagency Work Group to Study the Possible Long-Tehn Health Effects of Phenoxy Herbicides and Contaminants for an additional scientific review and recommendations to the White House as to whether the Air Force should conduct this study. This review was conducted in June 1980 and resulted in an affirmative recommendation. The White House subsequently directed that the study be formally started. C. The Military Use of Herbicides Research and development on phenoxy herbicides began in the early 1940s, when most of the initial phytotoxic screening programs and the development of application technologies were sponsored by the DOD. The herbicide, 2,4,5-T, was first commercially produced in the United States in 1944. During the years from 1961 through 1969, the DOD procured 53 million pounds of this herbicide (approximately 34 percent of the total US production) for use in the Republic of Vietnam (RVN). However, 8.9 million pounds of that amount were not sprayed in Vietnam, but were destroyed by at-sea incineration in 1977. The first sustained DOD operational use of herbicides was initiated during the Vietnam Conflict (Operation RANCH HAND) and the first shipment of herbicides used in RANCH HAND was received at Tan Son Nhut Air Base, (RVN), on 9 January 1962. The use of these compounds was intended to accomplish two objectives: (1) the defoliation of vegetation to improve visibility and thus decrease the risk of ambush, and (2) the destruction of enemy crops. Four 2,4,5-T-containing herbicides were used by the military during the period 1962-1970. These four included: (1) Herbicide Purple (used from 1962 through 1964) n-butyl n-butyl iso-butyl (2) 2,4,5-T 2,4,5-T 60% 40% Herbicide Gree.n (used from 1962 through 1964) n-butyl (4) 50% 30% 20% Herbicide Pink (used from 1962 through 1964) n-butyl iso-butyl (3) 2,4-D 2,4,5-T 2,4,5-T 2,4,5-T 100% Herbicide Orange (used from early 1965 through 15 April 1970) n-butyl n-butyl 2,4-D 2,4,5-T 50% 50% �Analyses of archived samples of Herbicide Purple suggest that the mean concentration of TCDD may have been approximately 33 ppm (Range: 17 to 47 ppm TCDD) while archived samples of Herbicide Orange had a mean concentration of approximately 2 ppm (Range: <0.02 to 15 ppm TCDD). In addition, two other herbicides were widely used in RVN. These were Herbicide Blue, an organic arsenical formulated from the sodium salt of cacodylic acid, and Herbicide White, a water soluble triisopropanolamine salt formulation of 2,4-D and picloram. The amounts of the various herbicides used in RVN from January 1962 through February 1972 are shown in Table 1. Table 1. ESTIMATED QUANTITIES OF HERBICIDES AND TCDD SPRAYED IN RVN, JAN 1962-FEB 1972 CHEMICAL POUNDS 2,4-D 2,4,5-T TCDD 55,940,150 44,232,600 368 Picloram Cacodylic Acid Herbicide Total 3,041,800 3,548,710 106,763,260 Ninety-six percent of the 2,4,5-T disseminated in RVN was contained in Herbicide Orange; the remaining 4 percent in Herbicides Green, Pink, and Purple. However, Herbicides Green, Pink and Purple contained approximately 40 percent of the estimated amount of TCDD disseminated in RVN. Green, Pink and Purple were sprayed as defoliants on less than 90,000 acres from 1962 through 1964, a period when only a sma1! force of U.S. military personnel were in RVN. Ninety percent of all the Herbicide Orange (containing 38.3 million pounds of 2,4,5-T and 203 Ib of TCDD) was used in defoliation operations on 2.9 million acres of inland forests and mangrove forests of RVN. Most of the herbicide used in RVN was sprayed from aircraft. RANCH HAND aircraft, the C-123, disseminated 88 percent of all herbicide. Helicopters and ground application equipment used by personnel from all branches of the U.S. Armed Forces applied the remaining 12 percent, primarily Herbicide Blue, to maintain visibility around base perimeters. Concurrent with the change to Herbicide Orange, the scope of aerial use shifted from four aircrews on temporary assignments, to 36 permanently assigned aircrews, and additional support personnel. Following the announcement in October 1969 that the administration of 2,4,5-T to pregnant rodents �caused an increase in the rate of congenital abnormalities, the DOD confined Herbicide Orange spray operations to nonpopulated areas and in April 1970, all uses of the 2,4,5-T containing herbicides were halted. Other non-2,4,5-T herbicides continued to be used until June 1971 and Operation RANCH HAND was officially deactivated in October 1971. In March 1972, all remaining stocks of 2,4,5-T-containing herbicides were removed from RVN, and transported to Johnston Island, Pacific Ocean, for open storage (Project PACER IVY), and eventual incineration at sea in 1977 (Project PACER HO). In 1979, the Environmental Protection Agency (EPA) suspended the use of herbicides containing 2,4,5-T because an epidemiologic study in the United States attributed abortogenic effects to its use. �Ill. Goal s of the Investigation The health goals of this investigation are: (1) to identify veteran and active duty individuals with adverse health effects (physical and psychological ) if any, which are attributable to herbicide exposure, and (2) to identify other individuals at risk of developing future adverse health effects, if such exist. Spinoffs from the primary health goals are clearly evident. Increasing media emphasis, in tandem with rising veteran concern and Congressional action, have caused numerous governmental agencies to pursue the issue from several scientific perspectives. The RANCH HAND study is an important part of the overall scientific mosaic, but in itself, may not be definitive in answering the herbicide question. Nevertheless, it is clear that data and conclusions arising from this investigation, whether positive, negative, or indeterminant, will be used as a substantative data base upon which government can formulate policy decisions. With numerous individual and class action lawsuits pending, currently totaling in excess of $44 billion, the primary governmental decision will concern compensation for attributable adverse health. As the award of compensation to any veteran is solely controlled by the Veterans Administration, this Air Force study in no way represents a "conflict" but rather constitutes another reaffirrnation that "The Air Force cares for its own." �IV• Synopsis and Discussion of J-i A. Overview More than 20,000 scientific articles relating to the phenoxy herbicides have been published since the 1940's. Many of the articles cite herbicidecaused health effects in a variety of animal species, but most early studies used a myriad of herbicide formulations and unknowingly dealt with physically and chemically impure compounds, and the assay technology was far short of today's state-of-the-art. Many human studies have ascribed cause and effect relationships but have suffered from problems of clinical empiricism or questionable methodology. The only consistent chronic clinical finding associated with exposure to 2,4,5-T herbicide and TCDD has been chloracne, recognized by most workers as the herald sign of overexposure to the herbicide and other chloracneigens. It is now recognized that the chloracne was caused by the presence of TCDD rather than 2,4,5-T. Sequelae from chloracne, localized or systemic, appear to be unusual according to the preponderance of the literature. It is appropriate to note that sustained, worldwide usage of herbicides for 30 years has not yet evoked a readily identifiable disease state. It is clear from the literature and the usage history of herbicides that if there are significant attributable long-term health effects, they are either reasonably rare, or of such nonspecific commonality that they blend unnoticeably into the symptoms, syndromes, or diseases associated with increasing age or other similar factors. B - Pharmacokinetics of2,4-D, 2,4,5-T and TCDD (1) LM The pharmocokinetics of 2,4-D have been well studied in animals. 2,4-D is readily absorbed after oral administration, and is initially distributed in high concentrations to the central nervous system and liver. Eventually, all tissues are involved, with the kidneys accumulating twenty times the concentration of the other tissues. The plasma half-life of 2,4-D is approximately 3 to 12 hours, with elimination from the body through the kidneys at a dose-dependent rate. Generally, high doses or repeated lower doses result in tissue accumulation. The majority of 2,4-D is eliminated unmetabolized; however, esters of 2,4-D have been shown to undergo hydrolysis prior to excretion. Muscle and fat show the lowest accumulation of 2,4-D on repeated exposure, whereas the kidneys and liver show the highest accumulations. Within 24 hours of ' ingle-dose administration of 2,4-D, 16.8% was present in the uterus, placenta, fetus and amniotic fluid in gravid rats. In addition, 2,4-D was found in the milk of lactating rats for up to six days following single-dose exposure. (2) ldi5-T The pharmacokinetcs of 2,4,5-T have been well studied in animals. In all animals, 2,4,5-T has been shown to be readily abosrbed upon oral administration. However, beyond this point, 2,4,5-T has shown marked variations in its pharmacokinetit.s depending on the species tested. These differences are thought to be due to variations in species, age, dosage levels, routes of administration �and chemical formulations used in the various studies. Generally, the distribution is ubiquitous throughout the body except in hamsters, which show no placental passage, and in mice, which show placental passage only in late gestation. Clearance from plasma and from the body varies greatly among species with rats showing faster clearance than dogs, mice and man. In addition, this clearance appears to be generally dose-dependent. The biological half-life of 2,4,5-T in rats, as estimated by tissue analyses and urinary clearance at administered dosages of 5 mg/kg, is 4.7 hours. However, at 200 mg/kg, the half-life in rats is prolonged to 25 hours. Excretion of 2,4,5-T is primarily via the kidneys. The elimination of 2,4,5-T at low doses is essentially achieved in an unmetabolized form. However, at higher or more chronic doses, elimination involves the liver in a more active role (i.e*, conjugation). Higher doses and repeated lower doses appear to result in accumulation in animal tissues. (3) Phenoxy Herbicides in Humans Relatively few studies have dealt with the pharmacokinetics of 2,4-D and 2,4,5-T in humans. Numerous reports of occupational exposures in industry and in commercial and private herbicide applications have supported percutaneous entry as a major route of exposure. Rapid absorption of 2,4-D and 2,4,5-T has been observed after oral administration. The primary mode of excretion of the phenoxy herbicides is via the urine with 74% of 2,4-D and 63%-72% of 2,4,5-T being cleared from the body within the first 96 hours. The majority of the herbicide is unmetabolised prior to excretion and the biological half-life of 2,4-D and 2,4,5-T in humans (as estimated by tissue analyses and urinary excretion) is 33 hours and 18 hourc, respectively. Tissue analysis has revealed an ubiquitous distribution of the herbicides after absorption. Limited studies on the accumulation of the phenoxy herbicides following repeated doses suggest that such accumulation in humans is unlikely. This is in contrast to numerous animal studies on 2,4-D and 2,4,5-T which show that such accumulation does occur. No specific data are available on the odor threshold of Herbicide Orange. Data are available however, on the odor threshold of a butyl ester formulation of 2,4,5-T. The odor threshold was found to be about 0.3 ppb 3 (the taste threshold was 1.3 ppb). A Threshold Limit Value (TLV) of 10 mg/m for both 2,4-D or 2,4,5-T has been adopted by the American Conference of Governmental Industrial Hygienists. Thp TLV is a time-weighted average concentration for a normal 8-hour workday/40-hour workweek to which workers may be repeatedly exposed, day after day, without adverse effect. Analysis of ambient air samples collected adjacent to and downwind from actual dedrumming operations involving Herbicide Orange were at least two orders of magnitude below the TLVs. (4) TCDD Information on the absorption, distribution and excretion of TCDD has been mostly derived from animal models. Studies in rats, mice and guinea pigs generally show that intestinal absorption of TCDD is relatively complete, with a large proportion being stored unmetabolized in the liver. The majority of this TCDD is assumed to be localized in the liver microsomes (centrifugation 8 �techniques). Initially, adipose tissue accumulates TCDD, followed later by accumulation in the liver, adrenals, kidneys and lungs. The level of TCDD in the liver and adipose tissue is about ten-fold greater than in other body tissues; however, significant species variability has been observed. The biological half-life of TCDD varies by species, but is reported to range from 12 to 50 days. The major route of excretion is via the feces with urinary excretion occurring at a much reduced rate. TCDD has three proposed mechanisms of action by effects, both documented and suspected, can be understood. able information in this area is derived from animal, models. The few human studies dealing with mechanisms are cal manifestation of chloracne. which its variety of All currently availplant, and bacterial limited to the clini- TCDD's ability to induce a variety of microsomal enzymes is well documented. The induction of aryl hydrocarbon hydroxylase, delta-aminolevulinic acid synthetase, and cytochrome P-448/P-450 associated enzymes has been implicated in the development of cutaneous porphyria. The induction of aryl hydrocarbon hydroxylase and other mixed function oxygenases/oxidases has been associated with carcinogenesis and tumorogenesis. In addition, TCDD has been shown to be a possible promoter or cocarcinogen of known carcinogens. In some nonhuman studies, TCDD produced a protective effect against endocrine tumors (e.g., pituitary, uterine, pancreatic, adrenal, and mammary tumors). TCDD's induction of UDP-glucuronyl transferase, an important enzyme in steroid metabolism, may explain this peculiar effect. The induction of DT-diaphorase and lysosomal acid proteinases has been implicated in TCDD's neuropathic effects. These and other biochemical alterations may account for TCDD's clinical manifestation of chloracne resulting from an over production of keratin in the sebaceous ducts. (2) DNA/TCDD Interact 1 on . Alterations in the structure and fidelity of transcription of DNA due to TCDD have been indirectly demonstrated, TCDD, because of its planar ring structure, may "intercalate" with DNA causing "frame-shift" mutations in a manner similar to that seen with the acridine family of compounds. A few laboratory studies with bacterial systems (Escherichia coli and SaJjroQeljla. typ n i mu rium) and one plant system (the Af ri can "Blood*" Li ly ) have implicated TCDD as being capable of producing chromosomal aberrations and perhaps a weak dominant lethal effect. This hypothesized DNA/TCDD interaction could explain the development of chloracne, as well as the suggested mutagenic and carcinogenic effects, if similar mechanisms occur in mammalian species, A nonspecific or as yet unspecified toxicity continues to serve as reasonable mechanism for TCDD's hepatic and thymus toxicity. TCDD has been �described by some as "one of the most potent, low molecular weight toxins known", with extremely low concentrations producing severe Jiver damage and death in various animal studies. The immune suppression effect of TCDD has been shown to result specifically from its T-cell (thymus) toxicity. If bioaccumulation and persistence of TCDD occur in human adipose tissue, it could be released into the circulation under situations of weight loss (e.g., life style modification, medical indications, or disease). Such hypothesized reemergence of the agent could result in low doses being either detectable and/ or toxic at some later point in time. If TCDD's primary toxicity results from low doses (e.g., a mutagenic/carcinogenic effect) rather than high doses (e.g., cellular poisoning and cell death), then the deposition of TCDD in the adipose tissue may have greater significance with respect to delayed effects on the long-term health of the exposed individual. This possibility raises a theoretical dose-response paradox which might "explain" the prevailing preponderance of symptoms in populations which may have been exposed to relatively low doses of TCDD (see Section IV D). However, persistence of TCDD in humans has not been demonstrated. Attempts to measure TCDD in human tissue are limited by technical difficulty in differentiating between the 2,3,7,8 isomer found in 2,4,5-T and the other 21 isomers from non-herbicide sources. There is also no reasonable method to determine whether tissue TCDD is from an RVN exposure, or from a more recent environmental source. D. Animal Studies A comparison of animal toxicity studies is difficult due to variations in experimental designs which include differences in (1) the species, age, and sex of animals used; (2) the level, route, and length of exposure to chemicals; (3) the purity of the chemicals used; and (4) the criteria measured and the time sequence of data collection. Animals have shown a wide range of toxic effects, but this range may serve as a guide to anticipate the potential toxic effects in humans following exposure to Herbicide Orange. A summarization of the literature is presented in Table A»l of the Appendix, "Section XV. It is apparent that the toxic effects of 2,4-D and 2,4,5-T are markedly different 1 rom the effects of TCDD. TCDD is approximately 1000 times more toxic in acute studies. In addition, the slower clearance time of TCDD may account for the significantly lower daily doses required to elicit chronic toxicity. A consistent finding in TCDD toxicity is depletion of the lymphoid tissues throughout the host. This is readily characterized by involution of the thymus in all species studied. In relation to the chronic maternal toxic dose, the embryotoxic dose is markedly lower for TCDD than for 2,4-D and 2,4,5-T. Both 2,4,5-T and 2,4-D appear to be very weak teratogens and/or carcinogens at best, but these evaluations are complicated by varying levels of contamination by various dibenzo-p-dioxins. TCDD appears to have significant teratogenic and carcinogenic potentials which appear to be species specific. The most striking observation noted in the literature is a marked variation in response among species. Examples of these variations are in the areas of acute toxicity (TCDD's LDso in guinea pigs is 1 yg/kg compared to 1000 yg/kg 10 �in dogs), excretion (2,4,5-T plasma half-life in rats in 4.7 hrs compared to 77 hrs in dogs), and oncogenicity. Even among strains of the same species (rats) variations in oncogenicity were noted following 2,4,5-T exposures. As noted earlier, this high variability between species is an important consideration in the designing1 of human studies. A second area of interest noted in the literature is a hypothetical dose-response paradox in nonhuman primates (rhesus monkey) following exposure to TCDD. Animals in a chronic exposure study fed a low level of TCDD in feed [e.g., 50-500 parts per trillion (ppt)] have shown signs of disease only after several months when total TCDD consumption was approximately 1 ug/kg body weight. Unfortunately, animals receiving comparable amounts of TCDD in a single-dose acute toxicity studies (LD50 determinations) have not been observed for the emergence of chronic effects. Therefore, it remains unclear whether the toxicity demonstrated in chronic exposure studies is dependent upon repetitive, cumulative exposure or whether similar toxicity would also be demonstrated following an equivalent single dose after a comparable observation period. Much concern has been raised over the potential of 2,4-D, 2,4,5-T or TCDD to induce genetic change in male animals which are subsequently passed on to the progeny of these exposed animals. In a recent experimental study by Lamb, Moore, and Marks, 150 male mice were exposed to various concentrations of the three chemicals in their food for eight weeks. Acute toxicity was evident with all dosages, as animals lost weight and had dose-related liver and thymus abnormalities, but these effects were reversed upon return to a normal diet. These exposures did not result in abnormalities in sperm concentration, motility or morphology. After the exposure period, the mice were mated, and no dose-related differences in mating frequency, fertility or reproducitve success were evident between the chemically exposed mice and their 50 nonexposed controls. Much of the medical literature on 2,4-D, 2,4,5-T and TCDD exposures in humans is based on individual case reports following acute exposures. Since most of the patients discussed in these reports were exposed to multiple chemical agents, it is difficult to determine which agents were responsible for specific symptoms. Nevertheless, the general areas of dermatologic and neuropsychiatric disease have been of primary interest in most investigations. Since the neuropsychiatric symptoms of herbicide exposure are numerous and largely subjective in nature, they have been extremely difficult to assess from a clinical 'tandpoint. In addition, hepatic dysfunction, and renal, gastrointestinal and cardiac disturbances have been "linked" to exposures to these chlorophenolic compounds. A multitude of symptoms have been attributed to 2,4-D and the ones reported most consistently are listed in the Appendix, Table A-2. Components of some of these selected symptoms/signs are described in Table A-3 of the Appendix. The asthenic syndrome, peripheral neuropathy, and hepatic dysfunction are 11 �of particular interest. Other symptoms of acute systemic toxicity occur, but with 2,4-D exposure has been extensively described. It has an early onset, causes prolonged disability of variable degree, and recovery has been incomplete in many cases. Electromyography in some patients has demonstrated denervation, and some studies have detected decreases in nerve conduction velocities. One autopsy study demonstrated a demyelination process within the brain of a 76year-old male who committed suicide by ingesting 2,4-D in kerosene. (2) 2.4.5-T/TCDD The human effects of 2,4,5-T are difficult to evaluate since the chemical is contaminated with TCDD in the manufacturing process. The effects of TCDD itself have been determined from studies of trichlorophenol workers, and from laboratory workers using TCDD. Symptom/sign complexes attributable to exposure to 2,4,5-T and TCDD are listed in Tables A-2 and A-3 of the Appendix. Chloracne usually begins in the zygomatic/temporal region and is often found on and behind the pinna of the ear. This is an oily acne-like skin condition characterized by comedones and inclusion cysts which may result in extensive scarring. In severe cases following heavy exposure, spread of lesions to the throat, back and inguinal areas has been noted. This skin condition is frequently preceded by erythema and blepharoconjunctivitis. Active lesions usually disappear within two years, but have been found 30 years after exposure. Porphyria cutanea tarda and hypothyroidism have also been linked to 2,4,5-T/TCDD exposure. Other symptoms such as asthenia, liver and re'nal dysfunction, neuropathy, and gastrointestinal and cardiac disturbances are probably due to mechanisms similar or identical to those of 2,4-D. With the exception of chloracne and possible disorders of porphyrin metabolism, all of these effects have been acute or subacute in nature. Numerous instances of alleged disease due to 2,4-D/2,4,5-T exposure have been the subject of heavy media attention, particularly an episode of alleged 2,4,5-T exposure in Globe, Arizona, in 1969. Despite extensive scientific review and analysis with negative findings, the Globe incident continues to be cited in news media presentations. An incident in Missouri in 1971 in which six children, two adults and numerous animals were exposed to TCDD-contaminated oil is frequently described as well. Many of the animals died and the humans developed chloracne and other ocute toxic effects; however, all humans were healthy after five years of follow-up study. A final prospective assessment of fertility, teratogenesis and carcinogenesis, in these individuals will probably be made in the future. F. Vete r a n Concerns The Veterans Administration provided the USAF with data on 46,771 patients participating in the Herbicide Registry. Numerous media presentations emphasizing both military and civilian herbicide exposures have described a remarkably wide spectrum of health effects being claimed by the veterans. Three compensation claims have been allowed for service-connected acneiform skin lesions (but not chloracne), 16 claims for other skin conditions, and an additional three claims for other diagnoses. A direct causal relationship between a disease and a specific exposure is not necessary to receive compensation. 12 �If the condition is shown to have occurred during active duty or within a reasonable time after separation, it is compensable, regardless of cause. Current Veterans Administration guidelines state that the only chronic residual of defoliant exposure has been chloracne. Table 2 summarizes the descriptive characteristics of 46,771 patients in the VA Herbicide Registry as of 31 August 1980. Table 3 summarizes symptoms from these patients by category. Table 2 SUMMARY OF DESCRIPTIVE CHARACTERISTICS OF PATIENTS IN THE VA HERBICIDE REGISTRY, AS OF 10 FEBRUARY 1981 Total Number of Registered Patients: 46,771 Branch of Service of Registered Patients: Army Marine Corps Air Force Navy Other 66.3% 18.9% 7.3% 5.9% 1.6% Table 3 VA HERBICIDE REGISTRY SYMPTOM REPORTING Number of Registered Patients: 46,771 Number of Symptomatic Patients: 34,145 (73%) Mean Number of Symptoms per Symptomatic Patient: 2.6 Symptom £ategory Number of Pati&nits Dermatologic Psychiatric/Psychological Headflche Peripheral Neuropathy Asthenia Gastrointestinal Sexual Dysfunction Other No symptoms Percent of Percent of Regi i ste_red Patjents Syjnpjtomatijc Patmits 18,675 11,745 6,021 5,729 5,637 5,454 2,105 20,702 12,626 39.9 25.1 12.9 12.3 12.0 11.7 4.5 44.3 — 13 54.7 34.4 17.6 16.8 16.5 16.0 6.2 27.0 �Study design implications that can be drawn from these data are limited because registered veterans may not be truly representative of the exposed population. The demonstrated lack of an easily identifiable symptom complex on review of the registry data clearly substantiates the need for a comprehensive evaluation of individual patients. 6. Epidemiologic Studies Epidemiologic studies of occupational groups have validated links between exposure to TCDD and the development of chloracne. Associations between TCDD and psychological abnormalities have also been suggested. A series of studies published from 1978-1980 by Hardell, Sandstrom, Axel son, and others in Sweden evaluated occupational exposure to chlorophenolic compounds in cancer patients. They found an association between cancer and exposure, but were unable to assess causality due to methodological limitations. Preliminary results of a case-control study of soft tissue sarcoma in New Zealand (Smith) did not detect any unusual clustering of occupations among the sarcoma cases. Tung (1973) reported an abnormal increase in the occurrence of primary carcinoma of the liver in Vietnam (26 cases per year during 1955-1961 versus 144 cases per year during (1962-1968). He attributed the increase to a suspected carcinogenic effect of TCDD. His published study, however, has been criticized for failure to contain sufficient data and descriptions of methodology to verify his conclusions, and the role of aflatoxin as an alternative cause of liver cancer was not addressed. His study is generally considered to be an empiric clinical observation. A study sponsored hy the EPA in 1979 in Alsea, Oregon, found a statistically significant increase in spontaneous abortion in areas where 2,4,5-T herbicide was routinely used in reforestation programs. The EPA concluded that "for all its complexity, this analysis is a correlation analysis, and correlation does not necessarily mean causation." Nevertheless, this study was used by the EPA to institute the ban on most uses of 2,4,5-T containing products. This report has been the subject of intense scientific criticism. Differences in the availability of specialty obstetrical care and in the patterns of health care delivery existed between the exposed and control areas; these differences were not taken into consideration by the researchers. Variations in the ascertainment of spontaneous abortions in each of the areas severely limited the validity of the data, and of the conclusions derived from them. A recent study conducted in Australia (1978) was unable to find an association between neural tube birth defects and the use of 2,4,5-T herbicide. A reproductive study of the wives of 370 2,4,5-T/TCDD exposed workers at the Dow Chemical Company in Midland, Michigan was recently completed (Cook and Bodner). No differences in fertility patterns, fetal wastage, or birth defects were detected. Epidemiologic studies are continuing in Seveso, Italy, where a population of 220,000 was potentially exposed to TCDD following an industrial accident in July 1976. These studies have involved investigations of more than 30,000 children, and detailed clinical examinations of 1,024 persons, including the most severely exposed children and adults. Recent data (Homberger, et al., 14 �1979) indicated that most cases of chloracne from this incident cleared rapidly. To date, the growth and development of newborn infants and children, iinmunological response, chromosomal aberrations, the response to the challenges of infectious diseases, and the morbidity and mortality patterns of the study population have not been significantly altered by TCDD exposure. Thirty-eight cases of birth defects were reported in early 1977, approximately 6-8 months after the industrial accident. However, the authors ascribe this increase to an artifact of surveillance. Analysis of surveillance data on the occurrence of spontaneous abortions after July 1976 is compromised by the lack of valid baselines for the pre-accident period. The social pressures operating in the Seveso population prior to the accident fostered underreporting of birth defects, while the atmosphere after the accident made the occurrence of a birth defect more socially acceptable. The post accident congenital malformation rate is not significantly different than the rate in similar areas of Central Europe. Another progress report on the aftermath of the Seveso accident (Pocchiari, et al. 1979) has revealed: (1) a decrease in the prevalence and severity of chloracne in the exposed population; (2) an increase in clinical and subclinical neurologic disease as demonstrated by delayed peripheral nerve conduction velocities; and (3) increases in the prevalence of hepatomegaly (8%) and alterations in liver function tests, which returned to normal over an 18 month period of follow-up. Thus far, immunologic, cytogenetic, and embryomorphologic analyses have been unable to detect significant differences between exposed and non-exposed individuals. A 2,4,5-T Dispute Resolution Conference was held in Arlington, Virginia, from 3 to 7 June 1979. Fifty-six recognized experts from the United States and seven foreign nations were actively involved in the deliberations of the conference. Human Exposure, Carcinogenicity/Mutagenicity, and Teratogenicity Working Groups independently reached the conclusions that there was no valid scientific evidence linking fetotoxicity, teratogenicity or carcinogenicity in humans in a cause and effect relationship to 2,4,5-T/TCDD exposures. The Human Exposure Working Group also concluded that there were no epidemiologic data associating TCDD with any long-term health effects in humans other than persistent chloracne. While they did not find evidence of serious long-term health effects, neither could they find strong evidence for "lack of effect. Most previous epidemiologic studies have not had sufficient statistical power to detect increased risks of low incidence/prevalence conditions in the observed populations, and the period of observation in many prospective studies has been less than ideal. Several potentially valuable epidemiologic studies are currently in progress. Two independent and comprehensive studies of workers exposed to TCDD at a Monsanto manufacturing plant in Nitro, West Virginia, are currently being conducted (Mt. Sinai Medical Center, New York, and the Kettering Laboratory, University of Cincinnati, Ohio). These chemical industry workers were exposed over long periods of time, and were previously evaluated in 1953 and 1956, following an industrial accident which occurred in 1949. Zack and Suskind of the Kettering Laboratory have reported a follow-up study of 122 workers, 28 years 15 �after heavy exposures to TCDD. There were 32 deaths in the group, and the relative risks of death were 0.69 for all causes, and 1.0 for malignancy; however, no firm conclusions can be drawn due to the small numbers involved. A Czechoslovakian study involving a 10 year followup of TCDD exposed workers, and a US National Cancer Institute (NCI) mortality study of 4,400 structural pest control workers are also underway. Preliminary results of a larger study of long-term morbidity by Suskind at the Nitro site have failed to reveal significant abnormalities other than persistent mild chloracne and decreased nerve conduction velocities, possibly associated with alcohol intake. These new studies, and the continuing evaluations of the Seveso, Italy, population, should continue to provide valuable data. The large study groups involved in the Seveso and NCI studies should provide good statistical power, and the Nitro, West Virginia, and Czechoslovakian efforts will evaluate the effects of exposure after prolonged periods of time (10-30 years). The results of these studies should fill major gaps in the knowledge of 2,4,5-T/TCDD epidemiology, and should prove to be useful in evaluating the long-term effects of these compounds on health and reproductive outcomes. 16 �V. Epidemiologic Study Desi gn: Matched Cohort A. DesjjnConjsicleratjions The goal of this study clearly mandates a comprehensive epidemiologic approach, incorporating mortality, and historical, current, and followup morbidity studies. Exposure to herbicides during the 1962-1971 time period may have initiated long-term health effects that may or may not be progressive. If such effects are detectable by a review of the subject's past medical history, and can be verified, direct links to compensation issues can be made. Current health status, as mirrored by a large number of recent VA claims and inquiries, is of major interest, because such claims and inquiries may indicate medical conditions that might be confirmed by a comprehensive physical examination. If analyses of both mortality and morbidity data yield only indeterminant or weakly suggestive findings, it may be that sufficient time has not yet passed for substantial emergence of longterm health effects. This dictates a requirement for a follow-up element to the study. Methodological shortcomings are inherent in each element of this comprehensive study. To some extent, the classical deficiencies of each particular epidemiologic approach are compensated by the concurrent use of the other elements. For example, the low chance of identifying a relatively uncommon disease solely by the use of a mortality study is offset by the inclusion of a current morbidity study. The relatively quick feedback that can be attained from current morbidity and mortality studies will serve to better define the follow-up study, and will help to alleviate problems that arise as a result of changes in diagnostic criteria and methods over time. Nevertheless, problems that can affect ascertainment of disease in all phases of the study will remain. Inaccurate patient recall of antecedent events, the distortion of information by knowledge of anticipated symptomatology, and participant or observer knowledge of their exposure status can only be corrected to a limited extent by review of records for symptom validation and "blind" assessment protocols. In addition, fundamental problems dealing with adequate selection of a control group and limiting loss to study can influence any comprehensive epidemiologic investigation. These and other pitfalls in study design will be discussed in more detail in Section VIII. The management of this project will be conducted through standard Air Force Research and Development procedures, including program monitors at Air Force Headquarters and Air Force Systems Command, and a Program Management office at Brooks AFB, Texas, Contract monitors will insure that all contractual efforts are conducted according to strict quality assurance procedures, and an on-site monitor will insure that the physical examinations are conducted in strict accordance with the study protocol. Since the study has three elements and confronts a health issue with incompletely specified or uncertain endpoints, strong potential bias, and severe time contraints, the following design represents the best overall framework for achieving validity. The design process is complex and in itself time dependent. 17 �B. Selection and Ascertainment of the Populations for Study (1) The Exposed Military Groups (a) Operation RANCH HAND Personnel Operation RANCH HAND personnel flew C-123 aircraft in RVN during 1962-1971. Data from hand-compiled lists obtained through the RANCH HAND Association (a reunion organization), Air Force personnel records, unit historical records, and actual C-123 flight orders, place the herbicide exposed population at approximately 1264 individuals. Of those personnel confirmed as RANCH HAND participants, 25% are still on active or reserve duty, with the remainder being composed of retired, separated, or deceased persons. To identify all RANCH HAND participants, an indepth search was conducted of all organizational records stored at the Military Records Division, National Personnel Records Center (NPRC), St. Louis, Missouri. Introductory letters will be sent to the last known address of all identified persons, and nonresponse will be pursued by cross-locator systems available within the government (e.g., Social Security Administration, VA, Internal Revenue Service). Significant efforts will be made to account for at least 99% of the total population (see Figure A-2, Section XV). Because of the limited number of RANCH HAND personnel, no subsampling of the exposed group is planned in any phase of the study. All members will be strongly encouraged to participate in all phases of the investigation. All RANCH HAND personnel are males currently ranging in age from 30-69 years (mean = 42.4 years). The normal C-123 crew composition was one pilot, one copilot/navigator (both officers), and one spray equipment console operator (enlisted) in the rear of the aircraft. The aircrew officerenlisted ratio is 2.2:1; however, the inclusion of RANCH HAND support personnel (predominantly enlisted) in the st-jdy will make the overall officerenlisted ratio 1:1.7. Approximately 98% of the officers and 92% of the enlisted men were Caucasian. Attempts have been made to identify all maintenance personnel assigned to the RANCH HAND units. Maintenance of the RANCH HAND aircraft was performed within a step-wise organizational structure. Routine daily maintenance (primary) was conducted by flight line support personnel who were often dedicated exclusively to RANCH HAND operations. More extensive maintenance (secondary) was carried out by consolidated support units at the base level, which were also responsible for non-RANCH HAND C-123s as well. Major aircraft overhauls and modification were conducted by maintenance units at Clark Air Base, Philippines. The maintenance personnel in these centralized units were not directly assigned to RANCH HAND, and their exposures to RANCH HAND C-123 aircraft and herbicide cannot be validated. From 1962 through 1964, the primary flight line maintenance teams were dedicated to RANCH HAND aircraft, and these individuals have been identified by the mechanisms described above. In 1965, flight line maintenance was performed by personnel of the centralized maintenance organization (secondary), and it is not feasible to adequately identify all of these individuals from available records. After 1966, the RANCH HAND organization transferred their 18 �base of operations to a new location, and primary maintenance was once again performed by personnel assigned specifically to RANCH HAND. These individuals have been readily identified. Thus, maintenance personnel directly assigned to RANCH HAND will be included in the study. These complexities are summarized in Table 4. Table 4 FEASIBILITY OF IDENTIFYING AIRCRAFT MAINTENANCE PERSONNEL (TOTAL POPULATION) EXPOSED TO HERBICIDE ORANGE Primary Maint Pers^ojinel Secondary Maint_ Personnel2 Jan 1962-Jul 1964 Yes No Aug 1964-Dec 1966 Yes/No3 No Jan 1967-Oct 1971 Yes No Time 1 individual assigned to RH; total number (denominator) known individual not assigned specifically to RH, although may have serviced the aircraft; denominator not ascertainable 3 other documents permit ascertainment of a portion of this group Because of the significant combat hazard associated with low, slow flying missions, some early RANCH HAND crewmembers were elite volunteers (see RiskTaking Bias, Section VIII, C). In fact, RANCH HAND crewmembers comprised one of the most highly decorated units during the RVN Conflict. Anecdotal stories reveal that most crew members were, on occasion, heavily exposed to Herbicide Orange due to normal or combat induced equipment malfunctions within the aircraft. Many former RANCH HAND personnel are expected to be currently employed in the aerospace industry as commercial airline pilots, airline managers, and flight mechanics. RANCH HAND personnel still on active duty are expected to be found in senior management positions. (^) Alternat^ Exjjosec^ Populations (I) Introduction The principal investigators, members of all of the peer review committees, and independent consultants have clearly recognized that the statistical power of this RANCH HAND study is suboptimal for the detection of specific uncommon conditions or diseases. This limitation is inherent because the size of the RANCH HAND population is fixed at approximately 1200 individuals, and it cannot be increased. 19 �A brief review of alternate military populations is in order to highlight the significant advantages of the RANCH HAND population. The desire to achieve more optimal statistical power by merely increasing the size of the population under study must be balanced with a careful analytic process which assesses the exposure level of alternate populations, and categorizes them as either additive or nonadditive to the RANCH HAND study population. (2) U.S. Army Ground Personnel Some U.S. Arrrjy personnel were undoubtedly exposed to herbicides during their duty in Vietnam; however, the objective ascertainment of exposed individuals is not possible. Any attempts to identify individuals assigned to combat units which may have been exposed would result in an unacceptable degree of misclassification since U.S. Army personnel records do not exist which would allow the accurate identification of soldiers below the battalion level. This lack of demoninator data, and the high degree of misclassification in determining the exposure status of Army troops makes this population unsuitable for inclusion in the framework of the RANCH HAND Study. (!) Ancillary Air Force Groups (Non-RANCH HAND Personnel) Air Force handlers of herbicide drums in RVN were exposed to herbicides because of drum leakage. As the drum handlers were ad lib participants, no personnel designator was assigned to these individuals, thus prohibiting computer tracking and identification. The size of this population is unknown, but it is expected to be small (less than 200), as the majority of drum handlers are known to have been Vietnamese. Additional groups such as U.S. Army helicopter crews, casual observers (both Army and Air Force), and experimental fighter-bomber personnel who may have occasionally conducted spray operations were also potentially exposed. However, population-at-risk determinations for all of these groups cannot be made, and any identification of individuals exposed in these situations must rely on self-selection or incomplete ascertainment. Also, the selection of suitable control groups for a study of these individuals is difficult if not impossible. ( ) U'S» Manne Corps Troops i On 16 November 1979, the GAO released a report which suggested that a herbicide-exposed population of nearly 22,000 U.S. Marine Corps troops could be identified, and that this identified group would be appropriate to study. Records exist which locate Marine Corps battalion headquarters near the C-123 spray paths. The GAO made several improper assumptions to conclude that all of the identified marines were in fact exposed. Specifically, all battalion troops were assumed to be located at the battalion headquarters. Further, the effect of prevailing winds on the direction of spray drift, and the photodegradation of the chemicals were not considered by the GAO. The National Research Council panel considered the GAO analysis, and proposed a study of 5900 marines who were "near" spray paths on the same day 20 �as the spraying. The "exposed" group was to be contrasted with the mortality experience of 212,000 presumably unexposed controls (also marines). The RANCH HAND study described in this protocol consists of approximately 1200 exposed individuals and 6000 controls for the mortality study phase. Despite the fact that the RANCH HAND Study involves a smaller sample size than the proposed Marine effort, the RANCH HAND Study is more powerful statistically. Specifically, lower exposure to herbicide by a conservative factor of from 1/10 to 1/1000 and misclassification in Marine exposure groups renders the Marine Study far less powerful than the RANCH HAND effort. As described in Section VI, misclassification and: decreased exposure are seen to be independent factors additively decrementing Marine Study statistical power. Even when all 21,900 marines within the herbicide spray paths up to 28 days following the spray operations are considered exposed, the RANCH HAND Study is noted to be significantly superior. (5.) Conclusions The Operation RANCH HAND participants are the most suitable of the military populations to study in evaluating the longterm effects of herbicide/dioxin exposures. The RANCH HAND group had a much higher level of exposure which was sustained over a prolonged period of time. This increased level of exposure implies that RANCH HAND personnel would be more likely to develop more acute and chronic symptoms from the exposure, and would manifest them sooner than the other exposed military personnel. The addition of significantly less exposed and/or misclassified groups to the RANCH HAND population for the attractive purpose of increasing statistical power would constitute an egregious dilutional error, (2) Control^ Group (Not exposed to Herbicide Orange) A review of all specialized flight units present in Southeast Asia during the RVN conflict, reveals clearly that there is no absolutely ideal control group for the RANCH HAND population, C-130 aircrew members and support personnel were selected because of sufficient population size, similar training profiles, and psychologic similarities to the RANCH HAND group. Total ascertainment of the C-130 population is being conducted by computer and hand selectipn for specific military flying organizations, and foreign country service> during the interval from 1962 thru 1970. Over 2.3 million personnel records have been reviewed, and the approximate C-130 population size is 23,978 individuals. Aircrew members who flew C-130 aircraft in Southeast Asia during 1962-1970 were selected as controls for the RANCH HAND aircrew population. The C-130 flight line maintenance population were ascertained from personnel records by similar mechanisms, and served as the specific control population for the RANCH HAND support personnel. The proportions on active duty, and non-active duty status are expected to parallel the patterns in the exposed group. Another possible control group» the non-RANCH HAND C-123 population, is known to be too small (approximately 3000) to provide adequate sampl i n g flexibility and replacement under the proposed matched variable concept 21 �(see below and Section VI). Also, many of the RANCH HAND aircraft were reconfigured for transport and insecticide missions and thus, the non-RANCH HAND C-123 crews responsible for these other missions may have been exposed to significant Herbicide Orange residue in these aircraft. Therefore, this group may not have been truly unexposed to herbicides, and was discarded as an appropriate control population. Crewmembers of C-7 transport aircraft were also considered as a potential control group; however, because of small sample size (1000-2000) and the fact that they served in RVN only during the post 1967 era, they were also dropped from consideration. The normal crew composition of a C-130 is three officers and two enlisted personnel. The control group is considered to be "pure" from the standpoint of lack of occupational exposure to herbicide. The entire control group will be considered "nonvolunteer" with respect to abnormally high combat risk. While in general they will possess lifestyle characteristics and socioeconomic backgrounds similar to the exposed group, their overall combat morbidity/mortality and the resultant stress influences upon general health may be slightly less than in the exposed group. For those separated and retired C-130 controls, similar proportions to the exposed group are expected to be employed in the aerospace industry. Known and estimated factors of the control and exposed populations are summarized in Table 5. (3) Matching Procedures and Rationale Each member of the exposed group has been computer matched to a set of C-130 controls comprised of approximately 10 individuals using three variables. Since the u;o groups are highly selected and inherently similar with respect to many variables, very close matches are feasible. This epidemiologic design incorporates a matched concept because: (1) a matched cohort design will provide maximum test power throughout the entire study, and (2) statistical intergroup comparisons may be made without normalization by three key variables known to effect symptom frequencies of interest, thus providing greater power for complex statistical testing. It is apparent that following the match, both exposed and control populations will be very nearly identical with respect to the three influencing variables so that a replacement concept is feasible (see F below). In the event that frequent match breaks occur, stratification techniques can be used. The selection of the control group produces an inherent match for equivalent SEA experience, and additional matching has been conducted for (1) age, by year of birth and closest month possible, (2) Air Force Speciality Code (AFSC) as an absolute match, and (3) race (Caucasian versus nonCaucasian) as an absolute match. Specific rationale for these variables is as follows: (1) the age match controls for the many clinical symptoms and signs associated with advancing age, (2) AFSC controls for officer-enlisted status (as well as crewmember-noncrewmember status), a variable strongly linked to educational background, current socio-economic status, and moderately linked to age (5 year median difference) and socio-economic background, and (3) race controls for differences in chronic disease development, socio-economic background, etc. 22 �Table 5 COMPARISON OF THE STUDY GROUP TO POSSIBLE CONTROL GROUPS BY KNOWN AND ESTIMATED FACTORS KNOWN FACTORS STUDY GROUP RANCH HAND C-123 POSSIBLE CONTROL GROUPS Non-RANCH HAND C-123 C-7 C-13( POPULATION SIZE 1264 3000 1200 23,97* OFFICER/ENLISTED RATIO 1:1.7 1:2 1:2 1:2 AIRCRAFT FUEL (AV-GAS) tvs CO YES (+JP-4)* YES (+JP-4)* YES NO (JP-4 OCCUPATIONAL HERBICIDE EXPOSURE YES YES/NO** NO NO OCCUPATIONAL INSECTICIDE EXPOSURE 2+ 0 0 0 COMBAT HAZARD 4+ 3+ 3+ 2+ RVN-IN COUNTRY ASSIGNMENT 4+ 4+ 4+ 2+ ESTIMATED FACTORS *In 1968, aircraft were modified with a JP-4 booster. **Contaminated aircraft reconfigured for transport may have resulted in exposure to non-RANCH HAND personnel �The inherent match for SEA experience controls for combat-induced physiologic, psychophysiologic, and other related morbidity and mortality disorders. Additionally, this inherent match may reflect the effects of alcohol consumption, the use of chemoprophylactic and/or illicit drugs, and the acquisition of tropical diseases associated with life in SEA. The comparisons of the exposed (RANCH HAND) subjects and their selected sets of controls are detailed in Appendix Table A-4. Only 4 of the ten categorical AFSC/case strata had less than ten controls for each exposed subject. The group of Caucasian pilots had a mean of only 9.5 controls per exposed subject, due to the extreme ages of several individuals, and the strata of Black pilots and other Black officers had means of 2.7 and 5.0 controls respectively. However, since there were only seven black officers in the exposed group and only thirty controls, high numbers of tight rnatchos could not be achieved. Black enlisted aircrewmembers had a mean of 9.8 controls each. (4) Computer Science and Statistical Petal 1s of the Matching Process As described above, the matching for this project has been performed using three variables: occupational category, race and age. Five occupational categories (officer/pilot, officer/navigator, officer/other, enlisted/flight engineer, and enlisted/other) have been used to reflect socioeconomic status and aeronautical rating. The variable of race has been dichotomized into black and non-black. Ten matched controls have been selected for each exposed subject, regardless of current vital status. The computer method applied to select the control subjects is an adaptation of a procedure studied by Raynor and Kupper (Nearest Neighbor Matching on a Continuous Variable, Technical Report, Department of Biostatistics, University of North Carolina, 1979). As the first step, the RANu! HAND and control groups were partitioned into ten strata using the categorical occupational and race variables. The Raynor and Kupper matching procedure was then applied iteratively within each of the strata to match for the continuous variable of age, given in months. The Raynor-Kupper procedure involves the following steps: STEP #1. The RANCH HAND cohort in a given strata is randomly permuted. STEP #2. The first RANCH HAND subject in the permuted set is selected for matching. STEP #3. The closest available control is assigned to the selected RANCH HAND subject using the absolute value of the difference between the months of birth of the RANCH HAND and the control subjects. If the closest available control is further than 60 months from the selected RANCH HAND subject, a blank is assigned. Tied assignments are broken randomly. STEP #4. Step #3 is repeated for all RANCH HAND subjects in the strata proceeding down through the permuted set, until the entire RANCH HAND cohort is exhausted. STEP #5. Steps #1 through #4 are repeated ten times for each RANCH HAND subject to construct a 1:10 study set. At the completion of the matching activity, the RANCH HAND - Control study matrices for each of the ten occupation-race strata can be diagrammatically represented as in Figure #1. 24 �Figure 1. MORTALITY ANALYSIS COHORTS 10 1 19 1 R "2 r2j 1 R 3 p U R r 4 3,l * ^Hjl I 5 t. 1j O • •*• 9 *•* v i ,io U r2 j 2 r ^2,3 r ^2,4'*** • P f ^3,2 P 3j3 P "3^4**** • p ^3 , 10 pt J /"* HjO ^i^^q.**** fj/ /* • ^2,10 "4 ,10 1200,3 1200,l 12 QO,10 Figure 2. MORTALITY MATRIX RANCH HAND COHORT Rl _,, , CONTROL COHORTS cr c i,i,m' c i,2,m' *** '10 15m 16 ••• ^ , 10 j,5,m R 1200 ,5,m' Ci2oo,i,m' C12oo,2,m' C i200,e' C 1200,10' In each row of this matrix the controls are ordered from nearest to farthest in terms of age of the matched RANCH HAND person. The next operation defining the control group involved randomization of all of the controls in each row of each stratum matrix to negate the ordering by age. Then, the first five members of each control set for each RANCH HAND person are identified as being subjects in the mortality portion of the study. The resulting occupation-race strata matrices now have the form shown in Figure 2. In Figure 2, Cj^' or Cj^^1 may be equivalent to any Cj^ of Figure 1 due to the randomization process. Table 6 summarizes the results of the matching process, and Appendix Table A-5 provides a more complete statistical description of the process. In these tables, the age difference between the month of birth of the control and the month of birth of the RANCH HAND person, (counting months from 1900) and the cumulative number of controls and the cumulative percentage with this difference are shown. 25 �Table 6. RESULTS OF THE MATCHING PROCESS (1:10) Age Difference (in Months) Cumulative Number of Controls Cumu1 ative Percent 0 8612 70.6 1 10287 84.3 2 10749 88.1 3 10984 90.1 4 11167 91.6 5 11322 92.8 6 11410 93.5 12 11688 95.8 24 11921 97.7 36 12028 98.6 48 12129 99.4 60 12197 100.0 (5) Study Group Selection Procedures (a) Mo rta1i ty An a 1/s i s A 50% random sample of each control set will be drawn and used to comprise a 1:5 mortality analysis, as described in section (4). The vital status of each subject in this sample and of all exposed subjects will be ascertained at a minimum frequency of every five years for the 20 year duration of the study. Those individuals dying of combat causes will be excluded from the mortality analysis as it is assumed that combat death is independent of herbicide effect. Further, the known differential combat death rate between the RANCH HAND and control groups can be attributed to the hazardous and unique nature of the RANCH HAND mission. Twenty-two RANCH HAND personnel (15 officers and 7 enlisted) died in combat. Medical record reviews will be accomplished to assess the illness experience of these individuals prior to combat mortality. (b) Historical Morbidity Study Retrospective or historical health data will be gathered on each exposed subject and from the first randomly selected mortality control from his set by questionnaire techniques. Living but noncompliant controls in 26 �the historical morbidity study will be replaced by a compliant control selected from the control set. In order to avoid an information gap for data on deceased individuals, surrogate interviews will be obtained from the first order next-of-kin of exposed and control subjects dying of noncombat related causes between the date of their assignment to Southeast Asia and the initiation of this study. Since the validity and accuracy of surrogate derived data may not be equivalent to data obtained directly from living study subjects and their spouses, these data will be subsetted for analysis. All available medical records, (military, VA, and civilian) will be reviewed for all subjects selected for this morbidity analysis. (c) Prospect 1 ve Morbidity Study A baseline physical examination and review of systems will be conducted, and a prospective or followup approach will be used to assess the current state of health of study subjects using a series of questionnaires and physical examinations over the next 20 years. Each living exposed subject and the randomly selected primary control will be incTuefe'd "in the questionnaire and physical examination phases. In this prospective study of morbidity, primary controls who are deceased, unaccountable or unwilling to participate in the followup studies, will be replaced by a willing subject from the remainder of the control set (Figure 3). The selected control for a RANCH HANDER dying of a noncombat cause will be retained throughout the questionnaire, physical examination, and followup phases of the study. Since the control's vital status and volunteerism should be independent of the matching sequence, many primary controls should enter the study. The remaining members of the control set will be used as replacement candidates for possible use later in the study (see section F below). All replacement controls will be clearly identified for the purposes of subset analysis so that population differences, if any, between the first randomly assigned selectees (noncompliant) and the replacements (compliant) can be assessed. Specific rules and procedures for study entry are found in Table A-6 and Figure A-3 of the Appendix. (d ) Thne Jnterreljitedness of the £pjn_pa_r1_soni Groups It should be clear from the foregoing discussion that the study populations of the mortality, historical morbidity, and prospective followup phases are highly related but different. Once selected, the mortality control cohorts will remain unchanged throughout the 20 years of observation. The population under study in the historical morbidity phase will initially be a randomly selected subset of the mortality comparison group; however, some of these primary controls may be decreased or noncompliant for the voluntary aspects of this phase of the study. In this phase, noncompliant controls will be replaced, but deceased controls will not, as surrogate interviews with the next-of-kin will be used to reconstruct morbidity data. The subsetting arid replacement procedures create the difference between the mortality and historical morbidity comparison groups. The population in the prospective morbidity phase is the comparison group from the retrospective phase plus additional replacements for the deceased controls. Thus, it is clear that the comparison groups are slightly different, but they would be identical if no deaths recurred since 1962 and all primary controls were compliant. 27 �Figure 3. SELECTION PROCEDURE FOR THE QUESTIONNAIRE, PHYSICAL EXAMINATION, AND FOLLOW-UP STUDY LIVING RANCH HAND INDIVIDUAL CONTROL INDIVIDUALS Randomly Selected Mortality Controls Ji ^ ** t t * ** - t Deceased Unwilling Volunteer Replacement Candidates C. 0verview of Stat i st i ca1 Methodology The design of the study is presented in schematic form in Figure 4. R' refers to RANCH HAND personnel and C" refers to the collection of all possible control individuals. As defined, P' and C" will contain individuals who are deceased of noncombat causes. Combat deaths 1are excluded from R 1 and C". Since C" is approximately 20 times larger than R , a randomized subsample C' and C" will be obtained. C 1 will be constructed from C" by computer selection of the ten matched controls for each exposed study subject. As previously noted, close matches will be nude for the variables of age, AFSC, and race. The matched controls will form tan cohorts, Ci through C 10 , as shown in Figure 1. A 50% random sample from each of the matched control sets of 10 will be selected for inclusion in the mortality assessment so that a group, C' is obtained that consists of 5 matched controls for each exposed subject. These controls will be designated as initial replacement candidates for the morbidity and follow-up studies. The remaining individuals in the control set will be additional replacement candidates in the event that replacement must occur beyond the members of the mortality set (see Figure 3). C' will be constructed without regard to whether the individual is currently living or dead so that an assessment of noncombat mortality can be accomplished. 28 �FIGURE 4 DESIGN SCHEMATIC EXPOSED R 1:5 ro MORTALITY ANALYSIS UNUSED PHYSICAL EXAMINATION �Referring again to Figure 4, R and C indicate living RANCH HAND members and primary matched controls. If ITIR' is the proportion of R' found to be deceased, then R » (1 - mR')R' The questionnaire will provide data concerning specific symptoms and other findings in the R and C groups. Thus, various questionnaire finding rates in R, SR, will be calculated and compared with the corresponding rates in C, so The questionnaire will allow allocation of RANCH HAND personnel into those with symptoms on questionnaire, indicated by RS, and those without, R?. Similarly, the control Jndividuals will be placed into symptomatic, indicated CS, and asymptomatic, C"$" groups. The physical examination performed on individuals from R and C will allow estimation and comparison of rates of physical findings in these groups. Rates of abnormal physical findings can be symbolically indicated as fR and fc for RANCH HAND and control groups respectively. Comparison of these rates is very important and details will be discussed below. Let fRs be the rate of physical findings among RANCH HAND personnel with findings by questionnaire and let fR$ be the rate of physical findings among RANCH HAND people with no findings on their questionnaire. For most disease processes it would be expected that fR$ should be a larger rate than fRS*. If fRS is observe1! to be equal to or less than fR$", an interpretation of over-reporting may be warranted, although the possibility of subclinical disease is recognized. Rates fcs and ^CS" w^l a^so be estimated, and comparisons between fR$, fc$> fR$ and fc$"will be accomplished. The eight rates IDR', me', SR, SQ, fRs» fRs", fcs, fCS; and their refinements fully characterize this study. As depicted in Figure 4, "vertical comparisons of these rates provide relative risks mR'/mc'' S S R/ C» fR/fc» fRS/fCS and fRS/fcs" which are of central importance in defining herbicide effects. "Horizontal comparisons" relate fR to SR, fRS to fRs, fc to sj and fcs to fcs**. Specifically, the ratio fR/SR is the ratio of physical findings to reported symptoms in the RANCH HAND population. This ratio may be contrasted with the ratio fg/sr, and if fR/SR is less than fc/sc over-reporting is suggested. Likewise, if fRs is less to than fR^", over-reporting is further suggested. A comparison of fRS/fRS fCS/^CS contrasts the odds of findings given symptoms in the RANCH HAND population with the odds of findings given symptoms in the control group. If these odds are lower in the RANCH HAND group, overreporting is again implied. Further discussion of these rates is presented in Section V.G. During the questionnaire and physical examination phases of this study, only one of the five randomly selected mortality study controls will be used for each RANCH HAND individual. If this control is unwilling to participate, another mortality study control will be used as indicated in Figures 3 30 �and 7. These replacements will be carefully labelled for purposes of statistical analysis. A detailed discussion of this replacement concept is found in Section VI. D. Mortality Study (1) Introduction The mortality s retrospective morbidity, and follow-up studies are components of a "non-concurrent" prospective study used in the observation of a specially exposed group or industrial population starting from some date in the past. The initial exposures occurred 11-19, years ago and varied in intensity and duration from one RANCH HAND member to another. Access to employment, medical, or other types of records is an obvious requisite for such a study. The classical "case-control", retrospective study is not operative in this protocol due to the lack of defined clinical endpoints. The mortality study will be conducted in two phases; a review of past mortality, and a continuing assessment of the death rate in the exposed and control cohorts over the twenty year duration of the RANCH HAND II project. Based upon USA vital statistics, 8.6% of the study subjects are expected to have died between completion of their Vietnam tour and initiation of this study. Of these deaths in the control group, approximately 30% should have been due to cardiac causes, 24% to neoplasia, 13% to accidents, 5% to cirrhosis, and 0.1% to leukemia. (2) Data Collection Methods The mortality status of the exposed cohort and the randomly selected controls will be ascertained using multiple techniques including: payments of Veterans Administration Death Benefits, Social Security Administration Records, Air Force Accounting and Finance Center wage and retirement payments, and interviews with subjects or their families. Death certificates, autopsy reports, and medical records will be obtained for each deceased subject. The International Classification of Disease, Ninth Revision, 1978, will be used for coding. At the time of the first followup examination, all participants will be asked to allow an autopsy to be performed at government expense at the time of their death, and have the tissues forwarded to the Armed Forces Institute of Pathology, and the results sent to USAFSAM. (3 ) Analysis of^ Mortality Data Considering the basic groups R' and C 1 in Figure 4, individuals will be classified into three categories: alive, dead, or unaccounted. If a large number of individuals in each group are unaccounted for, the study can obviously be severely biased. Thus, significant effort will be expended to reduce the unaccounted category as much as possible. At most, 1 to 3 percent of both groups can be allowed to remain unaccounted, with a 1% rate 31 �being preferred. If for example, the mortality rate in C 1 is 0.10, then an unaccountability rate of 0,01 could alter the mortality rate by as much as 10%. Whatever the unaccountability rates, the pattern of unaccountability must also be compared between groups R 1 and C'. For example, the possibility of age differences must be examined, particularly if the unaccountability rates are high. The following paragraphs discuss the analysis of mortality under the assumption that low unaccountability rates have rendered the mortality analysis meaningful. Multiple mortality assessments will be accomplished during the course of this study, one at the beginning of 1 the study, using available mortality data on the basic mortality cohorts in C and R 1 (5:1 ratio), and others using mortality data on R 1 and all controls used in the study (both C 1 and replacements) as controls accumulate prospectively. The procedures described here will be used in all of these assessments. Henceforth, within the protocol, the term "mortality data" does not distinguish between that data collected initially and that data collected in the future. The mortality data will be analyzed using several different approaches. Crude age-specific death rates will first be calculated and tabulated. Age will be d-jvided into k strata, and person-years will be observed for each strata as will be the number of deaths in each strata. In this manner a tabular display will be developed as shown in Table 7. Table 7 STRATIFIED FORMAT OF AGE-SPECIFIC DEATH RATES Ranch Hand Person Years Deaths 1 Pl2 Person Years m 12 PIS Death Rate PH 2 Controls m Deaths P2i P21 22 m22 r m P !3 "Ilk Death Rate P23 23 22 •"ik Since the death rates TIJ and r2j are Poisson variables, they can be contrasted directly. If the relationship of rjj to r2j is found to be consistent between age strata (within statistical variability), a summary mortality index may be calculated. One summary index that will be calculated is the Standardized Mortality Ratio (SMR) which is (Armitage, 1971): 32 �SMR * M x 100 k I m i M » "Classical" standardized mortality ratios using national mortality data as the reference will not be calculated for RANCH HAND II due to the effects of the healthy worker phenomenon. The term I mn is the total number of deaths observed in the RANCH HAND group while I Py r2j is the number of deaths that would be expected were the age-specific RANCK HAND death rates the same as the age-specific control death rates. Thus the concern is for an SMR greater than 100%. If a crude death rate for controls, dc, is calculated as k I then the standardized crude rate for the RANCH HAND group d^^ is d RH = An approximate statistical test would regard d^ as a Poisson random variable with mean dr,. An alternative approach to the provision of a proportionate mortality ratio is that of Breslow and Day (1975). In this treatment, a multiplicative model is employed, for example: Mjk = 9 ifj*k where \\jk is bhe mortality rate, 0-j is the contribution due to population differences (RANCH HAND versus Control), <)>j is the contribution due to age 33 �group, and ^ is the contribution due to length of time in RVN, etc. statistical approach here is via maximum likelihood. The Logistic models (Walker and Duncan, 1967) have been extensively studied at USAFSAM for application in cardiovascular disease analysis. These models, in the herbicide context would have the form P = [1 + exp(a + 3iA + 3eT + 3sR + Bi»E + 05AE + ...J]- . where P A T R E = = = = = probability of death age in years length of time in RVN indicator variable for race exposure variable and where a\t 3i, i=l,2,... are coefficients to be estimated from the data. Testing for a group difference can be accomplished by estimating 3^ and the interaction coefficients such as 35. If all interaction coefficients involving the exposure variable E are zero and E is treated as a 0/1 variable, Cox (1958a, 1958b) has shown that thd most powerful test for non-zero 34, in the setting of matched pairs, is McNemar's test. This latter test makes full use of the paired design of the study. For McNemar's test, the data are cast into a 2 x 2 table as shown in Table 8. In this table, "a" is the number of pairs in which both members have died, "b" is the number of pairs in which only the RANCH HAND person has died, etc. Using McNemar's test, the test statistic 2 |b-c| 2 x =J L b +c is calculated and referred to the chi-square distribution with one degree of freedom. Cox (1966) and Meittinen (1969) provided extensions of McNemar's test for R controls per exposed (R-to-1 matching). Of course the above analyses will be accomplished considering all deaths, and deaths by specific cause. As previously discussed, RANCH HAND personnel may be characterized as risk takers. This risk taking behavior may be associated with increased mortality from a variety of causes. On the other hand, herbicide exposure has caused neuropathy in the RANCH HAND personnel, one could anticipate that this disability would increase the probability of accidental death. Therefore, 34 �Table 8 FORMAT OF McNEMAR'S TEST CONTROLS RANCH HAND PERSONNEL DEAD ALIVE TOTAL Dead a b a+b Alive c d c+d Total a+c b+d n accidental death rates among RANCH HAND participants will be corrected for risk taking. This can be accomplished by including assessment of risk taking behavior in the questionnaire, indepth interview, and psychological evaluation. Both control and RANCH HAND mortality could be corrected using these measures, with the resultant rates being less biased and, therefore, a better indicator of exposed versus control effect. (b) Mortality analysis without cov.a Mates. The first step in the statistical analysis of survival data is descriptive, i.e., the construction of summary measures which provide a basis for comparing different exposure groups without any allowance for the effects of possibly confounding variables (e.g., age) except perhaps for some limited stratification. Since one must expect many "losses to follow-up", only methods which take full cognizance of this complication will be considered. It should be pointed out that all the methods described below assume independence between censoring (e.g., loss to follow-up) and death or morbid event, although some techniques permit different patterns of censoring in different exposure groups. The life table method can be adapted to obtain a step-function approximation to survival distributions in the presence of censoring (Chiang, 1968, Gross and Clark, 1975). The failure time distribution is the function F°(t) which provides the probability of death at or before time t in the study. The Kaplan-Meier estimator of F°(t) is F°(t) where F-°(t) - 1 - n [1 - l/R(Ti)] ieD(t) In this equation, D(t) is the "death set" at time t, i.e., the set of all indices i of individuals who were observed to fail before time t. R(T^) is the number of individuals who were at risk just before time T^, the time of death (or morbid event) of the i^n study individual in D(t). A nonparametric approach to testing the equality of survival distributions in a matched 35 �pair study has been developed by Wei (1980). His statistic is a generalization of the Gehan (1965a) statistic. A second test for homogeneity of survival distributions for discretized failure data is the test for marginal homogeneity in a KxK table due to Stuart (1955). Thirdly, the McCullough Model and test may be used on the KxK array to test for marginal homogeneity and stochastic ordering. (c) Morta1 i ty ana lysis wi th _cgva r i ates. These methods allow adjustment of mortality rates or morbidity rates using covariates such as age, race, length of time in RVN, AFSC, risk taking score, etc. Kor the purposes of this discussion it will be assumed that the covariables are categorical, that there are only two such covariables and the covariables do not interact in affecting the hazard of death or morbidity. These assumptions can all be relaxed using available methods. The hazard function h-j(t) for the 1th individual in the study is the function which provides the conditional probability of death or morbid event in the time interval (t, t+dt) given his survival up to time t. The function H-j(t) where t Hi(t) = / hi(T)dT 0 is called the cumulative hazard for the ith individual. that the failure time distribution F?(t) is given by: It is readily shown F°(t) = 1 - exp(-H.(t)) From this last equation it follows that hj and F° are transforms of each other, hence the dependence of F° on covariables may be modeled via h-j. This may be accomplished as follows. Let X-j(t) and Y-j(t) denote discrete valued stochastic processes pertaining to the ith individual and describing two covariates of interest (e.g., one may be an exposure variable and the other may be covariate such as age or crew position). A basic model for hazard is: hi(t) = exp [?Xi(t) + nYi(t)] where ? and n are "log-relative risks". This model may be extended to allow for any number of possibly interacting factors. Inference about log-relative risks may be drawn using either an approach derived from D. R. Cox (1972) by E. Peritz and R. Ray (1978) or using an approach described by Frank (1977). Another model, termed the proportional hazards model, is given by 36 �[3Xi(t)] The proportional hazards model has been discussed, for the special case that Xi(t) does not change with time, by Cox (1972). A test for the equality of survival distributions in a matched pair study which incorporates the proportional hazard model has been given by Breslow (1975). A test of fit for the proportional hazards model is given by Schoenfeld (1980). E. Morbidity Study (1) General Consrderations A vigorous attempt to determine the morbidity experience of all exposed subjects and their primary controls will be undertaken using questionnaires, indepth personal interviews, and physical examinations. A waiver will be requested from the U.S. Attorney General so that medical information collected during the conduct of this study may be exempted from subpoena into Federal Court. Total confidentiality of medical information will be granted to subjects who are not on active duty, and partial confidentiality will be given to active duty subjects with release of information to the DOD only in instances where there is a public safety or national security risk. The schedule and method of contact with the study subjects is depicted in the Appendix Table A-7. (2) Questipnnaire Methods All living exposed subjects and their primary controls will be offered a comprehensive personal and family health questionnaire administered in the subject's home by a civilian contractor. In addition to subject interviews, a face-to-face interview will be conducted with the current spouses of the subjects to obtain a more accurate and complete assessment of fertility and reproductive function. Reproductive Information that will be collected includes but is not limited to the number of live births, the number of still births, the number of miscarriages, the number of children conceived, the number of abnormal offspring, and the total years of marriage. Previous spouses of divorced or remarried subjects w i l l also be interviewed to obtain similar data. Interviews with the first order next-of-kin of deceased subjects will provide morbidity data on thesubject prior to his death. Whenever subjects, their spouses or next-of-kin will not consent to participate in a face-to-face interview, attempts will be made to elicit the information by telephone. The questionnaire is an important part of this study because noncompliance rates for the physical examination and its face-to-face medical interview are expected to be substantially greater than non-compliance with the initial questionnaire. The questionnaire serves a four-fold purpose: (1) to capture baseline personal and medical data on subjects who might be noncompliant for subsequent physical examinations, (2) to serve as a cross-reference 37 �source for objective data obtained at the time of physical examination, (3) to obtain a targeted medical inventory, independent of the physical examination process, and (4) to obtain health perception data to serve as a foundation for the replacement strategy. As depicted in the Appendix, Figure A-2, only an estimated 40% of the RANCH HAND: population will participate in the examination, while at least 65% will respond to the questionnaire. The information collected by questionnaire from; these additional 309 individuals and their controls will provide valuable morbidity data which would otherwise be lost. The questionnaire (see Section XI) will emphasize identification data, RVN tour history, dermatologic conditions, neuropsychiatric conditions, fertility aberrations, genetic defects in offspring, sensory defects, and personality factors. A targeted medical inventory will be included in the questionnaire, and will inventory symptoms prior to, during, and after duty in RVN as well as those currently manifested. It will take approximately six months to complete all initial questionnaires on both groups. The questionnaire will be "fieldtested" by the contractor on former Air Force personnel with RVN experience. Specific questions on the questionnaire will be directed to verifiable information, wherever possible. Questionnaire development and refinement, including specific response verification procedures have been pursued through civilian contract. Questionnaire data will be cross-linked and integrated with medical record information and physical examination findings. Questionnaire data from individuals not completing all phases of the study will not be discarded, but will be incorporated within the entire data base where statistically appropriate. Each participant will be asked to sign release forms so that all civilian health records, including those of dependents, can be obtained and reviewed as necessary. Attempts will be made to obtain pathological reports and specimens following surgical procedures. Federal health records on all family members on file in the NPRC will be retrieved. For retired members, and separated members with VA privileges, all available VA medical records will be obtained. All retrieved medical records will be reviewed, scored, compared to questionnaire data for reliability, and then be entered into a repository system. I lantified participants who are nonresponsive to questionnaire will be pursued to determine status, disinterest, moribund state or death, etc. These individuals will be cross-referenced in other federal record systems in an attempt to achieve total ascertainment. Death certificates and autopsy sports will be retrieved on all dead exposed and matched control subjects for the mortality analysis. Birth/death certificates will be sought for all offspring. (3) Physical Examination A voluntary comprehensive physical examination will be offered to all individuals in both the exposed and primary control groups within one year of questionnaire administration. The condition for entry into the examination phase of the study will be the completion of the baseline questionnaire. In the event that the primary control does not complete both the questionnaire and the physical examination, a replacement will be selected from the control set [See Figure 3 and Section F(3)J. Statistical testing will be conducted by a variety of techniques on both questionnaire and examination findings. At the time of physical examination, an extensive physical examination, medical 38 �history, and review of symptoms will be conducted. A standardized protocol will be used to insure comparability of data. This will provide crossreference data to the initial questionnaire and to medical record data, if retrievable. Specific response verification and bias indicator questions will be included in this interview as well. (a) Examinati on Parameters A comprehensive physical examination will be conducted on all willing participants. The examination will be structured as outlined below and in Section XII and will be performed at the earliest practical time following the completion of the questionnaire. The close sequencing of these study components will limit the development of major symptoms in the interval between the questionnaire and the examination. Examinations will be performed under contract at a single civilian medical center haying dermatologic, neurologic and electromyogram/ nerve conduction capabilities. Informed consent forms will be obtained for all procedures. Physicians and technicians will handle all participants without a knowledge of exposed or control status, and will conduct the examinations by standardized protocols to minimize variability. Medical students, interns, and residents will not be allowed to perform these examinations, and specialty trained neurologists and dermatologists will perform the appropriate portions of the examination. An onsite monitor will insure that the examination protocol is followed. All laboratory tests will be subject to rigid quality control. Laboratory and physical examination data will be measured on a continuous scale whenever possible in order to improve statistical power in the analysis. Under special circumstances, additional testing will be accomplished. Karyotyping of the individual and his family members will be considered if clinical history or physical examination findings are suggestive of this need. Most well conducted studies have shown that, when present, chromosomal abnormalities due to TCDD are transient. If on detailed analysis of the baseline examination and questionnaire, reproductive areas are heavily affected, routine karyotyping may be included in the test battery for the followup phases of the study. TCDD analysis on blood and urine will be considered in the future provided that (1) strong cause and effect relationships can be ascribed to Herbicide Orange and (2) high resolution mass spectrometry technology achieves 10 femtogram sensitivity with high isomeric specificity. Serum, urine, and semen specimens will be obtained from all participants, aliquoted, and preserved at -70°C for possible analysis in the future. These serum and/or urine specimens will also be used for analysis of porphyrin metabolites if analytic techniques make this a feasible diagnositc procedure. Extensive immunologic function analyses will be conducted on a randomly selected group of subjects. Physical examination and laboratory data will be placed in the member's coded master file for detailed cross-analysis to questionnaire data. Information identifiable to the subject will not be released without his consent in accordance with the Privacy Act. However, in accordance with Air Force regulations, active duty flying personnel and active duty air traffic controllers found to have conditions which are disqualifying for flying duty will be temporarily "grounded" pending resolution of the medical condition. 39 �Physi cal Examination Prof i1e General Physical Examination FBS, 2 Hr Post Prandial Urinalysis BUN/Creatinine Cholesterol/HDL Triglycerides Serum Protein Electrophonesis Hemoglobin Hematocrit White Blood Cell Count and Differential Platelet Count RBC Indices Sedimentation Rate Prothrombin Time CPK ECG Chest X-Ray VDRL/FTA Cortisol Differential Thyroid Profile (RIA) Pulmonary Function Studies Blood Alcohol Dermatologic Examination Urine Porphyrins Urine Porphobilinogen Delta-aminolevulenic Acid Neuro-Psychiatric Examination General Neurologic Examination Psychological Battery: Nerve Conduction Velocities MMPI WAIS WRAT Halstead-Reitan Wechsler Memory Scale Subtests Cornell Index Reproductive Examination LH, FSH, Testosterone Semen Analysis Neoplastic/Hepatic Examination SGOT Alkaline Phosphatase SGPT LDH (Isoenzymes 1f elevated) GGTP Hepatitis B Antigens/Antibodies Bilirubin, Total and Direct Additional Studies (Individuals with abnormal history or examination) Karyotyping Immunoelectrophoresis Hepatitis A Antigens/ Bilateral profile and fullAntibodies face photographs Skin Biopsy Anti-Nuclear Antibody Quantitative Immunoglobulins Additional Consultations as Required Immunologic studies (conducted on a randomly selected group of subjects) Enumeration of B and T Cells B and T Cell Function Enumeration of Monocytes 40 �(4) Analysi s of Questi onnai re an^d Physi cal Exami nati on Data The Questionnaire and Physical Examination will produce data of three types: (1) dichotomous, (2) polytomous and (3) continuous. Dtchotomous (e.g., present/absent) rates will be evaluated using the tools described above for mortality analysis. For example, the questionnaire will provide data concerning the first occurrence of disease states by age, and standardized rates and relative risks may be calculated. The occurrence of such findings can be related to age, time spent in RVN, exposure, and other variables using logistic models followed by McNemar's test where appropriate. These tests will examine the presence or absence of group effect and allow assessment of the statistical significance on non-unity relative risks. Polytomous findings will occur in both questionnaire and physical examination responses. As an example consider retinal findings categorized into four grades, and studied as a function of age and exposure group as represented in Table 9. In this table the x-jjk's are counts of occurrence. In analyzing tables such as these, techniques as described by Bishop, Fienberg, and Holland (1975) will be used. Specifically, if m-jj^ is the expected value of x-jjk, general log-linear models of the form + u2(j) + u3(k) + u 13 (ik) + u23(Jk) + u123(ijk) will be used, where u:(i) is the effect of RANCH HAND membership alone on cell frequency, u12(ij) is the effect of an interaction on RANCH HAND membership with retinal grade, etc. This model can work with dichotomous as well as polytomous data. Under appropriate conditions on expected values of entries in Table 9, the pairing in the study design can be used with the data being organized as shown in Table 10. In Table 10, NJJ is the number of pairs such that the exposed person has retinal grade i, and the control person has retinal grade j. Appropriate tests for this setting are indicated by Fleiss (1973) and McCullough (1978). With regard to continuous variables, the intended method follows Carpenter (1977) who found substantial gains in analysis efficiency by matching cases, subsequently employing covariance analysis to remove non-controlled effects. The conditional logistic regression model for relative risk, Holford, White and Kelsey (1978), is also applicable and will be used. 41 �Table 9 FORMAT OF CATEGORICAL REPRESENTATION OF RETINAL CHANGES RANCH HAND PERSONNEL CONTROLS ""--vAge Category Retinal Category^*""**-^ 1 2 3 4 1 X i ; 1 X 11.2 X 113 X 2 X 121 X 122 X 123 3 X 131 X 132 X 133 4 X 1U1 i ^ X 1 2 3 4 m X 211 X 212 X 213 X 21»v X 12H X 221 X 222 X 223 X 224 X 13"+ X 231 X 232 X 233 X 234 u «/ w " 1 U 9. " 1 U' 3 ^ l A *L t U T" XTi XT O \f w y ** Table 10 FORMAT OF PAIRING FOR GRADES OF RETINAL FINDINGS ^Control Grade RANCH HAND Grade" 1 Nil N 12 N 13 N11+ 2 N21 N22 N23 N2(t 3 N31 N 32 N 33 N34 4 N N N N (5) Analysis of Fertl11ty/Reproduction Data. The herbicides under consideration in this study have been alleged to effect fertility and/or reproductive functioning. An attempt will be made to address these allegations by analyzing at least three primary variables: the total number of conceptions since exposure in RVN:, the number of miscarriages in spouses since exposure in RVN and, the number of abnormal offspring since exposure in RVN. The interview with current and former spouses will provide much more accurate information on fertility and reproductive functioning than if similar data were obtained from the male subjects themselves. The study questionnaire will provide the numbers of miscarriages, abnormal offspring and of live births. The sum of the number of miscarriages, still births, and live births will provide an estimate of the total number of conceptions. If differing divorce rates are found in the RANCH HAND and control groups, this may render the average number of years of marriage and the distribution of the years of marriage different in the two groups. This will be investigated and adjusted 42 �for if need be, either by analyzing total number of conceptions divided by (or normalized by) the number of years of marriage, or by using a more detailed covariance analysis. Further, the ratio of the number of miscarriages to adjusted total conceptions will be calculated and compared, as will be the ratio of the number of abnormal births and adjusted total conceptions. In summary, the following statistics relating to fertility will be calculated and analyzed at the very least: TOTAL CONCEPTIONS = #Live Births + #Still Births + #Miscarriages NORMALIZED FERTILITY INDEX MISCARRIAGE FRACTION ABNORMALITY FRACTION TOTAL CONCEPTIONS YEARS OF MARRIAGE # MISCARRIAGES TOTAL CONCEPTIONS # ABNORMAL OFFSPRING TOTAL CONCEPTIONS F. Foilpw - up S t udy (1) Study Adaptations Following complete data analysis of the initial mortality and morbidity studies, adaptive or restrictive health surveys will be developed and administered to all follow-up study subjects three, five, ten, fifteen and twenty years after the initial questionnaire. Similarly, a condensed physical examination profile that will achieve adequate sensitivity and specificity for prospective diagnosis will be developed. The adaptive physical examination will be offered to all follow-up participants, and will also be conducted in years three, five, ten, fifteen, and twenty (see Appendix, Table A-5). An interim examination in year three is essential in this study because the age group under study is approaching that portion of the mortality/illness incidence curve with the steepest slope. A lapse of five years between the first two examinations could easily miss significant development of disease in the intervening years. Ample precedent for interim examinations can be found in the Framingham cardiovascular disease study, and in the follow-up evaluation of West Point graduates being conducted by the Air Force. 43 �(2) Entry Criteria All exposed or control individuals completing the baseline questionnaire and physical examination will be entered into the follow-up; further continuation will depend upon the member's willingness/ability to participate in additional health surveys and condensed examinations. Specific study entry rules are detailed in Table A-6 and Figure A-3 of the Appendix. (3) Loss to Study; Key Issues Loss of participants over time adversely affects any epidemiologic study in two ways. Ar the sizes of the study groups decrease, statistical power also declines, and bias is injected into the study if losses are not randomly distributed in the study populations. It is reasonable to assume that in this study, losses will be non-random with greater non-compliance among individuals who perceive their health as "well," since there is less incentive for this group to continue participation. As shown in Figure 5, such a differential pattern of loss will alter the population, and skew the frequency distribution curve. Most previous epidemiologic studies have approached the problem of declining statistical power by beginning the study with multiple controls per exposed subject, and passively allowing attrition to occur throughout the study period. However, this approach does not address the problem of bias. This study will take an active approach to both of these problems by using a replacement concept. As a control is lost to study, a replacement will be chosen from the original set of ten matched controls. The replacement will be selected from the control set, and will have a perception of health similar to that of the lost control (Figure 6). The replacement strategy will maintain statistical power and the integrity of the matched design despite loss to study in the control group, and will correct anticipated bias while minimizing the number of required physical examinat'ons. At the initiation of the follow-up study, loss of an exposed member will not be cause to cease surveillance of his primary matched control. In the event of a control loss (for reasons other than death), another control from the set will be brought to study (Figure 7), the comprehensive questionnaire will be administered, and a baseline physical examination performed. If a control is noncompliant for one portion of the study and is replaced by another control, the noncompliant individual will be approached at the time of subsequent questionnaires and examinations, and encouraged to reenter the study. If he reenters, both he and the replacement will be included in the evaluation. Similarly, noncompliant exposed subjects will also be aggressively recruited for all subsequent study phases. 44 �FIGURE 5 EFFECT OF NON-RANDOM LOSS TO STUDY IN THE CONTROL POPULATION N WELL ILL HEALTH SPECTRUM IF CONTROL LOSSES ARE ILL, A SPURIOUS EFFECT IS ATTRIBUTED TO HERBICIDE EXPOSURE. IF CONTROL LOSSES ARE WELL, A TRUE/VALID HEALTH EFFECT IS DILUTED. �riGUKt THE REPLACEMENT STRATEGY EXPOSED PRIMARY CONTROLS LOSSES REPLACEMENTS MATCHED FOR HEALTH STATUS {DEAD CONTROLS NOT REPLACED 01 �FIGURE 7 CONTROL REPLACEMENT FOR THE MORBIDITY AND FOLLOW UP STUDIES EXPOSED CONTROL YEAR 0 YEAR 1 YEAR 2 YEAR 3 YEAR 4 YEAR 5 YEAR 6 • QUESTIONNAIRE DATA O RECONSTRUCTED DATA LOSS TO STUDY PHYSICAL EXAMINATION DATA �For exposed and control individuals who drop out of the study but subsequently re-enter, medical data for the intervening years will be reconstructed from questionnaire and interview responses. IN ALL CASES OF LOSSTO-STUDY, INTENSIVE EFFORTS WILL BE MADE TO DETERMINE THE SPECIFIC REASONS FOR NON-COMPLIANCE, AND DATA FROM REPLACEMENT CONTROLS WILL BE REVIEWED TO ASSESS COMPARABILITY WITH THE LOST INDIVIDUALS. Medical record reviews of new entrants will continue throughout the follow-up period, (4) Study Length The follow-up st"dy is initially planned for 20 consecutive years. Procedures, progress, anJ interim results of the study will be monitored by an independent scientific review group, responsible to the Office of Science and Technology Policy in the White House. G. Determination of "Disease" (1) Introduction Since this study is dealing with an unknown clinical endpoint with unknown latency, determination of a disease state by statistical methodology is a prime scientific thrust of the investigation. From the literature, chloracne is the only generally accepted chronic disease associated with high exposure to dioxin. The questions of primary interest are: (1) Does a history of chloracne invariably lead to future disease? and (2) In the absence of chloracne, is there emergence of other attributable diseases? Under a broad concept of "spectrum of illness", either or both of these conditions are possible. The clarification of their respective contributions to the natural history of past or of subsequent "disease" is of significant interest. (2) Discussion Inferences about a disease state from this study can be derived from several logical approaches. The^e approaches can be grouped into two categories: (1) those dealing with symptoms which can be used to construct a symptom complex that may represent disease, and (2) those dealing with physical signs which in themselves rep/esent disease. In the former, one can form a subset of individuals that have symptoms (e.g., infertility) and study them during the morbidity and follow-up studies. Focusing on the overall patterns of alleged symptoms and categorizing them into a symptom complex may identify those individuals with a disease syndrome, or those at higher risk of developing disease (e.g., genetic disorders, cancer). In the latter approach, data on abnormal physical signs (e.g., genetic defects in offspring) and laboratory results can be compared between exposed and non-exposed groups in an attempt to again establish the presence or absence of disease. By putting this array of data into a logical decision-making scheme, specific relative risks can be calculated in the follow-up study, and specific response patterns can be inferred as shown in Figure 8. 48 �FIGURE INTERPRETATION OF HORIZONTAL COMPARISONS OVERT EFFECT lYi D i\ SUBCLINICAL OVER-REPORTING M R = MQ i ' « f* u MR = MC _ Cf — o l> Q~ JR « - sc SR > S c n S t\ F F R * Fc c D C -^ •r f*C " uo F RS* F F CS F RS" F R - Fc r\o RS" F MORTALITY/SYMPTOM/ SIGN REGRESSION ON EXPOSURE F R = FC F CS ft! 5 - F C S F CS RS = F CS SIGN REGRESSION NO REGRESSION ON EXPOSURE ON EXPOSURE SEEN F R = F RS S R + F RS C - S R ) �Again referring to Figure 8, at least three clinical patterns can defined. These patterns are delineated using relative risks (mr/mc, sr/sc, fr/fc etc., between group or "vertical" comparisons, referencing Figure 4) and using within group ("horizontal" study) comparisons such as regressing symtoms and findings rates against an index of herbicide exposure, and other comparisons. Specifically, an overt clinical effect would be marked by: an increased mortality rate in the RANCH HAND group (mr > me), an increased rate of symptom formation in the RANCH HAND group (sr > sc), and an increased rate of objective medical findings in the RANCH HAND group as compared to the control group (fr > fc). Further, the occurrence of physical or objective medical findings would consistently relate to symptoms in the overt case (that is, frs > fcs and frs > fcs\ and finally, in the classic instance, mortality, and symptom and sign formation would be seen to be increased with increasing herbicide exposure. A subclinical pattern is indicated in the central column of Figure 8. In this setting, one expects no statistically significant differences in mortality or symptom reporting between the two groups, exposed versus control. However, one expects a consistent predominance of medical signs in the RANCH HAND group with regression of the signs on increasing herbicide exposure. A pattern strongly suggesting over-reporting is presented as the right column of Figure 8. In this setting, there is no difference between the groups as regards mortality or medical sign incidence; however, more symptoms are reported by the RANCH HAND group. While in this pattern the RANCH HAND subjects are reporting more symptoms, objective medical finding rates are not consistent with symptom reporting. When no regression of symptoms on exposure level is found, over-reporting is clearly and strongly suggested. This discussion of response patterns has used regression on an exposure index in a central way. Development of such an index is discussed below. It is noted, however, that a direct index of exposure can be confounded by other factors such as cellular repair mechanisms or bioaccumulation in adipose tissue with release over time upon weight loss. Use of other factors, such as time since exposure, should help to overcome these confounders. The strength of any inferences made from these analyses is dependent upon the statistical power inherent in the study. In addition, due to the possibility of latency being a factor in this study, a negative analysis at any time within the study does not categorically imply lack of disease, since sufficient time for emergence may not have passed. H. Exposure Indices (1) Exposure Concepts A major concern in conducting this study is the lack of accurate exposure data. Although most personnel assigned to RANCH HAND squadrons were undoubtedly exposed to Herbicide Orange and TCDD, the exposures within the 50 �group must have varied widely. Exposure to herbicides and TCDD by RANCH HAND personnel occurred almost daily. Anecdotal information suggests that many had direct skin contact which was repetitive over a long period of time (one-year tour for most individuals). Further, it is also suggested that most RANCH HAND personnel felt that the herbicides employed in the operations were not toxic to animals and man, and hence, they did not exercise the caution in handling these chemicals that is recommended today. From a historical review of RANCH HAND operations, it appears most individuals can be classified into one of three groups based on their likely potential for exposure to the herbicides: (1) Pilots, Co-pilots and Navigators: (2) Crew Chiefs, Aircraft Mechanic, and other Support Personnel: (3) Console Operators and Flight Engineers: low potential moderate potential high potential The "pilot" group received most of their exposure during preflight checks as well as during the actual dissemination missions. The crew chief group experienced contact with herbicides during dedrumming and aircraft loading operations, as well as during on-site repair of the aircraft and spray equipment. The console operator group was exposed while supervising the loading of the aircraft, during ground testing of equipment, and by tank leakage during dissemination missions. The available historical records on Operation RANCH HAND indicate that personnel assigned to the project seldom had a "routine" work schedule or environment, thus complicating estimates of the level of herbicide and dioxin exposure. Since actual exposure data (e.g., mg of herbicide/kg body wt) are not available, an exposure index will be used. The exposure indices will be calculated for each RANCH HAND individual to obtain frequency distribution, and will be calculated by evaluating the known factors that would have influenced exposure. These will include such factors as: (1) Date of tour with RANCH HAND in Vietnam. (2) Number and lengths of tours in Vietnam with RANCH HAND. (3) Number of herbicide dissemination missions (as reflected by flying hours and air medals). (4) Herbicides employed (records are available that reflect the amount oP each herbicide sprayed each month and year). (5) Crew position. (6) Routes of exposure (the major route of exposure for most RANCH HAND personnel was probably percutaneous, although exposure through inhalation may have also been significant). 51 �A crude exposure index which is applicable to the entire RANCH HAND cohort is expressed with the following formula: E. - q, X T. In this formula, E-,- is the calculated exposure for the iiil RANCH HAND member, qj is the quantity of TCDD-containing herbicide sprayed from aircraft assigned to the iJ^l subject's base during his assignment, and T-j is the length of the illl subject's assignment (tour length). However, great care must be exercised when applying the above index. For example, the index should be used as an independent regression variable against clinical findings only within occupational strata, to avoid confounding occupational effects with exposure effects. Different degrees of regression between clinical findings and the exposure index can be expected in differing occupational groups since: (a) modes of exposure are likely to be different in different occupational categories, (b) socioeconomic correlates within occupational category could confound an herbicide effect, and (c) other exposures which could synergistically or antagonistically interact with TCDD-containing herbicide may be correlated with occupational category. Another factor which must be considered when applying this crude exposure index is the problem on confounding a possible herbicide effect with an effect associated with tour length. Being in a comabt zone is a major psychophysiological stress, and time spent in such an area may be significantly associated with changes in long term morbidity and/or mortality. This crude exposure index, when used alone, could result in a positive regression with disease incidence or prevalence which is not due to the herbicide exposure. An approach that will correct for this potential confounding is to regress observed medical findings on both Ej and T-j to differentiate the independent effects of herbicide exposure and combat zone experience. The values of q-,- and T-j needed to calculate E-j are generally available from government records. Specifically, tour dates are available from military personnel records, and the quantity of herbicide sprayed is available for the period January 1965 through April 1970 from the "HERBS TAPES." These tapes are comprised of computerized data obtained from actual spray mission reports. This material provides the date, base of mission origin, amount and type of material sprayed (Herbicides Orange, Blue, or White) and location of the intended spray target. Estimates of the amount of herbicide sprayed prior to 1965 may be available from procurement records for Herbicides Purple, Pink, and Green, which were sprayed exclusively from Tan San Nhut Air Base from 1962 through 1964. Animal data imply that TCDD is the most toxic component in the herbicides used in RVN. By using q-j, the amount of herbicide sprayed, one is using a variable that roughly correlates with TCDD exposure. However, it would be highly desirable to be able to analyze observed health effects in terms of specific TCDD exposure. The material sprayed from 1965-1970 had significantly lower 52 �TCDD contamination then did those herbicides manufactured and purchased prior to 1962 and used from 1962 through 1964, but due to data limitations from a scarcity of Herbicide Purple, Pink, and Green samples, TCDD concentration profiles for those chemicals cannot be quantitatively determined. However, it may be feasible to develop estimates of the degree of contamination based upon the TCDD concentration from military and manufacturers' data. As another approach to examining the effect of TCDD itself, one might consider stratifying the exposed cohort by date of assignment in Vietnam, expecting that those assigned earlier were more heavily exposed to TCDD. While it may well be true that earlier assignees were exposed to higher TCDD concentrations, it is unlikely that differences between "early" and "late" assignees, if they occur, can be reliably attributed to TCDD concentration changes, since several potentially confounding variables exist: (a) volunteerism among early assignees, (b) differing assignment patterns between early and late RANCH HANDers (TOY vs long term pattern) and (c) different RVN living conditions. It is preferable to use an exposure index which is more closely tailored to the specific individual than the crude index discussed above. While T-j is subject specific, q-j is a value which refers to all individuals on the base during the period of time represented by T-j. A refined index for ground crew can be expressed as: E. = F1 X q. X C X T. where, FJ = Average flights per day served by the ii!l ground crew member. qi = Average quantity of herbicide dispensed by flights served by the illl ground crew member. C = Estimated TCDD concentration of the herbicides in use during the il!l subject's tour of duty. T-J = Time spent in TVN in days for the ilQ. ground crew member, The variable FJ can be estimated by dividing the number of RANCH HAND flights per day by the number of crew chiefs during the time period T-j. All other variables are estimated as with the crude index. A refined index is also possible for aircrew members and is expressed as fol1ows: Ei = M1 X D. X q. X C X P. where, Mi = total number of missions flown by the iJiil air crew member. Di = average duration of missions flown by the iJEill air crew member. qi = average quantity of herbicide dispensed per flight served by the \th. air crew member. C = estimated TCDD concentration of the herbicides in use. P< = a cr-w p o s i t i o n weight ing factor. 53 �As with the refined ground crew index, this refined aircrew index cannot be directly calculated in a strictly quantitative sense using available government records, since records to specifically link missions with particular individuals are not available to objectively determine M-j and Dj. However, reasonably accurate estimates of these parameters may be feasible using questionnaire data. Also air medal awards may allow an indirect estimate of Mi. The crew position parameter P-j must also rely upon estimations. While the specific crew duties of each subject are known, the differential exposures associated with the crew positions within the C-123 aircraft were not determined during RVN spray missions. The 355th TAS/Spray Branch, Rickenbacher AFB OH is presently using the C-123 aircraft, configured with the A/A 45 Y-l Internal Dispenser and attempts to assess Pi can be made. Air flow measurement and herbicide simulant deposition studies conducted by Meek are performed during the course of four C-123 flights. However, difficulties with the measurement equipment limit the validity of the value of the data in an exposure index. Further work along these lines could yield a more quantitative position weighting factor, P-j, for each individual. Refined ground crew and air crew exposure indices can be used singly or in combination with the crude exposure index first presented; however, as with the crude index, confounding must be avoided when the refined indices are used in statistical analyses. The exposure indices listed above are, of course, only applicable to the Ranch Hand cohort. As mentioned, a positive regression of disease incidence or prevalence with increasing exposure index will strongly support herbicide causation. We do not wish to minimize however the role of RANCH HAND versus control group disease incidence/prevalence differences as indicators of a herbicide effect. A major component differentiating the RANCH HANDers from the controls is the increased residence of RANCH HANDers in the RVN itself. If within country time does not correlate with disease incidence, RANCH HAND versus control disease incidence differences may be strongly related to herbicide. If in-country time is significant as a disease correlate, this in itself will be valuable information with regard to assessment of the RVN experience. 54 �V I . Specijl ratns The previous discussion has outlined the general statistical approach followed by this protocol, and has outlined planned analytical methods and inferential strategies for the mortality, questionnaire and physical examination study phases. This section provides an indepth consideration of some special statistical study aspects. A* False Report ing/Mis rep resentatiion Since concern for compensation could unconsciously or consciously influence symptom reporting, and since press reporting itself can stimulate anxiety-based symptom formation, a discussion of false reporting is indicated. A data pattern indicating overreporting has already been discussed in Section V. The goal here is to understand the effect of misrepresentation on estimates of relative risk and the odds ratio. Let S stand for presence of a symptom, and IT denote its absence. This false reporting may be represented as in Figure 9. Figure 9 FALSE REPORTING/MISREPRESENTATION TRUE STATUS Total A A+B REPORTED STATUS C+D B+D A+C The proportion of correctly classified positives is defined by A/(A+C) and is called the sensitivity of the classification scheme; the proportion of correctly classifled negati ves D/(B+D) is called the spec i fi city. When there is non-differential misrepresentation, that is, when the sensitivity and the specificity are the same among the exposed and nonexposed, the bias induced in the estimate of relative risk will be toward the null value. The situation is summarized by Figure 10. 55 �Figure 10 MISREPRESENTATION IN RANCH HAND II TRUE STATUS Exposed S Nonexposed **• S TOTAL S S TOTAL S a b a +b e f . e + f S REPORTED STATUS c d c +d g h a +c b +d n e +g f + h g +h n Using this representation, the true relative risk is (a+c)/n * (e+g)/n, and the apparent relative risk is (a+b)/n * (e+f)/n. Figure 11 provides a graphic representation of how apparent relative risk varies as a function of specificity. For this curve, the true relative risk is 2 with the exposed population having a symptom incidence of 0,1 and the nonexposed population having a symptom incidence of 0.05 (Copeland et. al. 1977). The effect of nondifferential false reporting on the odds ratio is nearly as severe as that shown in Figure 11 for relative risk. A technique does exist for correcting the estimate of relative risk to account for false reporting, but the technique requires knowledge of the sensitivity and specificity of the classification scheme; knowledge that may not exist in this study. It should be noted that since the above remarks are concerned with relative risk, the number n of subjects in each group is irrelevant, as the results shown are independent of n. If the false reporting is differential, an estimate of relative risk that is biased away from the null value can result. This will occur in situations in which the RANCH HAND personnel and controls do not misrepresent their symptoms in the same manner (Copeland et. al. 1977). Thus the "true" outcomes of herbicide exposure may be distorted depending upon the degree and direction of misrepresentation. B. Adequ acy of Samp1e Sizes (1) Overview The size of the RANCH HAND cohort is approximately 1000 individuals. It is clear that a lethal effect of herbicide which occurs in only 1 out of 2000 controls will be quite difficult to detect unless the herbicide effect is very strong. For example, at a rate of 1 in 2000, 0.5 affected controls are expected. If the basic rate is doubled by herbicide to 2 per 2000, one affected RANCH HAND individual would be expected. At a rate of 1 per 2000 for 56 �Figure 11 APPARENT RELATIVE RISK VERSUS SPECIFICITY 2.0 Sensitivity = .9 1.8 1.6 APPARENT RELATIVE RISK •1.4 1.2 1.0 50 55 60 70 75 80 85 90 95 100 SPECIFICITY controls and a rate of 2 per 2000 for RANCH HAND personnel, the probability of observing no affected individuals in both groups is (1 - 1/2000)1000 (1 - 2/2000)1000 = .22 or, in other words, "there is a 22% chance" that no affected individuals will be found in this study. In a population of 100,000 exposed individuals, 100 cases would be expected, 50 of which would be due to herbicide. In short, since the size of the RANCH HAND group is fixed, this study has limited statistical power to define the relationship of herbicide to the rarer diseases. The power (1-3) of a study design is the probability that a specified difference between populations will be detected if it in fact exists. In general, power is a direct function of sample size; that is, for a particular study design, the more subjects measured the larger the study power. It is understood that this protocol makes use of the entire known RANCH HAND population (and excludes ancillary exposed groups for reasons previously cited); the exposed sample size cannot be increased. Power augmentation, therefore, can only be accomplished by the less efficient procedure of increasing the control group size which has statistical limitations as well as staggering financial and logistic considerations. Hence, considerable effort has been made to correct loss to study issues (by replacement and other techniques to induce participation) and to use the most powerful statistical design concepts. 57 �Essentially all previous animal and human studies concerning herbicide suffer from a lack of adequate consideration of study power. The following presents a preliminary analysis of study power for the case of continuous and dichotomous variables expected from the study. Also reviewed are alternative studies involving Marine samples. (2) Power in Conti nuous Vari abl e Case Assume that blood cholesterol levels are being compared between RANCH HAND and control groups, and that the coefficient of variation for cholesterol in the control group is 0.1, where the coefficient of variation is the ratio oc/yc. Assume ORH = ac. The symbol ex is the probability that the study will indicate an effect where none exists, and 1-p is the power as defined before. Consider that the RANCH HAND mean cholesterol PRH is shifted from the control mean yc. A natural question is to inquire about the study power as a function of available pairs (n) and mean ratio y = Table 11 POWER CALCULATIONS ASSUMPTIONS: a=0.05, ac/pc=0.1, Y= Power = 1-3 n=180 n=450 .20 1.01 .20 .20 1.02 1.05 .20 .55 >.995 .38 .88 >.995 .70 .70 .70 1.01 1.02 1.05 .86 >.995 >.995 .995 >.995 >.995 Power calculations are displayed in Table 11. Study power in the case of a matched pair design is strongly dependent on the degree of positive correlation produced between the involved groups by the matching procedure. Of course, the degree of correlation can be expressed by the correlation coefficient r which can take values between -1 (negative correlation) and + 1 (positive correlation), and two values of r have been employed in Table 11. From this table it is seen that if only 450 pairs are studied a 1% shift in mean (= 1.01) will not be reliably detected, but a 2% shift will be detected with a probability of 0.88 if r = 0.2 at least. From this calculation one can infer the need to examine at least 450 pairs to obtain the 2% shift, and to strive for more if possible. 58 �(3) Power in the Pichotomous yajijble^ Case There is significant discussion in the mathematical statistics literature concerning the efficacy of paired designs in the setting of dichotomous responses (Billewicz, 1974; Dry, 1975; Miettinen, 1970; and several others). Table 12 shows a set of calculations which are applicable to the present study. Table 12 POWER CALCULATIONS FOR THE DICHOTOMOUS VARIABLE CASE AS A FUNCTION OF EFFICACY OF PAIRED DESIGNS POWER = 1 - 3 PI Rel. PZ Risk r n= 250 n= 350 n= 450 .05 .01 5 0 .77 .82 .92 .04 .01 4 0 .61 .75 .85 .03 .01 3 0 .40 .51 .59 .10 .05 2 0 .61 .75 .85 .20 .10 2 0 .87 .94 .97 .05 .01 5 .1 .89/. 029 .94/.032 .98/. 064 .04 .01 4 .1 .72/.033 .87/.038 .88/.041 .03 .01 3 .1 .387.020 .68/.046 .71/.077 .10 .05 2 .1 .76/.055 .85/.048 .88/.048 .20 .10 2 .1 .94/. 043 .98/.046 t *. .99/.057 * V i * ^ *« = .050 **a = as indicated In this table, r is again the correlation coefficient indicating the degree of correlation induced between the involved groups by the matching procedure. The probability of the disease among RANCH HAND personnel is symbolized as PI, while P2 is the probability of the disease among the controls. Relative risk is the ratio Pi7p?.. With r = 0.1, sign test power tables were used as an exact version of McNernar's test, and therefore different a levels are shown under each power number. Table 12 shows the positive influence of effective 59 �pairing in the higher power levels noted. Also, it appears that for p2 = 0.01 and P! = 0.03, physical examination of 450 pairs (900 examinations) will disclose the three-fold relative risk with probability less than the minimum target .80. In other words, there is a greater than "20% chance" that a three-fold relative risk on a 1/100 disease state will go undetected in this study if only 350 pairs are examined and if low correlations occur. Once again the need to examine the maximum numbers of pairs in the study is seen. To present these dichotomous power calculations more clearly, calculations in the context of actual disease states have been accomplished. The diseases considered are cardiovascular disease and cancer, corresponding to high and low rate illnesses for the age groups presently under investigation. (a) Card iovas cu1 ar Pi sea s e A logistic risk function was fitted to data from 17,455 autopsies gathered in a WHO collaborative study in Czechoslovakia, Sweden and the USSR. The function fitted has the form P = [1 + exp(a + 3(x-.5) + Yty-.B))]-1 where p = the probability of a complicated coronary lesion x = age scaled linearly so that x = 0 is equivalent to 30 years, and x = 1 is equivalent to 58 years (the age span of the current study) y = 1 or 0 if the subject is exposed or not and a and 3 were obtained from the data. The function represents a fairly high rate disease in that at 40 years of age 7% of the group had the lesion, and at 60 years of age 20% had the lesion. The coefficient y> represents the exposure effect. Power calculations for Y = 3 and Y = .83 are shown in Table 13. This table suggests that if, as a cell toxin, herbicide exposure accelerates cardiovascular disease, this study has a good chance of detecting that acceleration if the herbicide effect is comparable to the age effect. A slight beneficial effect of pairing is seen in this hypothetical example. (b) Cancer A logistic risk function was fitted to breast cancer data presented by Breslow and Day (1975). The function fitted represents a low rate disease in that at 35 years of age only .000336 of the group had the lesion while at 70 years of age .00676 of the group will have the lesion. 60 �Using pairing to achieve a power of 0.80 in this setting, 1312 pairs would be needed, when the exposure effect is equal to the age effect. This exceeds the size of our RANCH HAND cohort, and reinforces the fact that herbicide exposure effects on rarer diseases will not have a high likelihood of being detected by this study, and again supports an attempt to examine as many pairs as possible. Table 13 POWER CALCULATIONS AS A FUNCTION OF HERBICIDE EFFECT ASSUMPTION: a = 0.05 T = Y = Number of Pairs Power Neglecting Pairing Power With Pairing Power Neglecting Pairing Power With Pairing 100 .93 .93 .64 .53 (a = .036) 160 >.97 .98 .81 .82 200 >.99 >.995 .86 .87 250 >.99 >.995 .93 .95 300 >.99 >.995 .96 .97 350 >.99 >.995 ,97 .98 Alternatiye Studies Usi^ng Marine Cohorts The GAO and the National Academy of Sciences have referred to specific Marine cohorts as candidates for a Herbicide Orange epidemiological study. In one suggested study configuration, 5900 marines who were within one half kilometer of a herbicide spray track on the day of spraying are called the exposed group, while 212,100 marines are considered unexposed. In a second suggested study configuration, 21,900 marines within one half kilometer of a spray path within 4 weeks of spraying are considered exposed, while a remaining 196,100 marines are considered unexposed. A mortality study was proposed in both of these study configurations. The mortality phase of this protocol involves approximately 1200 exposed and 6000 control individuals, so that, on the surface, the Marine studies would appear to be more powerful in a statistical sense due to larger numbers. However, in fact, two factors couple to render the marine studies less powerful than the RANCH HAND study detailed in this protocol. First, calculations show that a soldier standing directly 61 �under a spray track at the exact time of spraying receives approximately 1/1000 the dose received by RANCH HAND individuals repeatedly disseminating the mixture throughout the usual RVN tour. Thus even if the unlikely event of being directly under a spray path were repeated 10 times during a marine's RVN tour, the marine's dose would still be only 1/100 that of the RANCH HANDERS. The second factor impacting the Marine study power is the difficulty imposed by the fact that troop positions are only very inexactly known. The available data provide only the battalion headquarter's position relative to herbicide spray paths. Thus troops considered to be exposed could be very far from spray paths, and in fact, be unexposed. On the other hand, troops deemed unexposed in terms of their battalion headquarter's position could in fact have been near spray paths on the day of spraying. Thus, the Marine studies are limited by the problem of misclassification in addition to the fact that the marines received a lesser herbicide exposure than RANCH HAND personnel. It is possible to compare the RANCH HAND study described in this protocol with the Marine studies in a quantitative way. Results of such an analysis are set out in Tables 14 thru 17. In Table 14, the Marine study using 5900 exposed soldiers is contrasted with the RANCH HAND study considering a disease with an incidence of 0.001 in the control groups, and 0,004 in the RANCH HAND exposed cohort. With a relative risk of 4 against a control disease incidence of 0.001, RANCH HAND power is 0.87 while the Marine study power is much less for several combinations of Marine exposure and misclassification. The misclassification figures shown refers to the percentage inclusion of unexposed individuals into the exposed Marine group. For the calculations, disease incidence in the marine exposed group was assumed to be linearly related to exposure. Table 15 is strictly analogous to Table 14 except that the disease state studied has an incidence of 0.01 in the control groups and 0,02 in the RANCH HAND exposed cohort. Again the RANCH HAND study is seen to be significantly more powerful than the Marine study. Tables 16 and 17 directly parallel Tables 14 and 15, respectively, except that the Marine exposed group is considered to consist of 21,900 soldiers. Here again RANCH HAND study power is seen to be significantly superior. Figure 12 shows the RANCH HAND mortality study power as a function of relative risk, and disease incidence in the control group. Figure 13 shows marine study power versus marine exposure for zero to 25% misclassification and a control disease incidence of 0.001 and RANCH HAND relative risk of 4. For this circumstance it is clear that the marine study becomes competitive with the RANCH HAND power only if one assumes that the marines received approximately one half of the RANCH HAND exposure dose. Figure 14 is the same as Figure 13 except that 21,900 marines are considered exposed. Again the Marine study becomes competitive with the RANCH HAND study only if one can assume the exposed marines received 0.2 or more of the RANCH HAND exposure, an assumption which is not supported by the available data. C. The Replacement Concept In the mortality analysis, a randomly selected group of control Individuals will be compared to the RANCH HAND group, and the data gathered will be analyzed for evidence of herbicide effect. In the questionnaire and 62 �MORTALITY ANALYSIS FABLE 14 POWER COMPARISON OF THE RANCH HAND STUDY TO THE MARINE POPULATION CONSIDERING MISCLASSIFICATION AND RELATIVE EXPOSURE POWER TABLE RANCH HAND POWER 1-B MARINE STUDY POWER EXPOSURE LEVELS RELATIVE TO RANCH HAND % MISCLASSIFICATION en us 1/10 .87 ASSUMPTIONS: 0 10 25 RH STUDY POP. 1,200: 6,000 (1:5) MARINE STUDY POP. 5,900: 212,100 NORMAL INCIDENCE OF DISEASE 0.001 DISEASE INCIDENCE IN RH 0.004 LINEAR DOSE - RESPONSE MISCLASS. Of MARINE CONTROLS EXCLUDED 1/20 1/100 .18 .16 .15 .10 .09 .09 .06 .06 .06 1/1000 .05 .05 .05 �MORTALITY ANALYSIS TABLE 15 POWER COMPARISON OF THE RANCH HAND STUDY TO THE MARINE POPULATION CONSIDERING MISCLASSIFICATION AND RELATIVE EXPOSURE POWER TABLE % MISCLASSIFiCATION RANCH HAND POWER n I-B MARINE STUDY POWER EXPOSURE LEVELS RELATIVE TO RANCH HAND 1/10 1/20 1/100 1/1000 CD .92 0 .10 .06 .05 10 .17 .10 .06 .05 25 ASSUMPTIONS: .19 .14 .09 .06 .05 RH STUDY POP. 1,200: 6,000 (1:5) MARINE STUDY POP. 5,900: 212,100 NORMAL INCIDENCE OF DISEASE = 0.01 DISEASE INCIDENCE IN RH = 0.02 LINEAR DOSE - RESPONSE MISCLASS. OF MARINE CONTROLS EXCLUDED �MORTALITY ANALYSIS .E 16 POWER COMPARISON OF THE RANCH HAND STUDY TO THE MARINE POPULATION CONSIDERING MISCLASSIFICATION AND RELATIVE EXPOSURE * POWER TABLE ANCH HAND POWER 1-B MARINE STUDY POWER % MISCLASSIFICATION EXPOSURE LEVELS RELATIVE TO RANCH HAND 1/10 1/20 1/100 1/1000 .87 0 10 25 ASSUMPTIONS: RH STURY POP. 1,200: 6,000(1:5) MARINE STUDY POP. 21,900: 196,100 NORMAL INCIDENCE OF DISEASE =0.001 DISEASE INCIDENCE IN RH= 0.004 LINEAR DOSE - RESPONSE MISCLASS OF MARINE CONTROLS EXCLUDED .38 .33 .26 .17 .15 .13 .07 .06 .06 .05 .05 .05 INCORRECT POPULATION NUMERICS BASED ON ENVIRONMENTAL FATE OF TCDD �TABLE 17 MORTALITY ANALYSIS POWER COMPARISON OF THE RANCH HAND STUDY TO THE MARINE POPULATION CONSIDERING MISCLASSIFICATION AND RELATIVE EXPOSURE * POWER TABLE RANCH HAND POWER 1-B MARINE STUDY POWER MISGLASSIFICATION 1/10 01 01 EXPOSURE LEVELS RELATIVE TO RANCH HAND 1/20 1/100 1/1000 0 .41 .17 .07 .05 10 .36 .16 .07 .05 25 .92 .28 .13 .06 .05 ASSUMPTIONS: RH STUDY POP. 1,200; 6,000 (V.5) MARINE STUDY POP. 21,900: 196,100 NORMAL INCIDENCE OF DISEASE = 0.01 DISEASE INCIDENCE IN RH = 0.02 LINEAR DOSE - RESPONSE MISCLASS. OF MARINE CONTROLS EXCLUDED * INCORRECT POPULATION NUMERICS BASED ON ENVIRONMENTAL FATE OF TCDD �FIGURE 12 CD POWER VERSUS RELATIVE RISK, 1:5 MORTALITY STUDY BY THREE DISEASE INCIDENCES RELATIVE RISK �POWER CURVES OF THE MARINE STUDY CONSIDERING RELATIVE EXPOSURE AND MISCLASSIFICATION OF THE STUDY POPULATION 1.0 r 0.8 CO NO MISCLASSIFICATION 0.6 CO CSL 0.4 5,900 EXPOSED 25 % MISCLASSIFICATION 212,100 CONTROL C3 a. RH RR = 4.0 0.2 0.0 .001 p ,=.001 p 2 = -004 • .01 i i i i i 111 .1 MARINE EXPOSURE/RANCH HAND EXPOSURE LI 1.0 �IGURE 14 POWER CURVES OF THE MARINE STUDY CONSIDERING RELATIVE EXPOSURE AND MISCLASSIFICATION OF THE STUDY POPULATION 1.0 r .8 j- INCORRECT POPULATION NUMERICS BASED ON EVIRONMENTAL FATE OF TCDD to CO uu NO MISCLASSIFICATION .6 25% MISCLASSIFICATION .4 21,900 EXPOSED 196,100 CONTROL RH RR = 4.0 .2 p 1 =.001 p 2 =-004 0.0 .001 J_ .01 I I I I 1 ill J .1 MARINE EXPOSURE/RANCH HAND EXPOSURE I 1.0 �physical examination phases of this study, one of the mortality controls will be randomly selected for each RANCH HAND individual. During the physical examination phase, we must anticipate a significant degree of unwillingness to participate, particularly on the part of control personnel. This loss to study can result in significant bias and loss in statistical power; thus the replacement concept has been developed to mitigate these consequences. In this replacement strategy, we make use of the control individuals matched with each RANCH HAND person. As previously noted, this is accomplished using computerized data files and the matching parameters of age, AFSC, and race. With each RANCH HAND individual Ri there will be associated ten controls C^, C-j2> C-j3, ..., C-j10' The first of these controls, Ci! will be employed in the questionnaire and physical examination phases of the study. If C^ is alive, but unwilling to participate in the study, he will be replaced by another randomly selected participant with similar perception of health status. In order to avoid bias in morbidity analyses, no dead control will be replaced. It is important to emphasize that all replacement controls will be carefully flagged so that they may be treated separately in the statistical analysis. These replacements will be carefully compared to the lost controls to develop indicators of comparability (e.g., morbidity and mortality experience). The initial analysis will be performed on the intact exposed/control pairs. Additional analysis will be conducted on all pairs, both those intact, and those with replaced controls. If we consider RANCH HAND individual R-j, with living control C-j1} we can calculate the probability that control C-jk will be available for the 1st, 2nd and 3rd physical examinations. To examine this question, a small computer Monte Carlo simulation was required. A short BASIC language computer program and glossary are included in Appendix Table A-8. This simulation examines the effect of non-participation expressed as two probabilities PI and P2. Figure A-2 displays the expected participation by the RANCH HAND population, and control group participation is expected to be somewhat less. PI is the probability that when first asked to attend a physical examination, the control individual will not comply. P2 is the probability that a control individual who has agreed once to a physical examination, will not comply for a subsequent examination. In general, PI may be greater than P2. Note that the probabilities P! and P2 must reflect all causes of non-compliance including morbidity and mortality. Table 18 displays a representative simulation run, which provides the number of controls required to find willing matches for 1000 RANCH HAND personnel. The potential bias introduced by non-willingness in controls can be analyzed statistically. If Pq(x) is thesame probability density function for compliant individuals and PNC(X) is the function for non-compliant individuals, we have 70 �Table 18 CONTROL DISTRIBUTIONS BY EXAMINATION MATCHING 1000 RANCH HAND PERSONNEL (Pi = .70, P 2 = .25) EXAMINATION NUMBER CONTROL COHORT 1 2 3 Ci 318 237 177 C2 211 188 156 C3 131 133 136 C4 96 101 97 C5 74 89 90 CG 49 68 77 C7 34 43 59 C8 25 39 52 C9 16 18 33 C 13 20 35 33 64 88 10 Number of Matching Failures p(x) = apc(x) where p(x) is the probability density function for the entire population and x is a vector of important health parameters available on each person. Since / p(x)dx = J pc(x)dx = / PNc(x)dx = 1 it follows that a + 3 = 1 71 �and a and 3 may be viewed as coefficients which "mix" the two subpopulations. If Mc and MNQ are the means of the compliant and noncompliant subpopulations respectively, it can be shown that M = aMn + where M is the mean of the entire population. From this last equation, it is clear that as noncompliant individuals are lost (i.e., 3 tends to zero, a tends to one), M tends to Mc. Thus the maximum bias is the quantity McM. In this study we propose to replace non-compliant control individuals with matched RANCH HAND control individuals, that is with individuals drawn from a population with density equal to or at least similar to PNC(X)' Tne resulting new density is P"(x) such that p"(x) = a"pn(x) + 3"PNC(X) where a" + 3" = 1 M" = a"Mn + 3"?NC and where PNC(X) approximates PNC(X)« If 3" is chosen to be close to or equal to 3 above, it appears that M" can well approximate M, the true population mean. The difficulty in this approach will be to assure that the replacements are representative of the non-compliant individuals in all respects other than logistic factors impacting willingness to participate in the program. Our proposed approach is to obtain sufficient data on the unwilling personnel so that a discrimination function of the form D = can be derived. This function is envisioned to have the following properties: 72 �(a) larger values of D correspond to decreasing probabilities of compliance with the physical examination, (b) the factors hj relate to the subjects' health status, while the factors 1-j relate to logistic difficulties (distance, job) which tend to preclude attendance at the physical. Factors to be considered in the formulation of this function are displayed in Table 19. (c) D is an increasing function of each hi and of each lj, Table 19 FACTORS AFFECTING COMPLIANCE Health Status (hj) Loglst1c Pifflailties (1 j) Subjective Health Assessment (good/poor) Time Away from Family Current Utilization of LongTerm Health Care (Yes/No) Time Away from Job Absenteeism Pattern (Greater Than/Less Than Ten Lost Days in Past Six Months) Distance to Examination Site Active Pilot Income (Greater than/Less than $17,000) In the replacement scheme, controls substituted for noncompliant controls, should have identical health factors (h-j) as those individuals they replace. The only significant differences should be in the logistic factors Oi)« The replacement method should permit correction of non-compliance bias given that health factors h-,- and logistic factors 1-j are actually distinct. The determination of these two classes of factors will be made using data from the study itself. Specifically, the logistic factors 1-j will be independent of health status to the degree testable by the quantity of data available in the study. This replacement strategy has two major advantages: selection bias reduction/estimation and cost reduction. Were replacements not employed, one would be compelled to start the morbidity study with a 4 to 1 or 5 to 1 design in order to insure an adequate number of participating controls on the third physical examination (see Table 18). Such a large control group for physical examination is very costly with little 73 �corresponding gain in study power and with no correction of the selection bias. D. Statistical Analysis of Large Data Sets A large amount of data will be collected on each subject in this study. Testing at the 0.05 a level means that in 5 out of 100 instances where there has actually been no herbicide effect, a herbicide effect will be falsely inferred. This is the inverse of the power question which concerns the probability of detecting an event when it actually occurs. If 100 independent measures are taken from subjects one should expect, testing at the 0.05 a level, that five measures will be positive on the average. This awareness itself should help prevent over reaction to isolated findings. Further, the present protocol does not in fact have one hundred independent measures. Rather the data gathered are grouped into correlated batteries or systems of data. Findings with any given measure will be related to the values of other correlated variables to provide substantiation indicating an authentic finding. E. Tiroe-In-Stu dy Effects The study outlined in this protocol is expected to involve up to six examinations extending over a period of twenty years. It could be anticipated that participation in the study, by increasing the health awareness of the subjects, would tend to improve the health of the cohorts. The possibility of differential participation in the study by the exposed and control groups could bias against finding a herbicide effect if one exists. The control group could be less willing to participate in the study than will the exposed RANCH HAND personnel. Thus, if on the average, controls spend less time in the study than RANCH HANDERS, and under the supposition that increased time in study will correlate with better health, increased RANCH HAND participation would counterbalance any adverse herbicide health effect. The corrector for this time-in-study effect is simply to study the relationship between health outcome and participation in the RANCH HAND study by regression or other analogous statistical methods. Participation can be quantitated by such metrics as (a) number of physical examinations attended (b) age at physical examinations attended or (c) pattern of physical examination attendance. Special study design features do not need to be incorporated to properly evaluate time-in-study effects on questionnaire and physical examination portions of the study. However, the effects of differential time-in-study on the mortality analysis must be carefully considered. In order to detect time-in-study effects on mortality, individuals whose mortality are being tracked should have been in the study for the same length of time (both exposed and control individuals), or the distribution of time spent in the study should be similar in both groups. Because of anticipated differential participation between the exposed and control groups, one cannot assume that both cohorts will have equal time in study distributions. Steps must be taken to insure that a proper time-in-study distribution occurs in the control mortality group. Control over this distribution is possible through placement of the mortality cohort in the structure of the control group with 74 �respect to the replacement strategy. The following five designs have been considered: I. mortality subjects randomized over all ten control positions, and therefore called into the study randomly. II. mortality subjects in the first five control positions, and therefore called into the study first. III. mortality subjects in positions #1 and #2, with the three remaining subjects randomized into positions #3 through #10. IV. mortality subjects in positions #1, #2, #9, and #10, with the one remaining subject randomized in positions #3 through #8. V. mortality subjects in the first four positions and position #10. For each of these five designs, certain quantities were calculated. For testing a physical examination effect on mortality, one would require adequate numbers of mortality subjects having had all six physical examinations, and adequate numbers having had none. Therefore, assuming 1200 RANCH HAND subjects, El = expected number of mortality subjects having all six physical examinations. E2 = expected number of mortality subjects never asked to take the physical examination. E3 = expected number of mortality subjects physical examinations. having taken no For testing or modeling time-in-study effects, one would want adequate numbers of mortality subjects having only one physical, having exactly two physicals, etc. Hence, we calculate, for J = 1, 2, 3, 4, 5, 6: NJ = expected number of mortality subjects taking exactly J physicals (for example N3 is the number of mortality subjects who will have taken three physicals by the end of the study). and MJ = expected number of mortality subjects which will actually have taken examination J. The values of El, E2, E3, NJ, and MJ have been calculated for the five study designs outlined above using an adaptation of the Monte Carlo program 75 �shown in Appendix Table A8. Best case and worst case situations were considered. In the worst case, it was assumed that when first asked to participate, 75% of the subjects refused, while when asked after having once participated, 50% of subjects refused further contact. In the best case, the first time refusal rate was assumed to be 50%, and the refusal rate for a subject who had participated in a prior examination was assumed to be only 15%. Table 20 shows the calculated results. In examining this table it is of interest to note that the calculated values are not strikingly dependent on study design configuration. However, for both the worst and best cases, design 2 where the mortality subjects are placed in the first five control positions, appears superior and will be used in this study. Table 20. TIME-IN-STUDY EFFECTS V DESIGN PARAMETERSi\ 1 WORST CASE 2 3 4 5 1 BEST CASE 2 3 4 El 18 29 25 22 27 267 E2 765 194 580 865 493 3953 2409 3137 3478 2690 E3 4691 4548 4645 4716 4623 Nl 700 751 713 681 714 227 370 284 239 331 N2 340 374 347 328 355 185 307 284 239 331 N3 163 188 168 157 177 146 249 189 161 225 N4 71 84 76 72 79 116 203 157 136 188 N5 33 43 36 31 39 94 171 129 110 160 • N6 18 29 25 22 27 267 521 454 428 504 Ml 18 29 267 521 M2 14 14 46 77 M3 24 17 53 83 M4 40 25 62 94 M5 54 28 76 107 M6 80 34 85 116 76 521 454 428 5 504 4977 4204 4569 4739 4345 �VII. Data Repository Throughout the period of this investigation, data collection methods will be integrated by use of computer systems. A data repository will be established at the USAFSAM. Master files will be formed for each exposed member and for his matched control/controls. The individual master files will be keyed to one or more identifiers. Confidentiality of data will be maintained by the use of computer generated code numbers. Addresses and telephone numbers of all study subjects will be continually updated to insure proper follow-up. Individual data items and their sources are as follows: (1) Questionnaire a. Initial b. Indepth interview (during physical examination) c. Follow-up (2) Psychological Battery a. b. Initial Follow-up (3) Physical Examination a. b. Initial Follow-up (4) Medical Records a. Active duty b. VA c. Civilian d. Dependent (5) Historical Data a. b. c. Military personnel files Flight records Military unit (6) Death Certificates and Autopsy Reports a. b. Study members Dependents (7) Birth Certificates a. Dependents Mortality data will be obtained from individual medical records, VA records, the screening of personnel records, contact with family or personel physicians, and other available information sources. Date of death (verified by death certificate and available autopsy reports) will be obtained. Cause of death will be expressed as an ICDA number or numbers. The reliability of the mortality data coding will be evaluated by using a dual coding system based on underlying cause of death criteria in use by the National Center for Health Statistics. This will assure that the results of this study are compatible with data based on US mortality statistics. In addition to standard coding for the underlying cause of death, all diagnoses entered on the death certificates will be coded so that multiple cause of death analyses can be conducted. 77 �The computer software for the data analysis phase will be prepared to assure proper data conversion, quality control and standardization of testmeasurements. Quality control areas will include verification of identification data, range checks, and identification/correction of ambiguous or conflicting data. 78 �VIII. Recogni zed Study Pi f f i cu 1 ties a n d C o r r e ct 1 ye Mea s u r es A - Medical P^recedence (1) Problem A departure from the ususal methodological approach characterizes this particular epidemiological investigation. Clearly there is no historical "roadmap of methodology" to conduct this study. Most occupational exposure studies use the presentation of an unusual disease to justify the initiation of a comprehensive study. A rare disease or a common disease in an uncommon site, or one with an unusual presentation appearing in space-time clusters, (often in an unusual population or age group) usually generates the requirement for a new study. In the case of Herbicide Orange, the evidence for long-term human effects is tenuous and controversial . Despite the unique problems that this study possesses, such as the lack of clinically defined endpoints, there are many problems that it shares with other occupational ly related exposure studies. For example, the question of a latent period in the development of symptoms/signs, the lack of accurate dose-response relationships, and the possibility of a synergistic effect with other toxins/ carcinogens are all operating in this study. Since most cohort studies of occupational mortality use the general population as a standard for deriving the expected number of deaths, preempl oyment selection ("healthy worker" bias) affects the comparative experience. Age-standardized mortality ratios (SMR's) in general are 60-90 percent of the standard in the working population. Similar conflicting results can occur using the matched cohort method proposed in this study design. Statistical verification of the validity of utilizing such a control for a summary mortality index (e.g., SMR) has been infrequently attempted in the past. Inability to verify the validity of the more classical methods of comparing mortality will necessitate the use of multiplicative and/or logistic models to obtain a valid standardized mortality ratio. (2 ) Correcti y e_ Measures Study approaches generated by unprecedented occurrences of occupational ly related medical complaints require novel approaches, and reorientation beyond standard methods. The success key to this study design is a series of effective, progressive, and helpful peer reviews (all of which have occurred to date and have been incorporated herein). Beyond even the immediacy of the current study, is the growing problem of a myriad of occupationally-related exposures, both in the military and civilian sector, which will require similar epidemiological studies in the future in order to make some judgment as to whether or not an association is of causal significance. (1) Problem The numerous media presentations on "Herbicide Orange" issues have focused attention on the RANCH HAND group. Several attempts have been 79 �made to construct lists of former members of this group, and thus, the RANCH HAND population should be somewhat easier to locate and contact than the control population. This difference will be particularly evident with respect to reported mortality experience. The incentives for cooperation and study participation are likely to be greater in the exposed group than in the controls. Also, the close knit reunion association of former RANCH HAND personnel will lead to a more precise reporting of morbidity and mortality in that group. Such group identity tends to decrease the degree of unaccountability in the exposed group while its absence in the controls may lead to under ascertainment of mortality. This could then lead to the attribution of excess mortality in the exposed population. (2) Corrective Measures Unaccountability bias will be minimized by keeping the percentages of unaccounted for study subjects below 1% in both exposed and control groups. The morbidity and mortality status of all individuals selected for the study will be strongly pursued utilizing a variety of techniques previously described in this document. C. "Risk Taking" Behavior Bias (1) Problem The early RANCH HAND aircrew population was an exclusively volunteer group; the C-130 control population, while volunteers in the Air Force, were not volunteers for special hazardous missions. RANCH HAND mission conditions were considered to be more dangerous than those encountered in the normal combat environment. This suggests that some differences may exist in the psychological profiles of the two groups. A sensation seeking or risk taking psychological orientation may have altered the accident mortality or morbidity patterns of the exposed group. In addition, an accident rate affected by peripheral neuropathy could be masked by undetected risk taking behavior bias. (2) Corrective Measures In an attempt to correct for the unique psychological factors that affect the choice of an aeronautical career, and to adjust for the effects of combat stress, transport aircrew members were matched with crewmembers of similar transport aircraft. However, the volunteer nature of the pre-1965 RANCH HAND operation suggests that this basic matching (as an attempt to control for the psychological effects of combat stress) is not totally ideal. The factors of volunteerism and risk-taking behavior must be considered from both the individual and group perspectives. The assessment of individual risk-taking behavior has been quantified by psychological instruments such as the Sensation Seeking Scale (SSS) of Zuckerman, et al . and the Life Experience Inventory (Torrance). The SSS has been demonstrated to have considerable validity in measuring a variety of phenomena including volunteerism and participation in risky activities and has been applied to naval 80 �aviation trainees (Waters). This study was unable to demonstrate an increased accident-related mortality in this group of individuals. D. Response Bias (1) Problem False positive response is anticipated as the primary bias operating in this study. Compensation issues arising from individual claims to the VA or from class action suits, heightened health concern generated by extensive publicity, disenchantment with military service, and the simple desire to please the interviewer may introduce positive responses that exceed the study's ability to correct or adjust. False negative response will also operate, and such bias is even more difficult to assess than the spurious response in a positive direction. Significant factors in this direction include: issues of patriotism and loyalty, personal conviction as to the propriety of the defoliation program and their participation in it, the strong virility orientation of the pilot/aircrew population (particularly with reference to questions of libido and fertility), personal inconvenience caused by study participation, errors of memory, and fear of the adverse effects on career goals that abnormal physical examination results could produce (a significant problem for active civilian and military pilots). (2) Pending Retirement Bias The military retirement system also creates a potential source of bias when personnel who are approaching the end of their careers exaggerate their symptoms so that they may become eligible for disability benefits. (3) Corrective Mea s ure s The primary correction technique for questionnaire response bias will be a carefully constructed and standardized physical examination. Multiple verification and bias indicator questions will be designed and included in the initial questionnaire. Memory verification will be conducted by cross-referencing responses to medical and personnel records. Detailed statistical correlations between the questionnaire responses and the physical examination results 'will be conducted. All interviews and physical examinations will be conducted on a "blind" basis to the maximum extent possible. Self-administered and group-administered questionnaires, which would allow for uncontrolled response changes, will not be conducted. The payment of a $100 per day stipend to all eligible participants will be arranged to increase participation rates. Medical data will not be released to agencies such as the Federal Aviation Administration, and therefore civilian flying activities will not be adversely affected by participation in this study. Models of anticipated biases and their estimated impact on the study will be attempted prior to the final analysis of any phase in order to justify the analytic methods used. Conclusions drawn from this study will be predicted and coupled to a bias estimate. 81 �E. Interviewer Bias (1) Problem Voice inflection, speed of interview, intonation and ethnicity are recognized factors which can affect positive or negative interview response. These factors will definitely operate in this study. (2) Corrective Measures The questionnaire itself will be developed and refined by a civilian contractor. This contractor will assure that the instrument will elicit sensitive personal and medical information in an accurate and efficient manner, while minimizing discomfort to both the subject and the interviewer. All questionnaires will be administered by well-trained and experienced personnel employed by an opinion research organization under contract to conduct this aspect of the study. F. Changes to the Protocol (1) Problem The question of adverse health effects due to Herbicide Orange exposure in Vietnam has evoked many strong emotions. The actions of consumer groups, environmentalists, and other special interest groups have generated defensive responses on the part of some governmental agencies, and reactive decisions by others. Frequently, these responses have been based on unsubstantiated claims and/or scientific evidence of questionable validity. As a result of these governmental actions, the impact on the planning of this study has been substantial. Suggestions to increase the scope of the effort to include other "exposed" individuals or poorly defined ancillary groups continue to surface. However, problems of group ascertainment, exposure validation, control group selection, and control of additional bias make the inclusion of such individuals undesirable from a sound scientific perspective. If such decisions are made without regard for their scientific impact, compromise of study validity is assured. (2) Corrective Measures The scientific groups participating in the extensive peer review process agreed with these concerns. The formation of an effective scientific monitoring group will insure that scientific issues will take precedence over emotional pressures to alter the study design when such changes will limit the scientific validity of the study. The dilution of the scientific credibility of this effort by unscientific decisions will be diplomatically resisted. While all suggested improvements will be considered, any alterations or corrections to the study protocol will be based on sound scientific assessments of the proposed changes. Alterations of the protocol will be made only after careful review and analysis by the principal investigators and the monitoring group. 82 �G- Loss to Study/Statistica l a n d Bias Considerat ions (1) Problem Losses to study in the RANCH HAND group pose a major problem to the validity of the inferences that can be made from any subsequent comparisons between or within groups. The avenues of loss will conceivably arise from individual apathy (volunteer bias), lack of appropriate financial reimbursement for loss of salary, the presence or absence of illness (perception of health), and the lack of a desire for "treatment". Losses of matched controls during the questionnaire and physical examination phases of the study, though predictably greater than in the exposed group, may be managed by replacement from the predetermined set of controls. The estimated participation of individuals is shown in Section XV, Figure A-2. It is estimated that the overall response rate of the exposed group will be 65% in the initial questionnaire and 40% in the physical examination phase of the study. These high non-compliance estimates are expected to occur despite great efforts to keep the questionnaire at an acceptable length, and to coordinate questionnaire administration and physical examination with the subject's personal schedule. Losses to study in either the exposed or control groups will obviously lead to decrements in statistical power, and will raise the possibility of severe bias. Losses from the control group are expected to be greater than losses from the exposed set. Such losses would skew the distribution of controls, (Figure 5) and thus alter the characteristics of the population available for study. If differential losses in the control group occur (i.e., "well" controls dropout more frequently than "ill" controls), a "true" herbicide effect would be diluted (Figure 15). Conversely, if "ill" controls are differentially lost, a spurious effect would be attributed to herbicide exposure. To a lesser extent, losses in the exposed group could create similar effects; however, loss to study in the RANCH HAND population should be much less of a problem then in controls, due to their vested interest. (2) Corrective Me as u res The USAF is committed to expending maximal effort to encourage participation. Loss to study problems in the study participants will be avoided as much as possible by detailed and exhaustive efforts to contact and followup each identified participant. NON-PARTICIPANTS WILL BE STRONGLY ENCOURAGED TO RECONSIDER THEIR INITIAL DECISIONS. Design considerations have been made to minimize loss to study in both the exposed arid control populations. Although the USAF can not fully compensate study subjects for lost wages during the physical examination, transportation costs, per diem, and lodging costs will be reimbursed, and a $100 per day stipend will be paid to all eligible participants. The replacement concept will help to counteract the decrement in statistical power, and offset the bias created by differential patterns of loss. The exposed group is already of maximum size and cannot be increased, but non-compliant controls can be replaced. This w i l l maximize the degree of pairing between the two study groups. If a non-compliant control is replaced by a control with a similar perception of �FIGURE 15 RATIONALE OF REPLACEMENT DILUTIONAL BIAS EXPOSED PRIMARY CONTROLS LOSSES CO P(L/WJ > PjL/1) CONDITIONAL PROBABILITIES: P(L/W)>» L - LOSS W = WELL I = ILL �his own state of health, the alteration of the control group distribution is offset; (i.6i, an " i l l " control is replaced with an "ill" individual, and a "well" control with another "well" individual.) This concept of replacement, coupled with the payment of stipends, and extensive efforts to encourage compliance will minimize losses to study and offset the adverse effects of those losses that do occur. H. Statistical Power Limitations (1) Problem As discussed in Section VI, statistical power considerations are heavily dependent on loss to study rates. Since the design of the study is also limited by the size of the exposed population, statistical power for identifying the relative risk of an uncommon disease or symptom-complex (<1/100) is very low (<.50), (See Section VI. B.). This study will, to a greater extent, be able to detect increased risks in common diseases or symptom-complexes (>1/100). (2) Discussion The "herald sign" of TCDD exposure, chloracrie, is expected to have the greatest likelihood of achieving adequate statistical power in this study. Recent findings from Seveso, Italy, support the importance of chloracne as the primary marker symptom. The incidence of chloracne has been reported by Regglani (personal communication) and Homberger, et al., to be 14.9 cases per 1000 residents in the region of highest contamination of Seveso (Zone A) and 6 to 12 cases per 1000 in the Seveso community as a whole. These rates vary by age group, with children being at highest risk. Only 1 to 5 cases per 1000 were seen in other regions of Northern Italy (Milan, Como, and Lecco). The incidence of adolescent acne in all of these populations varies between 21% and 30%. These incidence rates probably place chloracne at the lower limit of adequate statistical power of this study. In the Nitro, West Virginia studies, residuals of chloracne, as well as exacerbations of previously active disease, continue to be seen 10 years after the most recent exposures, and 30 years after the industrial accident. Thus, it is likely that any chloracne in the exposed population may be detected, despite the intervening years since RANCH HAND exposures. In addition to chloracne, other recently reported human effects of TCDD exposure at Seveso, Italy, appear to fall within the capabilities of this study design (e.g., peripheral neuropathy, neuropsychiatric effects, and liver dysfunction). In general, with respect to statistical power, continuous data (clinical or laboratory measurements) even from relatively small samples fair much better than either categorical or dichotomous data (presence or absence of a given condition). Consequently, a concerted effort will be made to obtain physical examination data in a scored and/or continuous manner. 85 �I. Variabi1ity of Procedu res (1) Problem The variation of physical examination findings from differences in technique and the random errors inherent in laboratory testing are items of concern, particularly if attributable health effects are subtle or of low magnitude. Nonstandardized procedures and techniques are major contributors to this variance. (2) Corrective Measures Variability in examination procedures will be minimized by the use of standardized procedures, examination protocols, on-site monitors, and training. All laboratory procedures will be conducted at the examination center and quality control will be stressed at all times. (See Section IX) J- Confounding Exposure Factors (1) Problem While virtually all of the media attention has been directed toward the 2,4,5-T-containing herbicide formulations, other herbicides were applied concurrently by the C-123 aircrews in Vietnam. Herbicide Blue (Cacodylic acid with 15.4% pentavalent arsenic) and Herbicide White (2,4-D and Picloram) were used throughout the 1962-1970 time period. Any long-term health effects from these additional compounds may confound the results of the study. Peripheral neuritis, tremors, skin and lung cancer, loss of hair and nails, skin rashes, and gastric symptoms have been alleged after exposure to arsenical pesticides. The organophosphate insecticide Malathion was also sprayed by some of these same aircrewmembers when RANCH HAND duties permitted their temporary assignment to mosquito/malaria control missions. Many of these individuals were involved in the aerial spray application of these and other pesticides both before, during, and after their Vietnam service. Longterm effects from these chemicals would confound the study results. The small size of the RANCH HAND population will allow very little opportunity for analytic stratification for these confounding variables. Differing patterns of exposure to aircraft fuels in the study populations have been suggested as confounding factors. The C-130 aircraft were powered by turbo-prop engines which used jet fuel (JP-4), while the C-123 and C-7 aircraft were powered by standard reciprocating engines which used leaded aviation fuel (AV-GAS). After June 1968, many C-123s were modified by the addition of auxilliary jet engine boosters for added power on takeoffs and in emergencies. (2) Discussion and Corrective Measures While the extent of confounding caused by exposure to these other pesticides is undetermined at this time, assessment of its magnitude must rely on responses of the subjects to that portion of the questionnaire dealing with other occupational exposures. For this reason, information 86 �concerning exposures to other herbicides/insecticides used in Vietnam will be collected. Whenever possible, stratification techniques will be used to adjust for these confounding variables during data analysis. Variations in fuel between C-130 and C-123 aircraft would be significant factors if individuals in the study were heavily and repetitively exposed. However, the normal duties of the study participants did not involve aircraft refueling or other fuel handling activities. Thus, fuel exposures can be minimized as significant confounding factors. 87 �IX. Quality Assurance and Management Considerations A. Quality Control (1) Overview As in any major scientific effort the quality of the data and the comparability of the data over time are key factors in achieving valid results. Quality assurance in both scientific and management aspects of this study are planned, and will be fully integrated into each phase of the study. (2) Sc i entif i c As pect s (a) Protocol Development The Air Force scientific protocol has been under development for more than one year. It has been subjected to an unprecedented five stage independent peer review process to insure the highest quality and validity of its science. (b) Blind Assessment Protocols The exposed or non-exposed status of each individual will not be revealed to any of the Health Examiners. Each aspect of the physical examination will be conducted by rigid adherence to the examination protocol. Past medical history and review of systems will be obtained by individuals not associated with the examining process. (c) Population Ascertainment Quality Control The study/control populations for this effort were ascertained through extensive computer, and hard copy record searches. The matching variables for each individual were entered and verified with a computer program to minimize transcription errors. Data collection for both exposed and control populations was conducted using identical techniques, thus avoiding systematic bias in population ascertainment. (d) _P reel si on Mate hi ng Computer techniques will permit extremely close matching of the control participants to the RANCH HAND participants for three distinct variables. This will substantially enhance the analytic flexibility and validity of the study. (e) Questionnaire Techniques Detailed questionnaire methods are under development to provide comprehensive crosschecks between objective and subjective health information. Particular emphasis will be placed upon techniques to ascertain false positive information which might impact the validity of the study. 88 �(f) Laboratory J[)ua1Jt,y Control The contractor for acquisition of health data mandatorily must have a detailed in-house laboratory quality control program coupled with enrollment into the "CLIA" or "CAP" laboratory survey. In addition, randomly selected duplicate specimens will be sent to a central Air Force reference laboratory for verification. (g) Single Physical Examination Site All physical examinations conducted by the contractor will be performed at a single site by dedicated teams of health professionals to insure that data variability is at an absolute minimum. The contractor will be a fully accredited medical institution, and must provide organizational evidence of national/international preeminence. (h) Re r son n el Qu a 1 i f i ca t i on s All examining physicians will be certified and .accredited by a Medical Specialty Board. Paramedics, medical students and interns will not participate as examiners in this study. (3) Management Aspects (a) Informed Consent All participants will be fully informed as to the nature and purpose of all medical diagnostic tests and examinatjpns, .and will certify their complete understanding by signing specially designed informed consent forms. Release of medical data will be in strict accordance with Privacy Act determinations, and Air Force policies. Total confidentiality will ;b.e granted to subjects who are not on active duty. Active duty sutej.ects will be given limited confidentiality with release of medical informatl<Hi fco the ;DOP only in instances in which there is a risk to public safety or national A monitoring group of scientists and p,er$pnn£l outside the USAF will regularly review and assess the con.d:WCt of the RANC.H :HA'N.P study* [his group will interact closely with the Air Force principal investigators, and will provide written commentary and recommendat!ioifi;S directly /to the .White i louse Office of Science and Technology Policy,. Apprpxi'Kist<6ily equal representation will be maintained between government scientists., ^C-ademic •.scientists, and scientific personnel nominated by veterans advocacy groups, (c) Consultants In addition to the $truct<ur.e£l Air Farpe ma:ntgerDgnt system, outside management and scientific consultants will -be utiiliz-ed fop .provide assistance to the principal investigators upon request. 89 �(d) Contract Performance All data acquisition contracts will contain highly detailed schedule performance requirements. All statements of work will be coordinated with two procurement levels, appropriate Air Force program coordinators, and the outside monitoring group. (e) On-Site Contract Monitor An Air Force Medical Service officer will be assigned to the physical examination site to: (1) provide visible Air Force representation to all participants, (2) conduct detailed entry and exit briefings with all participants, particularly ensuring that the health assessment was conducted on a "blind basis", (3J review all medical data for completeness and accuracy prior to computer entry, and (4) examine all relevant features of the data acquisition process, and insure absolute compliance to the contract specifications. (f) Data Security All medical information obtained on each participant will be entered into a computer data respository. Access to "these data will be limited to key scientific investigators by master code numbers. B. Management Structure (1) General Organization Standard Air Force Systems Command research and development concepts and organization will be used to manage this study and assure effective control of all phases of the investigation. The organizational structure is outlined in Figure 16. (2) Functions (a) Program Element Monitor (PEM) of the Program primary support the Air The tasks of the PEM will be preformed by a representative USAF Surgeon General's staff. The PEM will serve as the Air Staff Monitor, and as such, he will represent the needs and interests of the investigators to the Surgeon General and the Air Staff. He will the needs of the study to the Deputy Chiefs of Staff, the Secretary of Force, the Secretary of Defense, and Congress. 90 �Figure 16 MANAGEMENT STRUCTURE Program Element Monitor HQ USAF/SG Representative AFSC Systems Officer HQ AFSC/SG Representative Program Office USAFSAM AMD Representative USAFSAM Principal Investigators (Technical Contract Monitors) Det 1 AMD or TOY Rotation On-Site Contract Monitor (Quality Control-Physical EXAMS) (b) Systems Officer (SYSTO) The SYSTO will serve as the Program Manager at the Air Force Systems Command level. In this capacity, he will monitor program status, key issues, and problems. He will also serve as coordinator and expediter between the PEM and the primary investigators. Additionally, the SYSTO will prepare program documentation, coordinate all aspects of the program, monitor obligations and expenditures, and Initiate reprograrnming actions to support unfunded study requirements. (c) Prog rami J3ffice The Program Office will be staffed by a representative of the primary investigators and an Aerospace Medical Division (AMD) representative. This office is responsible for implementation of the complete program 91 �management plan on a day-by-day basis. Routine periodic management assessments and program status information will be provided to the SYSTO. The office will assure that all professional and technical aspects meet the stringent quality requirements outlined in the study protocol. It is the responsibility of this office to insure that all schedules, milestones, and financial requirements are met. This office also interfaces with, and provides guidance and support to the onsite contract monitor(s). (d) USAFSAM Principal Investigators/Scientists This team is the leading technical resource for this program. Members of this team are responsible for the faithful execution of the protocol, and as such, approve/disapprove all protocol changes, working in concert with the outside monitoring group. The principal investigators are the tecnhical monitors on all contracts under the protocol. They are responsible for the security of all data, for all data analysis, and for all interpretation of analyses subject to review by the outside monitoring group. These investigators provide summary data to Air Force management personnel on request, to enable proper contract billing and program resource analysis. The primary flow of data, data analyses, and analysis interpretation from the principal investigators/scientist directly to the monitoring group is designed to obviate any appearance of Air Force management bias. (e) Onsite Contract Monitor (Physical Examination Contractor) The onsite monitor will act as the Air Force representative at the examination site. He will monitor and assess the quality and timeliness of the contractor's performance, and will advise the Program Office of any performance decrements, as well as other problems encountered at the examination site. He will be responsible for the quality control of all aspects of the examination process (physical examination, laboratory procedures, and psychological and physiological testing). He will also welcome each study subject, review the results of the complete evaluation, and debrief each subject at the conclusion of the examination process. 92 �X. Repprti ng Procedures Interim synoptic progress reports will be provided to the Surgeon General through Quarterly Management Reviews conducted each January, April, July and October. Key data analyses will be displayed, but inferences and conclusions will await full data analysis at the conclusion of each phase. A formal report for each of the three phases will be completed with forecasted submission dates of: Mortality Study, June 1982; Morbidity Study, June 1983; and Follow-up Study, June 1985, 1987, 1992, 1997, 2002. Findings and conclusions of each phase will be published in a journal of stature. Total study design, findings, and conclusions will be published in the USAFSAM Aeromedical Reviews or Technical Reports. 93 �XI. QUESTIONNAIRE The release of the actual questions within the questionnaire could possibly result in irreparable damage to the study from an avoidable source of responder bias. Consequently, this section provides a summary of the general subjects to be covered on the questionnaire and a brief discussion of those specific areas that will receive particular emphasis. The questionnaire will, of necessity, be lengthy, but it will be administered at a time convenient to the subject. Subjects who refuse to participate in a face-to-face interview will be encouraged to cooperate with modified questionnaires given by telephone. The questionnaire will verify personal identification data such as name, SSAN/AFSN, date of birth, address, telephone numbers, race, military status, effective date of status, location of military medical records, and marital history information. RVN tour information will be rechecked and expanded to include specific data such as date of tour, tour end date, AFSC, organization of assignment, PCS and TOY status, combat missions, and whether or not the tour was a RANCH HAND affiliated tour. Pre- and Post-RVN exposure information, both occupational and avocational, to asbestos, radiation, herbicides, pesticides, and carcinogens will be elicited. Data concerning the frequency and duration of these exposures are very important. RVN exposure to these chemical and physical agents will also be collected. Medical information obtained during this interview will include a statement of general health, smoking history, alcohol consumption history and longterm medication/drug use. In addition, questions dealing with infertility, birth defects of offspring, as well as the wife's obstetrical history (i.e., total conceptions, live births, miscarriages, stillbirths and premature pregnancies) will be obtained. A family history emphasizing cancer, heart disease, liver disease and inherited disorders in both the subject's and spouse's families will be collected. A comprehensive medical inventory will be included emphasizing the neurologic, dermatologic, reproductive, and hepatic systems. At the time of the physical examinations, each subject will be given a comprehensive face-to-face medical history which will expand and verify the health information that was obtained in the initial questionnaire and records review. An extensive review of systems will be covered at that time, including an extensive occupational and avocational exposure history. Just prior to the time of follow-up adaptive physical examinations, a preliminary telephone contact will establish the subject's current health status and his willingness to continue participation in the study. Appointments for the follow-up examinations will also be arranged. Adaptive questionnaires will be given emphasizing those symptoms and systems that were found to be significantly associated with the exposed population on analysis of earlier study results. If the subject expresses a desire to cease participation at 94 �this time, he will be encouraged to reconsider his decision, and the reasons for dropping out of the study will be sought. At the time of subsequent followup evaluations, subjects who have left the study will be given the opportunity to rejoin the study. 95: �XII. Physical Examination Design A. General Instructijgns This phase of Project RANCH HAND II is a cross sectional study of the subject's health at the time of examination. The physical examination and all required laboratory procedures will be performed by physicians and technicians at a major civilian medical center under contract to the Air Force. It is important that examiners remain unaware of the subject's status as a RANCH HAND participant or as a control subject. The physician examiner is tasked to examine and objectively record his findings. The examining physician is not, and cannot be expected to arrive at any definitive diagnosis, as the full history and laboratory results will not be available to him. Medical history, laboratory results, and physical examination findings will be evaluated by an independent diagnostician employed by the contractor. This diagnostician will formulate diagnoses and differential diagnoses, if appropriate. In addition, he will present a detailed analysis and debriefing to the study subject, and provide a copy of the analysis to the subject's personal physician, if so requested. If, during the examination, the physician discovers evidence of serious illness requiring immediate treatment, the normal emergency or urgent care procedures of the medical facility would apply. Such care will be arranged by the diagnostician and will be supplied by the contractor at Air Force expense. If during the examination, evidence of illness requiring nonemergency medical attention is found, the diagnostician should infonn the subject and offer to have forwarded pertinent information to the subject's physician. A clear record of any such advice and treatment should be recorded. The ultimate value of the RANCH HAND II Study will lie in the collection of complete, accurate and, whenever possible, quantitiative data permitting the most stringent and powerful statistical analysis. For that reason, the physical examination protocol requires exact measurements in many instances, and the use of defined meanings of semiquantitative indicators in other places. These examinations will define the health status of the subjects at a point in time, and will establish the presence or absence of abnormal physical findings. After statistical review of the study groups, these findings may permit definition of a chronic effect due to exposure. An inaccurate examination may lead to falacious study results in two ways: a presumed syndrome may be defined which does not in fact exist, or a syndrome which in fact exists may not be defined with enough validity to warrant further actions. 96 �B. Conduct of the Examination SUBJECT HVMBRH SECTION PHYSICAL EXAMINATION t. OENF.HAL APPEARANCE a . Appearance/Stated Age f*$ Younger Than £70Tder Than ' jd Same As b'. f^J Well-nourished f~J Obese £7 Under-nourished £j Older Than C. Appearance of illness or distress / 7 Yes £7 No " d- Hair Distribution / / Normal / 7 abnormal " " SPECIFY: 2. HIIHHT SITTING BLOOD PRESSURE RIGHT AHM. AT HEART LEVEL WEIGHT fUn Cltl j. PULSE RATE R E G U L A R : [ _ | YES 3YSTOLIT . . . . . . . j -, 01ASTOUCDescribe any IrregulaHttet. Q] NO a. Irregular /*/ b. Irregularly irregular / 7 " c. VPBs per minute _ «. EYE GROUNDS f*"] NORMAL f~T_ABNORMAL ARCUSSENILIS 8. EN T T.l P«^NT f~ J NORMAL Describe any vascular loslons, hcmoirbagen, exudatea, £7 Hemorrhage's p«P»u«d«««. /"7 Exudates ^7 Papilledema f~J Disk Pallor f~J f> Cupping £2 A-V nicking" /~7 ^ light reflex f~~f Arteriolar spasm 5 QABMHT [~] ABNORMAL «- Describe any abnomi*llty. Tympanic membranes intact /~7 Yes /~7 No Nasal ulcerations /~7 No /~7 Yes 7. NECK ffi.p.ci.ili- (dvro/d «|««4 |_| NORMAL Thyroid gland palpable f~7 Enlarged 7y Nodules /V Tenderness 7 8. THORAX AND LUNGS 9. M E A N T ( " J NORMAL R £V | J ABNORMAL.^ L /"7 Describe any abnomalUy. f~~f Parotid gland enlargement /7 R £7 L . ["] NORMAL Asymmetrical expansion Hyperreconance Dullness Abnormal OciOar^igmentation Q] ABNORMAL f~7. Wieezes /""7 Rales ['^ABNORMAL Descrtb* «ny abtiowullty, sspuctnlly bulUr rolos. qircumferena^ at nipple level Expiration ,qm Inspiration _^ cm Describe sny enlargement, Irregularity of rat*, munmure, «t Ihrtllf. Displacement of apical impulse £J No f~J Yes Precordial thrust /^7 No /"7 Yes Heart sounds normal /"7 Yes /~7 No /~7Sl /^7Sg £783 jContinuof? in I too 18 on Reverse) 10. AOOOMPN PI NORMAL £7 Heptomegaly " cm Liver Span £7 Splenomegaly M. rxTn£MirifS I"*] NORMAL £7 Absence, specify: £7 Edema " "[^"J ABt*olMAL D*«ciibe at>y dbnotwaUty V»H*i special attentlon-lo th? £7 Other mass - tpi««« •bdiiYw. Record waist measurement Specify: on attached form. £7 Tenderness /"7 Liver £7 Spleen /7 Other, specify: [fj] ABNORMAL f~J Pitting /~7 Non-pitting /~7 Loss of hair on toes Describe any edema of signs of vascular insufficiency, £7 Clubbing of nails £7 Varicosities 97 �SECTION PHYSICAL EXAMINATIOH <C<x,tlm»<l) 2. P E R I P H E R A L P U L S E S RADIAL FEMORAL POPLITEAL 3 O R 5 A L I S PCDI3 POSTFnion TlfJIAU ; ABNORMAL /7 ~ /7 ~ /7 ~ [~J t~~t indicate type and location of lesions on the attached anatomical . fiaure £ Hyperpigmentation 7 ^-^I'almar Keratosi* /~7 Jaundice £7 Petechiae £7 Spider angiomata £7 Ecchymoses £7 Palmar erythema / / Soles of feet l)ormatoj>raphia Comedones Acneiforra lesions Acneiform scars Depigmentation Inclusion cysts Cunis Rhomboidalis Full-Face and Bilateral profile photos taken '~'J^,ilS i _ i NORMAL '^"ABNORMAL £7 Muscle - Specify: / 7 Spine ~ £7 Weakness / 7 Scoliosis ~ £7 Tenderness / 7 Abnormal Consistency ~ / 7 Tenderness, ~ /~7 Atrophy Level Decreased range of motion Biopsy Taken 14. M U S C U L O S K E L E T A L / 7 Pelvic tilt ~ ^ Straight Leg 7 • 15. GENITOURINARY - RECTAL - HERNIA £7 NORMAL f~7 Inguinal hernia /~7R /^/L /~7 Tastes Absent Enlarged Atrophic /I7PCJ rj rj /7 !L £7 £7 £7 £T ABNORMAL. /~7 Varicocele /~7 Epididymis £7 Scrotal Mass ^ dia / 7 Hemorrhoids ~ / 7 Prostatic ~ Enlargement /7 Rectal mass 16. IA'>yPH MODES - CHECK ALL AREAS. / 7 NORMAL' £7 ABNORMAL - SPECIFY CERVICAL, ~ OCCIPITAL, SUPRACLAVICULAR, AXILLARY, SPITRAaiLEAR, INGUINAL, FEMORAL / 7 Enlarged ~ £7 Tender £~7 :!ard / 7 Fixed ~ £V Confluent 17. NERVOUS SYSTEM - SEE'ATTACHED FORMS OBSERVATIONS 18. HEART -W (Continued from Item 9) Murmur / 7 No f~J Yes Area £7 ~ s-/s £7 Ao £ pu 7 £7 Z 7 APex I £7 Dia 98 �CLINICAL RECORD NEUROLOGICAL EXAMINATION HEAD AND NECK - Normal to Palpations/Inspection £7Y £7N Specify Scar £7 Asymmetry £7 Carotid Bruit £7No £JR £7L Neck Range of Motion £7 Normal or Decreased to £7 Left Depression £7 £7 Right £7 Forward £7 Backward TRUNK MOTOR SYSTEM - Handedness Right £7 Left £7 Gait £7 Normal or £7 Broad Based £7 Ataxic £7Small Stepped £70ther-Specify Associated Movements £7Arm Swing £7Normal or Abnormal ,£7R £7^ Muscle Status (strength, tone, volume, tenderness, fibrillations) Bulk £7 Normal £7 Abnormal Tone Upper Extremities £7N0rmal or £7Increased £7Decreased £7Right £jLeft Lower Extremities £7Normal or £7"Increased ,£7TJecreased £7Right £7Left Strength - Distal wrist extensors £7NontlJi'l £70ecreased Ankle/Toe Dors/Flexors £7Normal £7Decreased £JR £7L Proximal Deltoids £7Normal £7Decreased £7& Hip Flexors £7Normal £7Decreased £7L £7& £7L Abnormal Movements (tremors, tics, choreas, etc,) Fas 1cu lotions .£7N.o Tenderness £7No £7Yes (Ii4+) Tremor £7No £7Yes (1-4+) £JYes - Specify Upper Extremity £7R £7D £7Resting £7Essen:t1a1 r£7;Intention ther Lower Extremity £7R £7L/O° Coordination (a) EquiTibratory - Eyes Open Eyes Closed - Romberg £7Positive (Abnormal) Right Foot (b) (c) Left Foot Monequilibratory (F to N; F to 'F; H to K:) :F:inger-to-no.SB-to-ftnger £7Nort"a' £7 Abnormal £7R:ight ^ OLeft £7Both. . . „„. . Heel-Knee-Shin £7Normal £7«bnornial £7:RighT ,T7Le'ft £7Both Succession Movements (including check, rebound, posture-ho'lcnng) If indicated, check £7Normal £7Abnorma1 £J.R ,£7R Rapidly alternative movements £7Normal S k i l l e d Acts ( ) Handwriting. ( ) £7Nega:ti'Ve (Normal) If indicated, £7Norma1 £7;flbn.orma'l ,£7R .-£7'£7Abnorma1 Speech (articulation, aphasia, a.gnosia) Grossly £7Normal £7Abnormal - Specify Dysarthria ./~7 Aphasia .£7 •99 �Reflexes (0-absent; 1-sluggish; 2-active; 3-very active; 4-transient clonus; 5-susta1ned clonus) R Deep L Deep R R Other L L Abnormal Babinski R L Patellar Biceps Triceps Achilles Remarks MENINGEAL IRRITATION Straight Leg Raising £7Norma1 NERVE STATUS (tenderness, tumors, etc.) £7Abnormal £71 £7Both SENSORY SYSTEM (tactile, pain, vibration, position. If positive sensory signs are present, summarize below and indicate details on Anatomical Figure, Std. Form 531) Light Touch £7Normal Pin Prick £7Normal ^/Abnormal ^Abnormal (Map on Anatomical Figure) Vibration (at ankle, 128 hz tuning fork): £7Normal OAbnormal £7R Position (Great toe): £7Normal /^Abnormal £JR £7L £7Both CRANIAL NERVES I R Smell II III /^/Present /^Absent L Smell £JPresent £JAbsent Fundus R Normal £7' Abnormal /~7 Disk Pallor/atrophy £7Exudate £7Papi 11 edema Fundus L Normal £7 Abnormal £7 Disk pallor/atrophy £7Exudate £7 Papi 11 edema £7Hemorrna9e Fields (to confrontation) Right £7Normal /^Abnormal Left £7Normal ZI7Abnoninal Normal ~ Specify Unequal £7 Difference mm oth er £j /"/R 7~7L Abnormal £j Pupils-Size (mm) Equal Shape, position Round Light, Reaction Normal Position of Eyeballs Movements R Nystagmus Rotary £7 (Draw position) Horizontal £7 100 Vertical £7 �XI Ptosis R£7 L £7 • V Motor R Clench Jaw - Symmetric £7 Sensory R Normal f~J L Normal £7 Corneal Reflex Deviated £7 Abnormal /~7 Vi/~7 Abnormal £7 V R R/~7 L£J Vo L Yes No VII Motor R Normal smile £7 /I7 L Normal smile O Ves O'No Palpebral Fissure £7Y.es £JNo Palpebral Fissure OYes £7No IX Palate and Uvula X Movement Normal £J Palatal Reflex Deviation to OR R £7Normal /^Abnormal L £7Normal /~7 Abnormal XII Tongue-Protruded-Central £7 R £J L £7 Atrophy /^JNo £JYes MENTAL STATUS (alert, clear, cooperative, etc.) £JYes - Specify 10! Gross abnormalities: �DIAGNOSTIC SUMMARY SYNOPSIS OF POSITIVE FINDINGS Medical History: Physical Examination: General Dermatologic Neurological Psychological Laboratory Results: Diagnosis: Differential Diagnosis, if applicable: Date Signature of Diagnostician 102 �C. Spec i a1 P rocedu res (1) Nerve Conduction Velocities (NCV) These studies have been determined to be an important parameter in long-term follow-up studies of persons thought to have been exposed to Herbicide Orange components. The Nerve Conduction Velocities should be performed by a physician or by a specialty qualified technician under the supervision of a physician trained in neurophysiological methods. (a) Specific NCVs ( ) Ulnar Nerve (one side only) 1 (a) motor (above elbow, below elbow) (JD) values recorded (i) distal latency (ii) NCV (2) Peroneal Nerve (one side only) (a), motor (b) values recorded (i) distal latency (ii) NCV (1) Sural Nerve (one side only) (aj sensory: orthodromic (b) values recorded: NCV (b) Methods Standardized, published methods will be used (e.g., Smorto, Marcio P., and John V. Besmajian; Elecj:rod1jgnos1 S; Harper and Row; NY, 1977). (2) Psychological Test Battery (a) General This battery yields objective numerical da.ta, and is wellstandardised and clinically validated. The individual tests were chosen to insure an adequate analysis of one of the major alleged manifestations of 103 �Herbicide Orange toxicity. Each test either validates the other tests or is considered to be a "definitive" test for analysis of a suspected psycho-neuropathic effect under study. Compared to the general civilian population, characteristic response tendencies are observed on the MMPI and Cornell Index among active duty aircrewmen being evaluated in an aeromedical setting. It is also important to consider the effect that pending retirement has exerted on the reporting of medical history and symptomatology. This may also alter responses to psychological testing. (b) Specific Tests (D Nechsler Adult Intelligence Scale (NAIS) Individually-administered collection of verbal and nonverbal intellectual measures; also useful for clinical inferences when combined with the neuropsychological battery below. (2.) Reading subtest of the Wide jange Achievement Test (WAIT) Individually-administered measure of word recognition ability. Important to rule-out reading inefficiency should the response to the personality instruments below be of questionable validity (e.g., high F scale on MMPI). (1) Halstead-Reitan Neuropsychological Test Battery Individually-administered collection of brain behavior relationship measures for establishing the functional integrity of the cerebral hemispheres. The battery must include the following subtests: Category, Tactual performance, Speech-Sounds, Seashore Rhythm, Finger Tapping, Trail Making, and Grip Strengths. The Aphasia Screening and Sensory-Perceptual Exams are considered optional in view of their redundancy with the clinical neurologic exam included in this project. Individualized test debriefing is conducted to clarify test performances in the WAIS and Neuropsychological Battery. (D Three subtests of the Wechsler Memory Scale I (MIS I) Individually-administered measures of immediate and delayed recall of verbal and visual materials. The Logical Memory, Associate Learning and Visual Reproduction subtests are to be administered in the standard, immediate-recall fashion initially. After 30 minutes has elapsed, the examinee is asked, without prior alerting, to recall as much as he can about the Logical Memory and Visual Reproduction subtest stimuli. Standard scoring is used for both test-retest administrations. 104 �(5) Cornell Index (CD Self-administered and standardized neurppsychiatric symptom and complaint inventory, including items involving asthenia, depression, anxiety, fatigue, and GI symptoms in lay language. Endorsement of items are to be explored and clarified in test-debriefing, (6) Minnesota Multiphasic PersonalIty Inventory (MMPI) 60 to 90 minute self administered clinical psychiatric screening instrument; also capable of estimating response biases (e.g., "fake good," or "fake bad"). The shortened version of Form R (i.e., items 1 to 399) may be substituted for the 566-item Long Form. Standard scoring and Minnesota norms are to be used, with the possible exception of active duty examinees where USAFSAM aircrew norms may be applied. Clarification of profiles showing response biases, questionable validity, and/or unusual item endorsements will be conducted in individual test debriefing. (3) 12-Lead E1 ectrpeardlogram (a) Procedures A standard 12-lead scalar electrogram is required. If an arrhythmia is observed, a one minute rhythm strip will be obtained. The electrogram will be done following a minimum fast of fpur hours, (b) Interpretation The electrocardiograms will be interpreted logists at the examining center, and then forwarded to Brooks AFB cians in the USAF Central ECG Library will compare the tracing individual ECG records in the case of rated (pilot or navigator) by cardiowhere physi*to previous subjects. (c) Disposition (USAF Central |CG Library) (1) Pilots and Navigators The original tracings will permanent record established for each individual. (2) Enlisted Subjects The original tracings will permanent record established for each individual, (4) be micrqfisched and be microflsched antf a Radi ographi c Exami nati on A standard 14x17 in., standing* rpentgenograrn position using small nipple markers will be accomplished. 105 In the PA �(5) Pulmonary Function Studies Standard evaluation of vital capacity and forced expiratory volume at 1 second will be performed. (6) Laboratory Procedures (a) Specific Tests to be Performed on all Participants (1) Hematocrit (2) Hemoglobin (J3) RBC Indices (4) While Blood Cell Count and Differential (5) Platelet Count (6_) Erythrocyte Sedimentation Rate (7_) Urinalysis (8_) Semen Analysis (Number, % Abnormal, Volume) (9_) Blood Urea Nitrogen (10) Fasting Plasma Glucose (11.) Creatinine (12) 2-hour Post Prandial Plasma Glucose (13) Differential Cortisol (0730 and 0930 hours) (14) Cholesterol & HDL (J.5) Triglycerides (16) SGOT (J7) SGPT (18) GGTP (19) Bilirubin, Total and Direct (20) Alkaline Phosphatase (21) LDH 106 �(22) Serum Protein Electrophonesis (_2J3) CPK (24) VDRL (2^) LH (26) FSH (27) Testosterone (28) Thyroid Profile (RI.A) (13, 74, TSH.FTI) (_29) Delta-aminolevulinie Acid (30) Urine Porphyrins (31) Hepatitis B antigen/antibodies (32) Prothrombin time (33) Blood Alcohol (b) To be performed on selected subjects (1) autoimmune disorders Anti-nuclear Antibody on subjects with indications of (2) Hepatitis A Antigens/antibodies for those with current or past history of VTver disease (3) Karyotyping for those fathering children with birth (4) Skin photography aind skin biopsy on subjects with defects suspected chloracne (J5) To be performed if medical history indicates a subject has an increase in infectious diseases; (a) Immunoelectrophoresis (b) Quantitative Imngnoglobylin Determinations 107 �() 6 To be performed on a randomly selected group of subjects (a_) Enumeration of B and T cells (b) Enumeration of Monocytes (c_) B and T cell function tests (7) Rationale for laboratory procedures (a) Studies on the toxicity of TCDD in animals have shown that the following organ systems are damaged: (1) Liver: Hepatic necrosis, liver enzyme changes, hypoproteinemia, hypercholesterolemia, hypertriglyceridemia. (2) Reticuloendothelial System: cellular immunity, decreased lymphocyte counts. (J3) Hemopoietic leukopenia, pancytopenia. System: (4j Endocrine System: cortex, hypothyroidism. Thymic atrophy, altered Anemia, thrombocytopenia, Hemorrhage and atrophy of adrenal (5) Renal: Increase in blood urea nitrogen. (j>) In addition, statistically significant increases in hepatocellular carcinomas (liver) and squamocellular carcinomas of the lung were found. (b) Studies on the toxic effects of TCDD in man have shown that the following organ systems are damaged: (1) Skin: Chloracne, hirsutism. (2) Liver: Porphyria cutanea tarda. Increased levels of transaminase and of 6GTP. Enlarged, tender liver, hyperlipidemia. (_3) Renal: Hemorrhagic cystitis, focal Pyelonephritis. ( ) Neuromuscular System: 4 Asthenia, i.e., headache, apathy, fatigue, anorexia, weight loss, sleep disturbances, decreased learning ability, decreased memory, dyspepsia, sweating, muscle pain, joint pain and sexual dysfunction. (J5) Endocrine System: Hypothyroidism. 108 �(c) Based upon the reports of toxic effects in animal and human exposures, the following organ panels were thus recommended: (1) Hemopoietic (2) Reticuloendothelial (3) Renal (_4) Endocrine (J5) Neuromuscular (dj Hemopoietic screening should include: (1) Hematocrit (2) Hemoglobin (3) RBC indices ( ) Erythrocyte sedimentation rate 4 (J5) Platelet count (6) Prothrombin time (e) Reticuloendothelial system: (_!_) White blood cell count (2) Differential (J3) Serum protein electrophoresis (_4) Selective use quantitative immunoglobulin determination of immunoelectrophoresis (5) B cell and T cell counts and functions (f_) Hepatic screen: (1) SCOT (2) SGPT GGTP 109 and �(£) Bilirubin, Total and Direct (j>_) Alkaline phosphatase (6) LDH (2.) Cholesterol (8) HDL (9) Triglyceride (10) Urine prophyrins (11) Urine porphobilinogen (12) Hepatitis B antigens/antibodies (g) Renal screen: (_1) Urinal ysis (2) BUN (3) Creatinine (h) Endocrine screen: (_!) Differential cortisol (0730 and 0930 hours) (2) Thyroid profile (RIA) (3) Fasting plasma glucose (i) Neuromuscular system: (I) CPK (j) Elucidation of sympoms of asthenia: (1) Testosterone (2) LH FSH 110 �(k) The following tests should be peformed only as follow-up for abnormalities in the history or physical examination findings: (1) HAVAB (IgG and IgM) (2) ANA 111 �XIII. PRINCIPAL INVESTIGATORS, CO-INVESTIGATORS AND CONTRIBUTORS A. PRINCIPAL INVESTIGATORS George D. Lathrop, MD, MPH, PhD Colonel, USAF, MC Chief, Epidemiology Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 William H. Wolfe, MD, MPH Lt Colonel, USAF, MC Chief, Disease Surveillance Branch Epidemiology Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 Richard A. Albanese, MD, GS-15 Chief, Biomathematical Modeling Branch Data Sciences Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 Patricia M. Moynahan, BSN, MS, Lt Colonel, USAF, NC Chief, Occupational Epidemiology Branch Epidemiology Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 B. CO-INVESTIGATORS Alvin L. Young, BS, MS, PhD Major, USAF Special Assistant on Military Herbicides Department of Environmental Medicine Veterans Administration 810 Vermont Ave NW Washington DC, 20420 Clarence F. Watson, Jr., MD, MPH Colonel, USAF, MC Chief, Clinical Sciences Divsion USAF School of Aerospace Medicine Brooks AFB, TX 78235 112 �James A. Wright, MD, MPH Major, USAF, MC Assistant Chief, Disease Surveillance Branch Epidemiology Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 Joel E. Michalek, PhD, GS-13 Mathematical Statistician Data Sciences Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 Phelps P. Crump, PhD, GS-15 Chief, Consultation & Training Branch Data Sciences Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 Richard C. McNee, MS, GS-14 Chief, Advanced Analysis Branch Data Sciences Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 Alton J. Rahe, MS, GS-13 Mathematical Statistician Data Sciences Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 Michael A. Sauri, MD, MPH & TM Major, US Army, MC Department of Internal Medicine Brook Army Medical Center Fort Sam Houston, TX 78234 Richie S. Dryden, MD, MPH Lt Colonel, USAF, MC Chief, Flight Medicine Branch Clinical Sciences Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 113 �C. CONTRIBUTORS Thomas V. Murphy, MBA, GS-9 Statistician Disease Surveillance Branch Epidemiology Division USAF School of Aerospace Medicine Brooks AFB, TX 78235 Robert A. Bottenberg, PhD, GS-15 Chief, Computational Sciences Division AF Human Resources Laboratory Brooks AFB, TX 78235 Jimmy D. Souter, BA, 6S-14 Chief, Analysis & Programming Branch Computational Sciences Division AF Human Resources Laboratory Brooks AFB, Tx 78235 Calvin C. Fresney, GS-12 Chief, Data Base Management Section Computational Sciences Division AF Human Resources Laboratory Brooks AFB, TX 78235 Henry W. Clark, BA, GS-12 Supervisory Computer Systems Analysist Computational Sciences Division AF Human Resources Laboratory Brooks, AFB, TX 78235 Phillip G. Brown, BS, MS Major, USAF, BSC Assistant for Bioenvironmental Engineering Office of the USAF Surgeon General Boiling AFB, MD 20332 114 �xlv A. • SELECTED BIBLIOGRAPHY (Key References) GENERAL REFERENCES Meek, Stephen L., An estimate of the relative exposure of U.S. Air Force crewmembers to agent orange. Masters degree thesis, University of Washington, School of Public Health and Community Medicine, Seattle, 26 February 1981. Young, A. L., J. A. Calcagni, C. E. Thalken and J. W. Tremblay. 1978. The toxicology, environmental fate, and human risk of Herbicide Orange and its associated dioxin. USAF Occupational and Environmental Health Laboratory, Brooks Air Force Base, Texas. Technical Report OEHL-TR-78-92. 247 p. International Agency for Research on Cancer. 1978. IARC Internal Technical Report No. 78/001. Coordination of epidemiological studies on the longterm hazards of the chlorinated dibenzo-dioxins/chlorinated dibenzofurans. Joint NIEHS/IARC Working Group Report. Lyon, France. 44 p. Report by the Comptroller General of the United States. April 7, 1979. Health effects of exposure to Herbicide Orange in South Vietnam should be resolved. General Accounting Office Pub. CED-79-2. 38 p. B. PHARMACOKINETICS OF 2,4-D Berndt, W. 0. and F. Koschier. 1973. In vitro uptake of 2,4-dichlorophenoxyacetic acid (2,4-D) and 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) by renal cortical tissue of rabbits and rats. Toxicol. Appl. Pharmacol. 26:559-570. Clark, D. E., J. S. Palmer, R. D. Radeleff, H. R. Crookshank and F. M. Farr. 1975. Residues of chlorophenoxy acid herbicides and their phenolic metabolites in tissues of sheep and cattle. J. Agric. Food Chem. 23(3):573-578. hlo, H. and P. Ylitalo. 1977. Substantial increase in the levels of chlorophenoxyacetic acids in the CNS of rats as a result of severe intoxication. Acta Pharmacol. Toxicol. 41:280-256. .'i\:, U. 1966. Distribution and elimination of chlorinated phenoxyacetic acids in animals. Acta Vet. Scand. 7:264-271. urunow, W. and C. Bohme. 1974. Uber den stoffwechsel von 2,4,5-T and 2,4-D bei ratten und mausen. (Metabolism of 2,4,5-T and 2,4-D in rats and mice.) Arch. Toxicol. 32:217-225. (German) Khanna, S. and S. C. Fang. 1966. Metabolism of 14 C-labelled 2,4-dichlorophenoxyacetic acid in rats. J. Agric. Food Chem. 14(5):500-503. 115 �International Agency for Research on Cancer. 1977. 2,4-D and esters. P. 111-199. ln_ IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Man. Vol. 15. Some Fumigants, the Herbicides 2,4-D and 2,4, 5-T, Chlorinated Dibenzodioxins, and Miscellaneous Industrial Chemicals. Lyon, France. C. PHARMACOKINETICS OF 2,4,5-T Bohme, C. and W. Grunow. 1974. Uber den stoffwechsel von 4-(2,4,5-trichlorophenoxy)-buttersaure bei ratten. (Metabolism of 4-(2,4,5-trichlorophenoxy)-butyric acid in rats.) Arch. Toxicol. 32:227-231. (German) Dencker, L. 1976. The herbicide 2,4,5-T: Early placental barrier and accelerated fetal uptake with advancing gestation. P. 59-79. In Tissue Localization of Some Teratogens at Early and Late Gestation RUTated to Fetal Effects. Acta Pharmacol. Toxicol. 39(Suppl 1): 131p. Ebron, M. And K. D. Courtney. 1976. Difference in 2,4,5-T distribution in fetal mice and guinea pigs. Toxicol. Appl. Pharmacol. 37:144-145. Fang, S. C., E. Fallin, M. L. Montgomery and V. H. Freed. 1973. The metabolism and distribution of 2,4,5-trichlorohenoxyacetic acid in female rats. Toxicol. Appl. Pharmacol. 24:555-563. Lindquist, N. G. and S. Ullberg. 1971. Distribution of the herbicides of 2,4,5-T and 2,4-D in pregnant mice: Accumulation in yolk sac epithelium. Experientia 27:1439-1441. Piper, W. N., J. Q. Rose, M. L. Leng and P. J. Gehring. 1973. The fate of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) following oral administration to rats and dogs. Toxicol. Appl. Pharmacol. 26:339-351. Sauerhoff, M. W., W. H. Braun, G. E. Blau and P. J. Gehring. 1976. The dosedependent pharmacokinetic profile of 2,4,5-trichlorophenoxy acetic acid following intravenous administration to rats. Toxicol. Appl. Pharmacol. 36:491-501. D. PHARMACOKINETICS OF TCDD Allen, J. R., J. P. Van Miller and D. H. Norback. 1975, Tissue distribution, excretion and biological effects of [^C] tetrachlorodibenzo-p-dioxin in rats. Food Cosmet. Toxicol. 13:501-505. Piper, W. N., J. Q. Rose and P. J. Gehring. 1973. Excretion and tissue distribution of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the rat. Advan. Chem. Sen. 120:85-91. Reggiani, G. 1978. The estimation of TCDD's toxic potential in the light of the Seveso accident. 20th Congress of the European Society of Toxicology, Berlin (West), June 25-28. 20p. 116 �Rose, J. Q., J. C. Ramsey, T. H. Wintzler, R. A. Hummel, and P. J. Gehring. 1976. The Fate of 2,3,7,8-tetrachlorodibenzo-p-dioxin following single and repeated oral doses to the rats. Toxicol. Appl. Pharmacol. 36:209226. Vinopal, J. H. and J. E. Casida. 1973. Metabolic stability of 2,3,7,8tetrachlorodibenzo-p-dioxin in mammalian liver microsomal systems and in living mice. Arch. Environ. Contam. Toxicol. 1(2):122-133. E. PHARMACOKINETICS OF 2,4-D and 2.4,5-T IN MAN Coutselinis, A., R. Kentarchou and D. Boukis. 1977. Concentration levels of 2,4-D and 2,4,5-T in forensic material. Forensic Sci. 10:203-204. Gehring, P. J., C. G. Kramer, B. A. Schwetz, J. Q. Rose and V. K. Rowe. 1973. The fate of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) following oral administration to man. Toxicol. Appl. Pharmacol. 26:352-361. Kohli, J. D. R. N. Khanna, B. N. Gupta, M. M. Dhar, J. S. Tandon and K. P. Sircar. 1974. Absorption and excretion of 2,4-dichlorophenoxyacetic acid in man. Xenobiotica 4(2):97-100 Kohli, J. D., R. N. Knanna, B. N. Gupta, M. M. Dhar, J. S. Tandon and K. P. Sircar. 1974. Absorption and excretion of 2,4,5-trichlorophenoxyacetic acid in man. Arch. Int. Pharmacodyon. Ther. 210:250-255. Nielsen, K., B. Kaempe, and J. Jensen-Holm. 1965. by 2,4-dichlorophenoxyacetic acid (2,4-D). Fatal poisoning in man Ramsey, J. C., T. L. Lavy, and W. H. Braun. 1979. Exposure of forest workers to 2,4,5»T: Calculated dose levels. Toxicology Laboratory, Dow Chemical Company USA, Midland, Michigan. Mim. Unpublished data, 32p. Sauerhoff, M. W., W. H. Braun, G. E. Blau and P. J. Gehring. 1977. The fate of 2,4-dichlorophenoxyacetic acid (2,4-D) following oral administration to man. Toxicology 8:3-11. •. CELLULAR MECHANISMS OF ACTION OF TCDD Crow, K. D. 1970. Chloracne. Trans. St. John Hosp. Derm Soc. 56:79-99. Courtney, K. D. and J. A. Moore. 1971. Teratology studies with 2,4,5trichlorophenoxyacetic acid and 2,3,7,8-tetrachlorodibenzo-p-dioxin. T o x i c o l . Appl. Pharmacol. 20:396-403. 117 �Green, S. 1977. Cytogentio By-Lines 7(6):292. effects of TCDD on rat bone marrow cells. FDA Hussain, S., L. Ehrenberg, G. Lofroth, and T. Gejvall. 1972. effects of TCDD on bacterial systems. Ambio 1 (1):32-33. Mutagenic Poland, A. and A. Kende. 1976. 2,3,7,8-Tetrochlorodibenzo-p-dioxin: enviromental contaminant and molecular probe. Fed. Proc. 35(12):2404-2411. Poland, A., W. Greenlee, and A. Kende, 1979. Studies on the mechanism of action of the chlorinated dibenyo-P-dioxins and related compounds. P. 214-230. _I_n Health Effects of Halogenated Aromatic Hydrocarbons. The New York Academy of Science. Vos, J. G. and 0. A. Moore. 1974. Suppression of cellular immunity in rats and mice by maternal treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Int. Arch. Allergy 47:777-794. Wasson, J. S., J. E. Huff and N. Loprieno. 1977/1978. A review of the genetic toxicology of chlorinated dibenzo-p-dioxin. Mutation Research 47:141-160. G. DETECTION OF DIOXIN INDUCED DNA ABERRATIONS Buckton, K. E., W. M. C. Brown and P. G, Smith. 1967. Lymphocyte survival in men treated with X-rays for ankylosing spondylitis. Nature 214:270272. Czeizel, E. and J. Kiraly. 1976. Chromosome examinations in workers producing klorinol and buvinol. P 239-256. Jto The Development of a Pesticide as a Complex Scientific Task. L. Banki (Ed). Medicina (Budapest). Green, S. and F. S. Moreland, 1975. Cytogenetic evaluation of dioxins in the rat. Toxicol. Appl. Pharmacol. 33(1): 161. Hay, A. 1977. Identifying carcinogens. Hay, A. 1978. Vietnam's dioxin problem. several Nature 269:468-470. Nature 271:597-598. Tenchini, M. L., R. Giorgi, C. Crimaudo, G. Simoni, F. Nuzzo, and L. de Carli. 1977. Approaches to examination of genetic damage after a major hazard in chemical industry: preliminary cytogenetic findings in TCDD exposed subjects after Seveso accident. Presentation at the Expert Conference on Genetic Damage Caused by Environmental Factors, Oslo, Norway, May 11-13, 1977. 118 �H. ANIMAL STUDIES Allen, J. R., D. A. Barsotti, J. P. Van Miller, L. J. Abrahamson and J. J. Lalich. 1977. Morphological changes in monkeys consuming a diet containing low levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Food Cosmet. Toxicol. 15:401-410. Kociba, R. J., D. G. Keyes, J. E. Beyer, R. M. Carreon, C. E. Wade, D. A. Dittenber, R. P. Kalrnins, L. E. Frauson, C. N. Park., S. D. Barnard, R. A. Hummel and C. G. Humiston. 1978. Results of a two-year chronic toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats Toxicol. Appl. Pharmacol. 46:279-303. Kociba, R. J., D. G. Keyes, R. W. Lisowe, R. P. Kalmins, D. D.Dittenbar, C. E. Wade, S. J. Gorzinski, N. H. Mahle and B. A. Schwetz. 1979. Results of a two-year chronic toxicity and oncogenic study of rats ingesting diets containing 2,4,5-trichlorophenoxyacetic acid (2,4,5-T.) Food Cosmet. Toxicol. (In Press.) Lamb, J., J. Moore, and T. Marks, 1980. Reproduction and fertility in treated male mice and evaluation of congenital malformations in their offspring. Congressional Record, August 6, 1980. P. S10904-11. McConnell, E. E., J. A. Moore and D. W. Dalgard. 1978. Toxicity of 2,3,7*8tetrachlorodibenzo-p-dioxin in rhesus monkeys (Macaca mu]Iatta) following a single oral dose. Toxicol. Appl. Pharmacol. 437175-18'T1. McNulty, W.P. 1977. Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin for rhesus monkeys: brief report. Bull. Environ. Contain. Toxicol. 18:108-109. Rowe, V. K. and T. A. Hymas. 1954. Summary of toxicological information on 2,4-D and 2,4,5-T type herbicides and an evaluation of the hazards to livestock associated with their use. Am. J. Vet. Res. 15:622-629. Schantz, S.L., D. A. Barsotti, and J.R. Allen. 1979. Toxicological effects produced in non-human primates chronically exposed to 50 ppt 2,3,7,8tetrachlorodibenzo-p-dioxin. Presented at Meetings of Society of Toxicology, New Orleans, March 15,1979. I. CASE STUDIES Berkley» M. C. and K. R. Magee. dimethyl amine salt of 2,4-D. 1963. Neuropathy following exposure to a Arch. Intern. Med. 111:351-353 Blei berg, J., M. Wallen, R. Brodkin, and I. L. Applebaum. ally acquired porphyria. Arch. Dermatol . 89:793-797. 119 1964. Industri- �Brandt, M. R. 1971. Herbatoxforgiftning, en Kort oversigt og et nyt tilfaelde, (Herbatox poisoning, a brief review and a report of a new case). Ugeskrift for Laeger 133:500-503. (Danish). Crow, K, 1978. Chloracne: the chemical disease. New Scientist 78-80. Jirasek, L., J. Kalensky, and K. Kubec. 1973. Acne chlorina a porphyria cutanea tarda pri vyrobe herbicid. (Acne chlorina and porphyria cutanea tarda during the manufacture of herbicides. Part I.) Ceskoslovenska Dermatologie 48(5):305-317. Jirasek, L., J. Kalensky, K. Kubec, J. Pazderova and E. Lukas. 1974. Acne chlorina, porphyria cutanea tarda a jine projevy celkove intoxikace pri vyrobe herbicid. (Acne chlorina, porphyria cutanea tarda and other manifestations of general intoxication during the manufacture of herbicides, Part II.) Ceskoslovenska Dermatologie 49(3):145-157. Oliver. R. M. 1975. Toxic effects of 2,3,7,8 tetrachloro-dibenzo-l,4-dioxinin laboratory workers. Br. J. Ind. Med. 32:49-53, Taylor, J., 1979. Environmental chloracne; update and overview. P. 295307. J^n Health Effects of Halogenated Aromatic Hydrocarbons. The New York Academy of Science. J. EPIDEMIOLOGIC STUDIES Alfred, J. E. (Chairman). September 1978. Report on the consultative council on congenital abnormalities in the Yarram District (Australia). Austalian Commission on Public Health. Document No. 47-12004/78, Government Printing Office, Melbourne, Australia. 55p. Axelson, 0., and L. Sundell, 1974. Herbicide exposure, mortality and tumor incidence. An investigation on Swedish railroad workers. Scan J. Work, Environ, and Health. 11 (1974) 21-28. Axelson, 0., and L. Sundell, K. Anderson, C. Edling, C. Hogstedt, and H. Kling, 1980. Herbicide exposure and tumor mortality: an updated epidemiologic investigation on Swedish railroad workers. Manuscript. Cook, R. R., and K. M. Bodner, 28 October 1981. effects. International Symposium on Dioxins Arlington, Virginia. Dioxin and reproductive and Related Compounds, Eriksson, M., L. Hardell, N. Berg, T. Moller, and 0. Axelson, 1979. Casecontrol study on malignant mesenchymol tumors of the soft tissue and exposure to chemical substances. Lakartidningen 76: 3872-72 (Swedish). Hardell, L., and A. Sandstrom, 1979. Case-control study: soft tissue sarcomas and exposure to phenoxyacetic acids or chlorophenols, British Journal of Cancer, 39: 711-717. 120 �Hardell, L., M. Eriksson, and P. Lenner, 1980. Malignant lymphoma and exposure to chemical substances, especially organic solvents, chlorophenols and phenoxyacids. Lakartidningen 7 7 ( 4 ) : 208-210 (Swedish). Homberger, E., G. Reggiani, J. Sambeth and H. Wipf. 1979. The Seveso accident: its nature, extent and consequences. Givaudan Research Co. Ltd. and F. Hoffman-LaRoche and Co. Ltd., 4002 Basle, Switzerland. Mim. Unpublished data, 80 p. Kramer, C. G. 1970, Revised 1974. Health of employees exposed to 2,4,5-T. Dow Chemical Co., Midland, Michigan. Mim. Unpublished data, 19 p. Lavy, Terry L., Scott Shepard, and John D. Mattice, 1980. Exposure Measurements of Applicators Spraying (2,4,5-Trichlorophenoxy) and acetic Acid in the Forest. J. Agric. Food Chem 28:626-630. Pocchiari, F., Silano, V., Zampier, A. 1979. Human health effects from accidental release of tetrachlorodibenzo-p-dioxin (TCDD) at Seveso, Italy. Ann. N.Y. Acad. Sci., p. 311-320. Poland, A. P., D. Smith, G. Metter and P. Possick. 1971. A health survey of workers in a 2,4-D and 2,4,5-T plant. Arch. Environ. Health 22(3): 316327. Regianni, G., 1980. Acute human exposure to TCDD in Seveso, Italy. cology and Environmental Health, 6: 27-43. J. Toxi- Smith, A. H. 28 October 1981. (Personal communication), International Symposium on Chlorinated Dioxins and Related Compounds, Arlington, Virginia. Theiss, A., and R. Frentzel-Beyme, 1977. Mortality study of persons exposed to dioxin following an accident which occurred in the BASF on November 13, 1953. Presented at the Fifth International Conference of Medichem, San Francisco, California, September 1977. Tung, T. T. 1973. Le Cancer primaire du foie au Vietnam. (Primary cancer of the liver in Vietnam.) Chirurgie 99(7):427-36. (French). Z.ack, J., and R. Suskind, 1980. The mortality experience of workers exposed to tetrachlorodibenzodioxin in a trichlorophenol process accident. J. Orcupational Medicine 22(1): 11-14. "eport of Assessment of a Field Investigation of Six-year Spontaneous Abortion Rates in Three Oregon Areas in Relation to Forest 2,4,5-T Spray Practices. Epiderniologic Studies Program, Human Effects Monitoring Branch, Environmental Protection Agency, Washington, DC, February 29, 1979. Mim. Unpublished data, 100 p. 121 �K. STATISTICAL REFERENCES Armitage, P. 1971. Further analysis of qualitative data. P 384-391. lr± Statistical Methods in Medical Research. John Wiley and Sons, New York Bishop, Y. M. M., S. E. Fienberg, and P. W. Holland. 1975. Models for three-dimension arrays. P 31-42. In_ Discrete Multivariate Analysis: Theory and Practice. The MIT Press, Mass. Breslow, N. 1970. A generalized Kruskal-Wallis test for comparing k samples subject to unequal patterns of censorship. Biometrika 57:579-594. Breslow, N. E., and J. Growley. A large sample study of the life table and product limit estimates under random censorship. Ann. Stat. 2:437-453. Breslow, N. E. and N. E. Day. 1975. Indirect Standardization and multiplicative models for rates, with reference to the age adjustment of cancer incidence and relative frequency data. J. Chron. Dis. 28:289-303. Breslow, N. E. 1975. Analysis of survival data under the proportional hazards model. Int. Stat, Rev. 43: 45-58. Carpenter, R. G. 1977. Matching when covariables are normally distributed. Biometrika 64:299-307. Chiang, C. L. 1968. Chapters 9-12, P 189-296. _In_ Introduction to stochastic processes in biostatisties. John Wiley and Sons, New York. Copeland, K. T., M. Checkoway, A. J. McMichael, and R. H. Holbrook. 1977. Bias due to misclassification in the estimation of relative risk. Am. J. Epi. 105:488-495. Cox, D. R. 1958. The regression analysis of binary sequences. Statist. Soc. -B. 20:215-232. J. R. Cox, D. R. 1958. Two further applications of a model for binary regression. Biometrika 45:562-565. Cox, D. R. 1966. A simple example of a comparison involving quantal data. Biometrika 53: 215-220. Cox, D. R. 1972. Regression methods and life tables. -B 34:087-220. J. R. Statist. Soc. Fleiss, J. L. 1973. Analysis of data from matched samples. P 11-80. _[n Statistical Methods For Rates and Proportions. John Wiley and Sons, New York. Frank, J. 1977. Survival analysis with time-dependent covariates. Thesis, University of California, Berkeley, California. 122 Ph.D. �Gehan, E. 1965b. A generalized two-sample Wilcoxin test for doubly censored data. Biometrika 52:650-653. Gehan, E. 1965a. A generalized Wilcoxin test for comparing arbitrarily censored samples. Biometrika 52:203-223. Gross, A. and V. Clark. 1975. Chapter 2, P 23-48. In_ Survival Distributions: Reliability Applications in the Biomedical Sciences, John Wiley and Sons, New York. Holford, T., White, C., and Kelsey, J., 1978. Multi-variate analysis for matched case-control studies. Am. J. Epi. 107: 245-256. Kaplan, E. and P. Meier. 1958. Nonparametric estimation observations. J.A.S.A. 53:457-581. from incomplete Karger, S. 1972. Cardiovascular studies, P 298 ln_ Medical Primatology, Part III. Basel, Munchen. Mantel, N. 1966. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemotherapy Reports 50:163-170. Mantel, N. 1967. Ranking procedures for arbitrarily restricted observations. Biometrics 23:65-78. McCullagh, P., 1978. A class of parametric models for the analysis of square contingency tables with ordered categories. Biometrika 65: 413-418. McKinlay, S. M. 1977. Pair-Matching: a reappraisal Biometrics 33:725-735. of a popular technique. McKinlay, S. M. 1975. The design and analysis of the observational study: a review. J.A.S.A. 70:503-520. Miettinen, 0. S. 1969. Individual matching with multiple controls in the case of all-or-none response. Biometrics 25: 339-355. Miettinen, 0. S. 1970. Matching and design efficiency in retrospective ".'.-idles. Am. J. Epi. 91:111-117. i eto, K. and J. Peto. 1972. Asymptotically efficient rank invariant test procedures. J. R. Statist. Soc. -A 135:185-198. Peritz, E. and R. Ray. 1978. Some statistical methods for survival analysis with rare failure. Comrn. Stat. A7:619-636. Schoenfeld, D., 1980. Chi-squared goodness of fit tests for the proportional hazards regression model. Biometrika 67: 145-153. 123 �Stuart, A., 1955. A test for promogeneity of the marginal distributions in a two-way classification. Biometrika 42: 412-416, Torrance, E. P. 1954. The development of a preliminary life experience inventory for the study of fighter interceptor pilot combat effectiveness. AFPTRC-TR-54-89. Air Force Personnel Training and Research Center, Lackland Air Force Base, Texas. Ury, H. K. 1975. Efficiency of case-control studies with multiple controls per case: continuous or dichotomous data. Biometrics 31:643-649. Walker, W. H. and D. B. Duncan. 1967. event as a function of several 54:167179. Estimation of the probability of an independent variables. Biometrika Wei, L. J., 1980. A generalized Gehan and Gilbert test for paired observations which are subject to arbitrary right censorship. J. Am. Statist. Assoc., to appear. Zuckerman, M. 1972. Manual and research report of the sensation seeking scale. Department of Psychology, University of Delaware, Newark, Delaware. 124 �XV. APPENDIX TABLE A-l SUMMARY OF 2,4-D, AND TCDD ANIMAL STUDIES TABLE A-2 "SYMPTOM COMPLEX" DERIVED FROM LITERATURE REVIEW OF CASE STUDIES EXPOSED TO 2,4-D, 2,4,5-T AND/OR TCDD TABLE A-3 DETAILED LISTING OF SYMPTOMS/ SIGNS BY MAJOR CATEGORY FROM LITERATURE REVIEW OF CASE STUDIES EXPOSED TO 2,4-D, 2,4,5-T AND/OR TCDD TABLE A-4 AGE COMPARISON OF EXPOSED SUBJECTS AND THEIR MATCHED CONTROLS TABLE A-5 STATISTICAL DESCRIPTION MATCHING PROCESS TABLE A-6 SPECIFIC STUDY ENTRY RULES TABLE A-7 SCHEDULE AND MODE OF CONTACTS WITH STUDY SUBJECTS TABLE A-8 MONTE CARLO SIMULATION FIGURE A-l 2,3,7,8-TETRACHLORODIBENZO-PDIOXIN (TCDD) FIGURE A-2 ESTIMATED PARTICIPATION POPULATION FIGURE A-3 OF STUDY DESIGN FORMAT 125 2,4,5-T OF THE IDENTIFICATION/ RANCH HAND �TABLE A-l SUMMARY OF 2,4-D, 2,4,5-T, AND TCDD ANIMAL STUDIES 2.4-D 2,4.5-T TCDD LD50 RANGE (ACUTE) 100-1000 mg/kg 100-1000 mg/kg 1-1000 g/kg CHRONIC TOXIC OOSE APPROACHES ACUTE LEVEL RAPID CLEARANCE 1/2 ACUTE LEVEL; VARIABLE CLEARANCE MARKEDLY LOWER LEVEL BIOACCUMULATION SIGNS OF ACUTE/ CHRONIC TOXICITY ANOREXIA ANOREXIA WEIGHT LOSS WEIGHT LOSS ATAXIA INVOLUTION OF THYMUS MUSCULAR WEAKNESS G.I. INJURY ALOPECIA IRRITATED G.I. TRACT LIVER CONGESTION EPITHELIAL CHANGES MINOR LIVER INJURY KIDNEY CONGESTION LIVER LESIONS (VARIABLE) ro CTl MINOR KIDNEY INJURY HYPOTHYROIDISM MINOR LUNG CONGESTION EMBRYOTOXIC DOSE APPROACHES TOXIC LEVEL APPROACHES TOXIC LEVEL MARKEDLY BELOW TOXIC MATERNAL LEVELS TERATOGENICITY QUESTIONABLE; WEAK AT BEST *LOW INCIDENCE ONLY IN MICE (CLEFT PALATES DILATED RENAL PELVIS) SPECIES VARIATIONS: YES MICE NO RATS CARCINOGENICITY QUESTIONABLE; WEAK AT BEST ONE STUDY: YES NUMEROUS STUDIES: NO EPITHELIAL CHANGES IN PRIMATES: YES IN RATS �TABLE A-2 "SYMPTOM COMPLEX" DERIVED FROM LITERATURE REVIEW OF CASE STUDIES EXPOSED TO 2,4-0; 2,4,5-T WD/OR TCDO 2.4,5-T (+ TCDO) TCDO CHLORACME CHLORACNE PORPHYRIA PORPHYRIA HYPERPiGMENTATIOD HYPERP^G^eNTATION ASTHENIA ASTHENIA ASTHENIA PERIPHERAL NEUROPATHY PERIPHERAL fBURQPATHY PERIPHERAL NEUROPATHY CARDIAC .CARDIAC DISTURBANCE 2.4-D SWEATING/FEVER CARDIAC ^DISTURBANCE R£mL OYSPUHCJIfiH UV£R DYSR»CII€« fii HVEft DYSFUNCTION GI DISTURBANCE GI OISTtKBANCE JtfAOACHjE CSf H¥POTHYROIOIS« CO»VULSiONS «EARIHffi/SJ»ELL OISTURBAHCES �TABLE A-3 DETAILED LISTING OF SYMPTOMS/SIGNS BY MAJOR CATEGORY FROM LITbRATURE REVIEW OF CASE STUDIES EXPOSED TO 2,4-D; 2,4,5-T AND/OR TCDD NEURO-PSYCHIATRIC ABNORMALITIES AESTHENIA PERIPHERAL NEUROPATHY ANXIETY HYPOREFLEXIA DEPRESSION WEAKNESS FATIGUE PARESTHESIAS APATHY EXTREMITY NUMBNESS LOSS OF DRIVE MYALGIA DECREASED LIBIDO GAIT DISTURBANCE IMPOTENCY "MILD" PARESIS SLEEPLESSNESS EMOTIONAL INSTABILITY ANOREXIA DIZZINESS DECREASED LEARNING ABILITY �TABLE A-3 (CONTINUED) DETAILED LISTING OF SYMPTOMS/SIGNS BY MAJOR CATEGORY FROM LITERATURE REVIEW OF CASE STUDIES EXPOSED TO 2,4-D, 2,4,5-T AND/OR TCOD DERMATOLOGIC DISEASE CHLORACNE PORPHYRIA CUTANEA TARDA HYPERPIGMENTATION HIRSUITISM (B005f) ALOPECIA OF THE SCALP OTHER DISORDERS HEPATIC DYSFUNCTION SEMAL DYSFUNCTION INCREASED CHOLESTEROL AW TRHa.*CERIO£ PROTEINURIA ItCREASES IN LIVES TESTS IBStlLAR fiiOMERULAR DEGENERATION RENAL GLUO)SORIA 61 PtSTORBANCE OUaiAC 01STUR6AMCE •NAHSEA B8ADYCARDIA VOMITING TACHYCARDIA DIARRHEA ATRIAL FIBRILLATION GASTRITIS ABDOMINAL PAIN �TABLE A-4 AGE COMPARISON OF EXPOSED SUBJECTS AND THEIR MATCHED CONTROLS AFSC/Race Strata Officer: Pi lot /Caucasian /Black Number of Exposed Subjects Mean Number of Matched Controls 349 6 9.5 2.7 Age Differenc 0-60 0-57 Nonpilot /Caucasian /Black CO 10.0 10.0 0-07 0-36 Other/Caucasian /Black o 78 2 25 1 10.0 5.0 0-27 0-54 187 15 10.0 9.8 0-35 0-58 528 51 10.0 10.0 0-48 0-06 Enlisted: Flying/Caucasian /Black Nonf lying/Caucasian /Black Killed in Action Of fleers /Caucasian /Black 14 1 ~ Enlisted/Caucasian /Black 7 0 — — — �TABLE A-5 STATISTICAL DESCRIPTION OF THE MATCHING DIST IN MONTHS - 1 - 2 «. "^ - 4 « 5 - 6 - 7 •• ft - 9 -10 -11 -12 -13 -14 -15 -16 -17 -18 -19 -20 -21 -22 -23 -24 -25 -26 -27 -28 -29 -30 -31 -32 -33 -34 -35 -36 -37 -38 -39 -40 -41 -42 -43 -44 -45 -46 -47 -48 -49 -50 -51 -52 -53 -54 -55 -56 -57 -58 -59 -60 COUNT 847 231 114 92 88 41 33 28 27 10 18 17 9 23 11 20 16 4 6 11 10 4 4 4 3 6 2 6 2 2 2 1 6 3 5 4 3 4 2 3 5 4 2 6 9 3 0 0 3 4 2 0 4 6 3 4 3 5 5 3 DIST IN MONTHS 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 COUNT 828 231 121 91 67 47 39 22 23 21 18 22 18 11 8 15 11 6 6 3 9 6 13 5 4 7 8 8 9 4 5 5 2 3 7 3 3 11 5 5 2 6 2 9 4 6 4 3 1 4 2 0 4 2 3 0 2 3 1 4 PROCESS 0 1 2 3 4 5 6 7 6 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 COUNT % 8612 1675 462 235 183 ABSOLUTE DIST IN MONTHS 70.6 13.7 n 3.8 1.9 1.5 1.3 0.7 0.6 0,4 0.4 0.3 0.3 39 27 0.3 15$ 88 $0 50 31 36 34 19 35 27 10 12 14 19 10 17 9 7 13 10 14 11 6 7 6 8 6 12 7 6 1? 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 7 8 7 10 4 IB 13 9 4 3 4 8 4 0 8 8 6 4 5 8 6 7 131 0.2 0.3 0.2 0.3 0,2 0.1 0.1 o.i 0.2 0.1 0. 0. 0. 0. 0. 0. 0.1 0.0 0.1 0.0 0.1 o.o 0. 0. 0.0 0. 0. 0. 0. 0. 0.0 0.1 0.1 0.1 0.0 0.0 0.0 0.) 0.0 0.0 0.1 0.1 0.0 0.0 0.0 0.1 0.0 0.1 CUMULATIVE TOTAL 8612 10287 10749 10984 11167 11322 11410 11482 11532 11582 11613 11649 11688 11715 11749 11768 11803 11830 11840 11852 11866 11885 11895 11912 11921 11928 11941 11951 11965 11976 11982 11989 11995 12003 12009 12021 12026 12034 12049 12056 12064 12071 12081 12085 12100 12113 12122 12126 12129 12133 12141 1214S 12143 12153 12161 12V67 12171 12176 12184 12190 12197 70.6 64.3 88,1 90.1 91.6 92.8 93,5 94,1 94.5 95.0 95.2 95.5 95.8 96.0 96.3 96.5 96.8 97,0 97.1 97.2 $7.3 97.4 97.5 97.7 97.7 0.8 97.9 98.0 98.1 $8.2 $8.2 98.3 98.3 96.4 §1.5 98.6 98.6 98.7 99,8 99.8 99.9 99.0 9,9.0 99! 2 99.3 99.4 99.4 99.4 99.5 0.5 99.6 99.6 99.6 99.7 99.8 99.8 99.8 99.9 99.9 100,0 �Table A-6 SPECIFIC RULES FOR ENTRY INTO THE MORBIDITY STUDY RULES CIRCUMSTANCES RANCH HANDER (RH) DIES FOLLOWING INITIAL DATA COLLECTION CONTROL FOLLOWED THROUGHOUT AND REPLACED AS NECESSARY RH DIES OF COMBAT CAUSE MEDICAL RECORDS REVIEWED; NO CONTROL SET.FORMED RH DIES OF NONCOMBAT CAUSE PRIOR TO INITIAL DATA COLLECTION 1ST ORDER SURROGATE INTERVIEW ACCOMPLISHED; CONTROL SELECTED AND FOLLOWED THROUGHOUT; AS NECESSARY RH NONCOMPLIANT FOR BASELINE QUESTIONNAIRE AND PHYSICAL CONTROL FOLLOWED THROUGHOUT THE STUDY; REPLACED AS NECESSARY RH COMPLIANT FOR QUESTIONNAIRE; NONCOMPLIANT FOR BASELINE PHYSICAL EXAMINATION CONTROL FOLLOWED THROUGHOUT THE STUDY; REPLACED AS NECESSARY RH NONCOMPLIANT DURING FOLLOWUP CONTROL FOLLOWED THOUGHOUT THE STUDY; REPLACED AS NECESSARY CONTROL DIES FOLLOWING INITIAL DATA COLLECTION NOT REPLACED IN THE PROSPECTIVE STUDY OF MORBIDITY CONTROL DIES OF COMBAT CAUSE MEDICAL RECORDS REVIEWED; EXCLUDED FROM FURTHER STUDY CONTROL DIES OF NONCOMBAT CAUSE PRIOR TO INITAL DATA COLLECTION INCLUDED IN MORTALITY AND RETROSPECTIVE MORBIDITY STUDIES; SURROGATE INTERVIEW ACCOMPLISHED. NOT INCLUDED IN PROSPECTIVE MORBIDITY STUDY AND REPLACED BY A LIVING COMPLIANT CONTROL. CONTROL NONCOMPLIANT FOR BASELINE PHYSICAL EXAMINATION CONTROL FOLLOWED THROUGHOUT STUDY REPLACE AS NECESSARY CONTROL NONCOMPLIANT DURING FOLLOWUP CONTROL FOLLOWED THROUGHOUT STUDY REPLACE AS NECESSARY NONCOMPLIANT CONTROL RETURNS TO STUDY BOTH PRIMARY AND REPLACEMENT CONTROLS WILL BE CONTINUED IN STUDY 132 �Table A-7 SCHEDULE AND MODE OF CONTACTS WITH STUDY SUBJECTS STUDY PHASE CONTACT MADE Morbidity Study Introductory Letters Morbidity Study Comprehensive Questionnaire Oct .81 -Mar 82 Baseline Physical Exam Dec 81-Sep 82 Adaptive Questionnaire Adaptive Physical Examination Oct 83-Mar 84 Dec 84-Jun 85 Adaptive Questionnaire Adaptive Physical Exarnlnation Oct 86-Mar 87 Dec 86-Jun 87 Adaptive Questionnaire Adaptive Physical Examination Oct 91-Mar 92 Dec 91-Juri 92 Adaptive Questionnaire Adaptive Physical Examination Oct 96-Mar 97 Dec 96-Jun 97 Adaptive Questionnaire Adaptive Physical Examination Oct 2001-Mar 2002 Dec 2001-Jun 2002 Follow-up Study 133 Oct 81 �Table A-8 MONTE CARLO SIMULATION PROGRAM 210 220 230 240 250 260 270 280 290 300 310 320 330 340 10 DIM C(10,3) 20 DIM A(10,3) 30 P2=.25 40 D1-.45 50 M=0 60 N=0 70 FOR 1=1 TO 10 80 FOR 0=1 TO 3 90 A(I,J)=0 100 C(I,JH 110 NEXT J 120 NEXT I 130 M=M+1 140 PRINT M 150 IF M=1001 THEN 330 160 F=l 170 1=1 180 J=l 190 C(I,J)=RND(1) 200 X=P2+F*D1 IF C(IJ) > X THEN 270 1=1+1: F=l IF I > 10 THEN 250 GOTO 190 N=N+1 GOTO 130 A(IJJ)=A(I,J)+1 J=J+1 IF J>3 THEN 320 F=0 GOTO 190 GOTO 130 STOP SELECT PRINT 215 350 FOR 1=1 TO 10 360 370 380 390 400 410 420 PRINT A(I,1), A(I,2), NEXT I PRINT PRINT "N(l)", N(l) PRINT "N(2)", N(2) PRINT "N(3)M, N(3) END GLOSSARY I = Control individual index J = Examination number index A(I,J) = Attendance array = number of times the ith control was used for the jth examination C(I,J) = Testing variable array N = number of times no control was available M = number of matches attempted of = preselected probabilities. Pj = Dl + P2 and Pg = P2 P2 RND = Random DIM = Dimension F = Flag 134 �Figure A-l 2, 3, 7, 8-TETRACHLORODIBENZp-p-DIOXIN (TCDD) • WOlECUkAR WEIGHT • MELWN& POINT g 321.8935 303-305 °C • DECOMPOSITION POH*T 980-lr0000C ORittQ-IWCHLOiOffENZENE 1.40 CHtW?OBENZENE 0.72 HERBICIDE o.58 CttL(MOFORM ACETONE METOANOL WATER 0.37 ft 11 Q.m 2*1Q~7 �l\-2 ESTIMATED IDENTIFICATION/PARTICIPATION OF THE RANCH HAND POPULATION RESPONSE ESTIMATE RANCH HAND POPULATION ESTIMATED NUMBER OF PARTICIPANTS 1200 ^-^^^-^_ CO en UNACCOUNTABLE < 1% ACCOUNTABLE _-^— — -—-' NON-PARTICIPANTS QUESTIONNAIRE DEAD/MORIBUND 10% PARTICIPANTS UNWILLING 25% ' ' ,. • BASELINE EXAM NON-PARTICIPANTS 40% PARTICIPANTS — 1 „ • NON-PARTICIPANTS 1st FOLLOW-UP 20% EXAM PARTICIPANTS . — 2nd FOLLOW-UP NON-PARTICIPANTS EXAM PARTICIPANTS 20% 99% 1188 65% 772 60% 463 80% 371 80% 297 �S - i j u r e A-3 STUDY DESIGN FORMAT MORTALITY STUDY 1:5, RETROSPECTIVE MORBIDITY STUDY 1:1, Rr l Ci 1 — t " I* 1 -1 lj|,2 Ci T Cl,4 f» - 1*1,5 * R . 2 L,,,. . + -i > •4- *?o < . C2 3 C° 4 /*-> tto '•7 R * cV,- .—>• nc " •>,£ -1 C-? 3-!J • "OiJ 4- T C-3 A /*, "3^9 * ~ 1 • P , Cj 5 V4,^ Ci 3 - nc CA A CAT, I/ 4jU !._!_ I * C4,8 1 ] 1 l 1962 1965 1970 . SPRAY I-*-OPERATIONS R = RANCH HAND c = CONTROL l t l f 1 l 1 i i 1 , 1381 1984 1986 1996 1991 BASELINE QUEST. | ADAPTIVE QUESTS. AND EXAMS AND EXAM nc = NONCOMPLIANT t = DECEASED L_L_L_ 2001 * = 5 OTHER REPLACEMENT CONTROLS �XVI. Examiner's Handbook A. General Instructions Project RANCH HAND II is a multiyear effort to determine whether or not C-123 aircrew members who were engaged in the aerial spraying of herbicides in Vietnam have developed significant adverse health effects from that exposure. Detailed surveys of the world's literature have been used in designing the history questionnaires, physical examination protocol, and laboratory procedures. This phase of Project RANCH HAND II involves a cross .sectional study of the subject's health at the time of examination. It is important that examiners remain unaware of the subject's status as a RANCH HAND participant or as a control subject. The physician examiner is tasked to examine and objectively record his findings. The examining physician is not, and cannot be expected to arrive at any definitive diagnosis as the full history and laboratory results will not be available to him. Medical history, laboratory results and physical examination findings will be evaluated by an independent diagnostician employed by the contractor. This diagnostician will formulate diagnoses and differential diagnoses, if appropriate. Additional procedures to treat or evaluate emergency or urgent medical conditions will be directed only by this physician. In addition, he will present a detailed analysis and debreifing to the study subject and provide a copy of the analysis to the subject's personal physician, if so requested. The physicians performing examinations fpr Project RANCH HAND II should be aware that the report of examination will become a permanent record. This report will be referred to not only in the near future as the cross sectional study is analyzed, but also at the time of the next review of the subject in the follow-up phases of Project RANCH HAND. These examinations will define the health status of the subjects at a point in time, and will establish the presence or absence of abnormal physical findings. After statistical review of the study groups, these findings may permit definition of a chronic effect due to exposure. An inaccurate examination may lead to falacious study results in two ways: a presumed syndrome may be defined which does not in fact exist, or a syndrome which in fact exists may not be defined with enough validity to warrant further actions. The examining physician is responsible for recording a complete and detailed report of the physical examination. In this role, the examining physician is tasked with collecting evidence of the presence or absence of physical signs of abnormality only. TheTformulation of diagnostic impressions by individual examiners is not requested nor desired. All items on the physical examination report form must be completed. It is imperative that the physician make such additional remarks as may be required to adequately describe existing physical and mental impairments. Since clinical endpoints have not been well defined following chronic exposure to Herbicide Orange, the examining physician and the diagnostician must not definitively ascribe abnormalities to herbicide exposure during the course of the examination or during the patient's debriefing. If, during the examination, the physician 138 �discovers evidence of acute serious illness requiring immediate treatment, the normal emergency or urgent care procedures of the medical facility would apply. Such care will be supplied at Air Force expense. If during the examination, there is evidence of illness requiring non-emergency medical attention, the diagnostician should inform the subject and offer to forward or have forwarded pertinent information to the subject's physician. A clear record of any such advice and treatment should be recorded. The ultimate value of the RANCH HAND II Study will lie 1n the collection of complete, accurate and, whenever possible, quantitative data permitting the most stringent arid ppwerful statistical analysis. For that reason, the physical examination protocol requires exact measurements in many Instances, and the use of defined meanings of semiquantitative indicators in other places, B. Conduct of the Examination (1) Upon arrival at the examining facility, the subject should be briefed by the on-site monitor on the appolntjjieflts which have been arranged, their times, and locations. (2) Collation and: forwarding of 0xam|nat1pn rejiiltft The monitor will complete a checklist fpr each study subject and review all medical information for quality and' completeness before forwarding to USAFSAM/EK, Brooks AFB, TX 7B235." 139 �C. ..Examination. Format_ PHYSICAL E X A M I N A T I O N SECTION 1. G E N E R A L A P P E A R A N C E . £3 Same As a . Appearance/Stated Age '/*y Younger Than £701der Than bi £7 Well-nourished £7 obese £7 Under-nourished £7 Older Than C. Appearance of illness or distress f~7 Yes £7 NO d- Hair Distribution /~7 Normal f~7 Abnormal e. Temperature SPECIFY: 2. Mf I « H T cm i 3. PULSE H A T E W E I G H T ;-lMdr..«.d) REGULAR: !_JYE5 SITTING BLOOD PRESSURE RfOHT ARM A T H E A R T L E V t L kg Describe any irregularities. [~J NO a. Irregular £^7" b. Irregularly irregular f~J c. VPBs per minute 4, EYE GROUNDS [~] NORMAL [~"; ABNORMAL £7 A-V nicking" £7 f light reflex /~~J Arteriolar spasm J. ARCUS SENILIS 6. CNT T<1[7J PRESENT f" j NORMAL Describe any vascular lesions, hemorrhages, exudates, £7 Hemorrhage*! p«pi»ed«n.. f~7 Exudates £7 Papilledema £7 Disk Pallor £J /• Cupping Q A8S«NT ^3. T~] ABNORMAL Abnormal Ocular Pigmentation /~7 Yes ^7 No Describe any abnonnality. Tympanic membranes intact £7 Yes £7 No Nasal ulcerations £7 No £7 Yes R £7 L £T 7. NECK (F.,p.ci,iir ihrntld timnd) '_J NORMAL !__• ABNORMAL, Describe any abnormality. Thyroid gland palpable £7 £7 Parotid gland enlargement Enlarged £*^ £7 R £7 L Nodules /~7 Tenderness 8. T H O B A X A N D L U N G S £7 Carotid pulses [~; NORMAL /™7 Asymmetrical expans'ion £7 Hyperresonance /"7 Dullness 9, H E A R T ' ~ ' NORMAL [ Q] ABNORMAL £7 Wheezes /~7 Rales 1 ABNORMAL Describe anv abnormality, especially basilar rales. Circumference at nipple level Expiration cm Inspiration cm Describe any enlargement irregularity of rate, murmur?;, or t h n l l s . Displacement of apical impulse £7 No £7 Yes Precordial thrust £7 No £7 Heart sounds normal £J Yes £J No £7Sl £JS2 £7S3 £734 Yes (Continued in Item 10 on Reverse) 10. ABDOMEN { ] NORMAL /~7 Heptomegaly ' cm Liver Span /~~7 Splenomegaly II. '-XTREMITIFS ;""! NORMAL ' ~j ] ABMOtMAL Deecribe any *«normality Vn«i special attention to the £7 Other mass - »pi««> •»<< liver. Record waist measurement Specify: on attached form. f~7 Tenderness >leen /"7 Other, specify: rj Liver n S£ Describe a n y ederna or signs of vascular insufficiency. \ ] ABNORMAL £7 Absence, specify: £7 Edema £7 Pitting /~7 Non-pitting £7 Loss of hair on toes £72 £7L £7 Clubbing of nails £7 Varicosities 140 �PHYSICAL EXAMINATION 56CTIOM io AM XL' PULSSS 6IMIN, H»D|*lFEMOftAL POPLITEAL OOKSALIS I}. SKIN i I NORMAL fndiS^ts ty|5S! and loeatidn of lesions on the / / OormatQgf'aphia /~7 Comedones /~7 Acneiforirt lesions f~7 Acneiform scars' /~7 Depigmentation C7 Inclusion cysts C3. Cutis Rhofflboidalis f k'ofatosis Futt-PaCe _....,.,. . p^'dfitfe phQtOS .«Mll. '.'-/ffiV. A7 Muscle - Specify: £7 weakness £7 Tenderness / Abnormal Consistency 7 £7 Atrophy tilt Inguinal hernia ,/yR Testes Absent Enlaorged /~7R /7 /^ ^ ^ Hemorrhoids £7 Prdstatic AtrO^hifc /^ effi 16. occipftAt, *7 Enlarged 17 7'^ERvo'Js' T?eri4e:r tem Pu Did OF ^_fiA» s..; itx». rrtass Confluent sVglffiM -"SEE ATTAfiMb '?!&" (Continued from ftectal �CLINICAL RECORD NEUROLOGICAL EXAMINATION HEAD AND NECK - Normal to Palpations/Inspection £JY £JN Specify Scar £7 Asymmetry £7 Carotid Bruit £7No £7R £7L Neck Range of Motion £7 Normal or Decreased to £7 Left Depression £7 £7 Right £7 Forward £7 Backward TRUNK MOTOR SYSTEM - Handedness Right £7 Left £7 Gait £7 Normal or £7 Broad Based £7 Ataxic £7Small Stepped £70ther-Specify Associated Movements £7An)1 Swing £7Normal or Abnormal £7R £JL Muscle Status (strength, tone, volume, tenderness, fibrillations) Bulk £7 Normal £7Abnormal Tone Upper Extremities £7Normal or £7Increased £7Decreased £7Right £7Left Lower Extremities £7Normal or /"/Increased £7Decreased£7Right £7Left Strength - Distal wrist extensors £7Normal £7Decreased Ankle/Toe Dors/Flexors £7Normal £7Decreased £JR £7L Proximal Deltoids £7Normal £7Pecreased £7& Hip Flexors £7Normal £7Decreased £7L £7R £7L Abnormal Movements (tremors, tics, choreas, etc.) Fasiculations £7N° Tenderness £7No £7Yes (1~^+) Tremor £7No /^JYes - Specify Upper Extremity £7R £7L>£7Resting : £7Essential Right Foot (c) £7Negative (Normal) Left Foot NonequiHbratory (F to N; F to F; H to K) Finger-to-nose-to-finger Normal £7Abnormal £7Right £7Left £7Both Knee-Shin £7Normal £7Abnormal"^£7Right £7Left /JBoth Succession Movements (including check, rebound, posture-holding) If indicated, check £7Normal £7Abnormal £JR £7R £ Rapidly alternative movements £7Normal Skilled Acts (a) Handwriting. (b) £7Intention tner Lower Extremity £7 R'£7y/I70 Coordination (a) Equilibratory - Eyes Open Eyes Closed - Romberg £7Positive (Abnormal) (b) £7Yes H-4-i-) If indicated, £7Normal ^Abnormal £7R £7L £7Botl1 £7Abnormal Speech (articulation, aphasia, agnosia) Grossly £7Normal £7Abnormal - Specify Dysarthria £7 Aphasia £7 142 �Reflexes (0-absent; 1-sluggtsh; 2-active; 1 . 3-very active; 4-translent clonus; 5-sustained clonus) R L R L Other L Abnormal L Deep ;R Deep : R. Babinskt Patellar Biceps Triceps Cremasteric Achilles Remarks MENINGEAL IRRITATION Straight Leg Raising ^Normal £7AbnormaT £7R £Jl NERVE STATUS (tenderness, tumors, etc.) SENSORY SYSTEM (tactile, pain, vibration, position. If positive sensory signs are present, summarize below and indicate details on Anatomical Figure, Std. Form 53T) Light Touch £7Normal £7Abnormal . , __,. . ,—,., ,, /Map on AnatpmlcaT Figure); n. Pin n Prick £7Normal £7Abnormal v ^ T ..T 3 -/ . Vibration (at ankle, 128 hz tuning fork:): ^NopnifT QAbnormal £3* Norma1 Position (Great toe): £7 Abnorma:1 £7 CRANIAL NERVES £7Absent I R Smell L II Smell £7Present Fundus R Normal r^Exudats [_ £7 Absent Abnormal £7 Disk. Pallor/atrophy i T1 edema £7Hemorrhag:e Fundus L Normal £7 Abnormal £7 Disk ., £7Exuc!ate £7 Papi 11 edema £7Hemorrhage Fields (to confrontation) Right £7Normal £7Abnormal I I Monnal £7 I l £7Abnorrnal: Left ^Abnormal - Specify Pupils-Size (mm) Equal £7 Unequal /~7 Difference mm Shape, position Round £7_ Other ff rjK n\. Light, Reaction Nprmal^_7 Abnormal £j £JK Position of Eyeballs Movements R Nystagmus Rotary rj (Draw position) L Horizontal £7 143 Vertical £7^ �XI Ptosis R£7 L £7 V Motor R Clench Jaw - Symmetric £7 Deviated £7 Sensory R Normal £7 Abnormal £7 V]£7 R V] L Normal £J Abnormal £7 V-| Corneal Reflex R L VII Motor R Normal smile £7Yes £JNo L Normal smile £7Yes £7No R£7 L£J V2£7 Palpebral Fissure £7Yes £7No Pa lPebfal Fissure £7Yes £7No IX Palate and Uvula X Movement Normal £7 Deviation to £7R £7L Palatal Reflex R £JNormal /^Abnormal XII • L ^Normal- /"/Abnormal Tongue-Protruded-Central £7 R £7 L £7 Atrophy £7No £7Yes MENTAL STATUS (alert, clear, cooperative, etc.) Gross abnormalities: £7No £7Yes - Specify 144 �DIAGNOSTIC SUMMARY SYNOPSIS OF POSITIVE FINDINGS' Medical History: Physical Examination: General Dermatologic Neurological Psychological Laboratory Results: Diagnosis: Differential Diagnosis, if applicable: Date Signature of Diagnostician 145 �D. Special Procedures (1) Nerve Conduct1 on Velocities (a) These studies have been determined to be an important parameter in long-term follow-up studies of persons thought to have been exposed to Herbicide Orange components. (b) The Nerve Conduction Velocities should be performed by a physician or by a specialty qualified technician under the supervision of a physician trained in neurophysiological methods. (c) Specific NCVs (See form included in F. Below) (1) Ulnar Nerve (one side only) (a) motor (above elbow, below elbow) (b) values recorded (i) distal latency (11.) NCV (2) Peroneal Nerve (one side only) (aj motor (b>) values recorded (i) distal latency (ii) NCV (1) Sural Nerve (one side only) (a) (b) (d) sensory: orthodromic values recorded: NCV Methods 146 �PERONEAL NERVE (_!) Active electrode is placed over the extensor digitorum brevis and reference over the little toe. Stimulating electrodes are placed over anterior distal leg 8 cm proximal to active electrode. Proximal site is distal to head of fibula. If entrapment is suspected at fibular head use a stimulation site of 12-18 cm more proximal to the fibular head. Anomalous innervation to the extensor digitorum brevis occurs in 1/5 patients (at least partially). Identified by inability to evoke a muscle action potential when stimulating at anterior ankle or a different shape (smaller) potential when stimulating here. This accessory nerve causes posterior to lateral malleolus so cathode should be placed here. NORMAL VALUES 49.9 ± 5.9 M/sec Distal latency: 4.5 ± .8 ms Proximal latencies have been determined for use in below the knee amputees, and neuromuscular diseases where extensor digitorum brevis action potential cannot be elicited. Active electrode is placed 1/2 way down leg over middle of dorsiflexor muscle group and stimulation at fibular head. NORMAL VALUES 5.5 - 7.2 ms (N = 217) 147 �SURAL NERVE (2) Active and recording electrodes are placed under lateral malleolus on lateral aspect of ankle. Sural nerve is stimulated as it pierces the gastrocnemius fascia just lateral to the midline of posterior distal calf, 10-18 cm proximal to active electrode, If leg is cold - a clue is prolonged latency of peroneal nerve - determine temperature. Subtract .1 ms (latency of activation) from the observed latency and divide into the distance. NORMAL VALUES (after LaFratta) (To Peak) Age 20-29 30-39 40-49 50-59 60 & over 44 38.80 36.70 37.20 35.00 ± ± ± ± ± 2.5 3.3 3.7 3.0 3.8 M/sec M/sec M/sec M/sec M/sec UJ > cc UJ 2 148 �ULNAR NERVE MOTOR CONDUCTION (3) Active erence radial trode. and 18 electrode is placed over center of abductor digiti quinti; refover proximal phalanx fifth digit. Stimulation (cathode) just to tendon of flexor carpi ulnaris 8 cm proximal to active elecProximal site of stimulation should be just below ulnar groove cm proximal to ulnar groove on medial aspect of humerus. N. B.: Elbow should be flexed to 70 degrees during procedure of stimulation and measurement to make more precise the actual length of ulnar nerve. More proximal stimulation sites include supraclavicular and C-8 root (see median nerve). SENSORY CONDUCTION Antidromic - ring electrodes over fifth digit separated by 4 cm. N.B. motor artifact may be interfering. Stimulate 14 cm proximal to active electrode at same site as motor stimulation. Orthodromic - reverse stimulation and recording electrodes. More proximal sites of stimulation may also be done. NORMAL VALUES 57 ± 62.7 56.7 54.9 4.7 M/sec ± 5.5 M/sec ± 4.2 M/sec ± 3.9 M/sec motor forearm segment - motor across elbow segment - sensory orthodromic (to peak) - sensory antidromic (to peak) Distal Latency: Motor: 3.7 ± .3 Sensory: 3.0 ± .25 Antidromic (peak) 3.0 ± .25 Orthodromic (peak) Muscle AP 8-20 mV Sensory AP 15-50 mV ADDENDUM For deep branch surface recording electrode should be over adductor pollicus (i.e. just medial to thenar eminence on palmar surface of web space). Additional latency is .5 ms. 149 �ACT REF \\ i ULNAH NERVE ! V4 REF 150 �(e) Disposition Forward the recorded results on the form attached to the examination package to the diagnostician. (2) Psychological Battery (a) General (1) This battery yields objective numerical data, and is well-standardized and clinically validated. The individual tests were chosen to insure an adequate analysis of one of the major alleged manifestations of herbicide toxicity. Each test either validates one of the other tests, or is considered to be a "definitive" test for analyis of a suspected psycho/ neuropathic effect. (2.) Compared to the general civilian population, characteristic response tendencies are observed on the MMPI and Cornell Index among active duty aircrewmen being evaluated in an aeromedical setting. It is also important to consider the effect that pending retirement has exerted on the reporting of medical history and symptomatology. This may also alter responses to psychological testing. (3) The battery requires approximately 5-1/2 to 6-3/4 hours to administer, depending on the speed of the examinee. An additional 1 to 2 hours of scoring and other clerical tasks will be required. Si rice test debriefing to clarify unusual performances, response biases, etc., is a crucial part of the psychologic evaluation, it is recommended that testing begin and be completed as early as possible during each examinee's stay at his respective evaluative facility. (b) Specific Tests (!) Nechsler Adult Intelligence Scale (MAIS): 60-75 minute individually-administered collection of verbal and nonverbal intellectual measures; also useful for clinical inferences when combined with the neuropsychological battery below. (2) Reading subtest of the Wide Range Achievement Test (WRAT): 10-minute individually-administered measure of word recognition ability. Important to rule-out reading inefficiency should response to personality instruments below be of questionable validity (e.g., high F Scale on MMPI). (3_) Halstead-Reitan Neuropsychological Test Battery: 150-180 minute individually-administered collection of brain behaviorrelationship measures for establishing the functional integrity of the cerebral hemispheres. The battery must include the following subtests: Category, Tactual performance, Speech-Sounds, Seashore Rhythm, Finger Tapping, Trail 151 �Making, and Grip Strengths. The Aphasia Screening and Sensory-Perceptual Exams are considered optional in view of their redundancy with the clinical neurologic exarn included in this project. Individualized test debriefing is conducted to clarify test performances in the WAIS and Neuropsychological Battery. (4) Three subtests of the Uechsler Memory. Scale I (HMS I): 30-minute individually-administered measures of immecRate and delayed recall of verbal and visual materials. The Logical Memory, Associate Learning and Visual Reproduction subtests are to be administered in the standard, immediate-recall fashion initially. After 30 minutes has elapsed, the examinee is asked, without prior alerting, to recall as much as he can about the Logical Memory and Visual Reproduction subtest stimuli. Standard scoring is used for both test-retest administrations. (5) Cornell Index (CI); 10-15 minute selfadministered and standardized neuropsychiatric symptom and complaint inventory, including items involving asthenia, depression, anxiety, fatigue, and GI symptoms in lay language. Endorsement of items are to be explored and clarified in testdebriefing. (6) Minnesota Multiphasic Personality Inventory (MMPI): 60 to 90 minute selT administered clinical psychiatric screening instrument; also capable of estimating response biases (e.g., "fake good," or "fake bad"). The shortened version of Form R (i.e., items 1 to 399) may be substituted for the 566-item Long Form. Standard scoring and Minnesota norms are to be used, with the possible exception of active duty examinees where USAFSAM aircrew norms may be applied. Clarification of profiles showing response biases, questionable validity, and/or unusual item endorsements will be conducted in individual test debriefing. (c) Examination Results Forward all test materials as scored with annotations, interpretations, and impressions to the diagnostician for inclusion in the subject's examination file. (d) Psychometric?; Special Instructions (1) For the Cornell Index and MMPI, each subject is instructed: (a) to answer carefully every item; and (b) that wherever applicable, his responses should reflect personal experiences, beliefs, preferences, etc., only for the time period between his combat tour in SEA and the date of testing. These instruments are not to be group administered and a reasonable amount of privacy should be provided. These instruments should not be completed at the subject's overnight quarters nor anywhere else outside the supervised confines of the evaluative facility. 152 �(2) If a subject's measured word recognition falls below the 6.5 Grade Level (Raw Score=40, Level H) according to the WRAT Reading subtest, the Cornell Index and MMPI are read aloud or administered via tape recording. In such cases, the subject retains the right to mark his answer sheet outside the view of the examiner or of others within hearing distance. (jB) All eleven subtests of the WAIS are administered, i.e., pro-rating of subtests is not allowed. The scoring of WAI$ subtest items, and the operations of summing, transferring, and finding Raw Scores, Scaled Scores, and Tabled IQ values are double-checked for accuracy by the Psychologist in charge (or his/her appointed representative) before the raw data are forwarded to the diagnostician. (4) Precautions similar to those in #3 above are exercised in the scoring and other clerical tasks associated with the Halstead-Reitan, WMS I, WRAT, Cornell, and MMPI. (5i) For the Halstead-Reftanj use as the preferred, or dominant, hand the one which the subject uses most in writing. If 1n doubt, administer a "Name Writing Test", where the subject Is simply asked to write his name in a normal manner as though sighing a personal check. The examiner measures the time for each hand to perform, (without alerting the subject to the timing), and assigns dominance to the quickest hand. (6) For the grip strength measure, report the average, 1n kilograms, of 3 brief, but maximum, squeetls of the dynamometer for the preferred and the non-preferred hands. Alternate hands between trials. (7) The Psychologist in charge will conduct a one-to-one test debriefing with" each subject to estimate the test-by-test and overall accuracy and validity of the test results. A prepared form is provided for this purpose, and should be filled out completely before forwarding, with the subject's raw data, to the diagnostician* If applicable, input from the testing technician utilized is encouraged. (3) Elect roca rd i og ram (a) A standard 12-lead scalar eleetrogram is required. If an arrhythmia is observed, a one minute rhythm strip 1s requested, in addition. The eleetrogram will be done following a minimum fast of four hours. Mounting; ' Mount the tracing in the usual manner of the laboratory for the recorded used. ( c ) Pis posit ion: Forward the mounted tracing and rhythm strip, if obtained, to the diagnostician. 153 �(d) Interpretation: The electrocardiograms will be interpreted by cardiologists at the examination center, and forwarded to USAFSAM/NG where physicians in the USAF Central ECG Library will compare the tracing to previous individual ECG records in the case of rated (pilot or navigator) subjects. (e) Disposition (USAF Central ECG Library); (1) Pilots and Navigators - The original tracings will be microfisched and added to the individual s permanent record. (2) Enlisted Subjects - The original tracings will be microfisched and a permanent record established for each individual. (4) Radiographic Exami nation (a) Examination A standard 14x17 in,, standing, roentgenogram in the PA position using small nipple markers will be accomplished. (b) Interpretation A board-certified radiologist at the examination center will interpret the roentgenogram and record the results and forward them to the diagnostician. (5) Pulmonary Function Studies Standard evaluation of vital volume at 1 second will be performed. (6) capacity and forced expiratory ""' • . Laboratory Procedures (a) General Instructions; First Day (I) The patient should report in the morning in a fasting state having had water only after midnight. The patient will have been requested to eat approximately 150 gms of carbohydrate each of the three preceding days and to consume no alcoholic beverages. Non-compliance is not a contraindication to drawing the blood specimens. However, a notation of extent of noncompliance should be made by the examining physician to aid in the interpretation of the results. (b) General Instructions; Second Day Serum hormone levels should be determined from specimens collected on the morning of the second day. Hormonal levels appear to oscillate rapidly in a random fashion. Distributions drift with time suggesting 154 �diurnal variations and some are affected by nonfasting state. Therefore, patients should be fasting prior to drawing blood for hormone analysis. (c) Specific Tests.to be Perforrned, on a11_ Participants (I) Hematocrit (2!) Hemoglobin (3) RBC Indices (4) While Blood Cell Count and Differential (5) Platelet Count (6) Erythrocyte Sedimentation Rate (7J Urinalysis (<3) Semen Analysis (Number, $ Abnormal, Volume) (9J Blood Urea Nitrogen (10) Fasting Plasma Glucose (jU) Creatinlne (J.2^) 2-hour Post Prandial! Plasma Glucose (_13) Differential Cortisol (0730 and 0930 hours) Q4) Cholesterol & HDL (15) Trlglycerides (JL6) Bilirubin, Total and Direct (17.) SGOT (18) SGPT Q9) GGTP Alkaline Phosphatase LDH Serum Protein Electrophoresls (23) CPK 155 �( 4 VDRL 2) •f.*' (25) • ' ' " ' ' ' ' LH FSH Testosterone (28) Thyroid Profile (RIA) (T 3 , T4, TSH.FTI) (29) Delta-aminolevulinic Acid .-> . (30) Urine Porphyrins (ID ," Hepatitis B antigen/antibodies (HBsAg, anti HBcAg, anti HBsAg) Prothrombin time • (33) * *• , . . ' . . Blood Alcohol ' ' • •'''••''. (d) Tests to be performed on selected subjects (1) Anti-nuclear Antibody on subjects with evidence of autoimmune disorders ;-, (2) Hepatitis A Anti gens /anti bodies for those with current or past liver disease Karyotyping for those fathering children with birth defects ~ (4) Skin photography and skin biopsy on subjects with suspected chloracne (5) For those whose medical history indicates an increase in infectious diseases (aj Immunoelectrophoresis (b^) Quantitative Immunoglobulin Determinations (6) To be performed on a randomly selected group of study subjects (a) Enumeration of B and T cells (b) Enumeration of Monocytes (c) B and T cell function tests 156 �(e) Rat i ona 1 e .for. 1 a borat ftry (I) studies on the toxicity of TCDD in animals have shown that the following organ systems are damaged: (a) Liver: Hepatic necrosis, Mver enzyme changes, •hypoprotelnemia, hypercho1esterd1em1a, hy pe rt r i g 1 y ce r 1 dertii a , (jb) Reticuloendothflial System: altered cellular immunltyT decreased lymphocyte; counts,. (c) leukopenia, pancytopenia."" Hemopoletic (d) Endocrine adrenal cortex, hypothyroTdisiti. (e) Renah Sy$tenii Systet; Thyiwfc IfropHy, Anemia, throiiibocytopenla, Hemorrhage lncr§a§6 |n blood urta andatrophy of nftrogen. (f) in addition, statistically significant increases in hepatocellular carcinomas (liver) and s4tJamd;Celtylar eareinomas of the lung were found. (2) Studies on the t0^|| effects of 1P00 In man hwe shown that the following ofgan §y§teffls are skin: Chloracne, hlrsutisffi. (b) Liver: PorpHyfia cutanea tarda,, levels of transamlnase and" of GGTP. Eftl a rpd, tender liver, hyperlipiderttia. (£) Renal i Hewof^aflc cystitis* foeal tis. (dj Neurornuscular Systetn: Asthenia, i*e., headache, apathy, fatigue, anorexia, weight loss* sl&ep disturbances, decreased learning ability, decreased memory, dyspepsia, sw&atirljj, muscle pain, Joint pain and sexual dysfunction. (ej Endocrine Systertl: (3) Based upon the repor^f-of toxic effects in animal and human exposures, the following organ panels ari* recommended: (a_) Heiiiopoietic (b_) Reticuloeridothelial �(cj Renal (d) Endocrine (e) Neuromuscular Hemopoietic screening should include: (aj Hematocrit (b) Hemoglobin (cj RBC indices (c[) Erythrocyte sedimentation rate (e) Platelet count (_f) Prothrombin time (5) Reticuloendothelial system: (a_) White blood cell count (b) Differential (c) Serum protein electrophonesis (d) Selective use of immunoelectrophoresis and quantitative immunoglobulin determination (e) (6J B cell and T cell counts and functions Hepatic screen: S60T SGPT (£) G6TP (d^) Bilirubin, Total and Direct (e) Alkaline phosphatase (f) LDH (£) Cholesterol 158 �(hi) HDL ( j Triglyceride j ( j Urine prophyrins j (kj Urine porphobilinogen (_]_) Hepatitis B antigens/antibodies (HB^Ag, anti1 HBCAG, anti HBsAg3) (7_) Renal screen: (a) Urinal ysis (b^) BUN (c_) Creatinine (8) Endocrine screen (a) Differential cortisol (0730 and 0930 hours) (b) Thyroid profile (RIA) (c) Fasting plasnif glucose Neuromuscular system: CPK Elucidation of sympoms of asthenia: (a) Testosterone ( ) FSH £ (11) The following tests should be pefprmed only as followup for abnormalities in the history or physical examination findings: (aj HAVAB (IgG and IgM) (b) ANA 159 �E. Forms Anatomical Figure (Anterior) Anatomical Figure (Posterior) Nerve Conduction Velocities Psychometric De-Briefing Form 160 �;AL RECORD ANATOMICAL FIGURE 161 �162 �NERVE CONDUCTION VELOCITIES SOCIAL S E C U R I T Y NUMBER GRADE NAME il.nsi, First. All C,ASg NR, :?A fr 0^" E X A M I N A T I O N MONTH DAY TEMP: DATE: £Z7L • Elbow 1. Ulnar (one side only) £ 7 Above ^ /**"7Below Nonnal Values for Laboratofy Latency / / /./ Distance I N.C.V. /___ I Stm. Curr. 2. I I / ms I mm I.I I m/s / / .../.. / mV /......LJ......-J Peroneal (one side only) /""'/R Normal Values for Laboratory Latency / / . / . / / ms /..,/,/,/,_ ,/ Distance / / / / mm /. /.... 7.... J N.C.V. I / / /./.../ m/s Stm. Curr. 3. I I I I. ../,/. / I m. V Sural (one side only) (If unobtainable, Median or Ulnar Sensdty recommend) / "/R Normal Values for Laboratory Latency / / Ld-—L ms /...../ . . y . ./ .. Distance / I I I m m N.C.V. / Stm. Curr. /.././ / / / m/s / / mi / / 163 LJ �Ranch Hand II: Psychometric De-Briefing Form _______„ R L Test Date EvaT FacTTity Handedness Subject: Name Psychologist/ De-Briefer : Title Name Degree Yes No Clin/Couns Cert/Lie Testing Technician: Name Same as above Degree Test/Experience (Yrs) jnstructions In the appropriate column below, indicate the test-by-test validity of the psychometric results based upon the Examiner's observations of the subject during testing and upon the Psychologist's evaluation of the data in test debriefing with the subject. Use the numbered factors below to indicate the reason(s) for questionable validity among any of the data. For datum thought to be of questionable validity, also provide an estimate of the subject's "true" score or result. Forward the completed form with the subject's raw data. Reasons for Questionable Validity 1. Poor reading comprehension 2. Fatigue 3. Neg attitude, angry, marginal cooperator 4. Careless, hurried responses 5. Examiner Error Test Score 6. Exaggeration of complaints ("fake bad") 7. Minimizing complaints ("fake good") 8. Disorganized personality (Psychotic) 9. Physically ill (flu, venipuncture effects, etc) 10. Other (Specify ) Valid Results WAIS VIQ PIQ FSIQ 2, WRAT Reading 3, Halstead-Reitan Category Test 164 Reason(s) for Questionably Valid Results Est of "True" Score/Result �Ranch Hand II: Test Score Psychometric De-Briefing Form Continued Valid Results Reason(s) for Questionably Valid Results Est of "True" Score/Result Tactual Performance Test Preferred Hand Non-Preferred Hand Both Hands Memory Localization Speech-Sounds Perception Seashore Rhythm Finger Tapping Preferred Hand Non-Preferred Hand Trail Making Test Part A Part B Grip Strengths Preferred Hand Non-Preferred Hand WMS I Logical Mem (immed) Visual Repro (immed) Associate Lrng Logical Mem (delayed) Visual Repro (delayed) 5. Cornell Index 6. MMPI (overall rating of protocol) WNL or ONL 165 �XVII. ANNEX 1 - EXPOSURE INDEX CONSTRUCTION When exposure concepts were Intlally discussed, the principal investigators were optimistic about the feasibility of developing an exposure estimate or index which was specific to the individual study subject. However, as the investigators became more familiar with the operational environment of the Vietnam War and the limitations of the personnel records system, it became obvious that a validated individual-specific exposure index could not be developed. This specific index was dependent on the availability of operational records containing individual flying time data and aircraft maintenance records containing the names of ground support crew members. Because this data was unobtainable, a less specific exposure concept was then developed. This index was to be base-specific rather than individual-specific, and is the index presented in the protocol in this technical report. However, further inquiry disappointingly showed once again that available data sources would not provide adequate information to support the construction of this index. The base-specific index relied upon records to provide a quantitative measurement of the number of missions and amount and type of herbicides sprayed by air crews from each base. This Index assumed that all personnel at a given base shared equally in the workload. Unfortunately, the "Herbs tapes" did not specify the base to which spray aircraft were assigned, and the military personnel records did not definitively specify the exact duty locations of all personnel. Thus, a still more generalized exposure concept was necessary. Although the more refined indices could not be validly applied, it was feasible to develop an exposure index for this study which can be validated, fulfills the requirements of the study design, and is fully supported by available data sources. A crude index can be developed and applied universally to all exposed subjects, regardless of their assigned duties in Vietnam. This index is based solely on the amount of dioxin disseminated throughout Vietnam each month from January 1962 through April 1970. The data to support this index are based on a comprehensive listing of herbicide missions, being developed by the Department of Defense, and estimates of the TCDD content of 2,4,5-T over time. These estimates are being developed and refined at USAFSAM at the present time. A refined exposure index for ground crew members is also feasible and is under development. This index builds on the crude index and takes the experience of ground crew members into account. It also assumes that each individual assigned to these duties in the Vietnam theater carried out his share of the workload in his specialty. This "experience factor" is constructed by dividing the total number of herbicide spray sorties flown during a subject's tour of duty by the number of individuals performing the subject's duties during the period of his tour. Similarly, a refined air crew index can be constructed. This index expands the concept of the ground crew index by including a factor reflecting the variable levels of exposure within the C-123 aircraft. Simulant studies to quantify these differences were conducted, and plans are underway to repeat these studies to revalidate the conclusions. 166 �Symbolic representations of these indices are shown below: Crude Exposure Index: WTT «-*L ( TCDD E l = sprayed in the / Vietnam Theater during) the 1th subject's tourl •u. *r Refined Ground Crew Exposure Index: GND Herbicide sorties performed in the Vietnam theater during the ltn subject's tour TCDD sprayed in the Vietnam theater during the 1th subject's tour • > Number of ground crew personnel (job specified) in the Vietnam theater I during the 1th subject's \ "our _J Refined Air Crew Exposure Index: /— AIR \ Herbicide sorties performed \ in the Vietnam theater during the 3*" subject's J tour H \TCDD sprayed in / I the Vietnam I {theater during V the 1th subject's /tour \ Number of airmen with subject's duties in the Vietnam theater during the ltn subject's tour ««-«/ V™, r X \Crew positioni <weight of the] /1th subject L. (total amount) X (experience) X (intensity) The data required to support these indices are either currently available or are in the final stages of development. These indices are feasible and will adequately support the analytic strategy of the study design. 1.67 �XVIII. ANNEX 2 - COMPARISON GROUP ^ELIGIBILITY A central element of epidemiologic research is study population ascertainment. Incomplete population ascertainment always carries with it the possibility of serious selection bias which cannot be corrected using statistical procedures. Complete ascertainment of the exposed and comparison populations occurred through a manual review of military personnel records from 1962-1964, combined with a computer tape generated by the Air Force Human Resources Laboratory (AFHRL). This computer tape was based on retrieval parameters identified to AFHRL by the United States Air Force School of Aerospace Medicine (USAFSAM) principal investigators. The retrieval process required computer searches of multiple Air Force Military Personnel Center tapes spanning the time period of January 1965 through December 1971. In November of 1980, AFHRL delivered to USAFSAM a tape that was thought to contain the total eligible study population. The study match was completed and the selected individuals were contacted to participate in the study. In December of 1981, Louis Harris and Associates, the questionnaire administration contractor, notified the USAFSAM investigators that several of the participants had reported no experience in Southeast Asia, suggesting that there had been overselection. Review of these participants' military personnel records clearly revealed that they were comparison subjects who had not had Southeast Asia experience. In order to maintain the integrity of the questionnaire implementation and the physical examination contract, it was necessary to implement a modification of the replacement strategy which had been originally designed for use with control subjects who refused to, participate in the study. It had been intended that the noncompliance questionnaire be given to both the replacement and the refusing subjects, and that they would be matched for equivalent health perception prior to implementing this strategy. However, the early requirement to replace these ineligible individuals did not allow the use of the noncompliant instrument. The eligibility of replacement candidates was verified and these valid subjects were entered into the study. Inappropriate subjects were informed of this selection error and excluded from further participation in the effort. Two hundred eleven inappropriate subjects had been interviewed, and 26 had been examined. This situation also necessitated an immediate manual review of the personnel records of all individuals for th.e comparison group. The review of records was completed in March of 1982 and the verification of this process was initiated. The objective of this quality control effort was to verify the eligibility of the comparison group by subsampling techniques and to insure that errors in excess of one percent ineligibility did not exist. The estimated error rate was found to be 0.00748% with confidence bounds of 0.00340% and 0.0312%. To further reduce this error rate, each replacement candidate's personnel records were re-evaluated prior to forwarding his name to the questionnaire contractor, thereby assuring that all replacements were absolutely eligible for the study. The overall review demonstrated that 18% of the 12,193 individuals in the original control population were erroneously included. These ineligible subjects were randomly distributed throughout the C1-C10 matrix. Two percent of this error was due to inaccurate data on the USAF personnel tape and 16% due to incorrect cohort selection specification and/or computer search implementation. All errors were in the direction of overselection, due to the inclusion of non-Southeast Asia C-130 units in the specifications. 168 �Following the removal of the ineligible subjects from the cohort matrix, the empty positions were then filled by valid comparison subjects with higher cohort numbers, thus constituting a leftward shift of the matrix. This process was reviewed by the subcommittee of the Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants and members of one of the other peer review groups prior to implementation, and its use was found to be totally acceptable. Its use resulted in a reduction of the study from 1:10 to a 1:8 design. Monte Carlo studies using current physical examination compliance rates showed this collapse to have not significant impact on statistical power in the followup phase of the study. Although the shift-left process constituted an unplanned use of the replacement strategy, it permitted the continuation of both the questionnaire and physical examination contracts without disruption and with total validity. 169 �XIX. ANNEX 3 - SUPPLEMENTAL ANALYSES '*• ' ' " ' : • • • . - : • The study is of distinct benefit to the herbicide-exposed group since it may provide the individuals with an early warning of herbicide effects if they are occurring; or if no herbicide effects are uncovered, the study can provide some peace of mind by contributing to settlement of the public controversy. The study, however, is also of very significant benefit to unexposed individuals participating in the effort as comparison subjects. These additional returns occur because of the nature of the study design and the analytic flexibility inherent in that design. Except for the skin condition called chloracne, none of the disease entities that have been related to herbicide exposure are unique to that exposure. Processes such as peripheral neuropathy, teratogenesis, and carcinogenesis have been reported in laboratory studies with animals or in epidemiologic studies of herbicide; but these processes also occur somewhat commonly in general populations without herbicide exposure. Thus, to determine the occurrence of a true herbicide effect, this epidemiologic study is gathering data on other factors known or suspected to produce disease, and which could obscure herbicide effects. Among these potentially confounding factors are several military and civilian occupational exposures to chemical, physical, and biologic agents including: asbestos, x-ray or nuclear radiation, industrial chemicals, insecticides or pesticides, and prior infectious disease processes. By studying possible correlations between these factors and disease processes, benefits accrue to both the herbicide-exposed and unexposed subjects. Correlations between disease incidences and other potentially causative factors will be sought using statistical data-processing techniques such as multivariate regression or analysis of variance. This approach will identify herbicide effects in a fair and equitable manner as described in the protocol, but it will also provide additional medical data of significant direct interest in its own right. 170 �Glossary of Abbreviations ABBREVIATION AFSC ALK PHOS AMD AV-GAS BUN C-7 C-123 C-130 CBC CPK CSF Det 1 AMD DNA DOD 2,4-D ECG EPA FBS FSH G.I. GAO GGTP HDL Herbicide Orange Herbicide Pink Purple Green JP-4 LDH LD50 LH NCI MMPI PACER HO PACER IVY RANCH HAND DEFINITION Air Force Specialty Code Alkaline Phosphatase Aerospace Medical Division, Brooks AFB Texas supervises all medical research activities within the Air Force Systems Command Leaded Aviation Fuel (Reciprocating Engine) Blood Urea Nitrogen USAF Cargo Aircraft 2 engine, Propeller, Reciprocating USAF Cargo Aircraft, 2 engine, Propeller, Reciprocating USAF Cargo Aircraft, 4 engines, Turbo-Propeller Complete Blood Count Creatine Phosphoktnase CerebTOSpinal Fluid Onslte physical examination contract monitor Deoxyrlbonuclefc add Department of Defense 2,4-dlicftlorophenoxyacetic actd Electrocardiogram Eiwironmefttal Protection Agency Fasting Blood Sugar Fol1 tele Stlmu1 at1ng Hormone Gastrointestinal General Accounting Office Glutaryl-glutamic Transpeptidase High Density Lfpid Mixture of 2,,4-B and 2,,4,,5-T contaminated with TCDD Other 2,4,5-T/TCDD-conta1n1ng herbicides Jet Fuel Lactose Dehydrogenase (Median) Lethal Dose for 50% of Tested Animals Luteinlztng Hormone National Cancer Institute Minnesota Multiphasic Personality Inventory Code Name for the Herbicide Incineration Project Code Name for the Movement and Storage of Herbicides at Johnston Island USAF Organizational Code Name for the Defoliation Operations in Vietnam 171 �ABBREVIATION DEFINITION RBC RIA RVN SEA SGOT SGPT SMR SSS SYSTO TCDD TOY TLV 2,4,5-T USAF USAFSAM Red Blood Cell Radio-immune Assay Republic of Vietnam Southeast Asia Serum Glutamic Oxaloacetic Transaminase Serum Glutamic Pyruvic Transaminase Standardized Mortality Ratio Sensation Seeking Scale Systems Officer 2,3,7,8-tetrachl orodi benzo-p-di oxi n Temporary Duty Threshold Limit Value 2,4,5-trichlorophenoxyacetic acid United States Air Force United States Air Force School of Aerospace USSR VA VDRL/FTA WAIS WRAT ; Medicine Union of Soviet Socialist Republics Veterans Administration Serdlogical Tests for Syphilis Wechsler Adult Intelligence Scale Wide Range Achievement Test 172 �trr Special i rjjjCw iiia.f (1020 �USoF SCVOOL OF AEROSPACE MEDICINE AEROSPACE MEDSCAL DIVISION (AFSCJ BROOKS AFB IX 78235 PEN/ THIRD-CLASS SULK RATc MAIL POSTAGE & FEES PAID USAF PERMIT NO. G-l � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 058 Folder The folder containing the original item. 1570 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. Patricia M. Moynahan Richard A. Albanese William H. Wolfe Description An account of the resource <strong>Corporate Author: </strong>United States Air Force School of Aerospace Medicine, Aerospace Medical Division (AFSC), Brooks AFB, Texas Date A point or period of time associated with an event in the lifecycle of the resource 1982-12-01 Title A name given to the resource Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Study Protocol Subject The topic of the resource Air Force Health Study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/6647a4ccc4bb3f04be3584b6842a41b3.pdf e50578ba488ff84f65b68f969f4f9a74 PDF Text Text Item ID Number 01557 Lathrop, George D. United States Air Force School of Aerospace Medicine, Report/Article TitlB Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Baseline Questionnaires Journal/Book Title Year 1982 Month/Day November Color CI NumDeroflniaQ.es 289 Descplpton Notes Wednesday, May 23, 2001 Page 1568 of 1608 �Report SAM-TR- 82-42 EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES: BASELINE QUESTIONNAIRES George D. Lathrop, Colonel, USAF, MC Patricia M. Moynahan, Colonel, USAF, NC Richard A. Albanese, M.D. William H. Wolfe, Lieutenant Colonel, USAF, MC November 1982 Initial Report for Period September 1981 — November 1982 Approved for public release; distribution unlimited. Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 USAF SCHOOL OF AEROSPACE MEDICINE Aerospace Medical Division (AFSC) Brooks Air Force Base, Texas 78235 �NOTICES This initial report was submitted by personnel of the Epidemiology Division and the Data Sciences Division, USAF School of Aerospace Medicine, Aerospace Medical Division, AFSC, Brooks Air Force Base, Texas, under job order 2767-00-01. When Government drawings, specifications, or other data are used for any purpose other than in connection with a definitely Government-related procurement, the United States Government incurs no responsibility or any obligation whatsoever. The fact that the Government may have formulated or in any way supplied the said drawings, specifications, or other data, is not to be regarded by implication, or otherwise in any manner construed,, as licensing the holder, or any other person or corporation; or as conveying any rights or permission to manufacture, use, or sell any patented invention that may 1n any way be related thereto. The Office of Public Affairs has reviewed this report, and it is releasable to the National Technical Information Service, where it will be available to the general public, including foreign nationals. This report has been reviewed and is approved for publication. 5 ATRICIA M. MO^AHAN, Colonel, USAF, NC Project Scientist ROY L. DEHART Colonel, USAF, MC Commander GEORGE D.TATHROP, Colonel, USAF, MC Supervisor �: UNCLASSIFIED SECURITY CLASSIFICATION "Of THIS PAGE (When Dale Entered) READ INSTRUCTIONS BEFORE COMPLETING FORM REPORT DOCUMENTATION PAGE 1. REPORT NUMBER •2. GOVT ACCESSION NO 3. RECIPIENT'S C A T A L O G NUMBER SAM-TR-32-42 ». TITLE (*** subtin*) EPIDEmCnrOinr INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES: BASELINE QUESTIONNAIRES 5. TYPE OF REPORT & PERIOD COVERED Initial report Sen 1981-Nov 1982 6. PERFORMING O^G. REPORT NUMBER 3. CONTRACT OR G R A N T NUMBERS) George b. Lathrop, Colonel, USAF, MC Patricia M, Moynahan, Colonel, USAF, NC Richard A. Albanese. M.D. William H. Wolfe. Lieutenant Colonel. USAF, MC 9. PERFORMING ORGANIZATION NAME AND ADDRESS USAF School of Aerospace Medicine '(EKO) Aerospace Medical Division (AFSC) Brooks Air Force Base, Texas 78235 H. 10. PROGRAM ELEMENT, PROJECT. TASK AREA 4 WORK UNIT NUMBERS 65306F 2767-00-01 CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATS November 1982 USAF School of Aerospace Medicine (EKO) Aerospace Medical Division (AFSC) rQnk'siin Air Force S.ranks' A j.,r FQt Base. Taxas 78235 . 13. NUMBER OF PAGES I*. . MONITORING AGENCY NAME 4 AOORESSf*/ diUerent titan Controlling Office) MONITORING AGE: 284 IS. SECURITY CLASS, (at this report) Unclassified l!«. OECLASSIPICATtON/DOWNGRADING SCHEDUUE is. DISTRIBUTION STATEMENT (at M* Approved for public release; distribution,unlimited. 17. DISTRIBUTION STATEMENT (oi ttiu tbetrlxt ontfred In Slock 30, It different from Report) ie. SUPPLEMENTARY NOTES 19. KEY WORDS (ConHnuv on reverse aide il neceiifty and identity by block number) Questionnaires Baseline Data Collection Herbicides Herbicide Orange Study Subject Questionnaire Spouse Questionnaire Dioxin Next-of-Kin Questionnaire Epidemiologic InvesMiniquestionnaire tigation Morbidity Phenoxy Herbicides 20. ABSTRACT (Continue on reverse aide II neceetery end Identity by block number) In 1979 the United States Air Force (USAF) made the commitment to Congress and to the White House to conduct an epidemiologic study of the possible health effects from chemical exposure in Air Force personnel who conducted aerial herbicide dissemination missions in Vietnam (Operation RANCH HAND). The purpose of this epidemiologic investigation is to determine whether long-term health effects exist and can be attributed to occupational 00 FORM 1 JAN 73 1473 EDITION OF 1 NOV SS IS OBSOLETE UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE (Whm Data Entered) �UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAOEfHTien Dmta Entantt) 20. ABSTRACT (continued) exposure to herbicides. The morbidity portion of this epidemiologic investigation includes a questionnaire and a physical examination. The questionnaires presented in this technical report are the field instruments used for the baseline data collection effort of 1981-1982. UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE (When Dftt Enttrad) �QUESTIONNAIRE PREFACE In 1979 the United States Air Force (USAF) made the commitment to Congress and to the White House to conduct an epidemiologic study of the possible health effects from chemical exposure in Air Force personnel who conducted aerial herbicide dissemination missions in Vietnam (operation RANCH HAND). The purpose of this epidemiologic investigation is to determine whether longterm health effects exist and can be attributed to occupational exposure to herbicides. The study protocol for this effort incorporates a matched-cohort design placed in a nonconcurrent prospective setting. The study approach includes mortality, morbidity, and follow-up elements. The morbidity portion of the study consists of an in-home interview of the study subject and his spouse, as well as a unique physical examination of the study subject and his matched comparison. The choice of the in-home interviewing method, as well as refinement of the unique physical examination, was significantly aided by extensive peer review of the scientific study protocol. The peer review agencies included: The University of Texas School of Public Health, Houston, Texas, the USAF Scientific Advisory Board, the Armed Forces Epidemiologic Board, and the National Academy of Sciences. In 1980 the Science Panel of the Agent Orange Working Group was created as an additional peer review agency. This group, redesignated the Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants* continues to monitor the conduct of this epidemiologic investigation. The questionnaires presented in this technical report are the field instruments used for the baseline data collection effort of 1981-1982. They are the result of a maturation process which began in 1979. In that year, contract number F41689-80-M-0174 was awarded to Research Statistics, Inc. of Houston, Texas. The purpose of this contract was to develop a Statement of Work (SOW) which would describe, in survey research terms, the requirement for the questionnaires necessary to support the epidemiologic study. Following refinement by the USAF principal investigators (PI'S) and management personnel, this SOW was used as the basis for a contract no. F41689-80-C-0059, with the National Opinion Research Center (NORC) of New York, New York. In this contract the USAF required the development of questionnaire instruments, procedures, forms, field manuals, training programs, and a pretest of developed instruments. At the core of the required questionnaires was a foundational questionnaire targeted at in-person administration to study subjects and their wives. It also had to be adaptable for use with the next of kin of deceased subjects. A brief noncompliance instrument was also required for use with those study subjects who declined participation. This miniquestionnaire was to contain questions concerning general health status and noncompliance factors. All questionnaires (study subject, spouse, proxy, and noncomplaint) were required to be adaptable to telephone as well as in-person administration methods. The NORC staff worked very closely with the USAF PI'S as well as their consultant staff to develop questionnaire instruments that would collect �quality health data that could be analyzed for health effects due to herbicides and that would capture data that could be lost through low compliance to the physical examination. Questions concerning specific health effects of phenoxy herbicides and dioxin were defined from the known human and animal effects found in the literature, as well as hypothetical effects found in biochemical and other biological systems. Additionally, veterans' complaints and the public's perception of the health effects of these chemicals were included. Wherever possible, portions of the questionnaire were taken from instruments NORC and other survey groups had previously field tested, thus maximizing instrument validity and reliability. Following an interviewer training program, NORC conducted an acceptability pretest in May 1981. Twenty-two study subjects, eighteen spouses, and two proxy subjects were interviewed. The questionnaires were found to be acceptable. Following modifications that resulted from the pretest the statement of work was developed for the implementation of the questionnaires. A competitive bidding process resulted in the award of the questionnaire implementation contract, No. F41689-81-C-0060, to Louis Harris and Associates, Inc. (LHA) of New York in September 1981. The purpose of this contract was to collect baseline data on the health, medical, demographic, social, and psychological conditions of the study population through the use of the developed questionnaires. Participation of the study subjects was to be on a completely voluntary basis. Letters from the Secretary of the Air Force and USAF Surgeon General were sent to each participant prior to the start of the interviewing process to encourage participation and to provide a brief overview of the general purpose and nature of the study. Louis Harris and Associates initially reviewed the NORC products and reformatted the instruments from a horizontal to a longitudinal format to better suit their interviewing style. The reformatting process allowed the addition of medical questions generated from recently published studies, as well as the inclusion of behavioral measurements not previously identified. Following the reformatting process, LHA trained 86 executive interviewers in a series of 11 training.sessions held throughout the United States and Europe. All LHA interviewers were required to have a minimum of one year prior experience in interviewing, with at least one experience in health data collection. Addresses of the study population were forwarded to LHA from the USAF and a locate algorithm was developed. During the approximately two-hour interview with the study subjects, the interviewers obtained written permission for government access to medical, hospital, personnel, and other records necessary to validate the questionnaire data. A Privacy Act Statement was signed as well. LHA was required to comply with the letter and intent of the Privacy Act of 1974 in collecting, storing, processing, and transferring personal and medical data. All questionnaire data were and continue to be treated with complete confidentiality. In September 1982, the LHA contract was extended to 15 November 1982 to permit the collection of baseline questionnaire data on the entire study population. 11 �ACKNOWLEDGMENTS The services of many staff members and consultants of the United States Air Force School of Aerospace Medicine are acknowledged. Special acknowledgement is made to the following co-investigators: Clarence F. Watson, Jr., M.D., M.P.H.; Alvin L. Young, B.S., M.S., Ph.D.; Joel E. Michalek, Ph.D.; Phelps P. Crump, Ph.D.; Richard C. McNee, M.S.; Alton J. Rahe, M.S.; Michael A. Sairi, M.D., M.P.H. & T.M.; Richie S. Dryden, M.D., M.P.H.; James A. Wright, M.D., M.P.H., who consulted and coordinated on questionnaire development. In addition, special acknowledgement is made to the United States Air Force School of Aerospace Medicine, Management and Air Training Command Procurement Personnel: Hugh F. Mulligan, Colonel, USAF, BSC, Chief, Program Acquisition Division; Charles T. Fuller, Major, USAF, Deputy, Program Acquisition Division; Donald F. Norville, Air Training Command Contracting Officer, Randolph AFB, TX, who coordinated the requirements for the questionnaire implementation contract. 111 ��CONTENTS CHAPTER I STUDY SUBJECT QUESTIONNAIRE Page I. INTRODUCTION 1-1 II. DEMOGRAPHIC 1-2 III. EMPLOYMENT HISTORY 1-4 IV. MARITAL/REPRODUCTIVE HISTORY 1-12 V. MEDICAL HISTORY 1-26 VI. SUBSTANCE USE HISTORY 1-40 VII. PSYCHOLOGICAL HISTORY 1-45 VIII. EXPOSURE TO TOXIC SUBSTANCES 1-77 IX. INCOME 1-83 X. SUPPLEMENTAL RECORDING BOOK XI. 1-85 ATTACHMENTS: 1. Introductory Letters: a. Secretary of the United States Air Force 1-135 b. Surgeon General United States Air Force 1-136 c. Louis Harris and Associates, Inc 1-140 2. Privacy Act Statement 1-141 3. Life Events Chart 1-142 4. Show Cards: a. Show Card "A": Academic Classification 1-143 b. Show Card "B": Industrial Classification 1-144 c. Show Card "C": Birth Control Techniques 1-145 d. Show Card "D": 1-146 Infertility History �Page e. Show Card "E", Anatomical Illustration 1-147 f. Show Card "F": Frequency Chart 1-148 g. Show Card "G": Chemical/Toxic Substances 1-149 h. Show Card "H": Protective Equipment/ Decontamination 1-150 i. Show Card "I": Income Table 1-151 5. Self-administered Form 1-152 6. Medical Provider Permission Form 1-154 7. Former Wife Introduction Letter 1-155 8. Interview Evaluation 1-156 9. Mailing Transmittal Form 1-157 CHAPTER II SPOUSE QUESTIONNAIRE I. INTRODUCTION 11-158 II. DEMOGRAPHIC 11-159 III. REPRODUCTIVE HISTORY 11-160 IV. SUBSTANCE USE HISTORY 11-169 SUPPLEMENTAL RECORDING BOOK 11-171 V. VI. ATTACHMENTS: 1. Introductory Letter, Former Wife 11-194 2. Privacy Act Statement 11-196 3. Show Cards: a. Show Card "C": Birth Control Techniques b. Show Card "D-l": 11-197 Infertility History 11-198 4. Medical Provider Permission Form 11-199 5. Interview Evaluation 11-200 6. Mailing Transmittal Form 11-201 vi �CHAPTER III NEXT OF KIN (PROXY) QUESTIONNAIRE I. INTRODUCTION 111-202 II. DEMOGRAPHIC III-203 III. EMPLOYMENT HISTORY 111-205 IV. MARITAL/REPRODUCTIVE HISTORY I11-212 V. MEDICAL HISTORY III-217 VI. SUBSTANCE USE HISTORY II1-228 VII. LEISURE ACTIVITIES III-230 VIII. MORTALITY DATA II1-231 IX. SUPPLEMENTAL RECORDING BOOK X. III-232 ATTACHMENTS: 1. Introductory Letter: Chief, Epidemiology Division, USAFSAM II1-266 2. Privacy Act Statement III-267 3. Life Events Chart III-268 4. Show Cards: a. Show Card "B": Industrial Classification III-269 b. Show Card "E": Anatomical Illustration IH-270 5. Medical Provider Permission Form II1-271 6. Interview Evaluation III-272 7. Mailing Transmittal Form 111-273 CHAPTER IV NON-COMPLIANT QUESTIONNAIRE I. INTRODUCTION IV-274 II. DEMOGRAPHIC IV-275 III. HEALTH IV-275 vii �IV. V. INCOME IV-276 PHYSICAL EXAMINATION IV-276 �CHAPTER I STUDY SUBJECT QUESTIONNAIRE The following Study Subject Questionnaire was used to collect baseline data for the Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicide Orange. This data was collected during 1981-1982. The questionnaire and supplemental recording book are the actual field instruments. They have been photocopyed and reduced for the purpose of this report. Additional field documents, such as show cards, are included as attachments to the questionnaire. In total, these documents demonstrate complete data collection methods for the Study Subject Questionnaire. Additional questions regarding reproductive experiences were added following the initial publication of the Study Subject Questionnaire. These questions are inserted where applicable in this instrument. Additional attachments include: Introductory Letters, Privacy Act Statement, Life Events Chart, Self-administered Sheet, Medical Provider Permission Form, Interview Evaluation, and Mailing Transmittal Form. The Study Subject Questionnaire, as used in the field, follows. �LOUIS HARRIS AND ASSOCIATES, INC. 630 Fifth Avenue New York, New York 10111 FOR OFFICE USE ONLY: O.M.B. NUMBER 0701-0033 Approval Expires 11/30/82 Case No. (12^17) Study No. 812039 Respondent #: (5-8) CONFIDENTIAL STUDY SUBJECT QUESTIONNAIRE This study is being conducted to collect information on the health of current and former Air Force personnel and their families. Since I will be asking you questions about your health, career, and personal history, we have prepared a Life Events Chart to help you remember when various events in your life occurred. The best way to use the Life Events Chart .is to first record when you were born in the Age Column, or how old you were in 1930, if you were born before 1930. Then, record your age at subsequent 5-year intervals in the Age Column. Next, note the year you graduated from high school and/or college in the next column. You can enter the year you joined the military in the next column. There are other columns to record any marriages or children you may have had, as well as other major events in your life. I will be asking you questions about each of these areas during the interview. If you will take a few moments to fill out Che Life Events Chart now, it will help you to recall dates and ages during the interview. First, I have a few background questions to ask you. 1. CARD 001 What is your date of birth? (WRITE IN DATE) DAY MONTH YEAR I I I I <22 ) (23) T (18) (19) 2. (20 ) (21) In what city and state were you born? I RECORD IN SUPPLEMENTARY RECORDING BOOK ON PAGE ll 3. What is your religious preference — religion, or no religion? Protestant Catholic is it Protestant, Catholic, Jewish, some other „ _-l ~-2 Jewish Other (SPECIFY) None. -5 What is the highest grade or year in high school that you completed? Less than 1 year of H.S...(j25( 1st year H.S. (9th Grade) 2nd year H.S. (10th Grade).... 3rd year H.S. (llth Grade).... 4th year H.S. (12th Grade) ~" -1 -2 -3 -A �812039 CARD 001 JHAND REgPONPENT CARD "A"I 5a. Please look at this card and tell, me yhich of these .regular academic school certificates, diplomas, or degrees you have obtained? .JMULTIP^E RECORD _BELOWT .(26( High school diploma. YEAR I -1 ' 3 ) (37) (6 YEAR High school equivalency diploma..... ( ( _... . -1 (38) ,2( (8 Associate of Arts (A.A.). Bachelor of Arts (B.A.) or Bachelor of Science (B.S.) Masters..... Doctorate. * * -1 (3Q( • ••• «1 ...-».(31 ( -1 wl (39) YEAR T T~" T 1 . 1 I (40) fair Ti YEAR ir ni ) (43) YfiAR T I I, . J T 1. T I YEAR I T -I .... I (47) Others (SPECIFY) YEAR -1 (1) YEAR (2) .(33( -1 YEAR (3) -1 YEAR No certificate, diploma, or degree (volunteered)....(35( •f 1 TEW"EACH t)ESBEE. PIPLOMA. OR CERTIFICATE.''ASjC Q . b 'j] Jb. In what year did you receive"your (CERTIFICATE/ ST17D1PLOHA/DEGREE)? (5 5T iRECORD AB"5vET �812039 CARD QQ6 6a. I am interested in training programs which prepared you for a major change in your occupation. First, I will ask about civilian job training programs. Besides the formal schooling you told me about, have you participated in any civilian job training programs that prepared you for a major change in your occupation? Yes.-.t1^ (ASK Q.6 b) Mo 13-1 A -1 -2 (SKIP TO Q.7) 1st Program b. 2nd Program For what kind of work was your f i r s t civilian training program preparing you? f. 3rd Program For what kind of work was your next civilian training program preparing you? j. For what kind of work was your next civilian training program preparing you? 05 ( (16( (16 ( (17( (17 ( (17 ( net (18( O8( (19( c. (16( (19( U9( In what month and year did you s t a r t this training? g. MONTH MONTH YEAR 1 1 I I 1 1 1 ' | l-l 1 1 (20) (21) (22) (23) d. In what month and year did you complete this training? I I e. 1 I I 1 l l I l l (20) (2H (22) (23) h. In what month and year did you complete this training? MONTH 1 1 -1 (ASK Q . 6 f ) L. I I l-l (24) (25) 1 1 MONTH YEAR 1 I I 1 1 1 l-l 1 1 (20) (21) (22) (23) In what month and year did you complete this training? YEAR 1 1 L. In what month and year did you start this training? 1 1 1 MONTH 1 1 1 1 Have you participated in any other civilian job training program that prepared you for a major change in your occupation? Yes.(28( ..j ( ASK Q.6j) 1 1 l-l m. 1 1 1 1 (26) (27) Have you participated in any other civilian job training program that prepared you for a major change in your occupation? —3 f <IK T P *m O 7 ^ .. YEAR 1 1 (24) (25) (26) (27) No. . Yes.(28( k. YEAR 1 I MONTH YEAR I I I I I l l I l l (24) (25) (26) (27) Have you participated in any other civilian job training program that prepared you for a major change in your occupation? In what month and year did you start this training? Ye«.(28( -l (fkeCOhP AbjHTfhMif fSitHINC PROGRAMS IN S.R.B, ON Pfit I*> -7 iOO TO 0.7) No 01 79-80 02 79-80 03 79-80 �CARD 007 812039 7a. Now, let's talk about military technical and specialized training programs that prepared you for a major change in your occupation. Besides the formal schooling (and the job training programs) you've told me about, have you participated in any military technical or specialized training programs that prepared you for a major change in your occupation? Yes...U2< (ASK Q.7b) No 13-14 -1 -2 (SKIP TO Q.8) 1st Program b. 2nd Program g. For what kind of work was your first military training program preparing you? 3rd Program For what kind of work was your next military training program preparing you? L. For what kind of work was your next military training program preparing you? ( 15( (15( (16( (16( (16( (17 ( (17 ( (17 ( (18( (18( (18( (19( (19( (19( ••: (IX c. What is the AFSC for that job? h. What is the AFSC for that job? (20 ( d. MONTH ! 1 l-l (21) ( 2 2) e. 1 1 1 1 ' j. M 1 1 l-l 1 1 <25) ( 6 2 ) (27) ( 8 2) Have you participated in any other military job training program that prepared you for a major change in your occupation? , -1 (ASK Q.7g) -2 (SKIP TO 0.8) 1 I k. MONTH 1 1 ' I I l-l (21) ( 2 2) YEAR 1 1 ( 3 (24) 2) (20 ( n. 1 1 In what month and year did you complete this training? YEAR 1 1 Yeg.(29( No In what month and year did you start this training? (3 (4 2) 2) MONTH f. 1 1 In what month and year did you complete this training? 1 1 i. YEAR 1 1 What is the AFSC for that job? (20 ( In what month and year did you start this training? 1 1 m. 1 1 o. MONTH YEAR 1 1 1 1 1 l l I l l (25) ( 6 2) (27) ( 8 2) Have you participated in any other military job training program that prepared you for a major change in your occupation? Yes.(29( No... -1 (ASK Q.7L) -2 (SKIP TO Q 8) In what month and year did you start this training? In what month and year did you complete this training? 1 I p. MONTH YEAR 1 I I 1 1 1 l-l 1 1 (21 ) (22 ) (23 ) (24 ) MONTH YEAR 1 M 1 1 l l I l l ( 5 (26) 2) (7 (8 2) 2) Have you participated in any other military job training program that prepared you for a major change in your occupation? 79-8(1 02 y<7-Kfi -1 (RECORD ADDI- No .,J .° Yes.(29( ING PROGRAMS IN S.R.B. ON PG. 15) -2 (GO TO Q 8) 0!) 7<I-RO �CARD 008 812039 8. Now I have some questions about working. Please tell me about all your jobs that lasted three months or longer since the first time you stopped going to school full time. Count changes of jobs for the same employer as separate jobs. Do not include jobs in the military. 13-U Second Job First Job 8a. In what month and year did you start your first job that lasted three months or longer? MONTH 1 1 In what month and year did you start your next job that lasted three months or longer? YEAR III 1 1 1 H 1 1 (15) (16) (17) (18) 8b. What (is/was) the name of your employer? IRECORD IN S.R.B. - PG i ] 8c. 9a. (Is/Was) the job fulltime or part-time? Full time..(19( Third Job 1 1 MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (15) (16) (17) (18) lOa. In what month and year did you start your next job that lasted three months or longer? MONTH 1 1 YEAR 1 1 T i l 1 l-l 1 1 (15) (16) (f7) (16) 9b. What (is/was) the name of your employer? lOb. What (is/was) the name of your employer? 9c. 10c. (Is/Was) the job fulltime or part-time? IRECORD IN S.R.B. - PG i 1 (Is/Was) the job fulltime or part-time? -1 Full time..(19( IRECORD IN S.R.B. - PG i 1 Full time..(19( -1 -1 8d. What kind of business is that —• what (do/ did) they make or do there? 9d. What kind of business is that — what (do/ did) they make or do there? lOd. What kind of business is that — what (do/ did) they make or do there? 8e. What ally what your 9e. What ally what your lOe. What ally what your IRECORD (do/did) you actudo on the job -.(are/were) some of main duties? IRECORD'IN S.R.B. - PG 1 I (do/did) you actudo on the job — (are/were) some of main duties? (RECORD IN S.R.B. - PG i I (do/did) you actudo on the job — (are/were) some of main duties? IN S.R.B. - PG 1 1 [HAND RESPONDENT CARD "B"| (HAND RESPONDENT CARD ^B" 1 (HAND RESPONDENT CARD "B"l 8f. Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. 9f. lOf. Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. (WRITE IN NUMBER) (WRITE IN NUMBER) 8g. 1 I I 1 1 1 (20) (21) In what month and year did this job end? 9g. MONTH YEAR 1 ' 1 I I 1 1 1 1 l-l 1 1 (23) (24) (25) (26) Current job..(27( (SKIP TO -1 Q.Ha) Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. I I In what month and year did this job end? 1 I MONTH YEAR 1 1 1 1 1 I l - l 1 1 (23) (24) (25) (26) Current job..(27( 8h. What was the main reason you stopped working on your job? 9h. I 1 I 1 (20) (21) -1 (SKIP TO Q.lAa) What was the main reason you stopped working on your job? <28( 01 79-80 1 1 I 1 I 1 (20 ) (21 ) lOg. In what month and year did this job end? 1 i MONTH 1 11 YEAR 1 1 r__i-i._i__JL (23) (2/, ) Current iob..(27( <2S ) (2f-> (SKIP TO -1 Q.lAa) lOh. What was the main reason you stopped working on your job? (28 ( (8 2( (29( (ASK 0.9a) (WRITE IN NUMBER) (2<> ( (29 ( (ASK Q.I la) (ASK 0. lOa) o? 79-HO ( H 7'i-HO �• ' . . . ' cARD 008 Fifth Job Fourth Job lla. In what month and year did you start your next job that lasted three months or longer? 8120.39 Sixtt) Job 12a. In what month and year did you start your next job that lasted three months or longer? 13a. In what month and year did you Start your next job that lasted three months or longer? 13-11 MONTH i 1 YEAR 1 1 1 1 H 05 ) ( 16) 1 I 1 I 1 1 ( 17) ( 18) MONTH 1 I I 1 .1-1 ( 15) ( 16) YEAR MONTH YEAR f 1 1 1 I I 1 1 1 1 I l l I l l (17) (IB) ( 17) ( 18) (15) (16) lib. What (is/was) the name of your employer? 12b. What (is/was) the name of your employer? [RECORD IN S.R.B. - PG 1 1 13b. What (is/was) the name of your employer? lie. (Is/Was) the job fulltime or part-time? 12c. (Is/Was) the job fulltime or part-time? 13c. (Is/Was) the job fulltime or part-time? Full time..( 19( Full time..(19( TRECORD IN S.R.B. - PG i 1 Full time..( 19( ...._-! Part time -2 IRECORD IN S.R.B. - PG i 1 -1 -1 lid. What kind of business is that — what (do/ did) they make or do there? 12d. What kind of business is that — what (do/ did) they make of do there? 13d. What kind of business is that — what (do/ did) they make or do there? lie. What ally what your 12e. What ally what your (do/did) you actudo on the job — (a re /were) some of main duties? 13e. What (do/did) you aetu• ally do on the job — what (are/were) some of your main duties? Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. 13f. Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. (do/did) you actudo on the job — (are/were) some of main duties? IRECORD IN S.R.B. - PG i | (RECORD IN S.R.B. - PG i 1 IRECORD IN S.R.B. - PG i 1 IHAND RESPONDENT CARD 1(B"T THAND RESPONDENT CARD ''B"| IHAND RESPONDENT CARD "B"| llf. Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. 12f. (WRITE IN NUMBER) (WRITE IN NUMBER) I I 1 (20) 1 1 ( 21) llg. In what month and year did this job end? MONTH 1 1 I I 1 l-l. (23 5 (24 > I I 1 (25 ) ( 26) I 1 I 1 (20) (21) 12g. In what month and year did this job end? YEAR 1 1 I 1 1 1 MONTH YEAR 1 I I 1 1 1 H 1 1 (23 ) (24) C>5 ) ( 26) Current (SKIP TO job.. (27 ( -1 Q.14a) Current job..(2X llh. What was the main reason you stopped working on your job? 12h. What was the main reason you stopped working on your job? -1 (SKIP TO Q.14a) (28 ( (28 ( (29 < '7"~"""(AS$: 6Vi2a) " M; • VV~R?) (29 ( (ASK Q.lSa} Oc> 79"-8.(i (WRITE IN NUMBER) 1 1 <20) (£l) 13g. In what month and year did this job end? 1 1 MONTH 1 1 1 1 l-l 63 ) ( 24) Current job..(27( -1 YEAR 1 1 I . I ( 25) ( 26) (SKIP TO Q.14a) 13h. What was the main reason you stopped working on your job? <28( - $?[- ' (RECORD ADDITIONAL JOBS IS S.R.B. - PG 16 ATO 17) 06 Y'T-SO �CARD OU 812039 14a. During the past six months, did illness or injury keep you from work, not counting work around the house? Yes 02 (_. No _ Retired (vol.). -1 (ASK Q.Ub) _~2\(SKIP TO Q.15) lAb. . Altogether, how many days did illness or injury keep you from work during the past six months? (REFERS if) "WORKING DAYS" OW.Y) (WRITE IN NUMBER) ; __, „ days (1'S) (15) 14c. What illnesses or injuries caused you to miss work? I L S (22 124 (. (25( (26( (27( -128 T29 (30 (31( X32 (33( �812039 CARP 015 15. Now I am going to ask y<>u about your years in the mili tarya. —rr I 1 MONTH YEAR I I I 1 1 l-l 1 ( 14) (15 ) (lb ) (17) b. What branch of the military was that? Air Force. (18 ( -1 Ndvv -5 MONTH YEAR I 1 1 1 1 1 l-l 1 1 1 (14) (15) U6) (IT ) MONTH YEAR 1 I I 1 1 1 l-l 1 1 04 ) ( 15) (16) ( 17) 1 1 g. What branch of the mili- L. What branch of the military was that? tary was that? -2 Coast Guard... k. In what month and year did you next enter the Armed Forces? f . In what month and year did you next enter the Armed Forces? In what month and year did you first enter the Armed Forces? c. Were you discharged or separated from the (BRANCH OF SERVICE)? Air Force. ( 18( Coast Guard... Air Force. 08 ( -1 -5 h. Were you discharged or separated from the (BRANCH OF SERVICE)? Coast Guard... -1 ~ -5 m. Were you discharged or separated from the (BRANCH OF SERVICE)? Discharged/ Discharged/ Discharged/ separated. (19 ( -1 (ASK separated.( 1% -1 (ASK separated. (19( -1 (ASK Q.15d) Q.lSn) Still in Still in Still in (MILITARY) -2 (SKIP TO (MILITARY)..... -2 (SKIP TO (MILITARY)..... -2 (SKIP TO ~ Q.16) ~ . Q.16) Q.16) d. In what month and year were you discharged/ separated from the (BRANCH OF MILITARY)? MONTH YEAR 1 I I I 1 1 I I l-l 1 1 (20) (21 > <22) ^23) i. In what month and year were you discharged/ separated from the (BRANCH OF MILITARY)? 1 1 n. MONTH YEAR 1 I I 1 1 1 l-l 1 1 (20) (21) (22) (23) In what month and year were you discharged/ separated from the (BRANCH OF MILITARY)? MONTH 1 I I 1 1 l-l (20 ) (21 ) YEAR I I 1 1 (22 ) (23 ) e. Following your separation j. Following your separation o. Following your separation or discharge in (DATE IN or discharge in (DATE IN or discharge in (DATE IN "d"), did you reenter the "i"), did you reenter the "n"), did you reenter the Armed Forces? Armed Forces? Armed Forces? Yes..(24( -1 79^80 (ASK Q.15f) Yes..(24( -1 (ASK Q.15k) Yes..(24( -1 No.. . . — 2 (SKIP TO Q 16) 02 79-80 No --2 03 79-80 (RECORD ADDITIONAL SERVICE PERIODS IN S.R.E. PG 18) (SKIP TO Q 16) �CARD pie. 812039 16. I would like .to ask you the names of all the countries you have been stationed in while on active duty in the Armed Forces. First Country Second Country Third Country a. Starting with induction, g. What was the next country m. What was the next country in what country were you that you were stationed that you were stationed first stationed while on in for more than 90 days in for more than 90 days active duty? Include while on active duty? while on active duty? 12-13 temporary duties of (14-15) (14-15) greater than 90 days. (14-15) ^ RECORD COUNTRY HERE AND IN S.R.B. PG 2 AND CONTINUE) (RECORD COUNTRY HERE AND IN S.R.B. PG 2 AND CONTINUE) No others. ( 16( /T57 b. (RECORD COUNTRY HERE AND IN S.R.B. PG 2 AND CONTINUE) -1 (SKIP TO No others. (16 ( -1 (SKIP TO Q.18) . Q.18) In what month and year h. In what month and year n. In what month and year did you begin and end acdid you begin and end acdid you begin and end active duty in (COUNTRY)? tive duty in (COUNTRY)? tive duty in (COUNTRY)? BEGIN BEGIN YEAR MONTH YEAR 1 M i l I I I I 1 1 1 l - l I I 1 1 l-l I . I ( 17) ( 18) ( 19) ( 20) ( 17) ( 1© . 09 )( 20) END END MONTH YEAR MONTH YEAR 1 I I 1 1 1 I I I 1 1 1 l-l 1 1 1 .1 l-l 1 1 ( 2J) (22) ( 235 (24) (21 ) ( 22) (23 )( 24) BEGIN MONTH YEAR 1 1 1 1 1 1 1 1 l-l 1 1 (17)08) ( 19) ( 20) END MONTH YEAR MONTH Cur rent... (25 ( -1 Current..'. (25 ( 1 I 1 1 1 1 1 1 H I 1 ( 21) ( 22) G>3 ) ( 24) Current... ( 25( -1 -1 c. What specific job assign- i. What specific job assign- o. What specific job assignments (do/did) you have ments (do/did) you have ments (do/did) you have in (COUNTRY)? Can you in (COUNTRY)? Can you in (COUNTRY)? Can you give me the AFSC? give me the AFSC? give me the AFSC? 1. ( 26-28)1. 66-28) 1. C6-28) 2. £9-31) 2. C9-31) 2. (29-31) 3. (32-3<M 3. 02-34) 3. 02-34) (Do/Did) your duties in j. (Do/Did) your duties in p. (COUNTRY) include flying? (COUNTRY) include flying? (Do/Did) your duties in (COUNTRY) include flying? d. Yes.( 33( No . . Yes.(35( e. How many flight hours did you log while in (COUNTRY)? 1 1 1 | 1 1 Hours 06 ) (37) (38) Other (SPEC1FYJ -1 Yes.(35( -1 No -1 -2 -2 No -2 k. How many flight hours did you log while in (COUNTRY)? 1 1 I I 1 1 1 1 Hours ( 36) ( 37) ( 38) Other (SPECIFY) q. How many flight hours did you log while in (COUNTRY)? 1 1 I I 1 Hours ( 36) ( 37) ( 38T Other (SPECIFY) . 3 ( -1 (9 .(39( -1 1 | 1 . 3 ( -1 (9 f. What specific letter and L. What specific letter and r. What specific letter and numerical ^Utoigniltionfs) numerical designation^ ) numerical designation^) did each aircraft have? did each aircraft have? did each aircraft have? 1. f4,0-43) 1. (40-43) 1. (40-43) 2. (44-47) 2. (44-47) 2. <4ft-4.7) 3. f48-5]V 3. (48-51) 3. (48-51) (52-55) 4. (52-55) 4. *• (ASK Q.16p) 02 79-80 01 79-80 1U (52-55) (ASK Q.17a) (ASK Q.16m) 0) 79-HO �CARD_Q16 812039 Question 17 iT3 Fourth Country Fifth Country Sixth Country a. What was the next country g. What was the next country m. What was the next country that you were stationed that you were stationed that you were stationed in for more than 90 days in for more than 90 days in for more than 90 days tfhit« Oil active duty? while on active duty? while on active duty? (14-15) \ RECORD COUNTRY HERE AND IN 6,R,». PC J AND CONTINUE) (RECORD COUNTRY HERE AND" IN S.R.B. PC 2 AND CONTINUE) (14-15) (RECORD COUNTRY HERE AND IN S.R.B. PC 2 AND CONTINUE) No others. (16 ( -1 (SKIP TO Q.18) No others. (16 ( -1 (SKIP TO Q.18) No others. (16 ( -1 (SKIP TO Q.18) b. (14-15) In what month and year h. did you begin and end active duty in (COUNTRY)? BEGIN BEGIN MONTH YEAR 1 1 I I 1 1 1 1 l-l 1 1 (17) (18) (19) (20 r MONTH YEAR 1 1 1 1 f 1 l-l . 1 1 1 ( l7> lit ) (19) (20 !> END MONTH YEAR 1 M i 1 . . , -1 1 (21) (22) (23) (24) -1 Current.. .(25 ( In what month and year did you begin and end active duty in (COUNTRY)? BEGIN 1 II MONTH 1 1 1 l-l YEAR 1 " t 1 1 (n ) as ) fly ) po ) END MONTH YEAR I I I 1 1 1 -1 i (21) (22) (23) (24) l 1 Current . , . (25 ( In what month and year n. did you begin and end active duty in (COUNTRY)? END' MONTH 1 1 YEAR I I l-l (21). (22) (23) (24) Current . . . (25 ( -1 -1 c. What specific job assign- i. What specific job assign- o. What specific job assign, merits (do/did) you have ments (do/did) you have ments (do/did) you have in (COUNTRY)? Can you in (COUNTRY)? Can you in (COUNTRY)? Can you give me the AFSC? give me the AFSC? give me the AFSC? (26-28) 1. O6-28) 1. (26.- 28) 2. U9-31) 2. (29-31) 2. (29-31) 3. (32-34) 3. (32-34) 3. (32-34) 1. • d. (Do/Did) your duties in j. (COUNTRY) include flying? (Do/Did) your duties in p. (COUNTRY) include flying? Yes. ( 5 ( 3 No e. How many flight hours did you log while in (COUNTRY)? 1 L Other (SPECIFY) Yes. (35 ( No -1 -2 k. How many flight hours did you log while in (COUNTRY)? 1 1 I I I I Hours (36) (37) (38) 1 1 1 1 Hours (36) (37) (38) Other (SPECIFY) 1 1 1 1 1 1 J J Hours (36 5 (37 ) (38 ) Other (SPECIFY) . 3 ( -1 (9 f. What specific letter and t. numerical designation(s) did each aircraft have? -1 -2 q. How many flight hours did you log while in (COUNTRY)? 1 1 . 3 ( -1 (9 (Do/Did) your duties in (COUNTRY) include flying? .(39; -l What specific letter and r. What specific letter and numerical designation(s) numerical designation(s) did each aircraft have? did each aircraft have? 1. (40-43) 1. (40-43) 1. (40-43) 2. (44-47) 2. (44-47) 2, (44-47) 3. (48-51) 3. (48-51) 3. (48-51) fll'-W 4. 4. (ASK Q.17g5 04 79-80 (52-55) (52- IS) 4. (RECORP ADDITIONAL COUNTRIES IN S.R.B. PC 19 AND 20) ftt 05 7 "' 0 9:8 (ASK Q. 17ml 11 �812039 . CARD 020 Now I would like to ask you about your marital history. 18. Have you ever been legally married? -1 (ASK Q.19) -2 (SKIP TO Q.24) 19. How many times have you been legally married? I T (WRITE IN NUMBER) I times 43 ) (14) FIRST/ONLY MARRIAGE 20a. In what month and year did you get married (th f i r s t time)? 1 I SECOND MARRIAGE 21a. MONTH YEAR 1 1 1 ! 1 l l I l l ^5 )" (16 ) ( 17) ( 18) THIRD MARRIAGE In what month and year did you get married the second time? 1 1 MONTH 1 I I 1 • .1-1 (15 > ( 16) YEAR I 1 22a. In what month and year did you get married the third time? MONTH I 1 YEAR I I 47 T (18) 45) (16F (17 ) ( 18) 20b. What (is/was) her current full name? 1 RECORD IN S.R.B. PG 2 f 21b. 20c. 21c. 22c. 20d. What is her date of birth? 21d. What is her date of birth? 22d. What is her date of birth? What was her maiden name? 1 RECORD IN S.R.B. PG 2 I MONTH (19 ) (20 ) 1 1 MONTH 1 1 1 1 ( 21) ( 2'J) -1 -2 21e. 1st 1 1 r 2 n dI 3rd 4 t What was her maiden name? IRECORD IN S.R.B. PG 2 [ YEAR 1 1 ( 21) ( 22) MONTH Yes...(_2| Ho YEAR I (19 ) (20 ) Gl ) ( 22) Have you ever had any children by (your/thin) wife? 22e. Have you ever had any children by (your/this) wife? -I -2 -1 Yes...(23( No... Did your wife ever have 22f. Did your wife ever have any pregnancies by you any pregnancies by you which ended in a miscarwhich ended in a miscarriage? riage? -1 (ASK Q.21g) Yes...(24( -1 (ASK Q.20g) Yes...(24( -i) V (SKIP TO Don't know -3J Q.20L) 20g. When was that? (PROBE: Any others?) No 1 1(19 ) ( 20) 20f. Did your wife ever have 21 f . any pregnancies by you which ended in a miscarriage? Yes... (24 ( 22b. What (is/was) her current full name? I RECORD IN S.R.g. PG 2 I What was her maiden name? (RECORD IN S.R.B. PG 2 1 20e. Have you ever had any children by (your/this) wife? Yen...(23( No cur- 1 RECORD IN S.R.B: PG 2 1 YEAR 1 1- What (is/was) her rent full name? -2) I (SKIP TO Don't know -3J Q.21D 21g. When was that? (PROBE: Any others?) No MONTH YEAR 1 I I 1 1 1 |-| | | 1st < 2 5 ) (26) (27) (28) MONTH YKAfl (SKIP TO -3j Q.22L) Don't know 22g. When was that? (PROBE: Any others?) MONTH YEAR 1 I I I 1 | |-| | I 1st (25) (26) (27) (28) MONTH _VM!L_ YEAR 1 T MONTH 1 | i 11 T"~~I . T™ T---TT r i l l I l l 2 n dI l l I l l 2nd (29) (30) (31) (32) (29) (30) (317T3ST' MONTH YEAR MONTH YEAR 1 1 1 1 " 1 I I 1 1 1" 1 ' 1 I 1 |-| | 1 3rd | I |-| | | 3rd (33) ( 3 4 ) (35) (36) (33) (34) (35) (36) MONTH YEAR MONTH YEAR 1 I I I 1 T 1 1 I I I 1 hI l l I l l 4 t hI l l I l l 4th ( 3 7 ) (38) (39) (40) ( 37) ( 38) ( 39> ( 40) (GO TO Q.20h) (GO TO Q . 2 1 h ) 12 I r T I I-1 (29) I (TO <2HT YKAIl (25) <Z6) (30) MONTH (31) I (J^ YEAR T (33TT34) MONTH (35) YEAR i—i—r1-1i—ii ri i (37) i (38) (39) (411) (GO TO Q.22h) �.-.-.i'.-. ,.V; FIRST/ONLY MARRIAGE I J. 1 1 1 1 Months (41 ) (42 ) Hssn't trying ( 43S 1 1 1 t 1 1 Weeks <44 > ( 4# 1 1: 1 1 812039 THIRD MARRIAGE 22h. How many months did it take your wife to become pregnant this time? 1 1 Months <41 ) (42> Wasn't trying (_A3C -1 20 i. How many weeks had your 21 i. wife been pregnant when, the (1st, etc.) miscarriage occurred? 1st cARC '020 21h. How many mpnths did it take your wife to become pregnant this time? 20h. How many months did it take your wife to become pregnant this time? I t .. . SECOND MARRIAGE 1 1 (41) (42) Wasn't trying ( 4X -I How many weeks had your wife been pregnant when the (1st, etc.) miscarriage occurred? 1st I 1 2nd l l 1 1 Weeks (44) <45) 1 1 Weeks ( 4.4) (45 ) let 2nd 1 1 1 Weeks <46 ) (47 ^ 3rd 1 1 1 1 I 1 Weeks (48) (49> 3rd 4th 1 1 1 'I 1 1 Weeks (50 J ( 55) 4th 1 1 I 1 2nd; 1 1 I I I | Weeks (48 ) ( 49) 20 j. Did a doctor tell you why this (1st, etc,) miscarriage might have occurred? -1 22i. Mow many, weeks had your wife been pregnant when the (1st, etc.) miscarriage occurred? I I I I | | Weeks <46 ) ( 47} I 1 1 Months I 1 " t 1 | Weeks <46) (47) 1 3 r dI 1 1 1 I Weeks (50 ) ( 51) 4th 21 j. Did a doctor tell you why this (1st, etc,) miscarriage might have occurred? 1 1 1 1 I | Weeks ft»8 ) ( 49) 1 I Weeks (50 ) ( 51) 22 j. Did a doctor tell you why this (1st, etc.) miscarriage might have occurred? Yes. (52 ( -1 (ASK Q.22k) Yes. (52 ( -1 (ASK Q.21k) Yes. (52 ( -1 (ASK Q.20k) No -2 (SKIP TO Q.20L) No.i.... -2 (SKIP TO Q.21L) No -2 (SKIP TO Q 22L) 20k. What did the doctor say .21k. What did the doctor Say caused the miscarriage? caused the miscarriage? 22k. What did the doctor say caused the miscarriage? 1st 1st 1st 2nd 2nd 2nd 3rd 3rd 3rd 4th 4th 4th 20L. Rid your wife have any 211.. Did your wife have any 22L. Pid your wife have any pregnancies by you, which pregnancies by you which pregnancies by you which ended in a stillbirth? ended in a stillbirth? ended in a stillbirth? Yes... (53 ( No ,. ~~ 5 -1 (ASK Q,20ti)> Yes... ( 3 (T -2 (SKIP TO Q. No 1.'' ' " ' 20q) 20m. When was that? Any others?) MONTH (PROBE: MONTH YEAR V" 1 1 2nd MONTH YEAR 1 ' 1 1 I 1 1 1 1 l-l 1 1 2nd (58) (59) (60) (61) MONXH YEAR 1 1 3rd 1 1 4t>t». I' |.X ( 1 (& (3 6 , 6 ) ( 4 ' "(6$) 6) MONTH YEAR " '1; l - l 1 7 ' 'I l-l 1 "1 1 4th YEAR "\ " ' M " ' 1 1 l-l . 1: MONTH 1 \ i t 1 1 1 1; I I l-l (58) (59> MOUTH 1 ~ '1 1 1st 1. 1 2nd ( 0 (61) 6) YE&R [: f l-l I' 1 3rd (62). ' 6 ) ' ( 4 "ftftf (3 6J MQNTH YMR, 1- i r i; ' • i 1 \-\ ! 79-80" 13 YEAR 1 i 1 l 1 l-l \ 1 (5« ) PS ) (36F<;57) " MONTH TJ6AR i ~n r • f r l-l 1 1 ( 8 (59 >' (60) £61 ) 5) MONTH YiEAR, F" 1 i; """I : l-t ' 6 ) (6i> (2 MO>ITR r 1 ' I, ( 6 « •' 'U-3J, ,') YEftR r I r ' i i- ' ' f 1 4th ( l-li 1: 1: (fef)');,' (67=) " » : (:6»> (») (GQ TO Q,Hft)i (.6ft? ( ( . . • ?§) (GO TO Q.2ln) (?ROBE; \ v •• '\ \--\-- -\ 156 > <5T) YEAR (6(>5 (f>7): '/GO to, Q.20n) 22m. When vas that? Any others?) (PROflE: (59 ( 5 ) 5' MONTH 1 1' -1. (ASK q.22m) 22q) 21m. When was that? Any others?) F ' T" '1 T~ "i "T 1 (54 )1 (55 ) (56 ) (57 Ut '"I 1, 1 ) 3r4 -1 (ASK Q.Hm)*e.»,.«(53l "-2 (SKIP TO Q. t: �ADDITIONAL RECORDING SHEET FOR NON-LIVE-BIRTHS FIRST/ONLY MARRIAGE MISCARRIAGES - Q.20h I 2nd I Months (12) (13) Wasn't trying (14( Don't know -1 SECOND MARRIAGE MISCARRIAGES - Q.21h 2nd r T I 3rd I I I Months (12) (13) Wasn't trying (14( -1 Don't know -2 T I 3rd 1 I I I Months (15) (16) Wasn't trying (17( Don't know -1 -2 THIRD MARRIAGE MISCARRIAGES - Q.22h 2nd I I Months '(12) (13) Wasn't trying (14( -1 Don't know -2 T I I I Months (15) (16) Wasn't trying (17( Don't know -1 -2 3rd I I Months " (18) (19) Wasn't trying ( 0 2( -1 Don't know -2 4th I 1 [ Months (18) (19) Wasn't trying ( 0 2( Don't know (GO TO Q.20i) [ Months (15) (16) Wasn't trying (17( Don't know I I 4th 812039 CARD 143 -1 -2 4th I -1 -2 T I Months (18) (19) Wasn't trying ( 0 2( -1 Don't know -2 (GO TO Q.2H) I (GO TO Q.22i) [AFTER Q.20k/21k/22k ASK FOR EACH MISCARRIAGE:f Were either of you using birth control at the time she bedame pregnant? IFOR ANY "YES" ASKTT IHAND RESPONDENT CARD "C"| Please look at this card nnd tell me all the numbers that apply to the types of birth control you used. 1st: J21-22) 1st: J21-22) 1st: J21-22) 2nd: _(23-24) 2nd: _(23-24) J23-24) 3rd: _<25-26) 3rd: J25-26) 3rd: (GO TO Q.21L) (GO TO Q.20L) mST/ONLY MARRIAGE STILLBIRTHS - Q.20n \ I I Months (32) (33) Wasn't trying (34( -1 Don't know -2 (35) (36) Wasn't trying (37( Don't know Months -1 -2 2 n d I T I I Months (29) (30) -1 Wasn't trying (31( Don't know I I r 4th I i I I Months (32) (33) Wasn't trying (34( -1 Don't know ~2 I I Months (35) (36) Wasn't trying <37( -1 Don't know -2 I I T T I 3rd Months (32) (33) Wasn't trying ( 4 3 ( -1 Don't know... -2 T I Months (35) ( 6 3) Wasn't trying (37( -1 Don't know "2 4th (GO TO Q.21o) (GO TO Q.20o) r I Months (29) (30) Wasn't trying (31( -1 Don't know -2 2nd 1 3rd T -1 -2 I J 4th THIRD MARRIAGE STILLTBIRTHS - Q.22n I Months (29) (30) Wasn't trying (31( Don't know 3rd (27-28) (GO TO Q.22L) SECOND MARRIAGE STILLBIRTHS - Q.21n T 2nd _(25-26) (27-28) 4th: (27-28) 4th: 4th: 2nd: I I I ! (GO TO Q.22o) I AFTER Q. 2Qp/ 21 p/22p; ASK ^FOR EACH STILLBIRTHjj_ Were either of you using birth control «t"the' lime sh« became pregnant? IFOR ANY "YES" ASK~ IHAND RESPONDENT CARD "c"| Please look at this card and tell me all the numbers that apply to the types of birth control you used. _(38-39) J38-39) 1st: 1st: _(40-41) (40-41) 2nd: (40-4)) 2nd: 2nd: J38-39) 1st: _(42-43) 3rd: (44-45) (44-45) 4th: 4th: (GO TO Q.20q) (GO TO Q.21q) . 14 (44-45) 4th: (GO TO Q.22q) �RECORpiNO SHgfc'T FQR 1 2nd 1 1 Months (46 ) (47) Wilsn't trying (48( Don't 1 -1 2nd Wa»n' I tryjrifi (51( Don't I -1 1 I 1 1 Months (46) (47) W a s n ' t trying (4_8J -1 3rd 1 I I 1 1 1 Months W a s n ' t trying (51( -1 . 1 1 Months 1 1 I ABORTIONS - 9.225 2nd I I I I I | Months (46) (47) W a s n ' t trying (48( -1 3rd I I '2 r i( i1 Months 4th •ri.ii^.1) SKCQNIt MAjMjACiK ABORTIONS - Q.21s 1 3rd 812039 CARP 143 KIRST/Oftl.Y MAHRtACE 'ABORTIONS - Q.20s 'ft 2") (555 Wasn't trying (56( Don't I -1 4th 1 1 1 I (G() TO Q.20t) 1 I Months W a s n ' t trying ( 5 l ( -1 -2 1 [ Months (52) (53) Wasn't trying (54( I I -1 -2 4th 1 1 I 1 I 1 Months (52) (53) Wasn't trying (54( (GO TO Q . 2 1 t ) -1 (GO TO Q.22t) 1 AFf^R §,20t/21t/22t ASK FOR EACtf ABORTION;! Were e Lttief of you using birth control at the time she became pregnant? IFOR 1 A W .''YES" ASK: 1 :'\ ..":. • ".' IHAN^ iUSjPbNDENT •' •' : CAfo "C'M Please look at this card and tell roe all the numbers that apply to the types of birth 1st: (55-56) 1st: (55-56) 1st: (55-56) 2nd: (57-58) 2nd: (57-58) 2nd: (57-58) 3rd: <59-60) 3rd: (59-60) 3rd: (59-60) (61-62) 4th; 4th: (61-62) 4th: (GO TO Q.20u) 01 79^80 (61-62) (GO TO Q.Zlu) (GO TO Q.22u) 02 W%& 79^15 15 �CARD FIRST/ONLY MARRIAGE 20n. How many months did it take your wife to become pregnant this time? 1 1 1 1 Months (U) U3) Wasn't trying (14( -1 (12) (13) Wasn't trying (14( 20o. Did a doctor tell you why this stillbirth might have occurred? 21o. Did a doctor tell you why this stillbirth might have occurred? Yes.(i5( No . . 220 2 In. How many months did it take your wife to become pregnant this time? I I 1 1 1 1 Months -1 (ASK Q.20p) Yes.(l5( No -2 (SKIP TO Q.20q) -1 -2 -1 (ASK Q.21p) (SKIP TO ( 21 q) ) 812039 THIRD MARRIAGE SECOND MARRIAGE 22n. How many months did it take your wife to become pregnant this time? 1 I I 1 I I Months (i/) (U) Wasn't trying O*( -1 22o. Did a doctor tell you why this stillbirth might have occurred? Yes.(15( -1 (ASK Q.22p) No -2 (SKIP TO Q 22q) 20p. What did the doctor say 21p, What did the doctor say 22p. What did the doctor say caused the (1st, etc.) caused the (1st, etc.) caused the (1st, etc.) stillbirth? stillbirth? stillbirth? 1st 1st 1st 2nd 2nd 2nd 3rd 3rd 3rd 4th 4th 4th 20q. Did your wife ever have 21q. Did your wife ever have 22q. Did your wife ever have any pregnancies by you any pregnancies by you any pregnancies by you which ended in abortion? which ended in abortion? which ended in abortion? Yes...(lft ( No , -1 (ASK Q.20r) Yes... 06 ( -2 (SKIP TO Q. No 20u) 20r- When was that? Any others?) 1st 2nd 3rd 4th -1 (ASK Q.21r) Yes...(16( -2 (SKIP TO 0. No 21u) (PROBE: 21r. When was that? Any others'?) (PROBE: 22r. When was that? (PROBE: Any others?) MONTH YEAR MONTH YEAR 1 I I I I I I I I I| | 1st 1 -1 1 1 1st (17) (18) (19) (20) (17) ( Id (19) (20) MONTH YEAR MONTH YEAR 1 1 11 1 1 I I I I 1 1 I 1 1-1 I I 2nd 1 l-l 1 1 2nd <21 ) (22) (23) (24 ^> <21> (22> ^23) (24) MONTH YEAR MONTH YEAR 1 1 1 ! 1 1 1 1 1 1 1 I l l - I l 3 rl d 1 1 l-l 1 1 3 r d (25) (26) (27) (28) < 25 > (26J (27) (28> MONTH YEAR MONTH YEAR 1 1 1 I I 1 1 1 1 4th 1 l-l 1 I 4th <29> (30) (31) (32) < 29 > (30 > (31) (32> 20s. How many months did it take your wife to become pregnant this time? 1 1 1 1 (33) (34) Wasn't trying <35( 21s. How many months did it t-ke your wife to become pregnant this time? 1 1 Months -1 1 1 Months (33) (34) Wasn't trying (35( -1 (ASK Q.22r) -2 (SKIP TO Q. 22u) -1 MONTH YEAR 1 I I 1 1 1 - I I I (I/) ( i»> (19) (20) MONTH YEAR 1 I I I 1 1 1 1 - I I I ( 2J> (22 ) (23 ) (24 ) MONTH YEAR I I I I I 1 1 - I I I (25 ) ( 20 (27 ) ( 29 MONTH YEAR 1 1 I I 1 1 1 1 l-l 1 1 (29) (30) (31) (32) 1 1 22s. How many months did it take your wi fe to become pregnant this time? T~ f 1 1 ' Mofctfift (33) (34) Wasn't trying (35( -I 20t. What was the main reason 21 1. What was the main reason 22t. What was the main reason for the (1st, etc.) for the (1st, etc.) for the (1st, etc.) abortion? abortion? abortion? 1st 1st 1st 2nd 2nd 2nd 3rd 3rd 3rd 4th 4th (CO TO Q.20v ) 4th (GO TO Q.21v) 10 (GO TO Q.22v) �.CARP 220 FIRST/ONLY MARRIAGE 20u. (IF ANY CONCEPTIONS ~ 21u. CHILD, MISCARRIAGE, STILLBIRTH, OR ABORTION: SKIP TO q.iOw THIRD MARRIAGE (IF ANY CONCEPTIONS — 22u. (IF ANY CONCEPTIONS — CHILD, MISCARRIAGE, CHILD, MISCARRIAGE, STILLBIRTH, OR ABORSTILLBIRTH, OR ABORTION: SKIP TO Q.2lw TION: SKIP TO Q.22W ALL OTHERS: ASK Q.2lu ALL OTHERS: ASK Q.22u) Did either you or your Did either you or your wife use birth control wife use birth control techniqueB regularly? techniques regularly? ALL OTHERS: ASK Q.20" Did either you or your Vif* U8e birth control techniques regularly? Veg..( 36( -1 (ASK <J.20v) No __-2 <ASK Q.20x) 812039 SECOND -MARRIAGE Yes..(36( -1 (ASK Q.Zlv) No _^-2 (ASK Q.21x) Y«»..(_36(- -1 (ASK Q.22v) NO -2 (ASK Q.22x) I HAND BESPOMPENT I HAN Cv. Please look at this car 21v. Please look et this car 22v. and tell me all the numand tell me all the nui berS that apply to the bers that apply to the types of birth control types of birth control you or your wife normal you or your wife normal ly used. ly used. Please look at this card and tell me all the numbers that apply to the types of birth control you or your wife normally used. 06. ( 42(_-l 01. (M -i Q7.C4f""-l oz.daT^-i 08.(44(,. -1 03.(_39( -1 0.tT9 ( 5 ~ l OA.(40(. -1 io.(M_ -i OS.(4l( '1 02. ( a l -1 O3.(j?9( .-1 OA.(40( .-1 OS.(4i( -1 12 (SPECIFY) 12 (SPECIFY) n.csn-1 12 (SPECIFY) 01.(M -1 06. (421 07. Q8.(44(" -1 Q9.(45( • -1 10.(46( • Ug _ -1 (SKIP TO Q.20x) (SKIP TO Q.20x) (SktP TO Q.20x) 20w. Did any of these preg- 21w. Did any of these preg- 22w. Did any of these pregnancies occur while nancies occur while nancies occur while either you or your wife either you or your wife either you or your wife wire practicing birth were practicing birth were practicing birth control? control? control? Yes....(49( No Yes....(49( No -1 20x. During this marriage, how many times were you living apart from your wife for more than moti.thg? 1 I I Times (50 ) (51 ) -1 -1 "-2 21x. During this marriage, 22x. During this marriage, how many times were you how many times were you living apart from your living apart from your wife for more than threi wife for more than three months? tfionths? T Never.. (52 { (9 4( Yes No -1 I T I . I Time* (So) pi ) (SKIP TO Never..(53 Q.20aa/bb .-1 (SKIP TO Q.21aa/bb I I I I I ^ I Times po ) (51) Never..(53( -1 (SKIP TO Q.22aa/bb) 20y. How many months did you 21y. How many months did you 22y. How many months did you live apart the (first/ live apart the (first/ live apart the (first/ next) time? next) time? next) time? I I I ].... L Months (53) (54] I "~'~"i 2nd 3rd J III T I I Months (55) (56) I I I !„„- , Months 2nd 3rd Months 4 t h I t 16:0 ( I t I I I | Months (39 ) (60) i—n—r 5Jh I 6tli (57) (58) I 6th i (GO TO 0.20?.5 Months (55) (So)' I I T L . ,l . . J Months ..r,.. (61) (62) i,^,-] (53 I I .1 1 Months (55) ( 50 3rd 1 Months l.s.t Months 2nd Months L .. J (57) (58) I 1. .....I . .J Months (53 ) (54) ! J. L . . . Months ..-[ ((,:)) (64) (GO TO 0. 17 t 2th I <S'9) I I I Months "' v......i r \ I I,^..L, T ,,,.I. Months I I .._.!_.....I Months ( h J ) (r()4"5" (GO TO Q . 2 2 z ) �812039 CARD 220 SECOND MARRIAGE FIRST/ONLY MARRIAGE 20z. As a result of (this/ these) separations, did you and your wife have fewer children than you wanted to have? Yes....(65( No... 21i. As a result of (this/ these) separations, did you and your wife have fewer children than you wanted to have? -1 -2 Yes,...65( THIRD MARRIAGE 22z. As a result of (this/ these) separations, did you and your wife have fewer children than you wanted to have? Yes....(65( No -1 IIF LAST MARRIAGE! IIF LAST MARRIAGE! llF ONLY MARRIAGE] -1 -2 20aa.Are you currently married and living with your wife, or are you divorced, widowed, or separated? 21aa.Are you currently married and living with your wife, or are you divorced, widowed, or separated? 22aa.Are you currently married and living with your wife, or are you divorced, widowed, or separated? Living with wife...(66( Living with (SKIP TO wife...fe_6( -1 Q.23) Living with (SKIP TO wife...fe6( -1 Q.23) Separated.... -3f Q.20cc) Widowed -4J Divorced -2) (SKIP TO Separated..,. ~"3f Q-Zlcc) Widowed . . . -4j ... (RECORD IN S.R.B. PO 2 1 I RECORD IN S.R.B. PG 2 I Divorced -m SKIP TO Separated,... -3f Q.22cc) Widowed -4| ~"°" J (RECORD IN S.R.B. PG 2 1 -1 (SKIP TO Q.23) (IF OTHER MARRIAGES! (IF OTHER MARRIAGES] IIF OTHER MARRIAGES | 21bb.How did that marriage 22bb.How did that marriage 20bb.How did that marriage end -- were you divorced end — were you divorced end — were you divorced or were you widowed? or were you widowed? or were you widowed? Divorced<67( Widowed ~J.[(ASK Q.20cc) Divorced(.67[ -A (ASK Q.2f:c) Divorced ^7j_j_- 1 ( ASK Q . 2l.c c ) | -21 Widowed -2l Widowed....'. -21 ~"~ j "*"— J (RECORD IN S.R.B. PC 2, 1 (RECORD IN S.R.B. PG 2 1 (RECORD IN S.R.B. PG 2 1 20cc.In what month and year were you (divorced/ widowed/separated)? 21cc.In what month and year were you (divorced/ widowed/separated)? 22cc.In what month and year were you (divorced/ widowed/separated)? MONTH YEAR 1 1 1 1 1 l l I l l (68) (69) (70) (71) (IF A SECOND MARRIAGE GO TO Q.21a) 01 79-80 T I MONTH YEAR 1 I I 1 1 1 l-l 1 1 (68) (69) (?6) (71) (IF A THIRD MARRIAGE GO TO Q.22a) 02 79-80 1 1 18 MONTH YEAR 1 I I 1 ] l l I l l (68) (b9) (70) (71) (RECORD OTHER MARRIAGES IN S.R.B. PG 21-25) 03 79-80 1 I �812039 CARD Q2 3 23a. Have you ever lived together as a partner for 3 month s or more with someone other than your ( w i f e / w i v e s ) ? Yes ( 12C -1 (ASK ().23b) _-) Refused -3 f j 23b. How many times di 1 you live as a partner with someone fo c 3 months or more? TO Q.25) I (WRITE I 1 NUMBER) 1 I 1 I I times 03 ) ( 14) FIRST PARTNER SECOND PARTNER 24n. In what month and year t1 Ml yon b e g i n l i v i n g w i t h a partner (the f i r s t time)? 1 1 24h. In what month and year i l i d ynn liegin 1 i v l n g w i t h a partner the second time? MONTH YEAR 1 I I 1 1 1 l-l 1 1 (15) 116) (17) (18) 1 1 24b. How old was she at that time? (WRITE IN AGE) 1 I 1 1 1 (WRITE IN AGE) | 1 I (WRITE IN AGE) 1 I Yes.( 26( -1 (ASK Q.24e) -2 (SKIP TO Q.24g) No 24L. 24e. When was that? 1st YEAR 1 1 1 1 ) (30 ) 1st MONTH YEAR 1 I I 1 1 2 n d I l l I l (31 ) (32 ) (33 ) (34 ) -1 (ASK Q.24L) -2 (SKIP TO Q.24n) When was that? 1 1 MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (27 ) (28 ) (29 ) (30 ) MONTH YEAR 1 I I 1 1 l l I l l2 n d I (31 ) ( 32) (33 )( 34) 1 1 I 1 I 24q. In what month and year did this relationship end? 1 I -1 24k. Did this partner ever become pregnant by you? 1 I ( 19) GO ) MONTH YEAR 1 1 1 1 1 l l I l l (21 ) (22 ) (23 ) (24 ) Current . . ( 25( ~1 MONTH 1 I I . 1 l-l (27 ) (28 ) (29 1 1 I | (19 ) ( 20) 24 j . In what month and year did this relationship end? 24d. Did this partner ever become pregnant by you? 1 1 MONTH YEAR 1 1 1 1 1 I l - l 1 1 ( 15) (16 ) (17 ) ( Ib) 24p. How old was she at that time? MONTH YEAR 1 I I 1 1 l l I l l (21 ) (22 ) (23 ) (24 ) Yes. (26 ( No . .. 1 I 241. How old was she at that time? 24c. In what month and year did this relationship end? Current . . (_£^ 24o. In what month and year diil you b « g i n l i v i n g with a partner the third time? MONTH YEAR 1 I I 1 1 1 l-l 1 1 (15) (16) ( 17) (18) (19 ) ( 2(J 1 I THIRD PARTNER MONTH YEAR 1 1 1 1 1 l l I l l (21 ) C22 ) G3 ) ( 24) Current.. ( 25( 24r. -1 Did this partner ever become pregnant by you? Yes. (26 ( -1 (ASK Q.24s) No -2 (SKIP TO Q.24u) 24s. When was that? 1st 1 1 MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (27 ) (28 ) (29 ) ( 30) l 2nd 1 1 MONTH YEAR 1 I I 1 1 1 l-l 1 1 01 ) ( 32) ~63 ) ( 34) 1 24m. What was the outcome of 24t. What was the outcome of that pregnancy? (What that pregnancy? (What was the outcome of the was the outcome of the second pregnancy?) second pregnancy?) First 1 Second First 1 Second 1 1 Live birth. (35 ( - l i < 3 6 ( -1 Live birth. ( 33C -ll(36 ( -1 Live birth. (35 ( - l | 0 6 ( -1 -2 Miscarriage... -2 1 -2 Miscarriage... -21 -2 Miscarriage... -2| 1 -3| __-3 Stillbirth.... -3| -3 Stillbirth. ...~-3l ~-3 Stillbirth 24f. What was the outcome of that pregnancy? (What was the outcome of the second pregnancy?) First 1 Second Abortion...... -4 1 -4 24g. Did you or your partner use b i r t h control regul a r l y to a v o i d pregnancy? 24n. Did you or your partner use b i r t h control regul a r l y to avoid pregnancy? 24u. Did you or your partner use birth control regul a r l y to avoid pregnancy? Yes.<'J7( No ... Yes. (37 < No Yes(37 ( No -l"l (CO TO NEXT -if PARTNER ~ J Q - 2 4 h ) 0, -1~) <GO TO NEXT -2 r PARTNER -J Q . 2 4 o ) cp 19 -1\ (RECORD ADDITION-2| AL PARTNERS IN J S . R . B . PC; 26) ( 79 �CARP 025 812039 25a. Do you know of any other pregnancies, in addition to those we have already discussed, that you have caused? Yes.(12( (ASK Q.25b) No 2Sb. -1 -2 (SKIP TO Q.26a) 2Sd. When was that? When was that? SECOND FIRST MONTH (13) (14) MONTH YEAR (15) (16) r ii 1-1 n 25f. ri ii Live birth. ( 7 1( 25e. What was the outcome of that pregnancy? -1 Miscarriage.... Stillbirth,,,., -2 -3 PROBE: Were there any others? (IF YES, ASK Q.25d) THIRD MONTH YEAR 1 1 I I 1 1 1 1 l-l 1 1 (23) (24) (25) (26) YEAR ii ii r1-1i ii (18) (19) (20) (21) 25c. What was the outcome of that pregnancy? Live birth. ( 2 2( Miscarriage.... Stillbirth 25g. What was the outcome of that pregnancy? _i -2 —3 PROBE : Were there any others? (IF YES, ASK Q.25f) When was that? Live birth. ( 7 ( 2 Miscarriage. . . ._ Stillbirth -i -2 -3 PROBE: Were there any others? (IF YES, GO TO S.R.B. PAGE 27 Q.156) 26a. Did you ever try for a period of a year or more to conceive a child without being able to? Yes.(28( • No,...,. 26b. When was that? -2 (PROBE: First time (ASK Q.26b) . (SKIP TO Q.27) Were there any other times?) Second time FROM Third time 1 I MONTH C9 ) ( 30) ( 31) ( 32) (49) (50) (51) (52) (39) (40) MONTH 65 ) (36) YEAR I i I I F^T («') <«°) TO YEAR I 63 ) (3A) YEAR 1 MONTH I YEAR (37) (3B) TO MONTH FROM MONTH 1 T l I l-l FROM 1AR MONTH TO YEAR T 1 I 1 ~ I I l-l I (41) (42) (43) (44) 26c. During (this period/any of these periods) did either you or your partner see a doctor to discuss any difficulties in conceiving children? Yes.(53( No...777" -1 "-2 �CARD 027 812039 I ASK ETORYONET I HAND RESPONDENT CARD "D" j 2 7 a . T h e r e are many reasons that some couples find it difficult or impossible to conceive a child. Please read this card and tell me the letter for each reason which ever applied to you or a spouse or partner. Any other reason? TASK Q.27b AND Q.27c FOR EACH REASON IN Q.27a7| 27b. Did reason (LETTER) apply to you or your spouse? 27c. I MULTIPLE RECORD BELOW I In what year did this occur or become known to you? _ Q.27c Q.27b Year (28) Q.27a (29 ) A. Respondent.... (_18( Sterility due to surgery......(12 ( -1 -1 I l._ . 1 f'31)I . 00 ) Sr>ouse/partner(19 B. Respondent.... (20 ( -1 C33) 021 Sterility due to injury, accident, or illness (SPECIFY) I I -1 Spouse/partnerfel ( -1 I I (4 3) T I f35) C. Re sponden t.... (22 ( Sterility due to unknown causes 1 06)(37) Spouse/pa rtner(23(__-l (385 (39) D. Impotence (15 ( -1 <40 E. Respondent....(24 ( -1 Other known medical or physical conditions (SPECIFY) -1 Spouse/partner (25_(_-l F. Respondent.... g6 ( -1 Some other reason (SPECIFY) -1 Spouse/partner(27 2] I I �812039 CARD 028 28. How many children have you had — that is, of how many children are you the natural father? Please include children who live with you, those who live elsewhere, and those who may no longer be living. T I children (WRITE IN NUMBER) (ASK Q.29) da ) ( 13) No children A ( -1 (SKIP TO Q.33) 29. Starting with your first child, what is the first and last name of the child as it appears on the birth certificate? RECORD FIRST AND LAST NAMES OF ALL CHILDREN IN S.R.B. - PAGE 3-4. WRITE IN THE FIRST NAME ONLY AT THE TOP OF THE APPROPRIATE COLUMN(S). THIRD CHILD SECOND CHILD FIRST CHILD NAME: NAME: 30a. How old is (CHILD) now? 31a. How old is (CHILD) now? 1 1 1 1 1 1 I 1 Age (15 ) (16 ) Child died. . ( I X -1 ( I? Male 1 1 -1 I I 1 1 1 1 Age (15 ) ( 16) Child died..(17( -1 (18 ( Male (18 ( -1 32c. How much did (CHILD) weigh at birth? • POUNDS OUNCES I I I 1 j l l I l l (19 ) fcb ) Cl ) ( 22) Don't know...(23( -1 32b. (Is/Was) (CHILD) male or female? -1 31c. How much did (CHILD) weigh at birth?. POUNDS OUNCES I I I I I 1 l-l 1 1 (19 ) (20 ) (21 ) ( 22) Don't know... (23 ( as ) <i6 ) 32a. How old is (CHILD) now? l 1 Age 31b. (Is/Was) (CHILD) male or female? -1 -2 30c. How much did (CHILD) weigh at bir?h? I 1 l 1 Child died..( IX 30b. (Is/Was) (CHILD) male or female? Male Female... NAME: -1 1 1 POUNDS OUNCES I I I 1 1 1 l-l 1 1 09 ) CO ) (21 )( 22) Don't know...( 23( -1 30d. What is (CHILD) 's birth- 31d. What is (CHILD) 's birth- 32d. What is (CHILD) 's birthdate? date? date? MONTH DAY YEAR MONTH DAY YEAR MONTH DAY YEAR 1 I I 1 I I 1 1 1 1 1T I I I I I 1 1 1 1 1 1 1 1 1 1 l-l 1 l-l 1 1 1 1 l-l 1 l-l 1 1 124) (25) ( 2d (27) £>8 ) (29 ) l2t> ) (25 ) ( 2Q (27 ) *28 ) (29 ) *24 ) (25 ) ( 2 © ( 2 7 ) (28) (29 ) 1 1 l-l ALSO RECORD IN S.R.B. -PC 3l 1 l-l 1 1 lALSO KECORD IN S.R.B. -PC 3| lALSO RECORD IN S.R.B. -PC 3| >0e. Was the child premature, 31e. Was the child premature, 32e. Was the child premature, f u l l term, or overdue? full term, or overdue? full term, or overdue? Premature. OO ( -1 Overdue -3 (CO TO Q.30f) Premature.( 3Q[ Full term (GO TO Q.3H) -1 -2 Premature. ( 30^ Full term.. . (GO TO Q . 3 2 f ) -1 —2 �CARP 026 91203J. SECOND CHILD FIRST CHILD THIRD CHILD 30f. Where are f (CHILD) 's birth registration records located? In what city and state is that? 1 RECORD IN S.R.B. PG 3 | 3lf. Where are (CHILD) 's birth registration records located? In what city and state is that? 32f. Where are (CHILD) 's birth registration records located? In what city and state is that? IRECORD IN S.R.B. PG 3 1 30g. Where ate (CHILD) 's current medical records located' I" wnat city and ptafe is that? (RECORD IK S.'R'.B. PG 3 1 31g. Where are (CHILD) 's current medical records located? In what city and state is that? ] RECORD IN S.R.B. PG 3 1 32g. Where are (CHILD) 's current medical records located? In what city and state is that? 30h. What was (CHILD) 's mother's full name? 1 RECORD IN S.R.B. PG 3 1 31h. What was (CHILD) 's mother's full name? [RECORD IN S.R.B. PG 3 f 32h. What was (CHILD) 's mother's full name? (RECORD IN S.R.B. PG 3 1 30i. How old was the mother when (CHILD) was born? 31 i. How old was the mother when (CHILD) was born? 32i. How old was the mother when (CHILD) was born? 1 1 1 1 1 1 (31) (35) TRECORD IN S.R.B. PG 3 I 1 1 Age 1 1 Age 1 1 (31) (32) 30 j. Were either of you usin: 31 j. birth control at the time she became pregnant with (CHILD)? Yes. (33(^-1 (ASK Q.30k) No 1 I IRECORD IN S.R.B. PG 3 1 1 1 1 1 Age (31 ) ( 33 Were either of you using 32 j. Were either of you using birth control at the birth control at the time she became pregnant time she Became pregnant with (CHILD)? with (CHILD)? Yes.(33( -2 (SKIP TO Q.301) No, Yes.(33j -1 (ASK 0-32k) . -1 (ASK Q.31k) -2 (SKIP TO Q.31L) No -2 (SKIP TO Q.32L) IHAND RESP^NDEST CARD' "c'M 30k. IHAND RESPONDENT' CARD "c"l IHAND RESPONDENT CARD "C"l Please look at this 31k. Please look at this 32k. Please look at this card and tell me all o1 card and tell me all of card and tell me all of the numbers that apply the numbers that apply the numbers that apply to the types of birth to the types of birth to the types of birth control you or your control you or your control you or your partner were practicpartner were practicpartner were practicing? ing? ing? 31.(34( 32.(35( 33. ( R 3 ).3{ 4C7 35.(38( -1 -1 -1 -1 -1 06.(39( 07.(TO~( O8.(4ir 09.CS2T 1.« 0(( 11. 0 5 -i -1 -1 -1 -1 -1 12 (SPECIFY) Ol.(34( 02. (35? 0.3( 3(5 04. ( 7 37 05. ( 5 3? -1 -1 -1 -1 -i 06.(39( -1 07. (Z67 -1 08. ( T 3 ( -1 09. ( 2 7 1 -1 10. ( ^ -i " 11. (W -i 12 (SPECIFY) .5 U( (GO TO Q.30L) .« (( -1 (GO TO Q.31D 23 -1 01,(34( 02.(55( 03. ("351 04. ( 7 3? 05.(3B( -1 -1 -1 -i -i " 12 (SPECIFY) 06.(39( -1 01. (KR -1 08. ( ^ -1 O9.(72l -1 10.(43( -1 11. (W -1 • UM (GO TO Q.32L) -1 �CARD 028 812039 THIRD CHILD SECOND CHILD FIRST CHILD 30L. How many months did it 31L. How many months did it 32L. How many months did it cake her to become pregtake her to become pregtake her to become pregnant with this child? nant with this child? nant with this child? 1 1 1 I Months (46) (47) (46) (47) Less than 1 month. (48 ( ( 1 1 1 Months "1 Leas than 1 month. (48 ( Wasn't trying .,,.... 1 1 Months (46) (47) 4 -1 Less than 1 month. ( 8 ( -2 Wasn't trying 30m. Did (CHILD) have any birth defects? 31m. Did (CHILD) have any birth defects? 32m. Did (CHILD) have any birth defects? Yes. (49 ( -1 (ASK Q.SOn) Yes. (49 ( -1 (ASK Q.31n) -1 -2 Yes. ( 9 ( - 1 (ASK Q.32n) 4 No,..,.. -2 (SKIP TO Q,30o) No -2 (SKIP TO Q.31o) 30n. What kind of birth defects did (s)he have? Any others? 31n. What kind of birth defects did (s)he have? Any others? 32n. What kind of birth defects did (s)he have? Any others? 30o. Was (CHILD1) ever diagnosed as having cancer? 31o. Was (CHILD) ever diagnosed as having cancer? 32o. Was (CHILD) ever diagnosed as having cancer? Yes.(50( Yes.(50( Yes.(50( -1 (ASK Q.30p) -1 (ASK Q.31p) No,...,. -2 (SKIP TO Q.31r) No 30p. In what month and year 31p. was the diagnosis made? MONTH , YEAR 1 1 1 . 1. I I (51) f52) (53) (54) -1 (ASK Q.32p) -2 (SKIP TO Q.32r) In what month and year 32p. In what month and year was the diagnosis made? was the diagnosis made? MONTH 1 1 (51) (52) YEAR I I 1 1 (53) (54) MONTH YEAR 1 -1 (51) (52) (53) (54) 30q. What kind of cancer was diagnosed? 31q. What kind of cancer was diagnosed? 32q. What kind of cancer was diagnosed? (55-56) (55-56) (55-56) Not sure.,( 5X (GO TO Q.30r) -1 Not sure..(57( (GO TO Q.31r) -1 Not sure..(57( (GO TO Q.32r> _i �812039 . CARD ,028 SECOND CHILD FIRST CHILD THIRD CHILD 30r. (Does/Did) (CHILD) have a 31r. (Does/Did)(CHILD) have a 32r. (Does/Did) (CHILD) have i diagnosed learning disdiagnosed learning disdiagnosed learning disability? ability? ability? Yes.(5g.(,. -1 (ASK Q.30s) Yes. (58 ( -1 (ASK Q.31s) No 30s. What kind of learning disability (does/did) (s)he have? Yes.(58( -2 (SKIP TO Q.31t) No 31s. What kind of learning disability (does/did) (s)he have? -1 (ASK Q.32s) -2 (SKIP TO Q 32t) 32s. What kind of learning disability (does/did) (s)he have? 30t. (Does/Did) (CHILD) have 31t. (Does/DidMCHILD) have 32t. (Does/Did) (CHILD) have any physical, mental, or any physical, mental, or any physical, mental, 01 motor impairments? motor impairments? motor impairments? Yes.(52J__-l (ASK Q.30\i) No Yes. ( 9 ( -1 (ASK Q.31u) 5 -2 (SKIP TO Q.30v) No...... » 30u. What kind of impairment (does/did) (s)he have? Yes.(59( -2 (SKIP TO Q.31v) No...... 31u. What kind of impairment (does/did) (s)he have? "1 (ASK Q.32u) -2 (SKIP TO Q 32v) 32u. What kind of impairment (does/did) (s)he have? V IF CHILD IS DEAD: CONTINUE OTHERWISE: SKIP TO NEXT CHILD 30*v. On what date did (CHILD) die? IF CHILD IS DEAD: CONTINUE OTHERWISE: SKIP TO NEXT CHILD 31v. On what date did (CHILD) die? IF CHILD IS DEAD: CONTINUE OTHERWISE: SKIP TO NEXT CHILD 32v. On what date did (CHILD) die? ) MONTH DAY YEAR MONTH DAY YEAR MONTH DAY YEAR 1 I I 1 I I 1 1 I I V I I I I 1 1 l-l 1 1 1-1 .1.1-1 1 , 1 1 -1 -1 1 6) 60) (61) (62) (63) (64) (65) (60) (61 ) (62) (63) (64) (65)( 0 (61) (62) (63) (64) (65) iOw, What was the cause of death? 31w. What was the cause of death? 32w. What was the cause of death? (Ox. Where ie (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. PC 3 1 31x. Where is (CHILD) 's death registered? In what city and state is that? (RECORD IN S.R.B. PG 3 I 32x. Where is (CHILD) 's death registered? In what city and state is that? (RECORD IN S.R.B. FG 3 (GO TO NEXT CHILD Q.31a) _ _ 79-80 (GO TO NEXT CHILD Q.32a) J32 79-80 (RECORD ADDITIONAL CHILDREN IN S.R.B. - PC 28-39) J)3 7<J-80 �812039 CARD 033 33. Now let's talk about your health. Compared to other people your age, would you say that your health is excellent, good, fair, or poor? Excel lent.... (12( Good ............. -1 -2 -3 Poor. 34a. Did a doctor ever tell you that you had pneumonia? Yes. (13 ( -1 No......___ _-2 34b. (ASK Q.34b) (SKIP TO Q.36a) How many times have you had pneumonia? (WRITE IN NUMBER) i I _ r J I times 04 ) ( 15) 35a. During what month? and years did you have pneumonia (the first time)? [RECORD IN S.R.B. PcTT IF BEFORE 1961, SKIP TO Q.35f. 35f. During what months and years did you have pneumonia (the second time)? IRECORD IN S.R.B. PC 5 1 IF BEFORE 1961, SKIP TO Q.35k. 35b. What is the full name 35g. What is the full name of the doctor who made of the doctor who made the diagnosis or the the diagnosis or the medical facility where medical facility where the diagnosis was made? the diagnosis was made? IRECORD IN S.R.B. PC 5 I 35c. What prescribed medicine did you take for the pneumonia you had that time? Third Time Second Time First Time [RECORD in S.R.B. PC 5 I 35k. During what months and years did you have pneumonia (the third time)? IRECORD IN S.R.B. PC 5 I IF BEFORE 1961, SKIP TO 35L. What is the full name of the doctor who made the diagnosis or the medical facility where the diagnosis was made? I RECORD IM S.R.B. PG 5 I 35h. What prescribed medicine did you take for the pneumonia you had that time? 35m. What prescribed medicine did you take for the pneumonia you had that time? 35i. Were you hospitalized for the pneumonia you had that time? 35n. Were you hospitalized for the pneumonia you had that time? 1. 2. 3. 35d. Were you hospitalized for the pneumonia you had that time? Yes.(l6( No 35e. -1 (ASK Q.35e) Yes.OJJ^-l (ASK Q.35i) Yes.(18( -1 (ASK Q.35o) -2 (SKIP TO Q.35f) No -2 (SKIP TO Q.35k) No _-2 (SKIP TO Q.36a) What was the full name of that hospital? IRECORD IN S.R.B. PG 5 I 35j. What was the full name of that hospital? IRECORD IN S.R.B. PC 5 I 35o. What was the full name of that hospital? I RECORD IN S.R.B. PG 5 I �036 36a. 812039 Did a doctor ever tell you that you had cancer? Yes..(;2( -1 (ASK Q.36b) No -2 (SKIP TO Q.37) 13-14 36b. In which parts of your body was cancer located? LIST EACH BODY PART BELOW. IF MORE THAN THREE BODY PARTS, USE S.R.B. - PACE 46 FOR ADDITIONAL PARTS. Part 1 Part 1 J6c. In what month and year 36i. vet cancer of the (BODY PART) first diagnosed? Part 3 In what month and year 36o. was cancer of the (BODY PART) first diagnosed? IRECORD IN S.R.B. PG 5 I RECORD IN S.R.B. PC 5 1 in what month and year was cancer of the (BODY PART) first diagnosed? (RECORD IN S.R.B. PG 5 j 36 j. What is the full name 36p. What is the full name 36d. What is the full name of the doctor or the of the doctor or the of the doctor or the medical facility where medical facility where medical facility where the diagnosis was made? the diagnosis was made? the diagnosis was made? IRECORD IN S.R.B. PC 5 | RECORD IN S.R.B. PC 5 1 IRECORD IN S.R.B. PG 5 1 I6e. What is the full name of, the doctor or the medical facility you last consulted about Cancer of the (BODY PART)? RECORD IN S.R.B. PG 5 | 36k. What is the full name of the doctor or the medical facility you last consulted about cancer of the (BODY PART)? IRECORD IN S.R.B. PG i" 36L. 6f. During what month and year did you last consult (NAME FROM Q.36e)? What is the full name of the doctor or the medical facility you last consulted about cancer of the (BODY PART)? IRECORD IN S.R.B. PC 5 I 36 r. During what month and During what month and year did you last conyear did you last consult (NAME FROM Q.36q)7 sult (NAME FROM Q.36k)? [RECORD IN S.R.B. PG 5 1 IRECORD IN S.R.B. PC 5 RECORD IN S.R.B. $ 5 1 36q. 36s. What treatments or 36m. What treatments or What treatments or medicines did you take medicines did you take medicines did you take for cancer of the (BODH for cancer of the (BODY for cancer of the (BODY PART)? PART)? PART)? (MULTIPLE RECOR6 BELOW I (MULTIPLE RECORD BELOW | [MULTIPLE RECORD BELOW 1 )6g. Radiation .^ ( ( -1 Chemotherapy. . . ( ° * ( -1 Radiation (15( -1 Chemotherapy. . . (16( -l Radiation (15( -1 Chemotherapy. . . (16( -1 Other (SPECIFY) Other (SPECIFY) ' Other (SPECIFY) .(18( .(18( -1 6h. During what month and year did you first receive (EACH TREATMENT CODED IN Q.36g) for cancer of the (BODY PART)? 36n. ~ -1 During what month and year did you first receive (EACH TREATMENT CODED IN Q.36m) for cancer of the (BODY PART)? .U8( -1 36t. During what month and year did you first receive (EACH TREATMENT CODED IN Q.36s) for cancer of the (BODY PART)? MONTH YEAR MONTH . YEAR MONTH YEAR adia1 i 1 T 1 1 RadiaT i l l RadiaI I 1 1 1 tion.... 1 1 l-i 1 I tion.... 1 1 1-1 1 1 tion.... 1 |. ... 1 J fl9> T20) hi!) (22> < 19) ( 20) ( 2 ) 62 ) : (19)( 20> < 21> < 22> MONTH YEAR MONTH YEAR MONTH YEAR hemo1 1 1 1 1 Chemo1 I 1 1 i Ch emoI 1 therapy. . 1 - 1 1 1 therapy. 1 1 1-1 1 therapy. 1 „ 1 (23) (24) (23) (56) ( 23) ( 24) ( 25) i 265 (23 ) ( 24> ( 25) ( 26) MONTH YEAR MONTH YEAR MONTH YEAR 1 1 1 1 1 1 1 1 1 1 ( 1 1 f 1 T uraery.. 1 1 1-1 1 1 Surgery.. 1 | I- I 1 | Surgery.. 1 1 1J | ( f 27) 66 ) ( H) ( 30) ( 27) ( 28) C29 ) ( 30) 27 > (28 > V29 > (30 > MONTH YEAR MONTH YEAR MONTH YEAR . 1 1 1 1 I 1 1 1 1 1 I 1 1 1 1 1 ther.... 1 l-l I I Other.... | | 1-| Other.... 1 1 H I . .1 ( 3lH J2) H3) ( 34) <3l ) (32 > (31 5 ~<J4 ) <-3! * <3!>> l 33> <34> (00 TO NKXT BODY PART) (GO TO NEXT BODY PART) (GO TO NEXT BODY PART IN 01 7'i-R() S.R.B. PACE 40) 02 03 7 9"- 80 79- HO 27 �CARD 037 812039 IIF LEUKEMIA NOT PREVIOUSLY MENTIONED. ASK:I 37a. Has a doctor ever told you that you had leukemia? Yes.l'(12( . -1 No 37b. -2 (ASK Q.37b) (SKIP TO Q.38) In what month and, year was your leukemia first diagnosed? I RECORD IN S.R.B. - PC 6 I 37c. What is the full name of the doctor or the medical facility where the diagnosis was made? TRECORD IN S.R.B. - PC 6 1 37d. P, What treatments or medicines have you taken for leukemia? I RECORD BELOW I MEDICINE/TREATMENT i. , _ E. FIRST'RECEIVED ,.^._..-_ ^____ r MONTH YEAR i Ii i-i Ii Ir n l4) (15) (16) MONTH 2. YEAR I J, I-I I, I (17) (18) (1?) (20) :YEAR 3. _ . . 1 1 T I I I I I, I-I I, I (21f ( 2 (23) 2) (24) 37e. During what month and year did you first receive (EACH TREATMENT OR MEDICINE IN Q.37d)? I RECORD ABOVE] 37f. What is the full name of the doctor or medical facility you last consulted about your leukemia? JRECORD IN S.R.B. - PC 6 j 37g. During what month and year did you last consult (NAME IN Q.37O? JJECORD IV S.R.B. - PC 6 I /25-60/ 28 �38a, had. I; yogld like to ask you some questions about other medical conditions you may have 1. Did you ever have diabetes? Yes No ( 1 ;_ 6( -I ("X" BOX ON PAGE 26) "-2' ' 2. Did you ever have thyroid problems? Yes (SPECIFY) .<62( No..,....., -1 T-r-,,.-,-,. "2 ( X BOX ON PAGE "" 3, Did you ever have anemia? Ye No (3 £( -1 -I ("X" BOX ON PAGE 26) A. Did you ever have a heart condition? Yes (SPECIFY) -I ("X" BOX ON PAGE 26) No...... . . . . ..... 5. Did you ever have an enlarged liver? Yea........<£5(_ No .......... -1 -2 <"X" BOX ON PAGE 2 . 6) 6. Did you ever have jaundice? Yea ........ ( , ( 66 , -1 No ............. -2 ( X BOX ON PAGE 27) "" 7. Did you ever have hepatitis? Yes ........ (67( -1 No .......... . . _ ; _..' -2 .. ("X" BOX ON PAGE 27) 8. Did .you ever have cirrhosis of the liver? Yes........( 6ft No......, ...... "1 -2 ("X" BOX ON PAGE 27) 9. Did you ever have intestinal parasites? Yes No, (9 $( -1 ("X" BOX ON PAGE 27) -2 10. Did you ever have gall bladder problems? Yes., No ( 7QL -1 ("X" BOX ON PAGE 27) '"™™.2 '_^*" 11. Did you ever have eny other liver condition? Yes (SPECIFY) .1 ("X" COX ON PAGE 2?) No 12. Did you ever have a respiratory condition other than pneumonia? Yes (SPECIFY) NO ,,, .7( (2 ~1 ("x11 BOX ON PACE 28) -1 <"X" 8PX ON PACE 28) ,..,,, .„,,;,.__•• 2 13. Did you ever have any other major condition' Yes (SPECIFY) ,<73J N o , , , , . , . , , . , , , , , , , ; , , , , , -_____-2 �CARD DIABETES I ASK Q.38b THROUGH Q . 3 8 h 1 IFOR EACH BOX "X"ED ON I IFF. 26-28 1 38b. 38c. When did a doctor first tell you that you had (CONDITION)? What is the full name of the doctor who made the diagnosis or the medical facility where the diag38d. 1 I THYROID PROBLEMS A HEART CONDITION ANEMIA I AN ENLARGED LIVER I IRECORD IN S.R.B.I IPAGE 7 1 (RECORD IN S.R.B.I (PAGE 7 1 IRECORD IN S.R.B.I IPAGE 7 1 IRECORD IN S.R.B.I IRECORD IN S.R.B.I IPAGE 7 1 IRECORD IN S.R.B.I IPAGE 7 1 I RECORD IN S.R.B.I IPAGE 7 1 IRECORD IN S.R.B.I IPAGE 7 I 1 PAGE 7 1 RECORD IN S.R.B.I IPAGE 7 1 1 1 RECORD IN S.R.B.I IPAGE 7 1 Do you have (CONDITION) -Yes.(l2( -1 (ASK Q.38e) Yes.( 12( No -2 (SKIP TO Q. No ~ 38g) 38e. I I 812039 038 Are you currently talcing any prescribed medicines for your (CONDITION) -Yes.(13( -1 (ASK Q.38f) Yes.( 13( -1 (ASK Q.38e) Yes.(12( -2 (SKIP TO Q. 38g) -1 (ASK Q.38e) Yes.(12( -1 (ASK Q.38e) Yes.(12( Uo -2 (SKIP TO Q. No. 38g) ~ 38g) -1 (ASK Q.36e) -2 (SKIP TO Q. -1 (ASK Q.38e) Yes.(13( -1 (ASK Q.38f) Yes.O3( (ASK Q.38f) -2 (SKIP TO Q. 38g) -1 (ASK Q.38f) No 38g) 38f. 38g. When did you last consult a doctor for (CONDITION)? 38g) 38g) 38g) What are the names of the medicines you are taking for (CONDITION)? 38h. What is the full name cf the doctor or medical facility you Last consulted about your (CONDITION)? - IRECORD IN S.R.B.I IPAGE 7 1 (RECORD IN S.R.B.I IRECORD IN S.R.B.I IPAGE 7 1 IRECORD IN S.R.B.I IPAGE 7 1 IRECORD IN S.R.B.I IPAGE 7 1 (RECORD IN S.R.B. | IRECORD IN S.R.B. I (PAGE 7 j IPAGE 7 1 (GO TO NEXT CONDITION "X"ED) (GO TO NEXT CONDITION "X"ED) (GO TO NEXT CONDITION "X"ED) 61 IPAGE 7 62 79-80 I 79-80 IRECORD IN S.R.B.I IPAGE 7 1 (GO TO NEXT CONDITIOt< "X"ED) 79-80 1 RECORD IN S.R.B. 1 IPAGE 7 1 1 RECORD IN S.R.B. ! IPAGE 7 1 (GO TO NIEXT CONDITION "X"ED) 65 �CARDO38 JAUNDICE IFOR EACH BOX "X^ED ON 1 1 1 1 ! 1 (PP. 26-28 I 1 1 1 "i.ASK O..38b THROUGH Q,38h I When did 8 doctor first tell yeu that you had (CONDITION)?, 38c. 38d, HEPATITIS 1 1 1 1 1 CIRRHOSIS OF THE LIVER 1 INTESTINAL PARASITES ! 1 1 1 1 t 1 1 ' ! 1 1 1 1 1 jRECOJtD IN S.R.B.1 What is the full name of the doctor vho made the diagnosis or the medical facility where the diag- (RECORD IN S. R.B.I IPAGE 7 IPAGE 8 I 1 RECORD IN S.R.B.1 IPAGE 7 1 I (RECORD IN S.R.B.1 IPAGE 8 I 1 RECORD IH S.R.B.1 IPAGE 8 1 1 RECORD IN S.R.B.1 IPAGE 8 1 J 1 1 1 1 RECORD IN S.R.B.1 IPAGE s 1 I RECORD IN S.R.B.1 IPAGE s 1 GALL BLADDER PROBLEMS 1 I i I IRECOR& IN s. R.B.I (PAGE S I (RECORD IN S.R.B. 1 IPAGE s 1 Po you have (CONDITION) .Yes! (12 ( -1 (ASK Q,38e) Yes. 02 ( -1 (ASK Q.38e) Yes. (12 ( -1 (ASK Q,38e) Yes.02 ( -1 (ASK Q.38e) Yes.Q2 ( -1 (ASK Q.38e) {jo -2 (SKIP TO Q. No -2 (SKIP TO Q. No. ... -2 (SKIP TO Q. -2 (SKIP TO* Q. Ho -2 (SKIP TO Q. No 38g) 38g) ~~ 38g) 38g) 38g) 38e. Are you currently taking X any prescribed medicines for your (COHWnOU) .Yes. (13 ( -1 (ASK Q.38f) Yes. ( 3 ( -1 (ASK Q.38e) Yes. (13 ( -1 (ASK Q.38f) 1 Yes. 03 ( -1 (ASK Q.38f) Yes. (13 ( -1 (ASK Q.38f) -2 (SKIP TO Q. No... -2 (SKIP TO Q. Ho -2 (SKIP TO Q. H . . 7 7 -2 (SKIP TO 0 . Ho o. 7 . -2 (SKIP TO Q No 38g) 38g) 38g) 38g) What are the names of the medicines you are taking for (CDtTOtTIOD)? When did you last consult a doctor for (CONDITION)? What is the full name of the doctor or medical facility you last consulted about your [RECORD IN S.R.B.} (PAGE 7 1 I RECORD IN S.R.B.1 IPAGE 8 1 (RECORD IN S.R.B.1 IPAGE 7 1 (RECORD IN S.R.B. I (GO TO NEXT CONDITION «X"EO) 66 79-HQ' IPAGE s 1 (GO TO NEXT CONDITION -X"ED) 67 79-BO (RECORD IN S. R.B.I IPAGE s 1 (RECORD IN S.R.B.1 (PAGE 8 1 (GO TO NEXT CONDITION "X"ED) 68 79-80' (RECORD IN S.R.B.1 (PAGE 8 1 (RECORD IB s. R.B.I (PAGE s 1 1 RECORD IN S.R.B.1 (RECORD IN S.R.B.T IPAGE 8 1 IPAGE s I (GO TO NEXT CONDITIOh "X"ED) 69 79-SU (GO TO NEXT CONDITION "X"ED) 70 79-80 �CARD 038 ANY OTHER LIVER CONDITION _____ I I I I (ASK Q.38b THROUGH Q.38h IFOR EACH BOX "X"ED ON IFF. 26-28 38b. When did a doctor first tell you that you had (CONDITION)? A RESPIRATORY CONDITION I OTHER. THAN PNEUMONIA I I I I I I I (RECORD IN s. R.B.I IPAGE 8 ! IRECORD IN s. R.B.I IPAGE 8 1 IRECORD IN s. R.B.I IPAGE 9 1 1 RECORD IN S. R.B.I 38c. What is the full name of the doctor who made the diagnosis or the medical facility where the diagnosis was made? IRECORD IN S.R.B. I IRECORD IN S.R.B.I IPAGE 9 1 (PAGE s 1 (PAGE 8 ( 12( -1 '(ASK Q.38e)Yes.( 12( No. . . -2 (SKIP TO Q. No .. 38g) 38e. I I I I ANY OTHER . MAJOR CONDITION 1 -1 (ASK Q.38e) Yes. 02 ( -1 (ASK Q.38e) -2 (SKIP TO Q. No...... -2 (SKIP TO Q. 38g) , 38g) Are you currently taking any prescribed medicines (13 < -1 (ASK Q.38f)Yes.( 13J -1 (ASK Q.38e) Yes. (13 ( -1 (ASK Q.38f) -2 (SKIP TO Q. No. . . -2 (SKIP TO Q. No -2 (SKIP TO Q. No .. 38g) ~ 38g) .8> 'g 38f. What are the names of the medicines you are taking for (CONDITION)? Any others? 38g. When did you last consult a doctor for (CONDITION)?. 38h, What is the f u l l nave, of the doctor or aedicat f a c i l i t y you last consulted about your (CONDITION)?.. 1 RECORD IN S.R.B. 1 IPAGE 8 1 IRECORD IN S.R.B. I IPAGE 8 1 IRECORD IN S.R.B.I IPAGE 9 1 (RECORD IN S.R.B. 1 IPAGE 8 1 IRECORD IN s. R.B.I IPAGE 8 1 IRECORD IN S.R.B.I TO NEXT CONDITION "X"E :D) 71 79-80 (GO TO NEXT CONDITIO •"X"ED) „ f 79-80 IPACE 9 I 73 79-80 812039 �812039 CARD 039 39. Have you ever had acne on your face':1 Yes..(12( _-l -2 No 40a. (ASK Q.40a) (SKIP TO Q.42) During what year did you last have acne on your face? I Year (WRITE IN YEAR) Before 1961 . .(15( _____.(ASK Q.40b) ~T13) TT4T Second Period First Period 40b. Think about the first time you had acne on your face **• when did it start? MONTH T~7T~-f T I l-l I I 06 )(17)( 18) (19) 40c. Until when did that last? •MONTH YEAR i n r (20) 22) (23) AOf. Think about the second time you had acne on your face -— when did it start? MONTH MONTH YEAR I I L (48) (49) (50) (51) I I (32) (33) (3A) (35) MONTH Third Period AOj. Think about the third time you had acne on your face — when did it start? YEAR AOg. Until when did that last? -1 (SKIP TO Q.A2) 40k. Until when did that last? YEAR I I MONTH YEAR I I (52) (537 ( 5 ( 5 5) 5) I I (36) (37) ( 8 ^ l ) 3 )t 9 40d. Please show me on this diagram where the acne was located (the first time). AOh. Please show me on this diagram where the acne was located. AOL. Please show me on this diagram where the acne was located. I HAND RESPONDENT CARD "E"l I HAND RESPONDENT CARD "E" I I HAND RESPONDENT CARD "E" I MULTIPLE RECORD BELOWT [MULTIPLE RECORD BELOW| [MULTIPLE REQORD BELQWJ templet;'. .'.... QA'r ^T Byes or eyelids. (25 ( ~-l Ears ............ (26( -1 Cheeks . .V(7 ( . ..2 -1 Nc.se Forehead ........ (29 ( Jaw, Chin, Other(30" Temples .TgiO( -1 Eyes or eyelids.(41 ( . -1 Ears (42 ( -1 Temples..» 66 ^ -1 Eyes or eyelids.(57 ( -1 Ears. Cheeks.......... $9 ( -1 Nose* .ftb ( -1 Forehead (ft3• ( . -1 Jaw, Chin, Otherl^T -1 AOe. Did you ever have another period of acne on your face? Yes.Ojj ( -1 (ASK Q . A O f ) No ...... _j-2 (SKIP TO Q.41a) :..:.M Cheeks -1 Nose Forehead........(45 (......-1 Jaw, Chin, Other(4T( -1 AOi. Did you ever have another period of acne on your face? 40m. Did you ever have another period of ache on your face? Yes.(63 ( -1 Yes. (47 ( -1 (ASK Q.AOj) -2 No _ -2 (SKIP TO Q.Ala) No. IF ANY" "Yls" TO TEMPLE, EYES, EYELIDS, OR EARS IN Q.AOd.AOh, OR AOL ABOVE: ASK Q.Ala. ALk OTHERS,! SKIi' TQ Q.42. Ala. Did you ever consult a doctor or medical facility about the acne on your (temples/eyes Of eyelids/ears)? -1 (ASK Q.Alb) -2 (SKIP TO 0.42) 4lb. When did you last consult a doctor about the aene on your (temples/eyes Or eyelids/ears)? s.R.ft. .7" Ale. What was the name of the doctor or medical facility y&u Consulted at the time? S.R.B. r.PC..9 33 �-29a- CARD 135 Aid. When you had this acne on your face did you also have it on your chest, back, shoulders, arms, or legs? Yes ..... (26( (ASK Q.Ale) No.........._ Ale. -1 -2 (SKIP TO Q.A2) Where was that? I CODE ALL THAT APPLY! Chest........(27<_ ,-1 Back ......... (T6( '''"'"'-! _ Shoulders. . . . 2 K_ (j .-1 Arms----.----( 30( _ -1 Legs ......... (3JT -1 Alf. When was that? FROM .MONTH YEAR "~I~ r r~~ T r i Ii I I nrr OAI TSFT TO MONTH (36) (37) YEAR (38) (39) �CARD 039 42a. 812039 Have you ever had (READ EACH COLUMN HEADING)? TlF "YES" TO ANY COLUMN HEADING. ASK Q.42b-h FOR THAT COLUMN I A. B. . Patches Of. .you*,Sfrifl. change color? Easier bruising of the skin than usual? Ycs.»v^( —1 J No ''~_ 2 C. Skin that was extra sensitive or seemed to hurt for no reason? Yes..(67(._-i Yes..(A9( . -1 No -2 Ir. On wlml p a r t of your body did you have (CONDITION)? Any other part? b. On what part 01 your body did you have (CONDITION)? Any other part? b. On what part of your body did you have (CONDITION)? Any other part? C. Did you discuss (CONDITION) with a doctor? c. Did you discuss (CONDITION) with a doctor? c. Did you discuss (CONDITION) with a doctor? Yes.( 66( -1 (ASK Q.42d) Yes.(68( No...... -2 (GO TO NEXT ~ CONDITION) No d. What was the diagnosis? d. e. e. What is the name of the doctor who made the diagnosis or the medical facility where the diagnosis was m?de? -2 (GO TO NEXT ~~ CONDITION) IRECORD IN S.R.B. - PG iol f. During what month and year was the diagnosis made? What is the name of the doctor or medical facility you last consulted about (CONDITION)? f. During what month and year did you last consult (NAME IN Q.42g)? IRECORD I'N S.R.B. - PG iol -1 (ASK Q.42d) No -2 (GO TO NEXT CONDITION) d. What was the diagnosis? e. What is the name of the doctor who made the diagnosis or the medical facility where the diagnosis was made? What is the name of the doctor who made the diagnosis or the medical facility where the diagnosis was made? IRECORD IN S.R.B. - PG iol During what month and year was the diagnosis made? f. IRECORD IN S.R.B. - PG iol g. IRECORD IN S.R.B. - PG iol h. What was the diagnosis? Yes.( 7 ( 0 IRECORD IN S.R.B. - PG 101 IRECORD IN S.R.B. - PG iol g. -1 (ASK Q.42d) What is the name of the doctor or medical facility you last consulted about (CONDITION)? IRECORD IN S.R.B. - PG idl g. IRECORD IN S.R.B. - PG iol h. During what month and year did you last consult (NAME IN Q.42g)? IRECORD IN S.R.B. - PG iol 35 During what month and year was the diagnosis made? What is the name of the doctor or medical facility you last consulted about (CONDITION)? IRECORD IN s.jt.B. - PG iol h. During what month and year did you last consult (NAME IN Q.42g)? IRECORD IN S.R.B. - PG iol �CARD 039 42a. Have you ever had (READ EACH COLUMN HEADING)? I IF "YES" TO ANY COLUMN HEADING. ASK Q.A2b-h FOR THAT COLUMN! D. A rash on your back caused by lower back pain? £. A short period of excessive hair growth caused by lower back pain? f i Yes.. ( 3 ( -1 7 No -2 Yes.. (71 ( -1 No, -2 On what part of your body did you have (CONDITION)? Any other part? b. On what part of your body did you have (CONDITION)? Any other part? c. Did you discuss. (CONDITION) with a doctor? c. Did you discuss (CONDITION) with a doctor? Ye s. (72 ( -1 (ASK Q.42d) Yes. ( 4 (^-1 (ASK Q.42d) 7 No No »•»••• b. -2 (GO TO NEXT ~~ CONDITION) d. What was the diagnosis? *2 d. What wa« the diagnosis? ' e. What is the name of the e. What is the name of the doctor who made the diagdoctor who made the diagnosis Or the medical nosis or the, medical facility where the diagfacility where the diagnosis was made? nosis was made? IRECORDJti S.R.B. - PG 101 (RECORD IN S.R.B. - PG 10| f. During what month and year was the diagnosis made? f. During what month and year was the diagnosis made? RECORD IN S.R.B. - PG 101 h. What is the name of the doctor or medical facility you last consulted about (CONDITION)? g. What is the name of the doctor or medical facility you last consulted about (CONDITION)? I RECORD IN S.R.B. - PG 10J g. IRECORD IN i.R.B. - PG 101 IRECORD IN S.R.B. - PG 10| During what month and year did you last consult (NAME IN Q.42g)? RECORD IN S.R.B. - PG 10| n. During what month and year did you last consult (NAME IN Q.42g)? IRECORD IN S.R.B. - PG 10! 812039 �CARD 812039 043 43a, Aside from injury, has there ever been a period of time when you had (READ EACH " COLUMN HEADING)? j: _ , ___ flF "YES" TO ANY COLUMN HEADING, ASK Q.43b-K FOR THAT COLUMNT B. Persistent tingling sensations in anv of your limbs? A. Persistent numbness in any of your limbs? b. When did you first notice (CONDITION)? When did you first notice (CONDITION)? MONTH YEAR ( 1 1 1 1 1 1 . H 1 1. c. Which limbs or muscles were affected? (CONDITION)? Any other part? c. Which limbs or muscles were affected? (CONDITION)? Any other part? Do you still have (CON- d. Do you still have (CONDITION)? DITION)? f. b. When did you first notice (CONDITION)? MONTH YEAR t 1 1 1 1 1 1 1 l-l 1 1 (43'. (443 ( 5 (16) 41 c. Which limbs or muscles were affected? (CONDITION)? Any • other part? d. Do you still have (CONDITION)? Yes.{47( Yes. (32 ( -1 No -2 Yes.(LZ!_-l No. . . . -2 .. e. NO....VTI '-2 MONTH YEAR 1 ! 1 1 1 1 1 1 l-l 1 'I (28 ) (29 > ( 30) (31 ) 1 ( 1} (14) (15) (16) d. Yes..^2( -! Yes.. (27 ( -1 Yes.. (12(^-1 b. C. Persistent deep burning sensations in any of your limbs? -i During what period was the (CONDITION) most intense? FROM MONTR YEAR 1 1 1 f 1 1 I I H 1 1 (18) <ig) <2Q) <21> TO MONTH YEAR 1 1 1 1 1 1 1 1 H I . I <22> <23) (24) (25) e. During what period was the (CONDITION) most intense? FROM MONTH YEAR 1 1 1 1 1 1 I l l - I l l (33) <34> ( 33 (36) TO MONTH YEAR 1 1 1 1 1 1 1 1 l-l 1 1 (37) (38) (39) (40) e. During what period was the (CONDITION) most intense? FROM MONTH YEAR 1 1 1 1 1 1 I I l-l 1, 1 &b ) 69 ) t 5d) I 51) TO MONTH YEAR 1 1 1 1 1 1 1 1 .1-1. 1 . 1 Did you see a doctor for (CONDITION)? f. Did you see a doctor for (CONDITION)? f. Did you see a doctor for (CONDITION)? CJ2') ('S3) (54) (55) V«t><_a6trJ-l (I* NO, GO TO Yes. ( 6 (_^1 (IF NO, CO TO 5 Yes.(u( -1 (IP NO, GO TO No...... -2 NEXT CONDITION) No...... -2 NEXT CONDITION) g. What was the diagnosis? g. What was the diagnosis? g. What was the diagnosis? h. What is the name of the doctor who made the diagnosis or the medical facility where . .the diagnosis was, made? 1 RECORD IN S.R.6. - PC 111 _ _...........______ _ [RECORD IV g.R.B. - PC 111 i . During what month and, year was the diagnosis made?. [RECORD IK S.R.jB. - "PC "ill' [RECORD IN S.R.B...'.- PC 111 - - _____ | RECORD IN S.R.B. - PC 111 I RECORD IN £.RJB. - PG 1.1 1 j. What is the name of the doctor Or medical facility you last consulted about (CONDITION ) ? TJEGOfiD IN ,.Sr^.,BA3^ k. . IN . S.R.B, - I)u^ri : n£_wh«t itiontVi apd ...year did you l a s t consult (NAME IN Q.4Jjg)? .............. .' j>j .;S»A ;.»•,,- PptllT .iRECOBi) JN .S.R.B. .-. PC ITT ' ' ' �CARD 443 812039 43a. Aside from injury, has there ever been a period of time when you had (READ EACH COLUMN HEADING)? ^ . I IF "YES" TO ANY COLUMN HEADING. ASK Q.A3b-K FOR THAT COLUMN I E. A reduction in grip strength? Persistent aches and pains in any of your limbs? Yes..( 1 2 < No., b. -1 -2 When did you first notice (CONDITION)? Yes..(j7( No b. MONTH YEAiR ! I I 1 1 1 H 1 1 (13) (14) (15) (16) -1 -2 When did you f i r s t notice (CONDITION)? MONTH YEAR 1 I I I I I C8 ) ( W, ( 30) ( 31) 1 1 c. Which limbs or muscles were affected? (CONDITION)? Any other part? c. Which limbs or muscles were affected? (CONDITION)? Any other part? d. Do you s t i l l have (CONDITION)? d. Do you still have (CONDITION)? Yes.G2 J Yes.(l7( i -1 No -2 e. During what period was the (CONDITION) roost intense? FROM MONTH YEAR ,1 I I I 1 :i e. l-l i i as ) (19 : GO ) ( 2i) TO MONTH YEAR 1 1 I I I 1 1 1 l-l 1 1 • , (22 > <23 •) (24 1 ( '251 f. Did you see a doctor fpr (CONDITION)? Yet,.C'-26< No g- During what period was the (CONDITION) most intense? FROM MONTH YEAR 1 1 1 1 1 1 1 1 l-l 1 1 (33) (34) (35) (36) TO MONTH YEAR 1 I I I 1 1 I l l - I l l (37) (38) (39) ftOJ f . Did you see a doctor for (CONDITION)? Yes.(41( -1 (IF NO, GO TO -2 NEXT CONDITION) What was the diagnosis? -1 NO...TTT -2 g. -1 What was the diagnosis? h. What is the name of the doctor who made the diagnosis or the medical facility where the diagnosis was made? 1 RECORD IN 'S.R.B . - PC 111 T ^RECORD IN S.IUB. - PC 11] i. During what month and year was the diagnosis made? |RECORD IN S.R.B. - PC llj I RECORD IN S.R.B. - PC UI j. What is the name of the doctor or medical facility you last consulted about (CONDITION)? • I RECORD IN S.R.B7 - PC YlT I RECORD IN S.R.B. - PC 111 k. During what month and year did you last consult (NAME IN Q.62g)? ' [RECORD IN E.IR,.¥. - PC 1 1 1 IRECORD ~IN S . K . B . - PC 111 38 �__; r_^.: ^ CARD Q4A 812039 44a. (Besides the prescribed medicines you told me about), are you currently taking any (other) medicines prescribed by a doctor? Yes (2 ( ^ No.,, "1 (ASK Q.44b) -2 (SKIP TO Q.45) 44b. For what conditions were the medicines prescribed? Any other conditions? (15C. 3:) �..CARP 045 45a. Have you ever smoked cigarettes regularly for a period of at least one month? Yes (12( -1 (SKIP TO Q.48a) In what month and year did you start smoking cigarettes on a fairly regular basis? MONTH YfAg"" .(13) ,(J4) 45c, (ASK Q.45b) -2 No 45b. 812039, Tl.5) In what month .and year did you ^aat smoke cigarettes on a fairly regular basis? MONTH YEAR 1 T I I (17) (18) (19) (20) 46a. When you started smoking cigarettes on a fairly regular basis in (START DATE), about how many packs per week did you smoke'? By "pack" we mean 20 cigarettes. I packs per week (22) (21) 46b. Until what month and year did you continue to smoke (NUMBER) packs per week on a regular basis? MONTH YEAR 1 | T 1 I T (IF DATE IS THE SAME AS Q.45c: SKIP TO Q.47a. I J J-l I | ALL OTHERS: CONTINUE) (23) (24) (25) (26) 46c, After that, about how many packs per week did you smoke? ~ P I I | I packs per week (27) (28) 46d. Until what month and year did you continue to smoke (NUMBER) packs per week on a regular basis? MONTH YEAR T | TT I "T (IF DATE IS THE SAME AS Q.45c: SKIP TO Q.47a. I j j-l || ALL OTHERS: CONTINUE) (29) (30) 731)132) 46e. After that, about how many packs per week did you smoke? ~ I I I I I packs per week (33) (34) 46f. Until what month and year did you continue to smoke (NUMBER) packs per week on a regular basis? MONTH YEAR I I TT I T (IF DATE IS THE SAME AS Q.45c: ASK Q.47a. I I J-l I I ALL OTHERS: RECORD ADDITIONAL PERIODS IN (35) (36> (37) (38) S.R.B. PAGE 41) 47a. You said Chat you (last smoked cigarettes/are currently smoking cigarettes) on a fairly regular basis (in DATE). On Yirj many days did you smoke cigarettes during the last three months (that you smoked on a fairly regular basis)? I I T I I I days (39) -(40) 47b. On the days that you smoked, about how many packs did you smoke per day? T J \ \ j \_ packs per day (41) ( 2 4) 47c. In general, did you inhale the smoke? Yes...(43(. No -1 -2 �. 48a. CARD 045 Have you ever smoked a pipe regularly for a period of at least one month? -2 No, A8b. 812039 (SKIP TO 0- 51a) In what month ami year did you start smoking a pipe on a fairly regular basis? MONTH YEAR 1 1 T T 1 1 1 1 1 1 (47) 48c. In what month and year did you last smoke a pipe on a fairly regular basis? MONTH YEAR 1 I I T l I 149) (50) I-1 I I T I (51) (52) 49a. When you started smoking a pipe on a fairly regular basis in (START DATE), about how many pipefuls per week did you smoke? F I I | I I pipefuls per week (53) (54) 49b. Until what month and year did you continue to smoke (NUMBER) pipefuls per week on a regular basis? MONTH YEAR 1 T^ T1 H T (IF DATE IS THE SAME AS Q.ASc: SKIP TO Q.SOa. J[ | |-j__ J I ALL OTHERS: CONTINUE) (55) (56) (57) (58) 49c. After that, about how many pipefuls per week did you smoke? iI r \ | I pipefuls per week "759 ) (60) A9d. Until what month and year did you continue to smoke (NUMBER) pipefuls per week on e regular basis? MONTH YEAR .1 [ T T1 I T (IF DATE IS THE SAME AS Q.ASc: SKI-P TO Q.SOa. j J I-| I j ALL OTHERS: CONTINUE) (61) (62) (63) (64) A9e. After that, about how many pipefuls per week did you smoke? I I T I (65) T [ pipefuls per week (66) 49f. Until what month and year did you continue to smoke (NUMBER) pipefule per week on a regular basis? MONTH YEAR 1 1 T1 1 T ( I F DATE I S T H E SAME A S Q.ASc: A S K Q.SOa. J L H I... I ALL OTHERS: RECORD ADDITIONAL PERIODS IN (67) (68) (69) (70) S.R.B. PAGE Al) 50a. You said that you (last smoked a pipe/are currently smoking a pipe) on a fairly regular basis (in DATE). On how many days did you smoke a pipe during the last three months (that you smoked on a fairly regular basis)? I I . I I I I days (71) (72) 50b. On the days that you smoked, about how many pipefuls did you smoke per day? T I 50c. I I I I pipefuls per (73) "(74) day In general, did you inhale the smoke? Yes... (/'-,( No -1 V T " -2 . �CARD 051 51a. Have you ever smoked cigars regularly for a period of at least one month? : Yes .(12( -1 (ASK Q.51b) , -2 (SKIP TO Q.54a) No Sib. In what month and year did you start smoking cigars on a fairly regular basis? MONTH YEAR "(13) (14) 51c. 812039 (15) (16) In what month and year did you last smoke cigars on a fairly regular basis? MONTH YEAR (17) (18) (19)(20> 52a. When you started smoking cigars on a fairly regular basis in (START DATE), about how many cigars per week did you smoke? iI I r I I cigars p e r week (21 ) (22) 52b, Until what month and year did you continue to smoke (NUMBER) cigars per week on a regular basis? MONTH YEAR I I TT I T (IF DATE IS THE SAME AS Q.Slc: SKIP TO Q.53a. I I . I-I I I A L L OTHERS: CONTINUE) (23) (24) (25) (26) 52c. After that, about how many cigars per week did you smoke? J j I cigars per week (27) (28) S2d. Until what month and year did you continue to smoke (NUMBER) cigars per week on a regular basis? MONTH YEAR T - | T1 I T ( I F DATE I S T H E SAME A S Q.Slc: SKIP T O Q.53a. I I I-J I I ALL OTHERS: CONTINUE) (29) (30) (31) (32) S2e. After that, about how many cigars per week did you smoke? I j I cigars per week (33) (34) 52f. Until what month and year did you continue to smoke (NUMBER) cigars per week on a regular basis? MONTH YEAR T IT1 1 T (IF DATE IS THE SAME AS Q.Slc: ASK Q.53a. J^ I l-j [_ j ALL OTHERS: RECORD ADDITIONAL PERIODS IN (35) (36) (37) (38) S.R.B. PAGE 42) 53a, You said that you (last smoked cigars/are currently smoking cigars) on a fairly regular basis (in DATE). On how many days did you smoke cigars during the last three months (that you smoked on a fairly rp'_,ular basis)? I 1 53b. I T \. I days (39) ( 0 4) On the days that you smoked, about how many cigars did you smoke per day? 1 I I I I I cigars per day (41) (42) 53c. In general, did you inhale the smoke? Ves...(43( No _-l -2 .12 �CARD 051 812039 54a. Now let's talk about drinking alcoholic beverages, that is, beer, wine, or hard liquor. Did you ever drink alcoholic beverages on a fairly regular basis? Yes (44 ( (ASK Q.54b) -2 No -1 (SKIP TO Q.57 ) 54b. In what month and year did you start drinking alcoholic beverages on a fairly regular basis? MONTH YEAR T I I I T l-l "(45) (46) 54c. T I T l - l (47) (48) In what month and year did you last drink on a fairly regular basis? MONTH YEAR I l-l I (49) (50) (51) (52) 55a. When you started drinking alcoholic beverages on a fairly regular basii in (START DATE), about how many drinks per week did you have? T I drinks per week (54) (53) 55b. Until what month and year did you continue to drink (NUMBER) drinks per week on a regular basis? MONTH YEAR T I T T 1 T ( I F DATE I S T H E SAME A S Q.SAc: SKIP T O Q.56a. I. I l-l I I ALL OTHERS: CONTINUE) (55) (56) (57) (58) S5c. After that, about how many drinks per week did you have? 1 J I I (59) I | drinks per week (60) 55d. • Until what month and year did you continue to drink (NUMBER) drinks per week on a regular basis? MONTH YEAR I T I I I ( I F DATE I S T H E SAME A S Q.54c: SKIP T O Q.56a. I ALL OTHERS: CONTINUE) (61) (62) (63) (64) 55e. After that, about how many drinks per week did you have? I J T I j [ drinks per week (65) ( 6 6) S5f. Until what month and year did you continue to drink (NUMBER) drinks per week on a regular basis? MONTH YEAR T I TT ~\ T (IF DATE IS THE SAME AS 0.54c: ASK Q.56a. I I ,1-1 I I ALL OTHERS: RECORD ADDlfldNAL PE'SIODS IN T67) (68) (69) (70) s.R.B. PAGE 43) 56a. You said that you (last drank/are currently drinking) alcoholic beverages 6H a fairly regular basis (in (END DATE)). On how many days did yo'u drink during the last months (that you drank on a fairly regular basis)? 1 566. I J_ days (71) (72) On the days that you drank, about how many drinks did you have per day? J I |_ drinks per ( 73) T74) day 53c. During these months which one of the following beverdges did you drink most -hard liquor, beer or ale, or wine or champagne? Hard liquor.. ( 5 7j Beer or ale Wine or champagne Combination 43 _-l -2 L-3 -4 �CARD 057 57. Have you ever tried smoking marihuana? Yes (12 ( Vb.'.....'..'... 57a. -2 (i3(. (SKIP TO Q.60) . -1 (ASK Q,57b) -2 No (SKIP TO Q.60) In what month and year did you start smoking marihuana on a fairly regular basis? MONTH YEAR 1 I T T 1 I l-l (14) (15) 57c. -1 (ASK Q.57a) Have you ever smoked marihuana regularly for a period of at least one month? Yes 57b. 812039 T ~T I I (16) (17) In what month and year did you last smoke marihuana on a fairly regular basis? MONTH YEAR T i T l I T <18> (19) (20) (21) 58a. When you started smoking marihuana on a fairly regular basis in (START DATE), about how many joints per week did you smoke? T I I | T I joints per week "(22) (23) 58b. Until what month and year did you continue to smoke (NUMBER) joints per week on a regular basis? MONTH YEAR T 1 T1 I T ( I F DATE IS T H E SAME A S Q.57c; SKIP T O Q.59a. I I hi I I ALL OTHERS: CONTINUE) (24) (25)(26) (27) 58c. After that, about how many joints per week did you smoke? I I T I I I joints per week (28) (29) 58d. Until what month and year did you continue to smoke (NUMBER) joints per week on a regular basis? MONTH YEAR 1 1 T1 T T (IF DATE IS THE SAME AS Q.57c: SKIP TO Q.59a. I I j-l | I ALL OTHERS: CONTINUE) "(30)01) (32) (33) 58e. After that, about how many joints per week did you smoke? I I r \ I \_ joints per week T34) (35) 58f. Until what month and year did you continue to smoke (NUMBER) joints per week on a regular basis? MONTH YEAR T ~] T1 T T (IF DATE IS THE SAME AS Q.S7c: ASK Q.59a. 1 I l-l I I ALL OTHERS: RECORD ADDITIONAL PERIODS IN (36) (37) (38) (39) S.R.B. PAGE 43) 59a, You said that you (last smoked marihuana/are currently smoking marihuana) on a fairly regular basis (in DATE). On how many days did you smoke marihuana during the last three months (that you smoked on a fairly regular basis)? T days (40) (41) 59b. On the days that you smoked, about how many joints did you smoke per day? I I I I I I joints per ~(42) (43)~ day 44 �CARD 057 812039 60. In your lifetime, have you ever had two weeks or more during which you felt sad, blue, depressed, or when you lost all interest and pleasure in things that you usually cared about or enjoyed? Yes No UA( ^-iV (ALSO RECORD ON S.R.B. PAGE 12) -2J 61a. Have you had two years or more in your life when you felt depressed or sad almost all the time even if you felt O.K. sometimes? Yes (45(__ (ASK Q.61b) -2 No -1 (SKIP TO Q.62) 61b. Did you tell a medical doctor about feeling depressed during this period? The term "medical doctor" includes psychiatrists, osteopaths, and medical students. Yes (6 4( (SKIP TO Q.62) -2 No -1 (ASK Q.61e) 61c. Did you tell any other professional about feeling depressed during this period? The term "other professional" includes psychologists, counselors, members of the clergy, and chiropractors? Yes (7 4( (SKIP TO Q.62) -2 No -1 (ASK Q.61d) 61d. Did you take medication more than once, either prescribed or nonprescribed, for feeling depressed during this period? Yes (8 4( (SKIP TO Q.62) -2 No -1 (ASK Q.61e) 61e. Did being depressed during this period interfere with your life and activities a lot? Yes No (4<j( 45 -1 -2 �812039 CARD 057 62a. Has there ever been a period of two weeks or longer when you lost your appetite? Yes (Q j( 62b. (ASK Q.62b) -2 No -1 (SKIP TO Q.63 a) Did you tell a doctor about your loss of appetite? Yes..'....($!_( No.. 62c. ,-* -2 (ASK Q-62cy (SKIP TO Q.62d) When, you told the doctor, what was his diagnosis? IF "NERVES, STRESS, ANXIETY": CIRCLE "5" BELOW AND SKIP TO Q.feyT IF "NOTHING DEFINiTE"/"DON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.63a. IF "PHYSICAL ILLNESS OR INJURY," SKIP TO Q.62e. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP TO Q.62e. 62d. What was the cause of your loss of appetite? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "5" BELOW AND SKIP TO Q.63a. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.62e. 62e. Has there ever been a period when you lost your appetite for two weeks or longer for any reason other than (READ'RESPONSE FROM Q.62c OR Q.62d)? Yes. -1 (ASK Q.62f) No 62f. .5( (2 ,...._ -2 (SKIP TO What was the cause of your loss of appetite in that period? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE'^OR "DON'T KNOW": CIRCLE "5" BELOW AND GO TO Q.63 a. ALL OTHERS: GO TO Q.63 a. FOR OFFICE USE ONLY 1 2 3 A (53) . 5 IF "5" CIRCLED RECORD IN S.R.B. �CARD 057 63a. Have you ever lost weight without trying to — several weeks (or as much as 10 pounds altogether)? as much as two pounds a week for Yes -1 (ASK Q.63b) No 63b. -2 (SKIP TO Q.64a) Did you tell a doctor about your weight loss? Yes (5 5( -1 (ASK Q.63c) -2 No 63c. (II ? ) ' (!) (SKIP TO Q.63d) When you told the doctor, what was his diagnosis? IF "NERVES, STRESS, ANXIETY"": CIRCLE "5" BELOW AND SKIP TO Q.64. IF "NOTHING DEFINITE"/"DON!T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.64a. IF "PHYSICAL ILLNESS OR INJURY," SKIP TO Q.63e. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP TO Q.63e. 63d. What was the cause of your weight loss? IF "NERVES, STRESS, ANXIETY11 OR ^'NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "5" BELOW AND SKIP TO Q.64a. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.63e. 63e. Has there evfer been a period when you lost weight without trying to — as much as two pounds a week for several weeks (or as much as 10 pounds altogether) for any reason other than (READ RESPONSE FROM Q.63c OR Q.63d)? Yes. .5( (6 -2 No.. 63f. _-l (ASK Q.63f) (SKIP TO Q.64a) What was the cause of your loss of weight in that period? IF ''NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "5" BELOW AND GO TO Q.64a. ALL OTHERS: GO TO Q.64a. FOR OFFICE USE ONLY 1 2 3 A (57) . 47 5 IF "V CIRCLED RECORD IN S.R.B. �CARD 057 812039 6Aa. Have you ever had a period when your eating increased so much Chat you gained as much as two pounds a week for several weeks (or 10 pounds altogether)? Yes. .5( (8 _-l -2 No.. 64b. (SKIP TO Q.65a) Pid you tell a doctor about your increased appetite and weight gain? Yes <5gL -1 (ASK Q.64c) No 64c. (ASK Q.64b) -2 (SKIP TO Q.64d) When you told the doctor, what was his diagnosis? IF "NERVES, STRESS, ANXIETY": CIRCLE "5" BELOW AND SKIP TO Q.65. IF "NOTHING DEFINITE'V'DON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL -ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.65a. IF "PHYSICAL ILLNESS OR INJURY," SKIP TO q.64e. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP TO Q.64e. 64d. What was the cause of your increased appetite and weight gain? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE'^OR n'DO(rT KNOW": CIRCLE "5" BELOW AND SKIP TP Q.65a. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.64e. 64e. Has there ever been a period when your eating increased so much that you gained as much as two pounds a week for several weeks (or 10 pounds altogether) for any reason other than (READ RESPONSE FROM Q.64c OR 64d)? Yes. No., -1 (ASK Q.64f) -2 (SKIP TO Q.65fl) .6( (0 64f. What was the cause of your increased appetite and weight gain in that period? IF "NERVES, STRESS, ANXIETY" irOR "NOTHING DEFINITE'^W "DON'T KNOW": CIRCLE "5 "BELOW AND GO TO p.65a. ALL OTHERS: CO TO Q.65 a. FOR OFFICE USE ONLY 1 2 3 4 (61) . 5 IF "5" CIRCLED RECORD IN S.R.B. DA^F 1 1 �812039 CARD 057 65a. Have you ever had a period of two weeks or yore when you had trouble falling asleep, staying asleep, or with waking up too early? Yes. No 6M». (ASK Q.65b) .6( (2 _ -2 (SKIP TO Q.66a) Did you tell a doctor about your trouble sleeping? Ves ...... ( 3 6( (ASK Q.65c) No..........._ 65c. -1 -2 (SKIP TO Q.65d) When you told the doctor, what was his diagnosis? IF "NERVES, STRESS, ANXIETY": CIRCLE "5" BELOW AND SKIP TO Q.66. IF "NOTHING DEFINITE'V'DON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.66a. IF "PHYSICAL ILLNESS OR INJURY," SKIP TO Q.65e. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP TO Q.65e. 65d. What was the cause of your sleeping problem? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "5" BELOW AND SKIP TO Q.66a. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.65e. 65e. Has there ever been a period of two weeks or more when you had trouble falling asleep, staying asleep, or with waking up too early for any reason other than (READ RESPONSE FROM Q.6Sc OR Q.65d)? Yes. (ASK Q.65f) No.. 65f. -1 -2 (SKIP TO Q.66a) What was the cause of your 'Sleeping problem in that period? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "S^ELOW AND GO TO Q.66e. ALL OTHERS: GO TO Q.66a. FOR OFFICE USE ONLY 1 2 3 4 (65) . 5 IF "5" CIRCLED RECORD IN S.R.B. Dirr i o �812039 CARP 057 66a. Have you ever had a period of two weeks or longer when you were sleeping too .much? Yes. (6 6( No 66b. ...: -2 (ASK Q.66b) (SKIP TO Q.67 a) Did you tell a doctor about your sleeping too much? Yes No 66c, "I (67( ,. -1 _-2 (ASK Q.66c) (SKIP TO Q.66d) When you told the doctor, what was his diagnosis? It "NERVES, STRESS. ANXIETY": CIRCLE "5"'BEtOW AHt) SKIP T6 6..fe7.a. IF "NOTHING DEFINITE"/"DON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.67 a. IF "PHYSICAL ILLNESS OR INJURY," SKIP TO Q.66e. IF "MEDICATIQN, DRUGS, OR ALCOH6L," SKIP TO Q.66e. 66d. What was the cause of your sleeping too much? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE11 OR "DON'T KNOW": CIRCLE "5" BELOW AND SKIP TO Q.67 a. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.66e. •* 66e. Has there ever been a period when you were sleeping too much for two weeks or longer for any reason other than (READ RESPONSE FROM Q.66c OR Q.66d)7 Yes. No.. 66f. .6( (8 _-i (ASK Q.66fj -2 (SKIP TO Q.6?a) What was the cause of your sleeping too much in that period? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "S^BELOW AND GO TO Q.67 a. ALL OTHERS: GO TO Q.67 a. FOR OFFICE USE ONLY 1 2 3 4 (69) 5 IF "5" CIRCLED RECORP IN S.R.B. rAQfe } / �812039 CARD 057 67a. Have you ever had a period lasting two weeks or more when you felt tired all the time? Yes (70( -1 -2 No 67b. (SKIP TO Q.68a> Did you tell a doctor about your feeling tired out all the tine? Yes (?1( No 67c. (ASK Q.67b) -1 ,.,,..ro.3i-2 j (ASK Q.67c) (SKIP TO Q.67d) When you told the doctor, what was his diagnosis? IP" "NERVES, STRESS, ANXIETYtr: CIRCLE "5" BELOW AND SKIP TO Q.689IF "HOTHING DEFINITE"/"DON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.68aIF "PHYSICAL ILLNESS OR INJURY," SKIP TO Q.67e. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP TO Q.67e. 67d. What was the cause of your feeling tired out all the time? IF "NERVES, STRESS, ANXIETY" OR NOTHING DEFINITE" OR^'DON'T KNOW": CIRCLE "5" BELOW AND SKIP TO Q.68a. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.66e. 67e. Has there ever been a period when you felt tired out all the time for two weeks or longer for any reason other than (READ-RESPONSE FROM Q.67c OR Q.67d)? .(72( Yes. No.. 67f. (ASK Q.67f) -2 , -1 (SKIP TO Q.68a) What was the cause of your feeling tired out in that period? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "5" BELOW AND GO TO Q.&8a. ALL OTHERS: GO TO Q.68 a. FOR OFFICE USE ONLY 1 2 3 A (73) . 51 5 IF "5" CIRCLED RECORD IN S.R.B. PARF 1? �812039 -CARP 057 689- Has there ever Veen a period of two weeks or more when you talked or moved more Blowly than is normal for you? ""*" ' ' '. " "'' • Ye6,M.-.(Zi< "l (ASK Q.68b) No..........._ -2 (SKIP TO Q.69 a) 68b. Did you tell a doctor about your slowed speech or movement? Yes......(7$( -1 USK Q.68c) N...... o....._ -2 (SKIP TO Q.68d) 68c. When you told the doctor, what was his diagnosis? IF "NERVES, STRESS, ANXIETY": CIRCLE "5" BELOW AND SKIP TO Q.699. IF "NOTHING DEFlNiTE"/"pQNlT KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "V BELOW AND SKIP TO Q.69 a. IF "PHYSICAL lUNESS OR INJURY," SKIP TO Q.68e. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP" TO Q.68e. 68d. What was the cause of your slowed speech or movement? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE^ OR "DON'T KNOW": CIRCLE "5" BELOW AND SKIP TO Q.69 a. IF "PHYSICAL ILLNESS OR INJURY" £R "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.68e. 68e., Has there ever been a period when you talked or moved more slowly than is normal for you for two weeks or longer for any reason other than (READ RESPONSE FROM Q.68c OR Q.68d)T ' Yes. .7( (6 -V (ASK p.68f) -2 (SKIP TO Q.698) No.. 68f. What was the cause of your slowed speech or movement in that period? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW"; CIRCLE "S^BELOW AND GO TO Q,69a. ALL OTHERS: CO TO Q.69a. FOR OFFICE USE ONLY 1 2 3 4 (77) 52 S IF "5" CIRCLBP RECORD IN S.R.B. �812039 CARD 069 69a. Hap there ever been a period of two weeks or more when you had to be moving all the time — that is, you couldn't stand still and paced up and down? Ye No 69b. _-l -2 (SKIP TO Q.7CO Did you tell a doctor about your need to be moving all the time? Yes No 69c. (ASK Q.69b) (13 ( -1 (ASK Q.69c) -2 (SKIP TO Q.69d) When you told the doctor, what was hi* diagnosis? IF "NERVES, STRESS, ANXIETY": CIRCLE *5" BELOW AND SKIP TO Q.70S. IF "NOTHING DEFINITE"/"DON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.7CaIF "PHYSICAL ILLNESS OR INJURY," SKIP TO Q.69e. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP TO Q.69e. 69d. What was the cause of your moving all the time? IF "NERVES, STRESS, ANXIETY11 OR "NOTHING DEFINITETrOR "DON'T KNOW1': CIRCLE "5" BELOW AND SKIP TO Q.70a. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.69e. 69e. Has there evej: been a period for two weeks or longer when you had to be moving all the time — couldnT stand still and paced up and down for any reason other than (READ RESPONSE FROM Q.69c OR Q.69d)? Yes. (ASK Q.69f) No.. 69f. -1 -1 (SKIP TO Q.70a> What was the cause of your moving all the time in that period? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "57r~BELOW AND GO TO Q.70a. ALL OTHERS: GO TO Q.70a FOR OFFICE USE ONLY 1 2 3 4 (15.) . 5 IF "5" CIRCLED RECORD IN S.R.B. �CARD Q69 81203_9_ 70a. Was there ever a period of several weeks when your interest in sex was a lot than usual? . Yes (16(r -1 (ASK Q.70b) No -2 (SKIP TO Q.71a) 7Ob. Did you tell a doctor about your diminished interest in sex? Yes (17( No -1 (ASK Q.70c) ^-2 (SKIP TO Q.70d) 70c. When you told the doctor, what was his diagnosis? IF "NERVES, STRESS, ANXIETY": CIRCLE "5" BELOW AND SKIP TO Q.71a. IF "NOTHING DEFIN1TE"/"DON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.71»IF "PHYSICAL ILLNESS OR INJURY," SKIP TO 9.70h. IF "MEDICATIQN .DRUGS, ORfltCOHOL'," SKIP TO Q.70h. 70d. Did you consult with any other professional, such as a psychologist, marriage counselor, minister, or nurse abput your diminished interest in sex? Yes....,.(jfl( -1 (SKIP TO Q.70g) No.,.,, -2 ._ (ASK Q.70e) 70e. Did you take medication more than once for your diminished interest in sex? Yes U9( No...... -1 (SKIP TO Q.70g) -2 (ASK Q.70f) 70f. Did your diminished interest in sex interfere with your life or activities a lot? Yeg......(20( , (ASK Q.70g) No 70g. -1 -2 (SKIP TO Q.71 a) What was the cause of your diminished interest in sex? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "5" BELOW AND SKIP TO Q.71 «•• IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.70h. "*" 70h. Has tl ere gver been a period when your interest in sex was diminished for two weeks or longer for any reason other than (READ RESPONSE FROM Q.70c OR Q.70d)? Ye (21t( (ASK Q.70i) -2 No.... -1 (SKIP TO Q.7U) 70i. What was the cause of your diminished interest in sex in that period? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "S^BELOW AND GO TO Q.71a. ALL OTHERS: GO TO Q.7U. FOR OFFICE USE ONLY 1 2 3 4 (22) 54 5 IF "5" CIRCLED RECORD I N S . R . 8 . �CARD 069 812039 71a. Has there ever been a period of two weeks or more when you had a lot more trouble concentrating than is normal for you? Yes. No.. 71b. (ASK Q.71b) (SKIP TO Q.72 a) Did you tell a doctor about your trouble concentrating? Yes (4 2( -1 (ASK Q.71c) -2 No /!<:. _-l _-2 .(23( (SKIP TO Q.71d) Wlmn you told Che doctor, what wag hie diagnosis? IF "NERVES. STRESS, ANXIETY": CIRCLE "5" BELOW AND SKIP TO Q.72a. IF "NOTHING DEFINITE"/"DON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.72a. IF "PHYSICAL ILLNESS OR INJURY," SKIP TO Q.71e. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP TO Q.71e. 71d. What was the cause of your trouble concentrating? IF "NERVES, STRESS, ANXIETY" OR "NOTHiNG'DEFINITE11 OR "DON'T KNOW1': CIRCLE "5" BELOW AND SKIP TO Q.72 a. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.71e. 71e. Has there ever been a period when you had more trouble concentrating than is normal for you for two weeks or longer for any reason other than (READ RESPONSE FROM Q.71c OR Q.71d)? Yes. .2( (5 71f. (ASK Q.71f) -2 No.. -1 (SKIP TO Q.72a) What was the cause of your trouble concentrating in that period? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "5" BELOW AND GO TO Q.72a. ALL OTHERS: GO TO Q.72a. FOR OFFICE USE ONLY 1 2 3 A (26) S5 5 IF "5" CIRCLED RECORD IN S.R.B. �CARD 069 812039 72a. Has there ever been a period of two weeks or more when your thoughts came much glower than usual or seemed mixed up? Yes. .(27< -1 (ASK Q.72b) -2 (SKIP TO Q.73) No.. 72b. Did you tell a doctor about your thoughts coming much slower than usual or seeming mixed up? Yes. .(2g( No -1 (ASK Q.72c) _-2 (SKIP TO Q.72d) Tic.. When you told the doctor, what was his diagnosis? IF "NERVES, STRESS, ANXIETY":CIRCLE "5" BELOW AND SKIP TO Q.73. IF "NOTHING DEFINITE"/"DON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS, IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.73. IF "PHYSICAL ILLNESS OR INJURY," SKIP TO Q.72e. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP TO Q.72e. 72d. What was the cause of your your thoughts coming much slower than usual or seeming mixed up? • IF "NERVES, STRESS, ANXIETY" QR "NOTHING DEFINITE" OR "DON'T KNOW": CIRCLE "5" BELOW AND SKIP TO Q.73. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS, OR ALCOHOL": ASK Q.72e. 72e. Has there, ever been a period when your thoughts came much slower than usual or seemed mixed up for two weeks or longer for any reason other than (READ RESPONSE FROM Q.72c OR Q.72d)? Yes. .2( (9 No. -1 (ASK Q.72f) -2 (SKIP TO Q.73) 72f. What was the cause of your your thoughts coming much slower than usual or seeming mixed up in that period? IP "NERVES, STRESS. AHXIETV" OR "NOTHING DEFINITE" OK "DON'T KNOW": CIRCLE "5" BELOW AND GO TO Q.73. ALL OTHERS: GO TO Q.73. FOR OFFICE USE ONLY 1 2 3 4 (30) 56 . 5 IF "5" CIRCLED RECORD IN S.R.B. Dir.f 1 1 �CARD 069 . 812039 "'YES IN p. 73 THROUGH o.77 IN"S.R".¥". PAGE 12 1 73. Has there ever been a period of two weeks or more when you felt worthless, sinful, or guilty? Yes ...... (31( No..........._ -1 -2 Ik. Has there ever been a period of two weeks or more when you thought a lot about death — either your own, someone else's, or death in general? . No...........__ -2 75. Has there ever been a period of two weeks or more when you felt like you wanted to die? Yes ...... (33( No..........._ -1 -2 76. Have you ever felt so low you thought of committing suicide? Yes ...... ( 4 3 ( ._.-! No..........._ -2 77. Have you ever attempted suicide? Yes ...... (35(________-1 No ........... -2 INTERVIEWER INSTRUCTIONS: IF LESS THAN THREE BOXES CHECKED IN S.R.B. PAGE 12 FOR Q.62-77: SKIP TO Q.88a. IF THREE OR MORE BOXES CHECKED IN S.R.B. PAGE 12 AND "YES" IN Q.60, ASK Q.78a. IP THREE OR.MORE. BOXES CHECKED IN S.R.B. PAGE 12 AND "NO" IN Q.60, SKIP TO Q.79a. �. ^CARD.069 812039 78a. You said you've had a period of feeling (depressed or blue/OWN EQUIVALENT) and also said you've had some other problems like (LIST ALL ITEMS CHECKED IN S.R.B. PAGE 12). Has there ever been a tine when the feelings of depression and some of these other problems occurred together — that is, within the same month? Yes (6 3( (SKIP TO Q.80) -2 No -1 (ASK Q.78b) 78b, So there's never been a period when you felt sad, blue, or depressed at the same time you were having some of these other problems? Has been a period..(3_7j -1 (ASK Q.80) Never been a period -2 (SKIP TO Q.88a, PAGE 57) 79a. You said you have had periods when (LIST ALL ITEMS CHECKED IN S.R.6. PAGE 12). Was there ever a time when several of these problems occurred together — t h a t is, within the same month? Yes (g 3( (ASK Q.79b) -2 No -1 (SKIP TO Q.88a, PAGE 57) 79b. When you were having some of these problems at about the same time, were you feeling okay, or were you feeling low, gloomy, blue, or uninterested in everything? Gloomy, low, etc..(39( Okay -1 -2 (ASK Q.80) (SKIP TO Q.88a, PAGE 57) 80. What's the longest spell you've ever had when you felt blue and had several of these other problems at the same time — that is, how many weeks did it last? IF "WHOLE LIFE" OR MORE THAN 19 YEARS, ENTER "996" AND CONTINUE. IF LESS THAN 2 WEEKS, CODE "001" AND SKIP TO Q.88a, PAGE 57. iI . l1 1l .... rI ( 0 (41) 4) weeks (42) 81. Now, I'd like to ask about spells when you felt both (depressed/OWN EQUIVALENT) and had some of these other problems like (LIST 3 ITEMS CHECKED ON S.R.B. PAGE 12). In your lifetime, how many spells like that have you had that lasted two weeks or more? JIF MORE THAN 90 SPELLS, ENTER "907*7 I I I ( 44) spells 58 �CARD 069 82a. 812039 Did you tell a doctor about (that spell/any of those spells)? Yes (j 4( 82b. (SKIP TO Q.83) -2 No -1 (ASK Q.82b) Did you tell any other professional about (it/any of them)? Yes (6 4( (SKIP TO Q.83) -2 No -1 (ASK Q.82c) 82c. Did you take medicine more than once becaune of (that »p*ll/»ny of those «pellt)7 Yes U?( (SKIP TO Q.83) -2 No -I (ASK Q.82d) 82d. Did (that spell/those spells) interfere with your life or activities a lot? Ves...:...<48< No -1 -2 (ASK Q.83) 83. How old were you the first time you had a spell for two weeks or more where you felt sad and had some of these other problems such as (PROBLEMS CHECKED IN S.R.B. PAGE 12)? AGE: I I l I l I (49) (50) years 8Aa. Did (this spell/any of those spells) occur just after someone close to you died? Yes ' . (5_l( No -1 (ASK Q.SAb) ' -2 (SKIP TO Q.85) 84b. Have you had any spell of depression along with these other problems such as (PROBLEMS CHECKED IN S.R.B. PAGE 12) at times when it wasn't due to a death? No, only due to death (52( Yes, other times not due to death... -1 -1 85. Are you now in one of these spells of feeling low or disinterested and having some of these other problems? Yes (53( 85b. (SKIP TO Q.85d) -2 No -1 (ASK Q.85b) When did your last spell like that end? •^ Within last Between two Between one Between six two weeks (54( ,4( weeks and one month ago... month and six months ago.. months and one year ago... More than one year ago 85c. How old were you then? AGE: I I I years C>.r>) (56) 59 -1 -2 •(SKIP TO Q.BSd) -3 -I* -5 (ASK Q.85c) �CARD 069 812039 INTERVIEWER: DO NOT READ Q.85d AND Q.85e TO RESPONDENT. 85d. IS MORE THAN ONE SPELL CODED IN Q.81? Ys...(7 e....5( No 85e. -1 -2 (SKIP TO Q.86) (ANSWER Q.85e) ARE 52 OR MORE WEEKS CODED IN Q.80? Yes. No.. .51 (8 -1 ~-2 (ASK Q.86) (SKIP TO Q.87) 86. Now I'd like to know, about the time when you were feeling depressed for at least two weeks and had the largest'number of these other problems at the same time. (IF CAN'T CHOOSE: Then pick one bad spell.) How old were you at that time? I AGE: years (59) (0 6) 60 (ASK Q.87) �CARD 069 812039 fl flEin"X BELOW ALLlPROBLfiMS RECORDED ON S.R.B. PAGE 12, AND ASK Q.87 FOR THOSE CONDITIONS. 87. During this spell of depression when you were (AGE IN Q.86) years old ... IASK ONLY FOR CHECKED CONDITIONS I 1 .. j t>id you lose your appetite?.............................(61(_ -1 _ Did you lose weight without trying to — as much as ten pounds altogether?......................................(62< -2 -1 _ -2 II 111 Ytiiii "sling irirt-oKiso an rolKli that you B«ined ten pounds altogether'/.. .................................... (6ft( «t _ __ __-2 Did you have trouble falling asleep, staying asleep, or waking up too early?................................ . 6 (_ (4 -1 _ -2 V I Were you sleeping too much?.............................(6ft(_ -1 _ -2 ] ' I Did you feel tired out all the time?................. . . 6 ( .(6 -1 _ -2 j ] Did you talk or move more slowly than is usual for you? . ( 67 (_ j _ | Did you have to be moving all the time — that is, you couldn't sit still and paced up and down?...............( 8 _ 6( | | Was your interest in sex a lot less than usual?.........( 9 _ 6( I I Did you have a lot more trouble concentrating than ia usual for you?..........................................( 0 _ 7( -1 _ -2 I I Did your thoughts come much slower than usual or seem mixed up?..........................„......................( j _ 7( -1 _ -2 ] | Did you feel worthless, sinful, or guilty?..............( 2 7( -1 _ -2 I I Did you think a lot about death — either your own, someone else's, or death in general?....................( 3 7( y°u fee * like you wanted to die? .................... ( 4 7( T ~"[ '' Did you feel so low you thought of committing suicide?. . (_7jj( I Did you attempt suicide? I (7A( -1 _ -1 _ -2 -1 _ -1 _ -1 -1 -1 -2 -2 -2 -1 -2 -2 �CARD 088 88a. Have you ever considered yourself a nervous person? Yes (12 ( Ho 88b. _-2 (ASK Q.88b) (SKIP TO Q.89a) At what age did this nervousness begin? AGE: 1 I Whole life.. \ I I | years (13) (14) (15 ( (SKIP TO Q.89a) -1 (SKIP TO Q.89a) -2 Not sure 88c. -1 (ASK Q.88c) Do you think it began before or after you were 30? Before 3 . . ( 6 0..1J After 30 -1 -2 Still not sure... -3 812039 �CARD 088 812039 8$a. Have you ever had a spell or attack when all of a sudden you felt frightened, anxious, or very uneasy in situations when most people would not be frightened? Yes..( 17( -1 (ASK Q.89b) No -2 (SKIP TO SELF-ADMINISTERED SHEET AFTER Q.94b, PAGE 60) 89b. Did you tell a doctor about your feeling frightened, anxious, or uneasy? Yes de( No 89c. -1 (ASK Q.89c) -2 (SKIP TO Q.89d) When you told the doctor, what was his diagnosis? IF "NERVES, STRESS, ANXIETY": CIRCLE "511 BELOW AND SKIP TO Q.90. IF "NOTHING DEFINITE'VDON'T KNOW": ASK IF DOCTOR'S EXAMINATION OR TESTS INDICATED ANY PHYSICAL ILLNESS. IF "NO": CIRCLE "5" BELOW AND SKIP TO Q.90. IF "PHYSICAL ILLNESS OR INJURY," SKIP TO Q.89H. IF "MEDICATION, DRUGS, OR ALCOHOL," SKIP TO Q,89h. 89d. Did you consult with any other professional, such as a psychologist, marriage counselor, minister, or nurse about your feeling frightened, anxious, or Uneasy? Yes. _-l No.. -2 (SKIP TO Q.89g) (ASK Q.89e) 89e. Did you take medication more than once for your fear, anxiety, or uneasiness? Yes. .(20C j-1 -2 No.. (SKIP TO Q.89g) (ASK Q.89f) 89f. Did your fear, anxiety, or uneasiness interfere with your life or activities a lot? Yes.(21( No 89g. -1 (ASK Q.89g) -1 (SKIP TO SELF-ADMINISTERED SHEET AFTER Q.94b, PAGE 60) What Was the cause of your fear, anxiety, or uneasiness? IF "NERVES, STRESS, ANXIETY" OR "NOTHING DEFINITE" OR CIRCLE "5" BELOW AND SKIP TO Q.90. IF "PHYSICAL ILLNESS OR INJURY" OR "MEDICATION, DRUGS ALCOHOL": ASK Q.89h. , i - . .. . - - . i i—i ..- ''DON'T KNOW": , ob ' .....if. ..»•.»•• 89h. Have you ever had a period of fear, anxiety, or uneasiness for two weeks or longer for any reason other than (READ RESPONSE FROM Q.89c OR 89g)? Yes ( 2% 89i. (ASK Q.89i) -2 No -1 (SKIP TO Q.90) What was the cause of your fear, anxiety, or uneasiness in that period? IF "NERVEFrSTlES^AMiT;TTr1r'W~ItNOTrifiG"' DEFINITE""OK "DON'T KNOW": CIRCLE "5" BELOW AND GO TO Q.90. ALL OTHERS: GO TO Q.90. FOR OFFICE USE ONLY 1 2 3 4 63 �CARD 088 812039 90. During one of the worst spells of suddenly feeling frightened or anxious or uneasy, did you ever notice that you had any of the following. During this spell... Yes No A. Were you short of breath — having trouble catching your breath?.(24( ~1 ~2 B. Did your heart pound? ( 25( -1 -2 C. Were you dizzy or light-headed? (26( -1 -2 D. Did your fingers or feet tingle? ( 27( -1 -2 -1 -2 E. Did you have tightness or pain in your chest? . . . ?R( ..( F. Did you feel like you were choking or smothering? (29( -1 -2 G. Did you feel faint? (JQ( -1 -2 H. Did you sweat?.. (31( -1 -2 1. Did you tremble or shake? (32( -1 -2 J. (33( -1 -2 (" •( -1 -2 -1 -2 Did you feel hot or cold flashes? K. Did things around you seem unreal? L. Were you afraid either that you might die or that you might act in a crazy way? . (35( 91a. How old were you the first time you had one of these sudden spells of feeling frightened or anxious? I AGE: I T I I I years (36) (37) Whole life....!.... (T&(_ Not sure..............._ 91b. (SKIP TO Q.92) -1 -2 (SKIP TO Q.92) • (IF RESPONDENT IS UNDER AO, CODE "01" IN AGE BOX AND GO TO Q.92. IF RESPONDENT IS AO OR OVER, ASK Q.91b) Would you say it was before or after you were AO? Before 40,(3_9j_ -1 After AO ...... __ -2 Not sure ...... -3 92. Have you ever had three or more spells like this close together — three-week period? Yes......( 0 _ 4( No ........... say, within a -1 -2 93, Have spelts like this occurred at least nix different weeks of your life? Yes......(41 < _ -1 No..........._ -2 �CARD 088 94a. 812039 When did you last have a spell like this? Within last Between two Between one Between six two weeks or current..(42( weeks and one month ago... month and six months ago.. months and one year ago... ~ More than one year ago 94b. -2 I (SKIP TO SELF-3 J ADMINISTERED SHEET -4 I AFTER Q.94b) J -5 (ASK Q.94b) How old were you then? AGE: I I I (43) (44) years HAND SELF-ADMINISTERED SHEET TO RESPONDENnWDSAY: "The next questions are about how you have felt during the last three months. For each question, please circle the number corresponding to the answer that comes closest to the way you have been feeling. AFTER RESPONDENT COMPLETES SELF-ADMINISTERED SHEET, COLLECT SHEET AND RETURN TO Q.95. You will need to refer to Card "F" for many of these questions. I HAND RESPONDENT CARD "F"l 95a.In the last 3 months, how often have you kept losing your train of thought — would you say very often, fairly often, sometimes, almost never, or never? Very often..(45( Fairly often.... Sometimes Almost never "'_ -ij.(ASK Q.95b) -4 h(SKIP TO Q.96a) Never 95b. During what month and year did you begin losing your train of thought? i MONTH r i YEAR r i i i (46)i (47)i-i (48)i (49)i Don't remember (50 ( -1 | C ONT INUE _ VflTH C ARD~lnF'T 9 6 a . I n the last 3 months, how often have you felt unable to get things done? Very often..(5l( Fairly often.... Sometimes Almost never Never 96b. -l"l (ASK Q.96b) -4}.(SKIP TO Q.97a) During what month and year did you first feel unable to get things done? MONTH (52) (53) YEAR | T I I ( 54) ( 55) Don't remember CS ( 56( -1 �CARD 088 812039 I CONTINUE. WITH CARD " T T 97a. In the last 3 months, how often have you had trouble concentrating or keeping your mind on what you were doing? Very often. .(5T( Fairly often.... Sometimes Almost never.... Never 97b. -ll (ASK Q.97b) -2j -3 1 4 Vl -A V(SKIP TO Q.98a) -5J During what month and year did you first have trouble concentrating? 1 I MONTH YEAR 1 T l I T I l-l I I (58) ( 59) ( 0 (61) 6) Don't remember. . 6 ( (2 -i [CONTINUE WITH CARDi"F"T 98a. In the last 3 months, how often have you found yourself having to redo work that you had already done? Very often..fegj Fairly often.... -l) (ASK Q.98b) -2j Sometimes.. Almost never.... Never -3] -41 (SKIP TO Q.99a) -5j 98b. During what month and year did you begin having to redo work you had already done? MONTH J (65 . ) YEAR toeT ( 67) Don't remember. . 6 8 (.; -1 I CONTINUE WITH CARD "F"I 99a. In the last 3 months, how often have you found yourself unable to handle a task which at one time you could perform with little difficulty? Very, of ten.. ( J , 6): Fairly often.... -1\(ASK Q.99b) -21 Sometimes Almost never.... Never -3) -4V(SKIP TO Q.lOOa) -5J 99b. During what month and year did you first find yourself unable to handle such tasks? MONTH YEAR T i l l I I I I l-l I I "(70) (71) (72) (73) Don't remember. . 7 ( (4 -1 jCONTINUE UITH CARD "F"T lOOa. In the last 3 months, how often have you had trouble remembering things? Very often..(75j Fairly rften.... -l\ -2J (ASK Q.i00b) Sometimes -3 J Almost never.... -AMSKIP TO Q.10U) Nave r _»^_m~ -1 1 lOOb. During what month and year did you first have trouble remembering things? MONTH I YEAR l l I l l (70) (77) (78) (79) Don't remember..( 8<l__- 1 �...... .. •; ., CARD 101 812039 .'.CARD. "F" I 101 a. In the last 3 months, how often have you found yourself unable to handle large tasks efficiently? Very often..(.12( Fairly often.... -A. (ASK Q.iOlb) -2J Some t ime s Almost never.... Never -3 J -4l(SKIP TO Q.102a) -5j lOlb. During what month and year were you first unable to handle large tasks efficiently? MONTH YEAR (13) (14) (15) (16) I Don't remember. .(17( -1 1 CONTINUE WITH CARD " T F 102a. In the last 3 months, how often have you experienced difficulties when trying to solve some type of problem? Very often..O8( Fairly often.... -1\. (ASK Q.102b) -2| Sometimes -3 j Almost never.... -4V (SKIP TO Q.103a) Never -5J 102b. During what month and year did you begin having difficulty solving problems? MONTH YEAR I (19) (20) I (21) (22) Don't remember..(23( -1 TCONTINUE WITH CARD "FT 103a. In the last 3 months, how often have you felt confused and had trouble thinking? Very often..(24( Fairly often.... - \ (AgK Q.103b) ] -2J Sometimes....... Almost never.... Never -3| -4V (SKIP TO Q.104a) -Sj 103b. During what month and year did you first feel confused and have trouble thinking? MONTH YEAR (25) (26) (27) (28) Don't remember..(29( -1 [CONTINUE WITH CARD "F"I 104a. In the last 3 months, how often have you found yourself unable to perform tasks as quickly as you wanted to? ..(3pJ_____-lL (ASK Q.iQ4b) (0 -1\ Very often. . 3 ( Fairly often... en.... -2j Sometimes Almost never.... Never -31 -4>(SK1P TO Q.lOSa) -51 104b. During what month and year did you first have trouble performing tasks as quickly as you wanted to? MONTH T" 1 I t YEAR 1 I Til) 02) fn) O4 ) 67 Don't remember. . 0_5j____-l �CARD 101 812039 I CONTINUE WITH CARD " ' FT 105a. In the last 3 months, how often have you had a hard time getting going when you wake up? Very often..(36< -11 (ASK Q.lOSb) Fairly often... -0 Sometimes Almost never Never 105b. -3) -4y(SKIP TO Q.106a) -5 I During what month and year did you begin having a hard time getting I 1 Don't MONTH 1 I (37) (38) 1 1- •V YEAR 1 1 1 1 • (ASK Q.lOSc) (39) (40) (41( Within the last 12 months... -1 . -2 (SKIP TO Q.106a) 105c. During what period in your life was this most serious? MONTH YEAR MONTH YEAR (42) (A3) I I ( 4 (45) 4) TO I 1 1 - 1 1 . I (46) (47}(48) (49) (50( No period Not sure -1 __-2 TCONTINUE WITH CARD "F"! 106a.In the last 3 months, how often have you had uncontrollable feelings of anger? Very often..(.5l( Fairly often.... -1 J.(ASK Q.106b) -2J Sometimes Almost never Never -3 | -4 \- (SKIP TO Q.107a) -5J 106b. During what month and year did you first have uncontrollable feelings of anger? MONTH YEAR ^, 1 I I l T l (52) (53) Don't remember l - I I T l l (54) (55) (56( Within the last 12 months.. V (ASK Q.106c) -U -2 (SKIP TO Q.107a) 106c. During what period in your life was this most serious? MONTH YEAR MONTH YEAR 1 I T T T T T I Tl I T I I l-l I I TO I I l-l I I ( 57) (58) (59) (60) (61) (62) (63) (64) No period Not sure ( 65( 68 -1 -2 �...CARD. .107 812039 TCONTINUE WITH ...CARD :rVl 10 7«. In the last 3 months, how often have you been bothered by tiring out easily? Very often..(_12(_ Fairly often.... _ -l"l (ASK Q.io7b) -2J Sometimes......._ Almost never----_ Never ........... -3| -AY (SKIP TO Q.108a) -5j lt)7b. During what month and year did you begin tiring out easily? MONTH __ YEAR 1 I TT T T -I ....,.(. I-L , 1 . J >-(ASK Q.107c) (13) (i'i) ?15) (16) Don' t retnambei. .... , ...... « ( jf / ( i -1 • ~ •-'•"---- *J Within the last 12 months.. _ _-2 (SKIP TO Q.108a) 10.7c. During what period in your life was this most serious? MOMTH, YEAR MONTH __ YEAR .j u-p j. ^ | p ^ J..J ^. j. 1 I I l-l I I TO I I l-l I _ I (18) (19) (20) (21) ( 2 2 ) (23) (24) (25) No period.. ..... (26( Not sure ............ -1 -2 , WITH CARD. "F" I 108'i.In the last 3 months, how often has tiredness caused ydu to cut back your hobbies ot leisure activities? v Very often..(.22( -* V (ASK Q.lOSb) Fa,irly often.... "2J Sometimes Almost never Never -31 -AW (SKIP TO Q.109a) -Sj 108b. During what month and year did you begin cutting back your hobbies or leisure activities because of tiredness? MONTH YEAR \ I HT l-l (28) (29) Don't remember I .. I I (30) (31) ( 32( Within the last 12 months.. V (ASK Q.108C) -1 _-2 (SKIP TO Q.109a) 108c. During what period in your life was this most serious? MptlTH, YEAR MONTH YEAR T I T] I T I . I l-l I I (33) (34) (35) (36) T I I I"! I TO (37) (38) (39) (40) No pefiod.......(41(- . _-l Not sure -2 �CARD 107 812039 ]CONTINUE WITH CARD "F" | 109s. In the last 3 months, how often have you felt like a powder keg ready to explode? Very often..(.42( Fairly often.••• -ll(ASK Q.109b) ~2J Sometimes Almost never.... Never . . . -3 | -4 4-(SKIP TO Q.HOa) -5J 109b. During what month and year did you first feel like a powder keg ready to explode? 1 1 1 I I 1 1 1 l-l 1 1 (43) (44) (45) (46) • (ASK Q.109c) ( 47( Within the last 12 months.. -1 -2 (SKIP TO Q.llOa) 109c. During what perioc in your life was this most serious? MONTH YEAR MONTH YEAR 1 1 1 I I I I 1 1 1 I I 1 1 l-l 1 1 1 TO 1 1 1 l-l (52) (53) (54) < 55) (48) (49) (50) (51) No period ( 56( -1 JCONTINUE WITH CARD "F"| UOa.In the last 3 months, how often have you been troubled by feeling tired all the time? •> yery_often..(.57( Fairly often...._ -1 L (ASK Q.llOb) Sometimes......._ Almost never----_ Never...........~__ -4 M >H SKIP TO -5j Q.llla) llOb. During what month and year did you begin feeling tired all the time? MONTH YEAR I I I I1 I • H (58) (59) Don't remember (60) I I I I (ASK Q.flOc) (61) ( 62( ~1 Within the last 12 months.. _ -2 (SKIP TO Q.llla) llOc. During what period in your life was this most serious? MONTH __ YEAR MONTH __ YEAR 1 ^p ^ | j. j . | p ( 1 1 I I l-l I I TO I I . 1-1 . 1 . I ( 63) ( 64) ( 65) ( 66) ( 67) ( 68) ( 69) ( 70} No period ....... (71(_ Not sure............_ 70 -1 -2 �CARD HI 812039 I CONTINUE WITH CARP "F"I Ilia. In the last 3 months, how often have you felt too tired to walk up a flight of stairs? Very often..( 12(_ _-1 I (ASK Q.lllb) Fairly often.... Sometimes Almost never.... Never (SKIP TO Q.112a) lllb. During what month and year did you begin feeling too tired to walk up a flight of stairs? MONTH i I YEAH T.T~ T""~r T-l (ASK Q.lllc) ( 13) ( 14) ( 15) (16) Don't remember ( -2 Within the last 12 months.. (SKIP TO Q.112a) lllc. During what period in your life was this most serious? MONTH YEAR MONTH YEAR I (22) (23) I l-l ( 18) ( 19) (20) (21) (24) (25) No period Not sure.. -2 I CONTINUE WITH CARD "F"I 112a. In the last 3 months, how often have you found yourself powerless to control your temper? Very often..(27( -1 L(ASK Q.ll2b) Fairly often.... -21 Sometimes _ Almost never. . . ." Never (SKIP TO Q.113a) 112b. During what month and year did you first find yourself powerless to control your temper? I MONTH i YEAR Ti i r r(ASK (28) (29) Don't remember Q.112c) (30) (31) (. -2 (SKIP TO Q.113a) Within the last 12 months.. 112c. During what period in your life was this most serious? MONTH YEAR MONTH YEAR T I I (33) ( 34) I (35) ( 36) ( 37) '( 38) ( 39) (40) No period (_< l(_ Not sure 71 -1 �CARD 111 812039 IgONTlNUE WITH CARD 'Tfw\ 113a. In the last 3 months, how often have you felt too exhausted to perform your usual duties at work or at home in a competent manner? Very often..(42< Fairly often.... -1 I (ASK Q.U3b) -2j Sometimes Almost never.... Never -3 J -4 t(SKIP TO Q.114a) -5J 113b. During what month and year did you begin to feel too exhausted to perform your duties competently? MONTH YEAR 1 I I V(ASK Q.113c) (A3) (44) (45) (46) Don' t remember. ( 47( Within the last 12 months.. -2 (SKIP TO Q.114a) 113c. During what period in your life was this most serious? MONTH YEAR MONTH YEAR '_ T I T T 1 T I I l-l I I TO (48) (4$) (50) (51) (52) (53) (54) (55) No period Not sure. ( In the next series of questions, we are no longer referring specifically to the last three months. 114a. In general, do you speak to close friends — either in person or on the phone — much more often, somewhat more often, just as often, somewhat less often, or much less often than you used to? Much more often. Somewhat more often... Just as often. -2 f>(SKIP TO Q.llSa) Somewhat less often... Much less often -4 |(AgK Q.114b) ' *" 114b. During what month and year did you begin speaking less often to your close friends? MONTH YEAR 1 I T l I T I I 1-1 I I (58) (59) (60) (61) Don't remember..(62( -1 72 �CARD 111 811039 THAND RESPONDENT CARD "F"| 115a. How often has losing your temper created strains in your family relationships? Very often..(63j Fairly often. ... Sometimes Almost never.... Never 115b. -l~l (ASK Q.llSb) ~^J -3 J -it (SKIP TO Q.116) -5 I During what month and year did losing your temper begin creating in your family relationships? MONTH I I YEAR I I I I I I l-l .1 I (64) ( S 6 T (66) (67) Don't remember (8 6( (ASK Q.llSc) -1 ~2 Within the last 12 months.. _ 115c. (SKIP TO Q.116a) During what period in your life was this most serious? MONTH .. ^—-^I. I (70) __ ^- ^ l l YEAR ! I ___ ^ I (71) (72) TO MONTH ^I YEAR | -^ ^ ^ 1 1 1 I I 73) (741 (75) (76) No period.......(77 ( __-! Not sure ............ -2 116. In choosing your friends, how important to you are things like their religious or political beliefs. Would you say they are always very important, usually important, sometimes important, hardly ever important, or not important at all? Always very important. (78 (_ -1 Usually important......... _ -2 Sometimes important......._ -3 Hardly ever important....._ -A Not important at all......_ -5 [HAND; RESPONDENT CARD"F" I 117. How often have you deliberately said something that hurt someone's feelings? Very often..(79( Fairly often.... Sometimes Almost never.... Never.... -1 -2 -3 -4 -5 .KITH. CARD "F" I 118. of? How often have you done something of a sexual nature that society does not approve Very often.. (8_pJ __ Fairly often.... _ Sometimes. ......_ Almost never.... _ Never........ . .. 73 -1 -2 -3 -4 -5 �CARD 119 812039 119a. Do yon confide in close friends and relatives much more often, somewhat more often, just as often, somewhat less often, or much less often than you used to? Much more often...(12( Somewhat more often... Just as often/no difference -1 -2 Somewhat less often... Much less often -*\ (ASK Q.119b) (SKIP TO Q.120) -3 119b. During what month and year did you begin confiding less often in close friends and relatives? 1 I MONTH YEAR 1 F 1 I I I l-l I I (13) U 4 ) ( 1 5 ) (16) Don't remember.. (17( -1 120. Are your table manners at home just as good as they are when you are invited out to dinner? Would you say your table manners are always just as good at home, usually as good, usually not as good, or never as good at home? Always just as good at home...(18( Usually as good Usually not as good Never as good at home -1 -2 _-3 -4 121a. Do you find your current involvement in community activities to be much greater than it used to be, somewhat greater, just as great, somewhat less, or much less than it used to be? Much greater than it used to be Somewhat greater Just as great/no difference (19( -2 f (SKIP TO Q.122a) Somewhat less... Much less than it used to be "^(ASK Q.121b) 121b. During what month and year did you begin involving yourself less in community activities? MONTH YEAR I T (20) ( 2 l 5 ( 2 2 ) (23) Don't remember. . 2 ( (4 I HAND RESPONDENT CARD "F"I 122a. How often do you find you have trouble keeping track of bills — very often, fairly often, sometimes, almost never, or never? -1 would you say Very often..(_25< Fairly often.... -1 1 (ASK Q.122b) -2 J Sometimes Almost never Never -3 | -4 L(SKIP TO Q.123a) -5 ] I22h. During what month and year did you begin having trouble ket-pinp. truck of hills? MONTH YEAR T—i—n—i—r I _ L. l -(28) (29)I l. (275 I Don't remember. . (30( -1 �CARD 119 812039 TCONTINUE WITH CARD "F"I • 12.3a, How often do you find that you are unable to balance your checkbook? Very often..(3l( Fairly often.... .... \ (ASK Q.123b) ...l -2 J Sometimes Almost never Never -3[ -4V (SKIP TO Q.124) -5J 123b. During what month and year were you first unable to balance your checkbook? MONTH YEAR (3r2) (33) (34) (35) Don't remember. . 3 ( (6 I CONTINUE WITH CARD "rT 124. How often do you eat too much? Very often. ,(_37( Fairly often..... Sometimes Almost never.... Never -1 -2 "-3 -4 -5 125. In general, would you say your morals have been definitely above reproach, probably above reproach, probably not above reproach, or definitely not above reproach? Definitely above reproach Probably above reproach Probably not above reproach Definitely not above reproach (38 ( -1 -2 -3 -4 THAND RESPONDENT CARD "F"T 126a. How often has losing your temper created problems for you at work? Very often..(39 ( Fairly often..«._< Sometimes Almost never Never. -1 J. (^v Q.i26b) ______ -4^(SKIP TO Q.127a) 126b. During what month and year did losing your temper begin creating problems for you at work? MONTH YEAR rI i I \ l-lT I\ \I 140) (41) Don't remember (42) (43) . . . 4/j( ,.( Within the last 12 months.. 126c. • (ASK Q.126c) -1 -2 (SKIP TO Q.127a) During what period in your life was this most serious? T MONTH i ri __ YEAR i r I ~ I l-l I I U5) ( 6 4r (47) ( 8 4) i TO MONTH I I i ri YEAR i (49> (SO) ~T5i>" (52) No period.......(53( _ ^ -I Not sure ............ -2 75 r �CARD 119 812039 [CONTINUE WITH CARD "F"T 127a. How often has losing your temper resulted in a friendship breaking up? Very often..(j^J Fairly often.... -ll(ASK Q.127b) -2J Sometimes Almost never Never -3 | -4V (SKIP TO Q.128) -5J 127b. During what month and year did losing your temper first result in a friendship breaking up? MONTH YEAR 1 I I I l-j I I (ASK Q.127c) (55) (56) (57) (58) Don't remember ............. ( 9 5( -1 Within the last 12 months.. _ 127c. iI (60) (SKIP TO Q.128a) During what period in your life was this most serious? MONTH _ -2 i 1 YEAR 1 (61) n 1 1i - MONTH __ YEAR r i I ' n Ii i M I TO I (62) ( 3 6) (bit) (65) No period ....... (fj8( Not sure............_ I r ( 6 (67) 0) -1 -2 — — — — — — — — - • — • — • — - — — — - — — • " — — — — — ' — * — " " — • — - • - » . • * . . • • - • - * • • • • . • ! • • • . . • • » • . . • . . — —•. • • » • •• • • » • • — • — - • • • • — - . • • - • • « • • « •••—.••«.—.••• Now I'm going to ask you a few questions about recreation and leisure activities. 128. What are some of the hobbies and sports you participate in on a regular basis? Any others? 3. A,. 5. 6. 129. Have you participated three or more times in (READ EACH ITEM)? ANY ITEM MENTIONED IN Q.128 AND DO NOT READ THAT ITEM) (CODE "YES" FOR _ 1. _ No Scuba diving 2. Auto, boat, or motorcycle racing..............................(70 ( A. Mountain climbing..........................................•••'i2-i_ 5. Hang gliding..................................................(73 ( 6. Plane racing or plane acrobatics, not including flight training or any assignments for the Armed Forces..............(74 (_ 76 -1 _ "2 �ITISTESTtU-EMJLOVEaS AMD' DUTIES'TROH S.R.B. - ?ACE 1 AT tut TOE> 6F THE JOB cOUMiS OTT PACES 72" AND 73. IF So JgBffeNTEREp"lN S ^ IF MOKE THAN SIX JOBS ON S.R.B. PAGE 1, RECORD ADDITIONAL EMPLOYERS AND DUTIES AT TOP OF COLUMNS ON PAGES 44-45 IN S . R . B . ! - . PAcn'r'SK'iP TO Q . U T . T . 130. I would like Co asfc you about your experience with certain chemicals or toxic su bstances. m i li tarv . FIRST JOB 1'H'AND RESPONDENT CARD "G" 1 130a. While working at (EMPLOYER) as (DUTIES), d o / d i d you come in contact with any of the subs tances on t h i s card? By c o n t a c t , I mean that you inha led, t a s t e d , had s k i n contact w i t h , or were radiates by any of these substances? I MULTIPLE RECORD I Ol..(12( 02. .( 13( 03.. (TS -1 -1 -1 05. (1« 06. ( 17( O7.(l8( 04.. ( iy. -i SECOND JOB 01..(12( O2..(l3( 04.. ( 15( (IF "07," -1 -1 O5.(l6( 06.(T7( -1 1 1 ! 1 1 THIRD JOB -1 -.1 C1..(12( O2..(13( -1 -1 O4..(15( (IF "07," SKIP TO NEXT JOB) SKIP TO NEXT JOB) TASK Q.i30b FOR EACH SUBSTANCE CODED IN q.i30a.T 130b. tn general, how many days a month ( d o / d i d ) you come in contact with (SUBSTANCE)? -1 -1 -1 1 First, l e t ' s talk about your employment outside of the -1 05.(16( O6.(i7( -1 -1 (IF "07," SKIP TO NEXT JOB) 01..! I I (19H20) 1 1 1 1 1 1 01.. 1 1 1 04.. I I 1 (19K20) (21K22) I I I I I 01. .1 1 1 04.. I I (19)(20) (21M22) I I I 02. .1 1 1 (23X24) I I I 05. .1 1 1 (25) (26) 1 1 I I I I 02. .1 1 1 05. .1 1 1 (23) (24) '.25) (26) I I I I I 02. .t 1 1 05.. 1 1 (23) (24) (25) (26) I I I 03.. I I I (27)(28> !lF ANY SUBSTANCE CODED IN Q.130b, ASK Q.130cl 13Gc. While you were on that job, how o f t e n di d you wash to remove the (SUBSTANCES) or use pro" tective gear —> would you say all of the time, some of the time, or never? 04. .1 I (21H22) I I I 06.. 1 1 1 (29) (30> I I I I I I 03.. I 1 1 06..! 1 I (27) (28) (29) (30) I I I 1 1 03.. 1 1 1 06.. 1 1 (27) (28) (29X30) All the time(3l( Some of -IKASK Q. f 130d) [HAND RESPONDENT CARD "H'M 130d. Which of the following did you use on that job? [MULTIPLE RECORD IF NECESSARY 1 Air f i l t e r ( 32( -l\ Goggles (_33(__~1 (GO TO Face shield ( 34( -1 NEXT JOB) Special clothing. .(_3<__-l Washing f acilitiesC 3ft -1 01 79-80 -ll(ASK Q. \ 130d) All the t i m e . ( 3 l ( Some of -l) (ASK Q. > 130d) Never JOB) All the time.(3l( Some of -3 (GO TO NEXT JOB) Never -3 (CO TO NEXT JOB) Air f i l t e r ( 32( Goggles • . •• ( 33( Face shield (~34( Special clothing.. ( 35( Washing faelllties( 36( Air f i l t e r (32( -l| ( 33( — 1 (GO TO Goggles Face shield (~34( -1 ' N E X T -1 JOB) Special clothing. .( 3s( -1 Washing £acilities( 36( 02 79-80 -1 -t (GO TO -1- NEXT JOB) -1 -1 03 79-80 �CARD 130 FOURTH JOB I HAND RESPONDENT CARD ''c'1 1 130a. W h i l e working at (EMPLOYER) as (DUTIES), d o / d i d vou come in contact with any of the substances on t h i s card? By contact, I wan that you i n h a l e d , t a s t e d , had skin contact with, or were radiated by any of these substances? IMULTIPLE RECORDl IASK Q.130b FOR EACH SUBSTANCE CODED IN Q.130a.| 130b. In general, how many days a month ( d o / d i d ) you come in contact vith (SUBSTANCE)? Ol..(12( 02. . ( J 3 ( 03..(T31 O4..(15( -1 -1 -1 -1 SIXTH JOB FIFTH JOB 05-(16( -1 O 6 . ( l 7 ( -1 O7.(18( -1 (IF "07," SKIP TO NEXT JOB) 01..(12( -1 05.(16( -1 03.. (TT( O4..(15( -5 -1 01..(12( O2..(l3( 03..U4C Q6..(15( 07. (T8T -1 (IF "07," SKIP TO NEXT J O B ) -1 -' -1 -1 05. ( I f , I -1 06. (j 7', -1 O 7 . ( 1 8 ( -1 (IF " 0 7 , " S K I P TO NEXT J O B ) 1 . 1 04.. 1 1 (21M22) 1 1 1 I I I 01. .1 1 I 04. .1 I 1 (19) (20) .(21) (22) I I I i i ! 01. .1 1 1 04..! ! i (19 ) ( 2 0 ) (21 )(22 i 1 1 1 02. .1 1 I (23K24) 1 1 1 05. .1 1 1 (25)(26) I I I I I I 02.. 1 I I 05.. 1 1 1 (23) (24) (25) (26) I I I 1 i 1 02. .1 1 1 05. .1 1 1 (23) (24) (25) (26) I I I 03.. 1 I I (27)(28) I I F ANY SUBSTANCE CODED IN Q.ljOb, ASK Q.130cl 130c. While you were on that job, how o f t e n di d you wash to remove the (SUBSTASCES) or use protective gear — would you say all of the time, some of the time, or never? 1 1 1 01.. 1 I I (19H20) 1 1 1 06.. I 1 I (29)(30) 1 1 1 I I I 03.. 1 I 1 06. .1 1 1 ( 2 7 ) (28) ( 2 9 ) (30) I I I 1 i ! 03.. 1 1 1 06.. 1 1 1 ( 2 7 ) (28) (29H30) -l)(ASK Q. J. 130d) All the time. (31 ( Some of -f)(ASK Q. f 130d) All the time. (31 ( Some of the t ime -l)(ASK Q. r 130d) -2 1 Never iHAND RESPONDENT CARD "H" 1 130d. Which of the following £id you use on t h a t iob 7 IMULTIPLE REC057* IF NECESSARY) All the time(3l( Some of -3 (GO TO NEXT JOB) Never -3 ( G O T O NEXT JOB) Never -3 (GO TO NEXT JOB) Air f i l t e r (32( -?) • Goggles (33 ( -1 (GO TO Face shield (34 ( -1 NEXT JOB) Special clothing.. (35( -1 Washing f a c i l i t i e s ( ^ f t ( -1 04 Air f i l t e r (32( -l) Goggles. (33 ( -l|(GO TO Face shield (34 (' - I T NEXT Special clothing. .(35 ( -ll JOB) Washing f a c i l i t i e s U b ( -ll 05 Air f i l t e r (32 ( _-l (IF MORF Goggles (33 ( -1 THAN' ft Face shield (34 ( -1 JOBS, GO Special clothing. . (js ( _-l TO PAGE W a s h i n g f a c i l i t i e s U b ( , -1 44 IN S.R.B.) 06 �CARD 131. 131 812029 Have you ever had any other jobSj such as a temporary job or a job while you were in school, outside of the military? Yes ( 12( -1 (ASK 0.132a) No -2 (SKIP TO 0.133) lHAND RESPONDENT CARD "G"l 132a. In those jobs, did you ever come in contact with any of the substances on this card? X-Ray Or Nuclear Radiation Asbestos Q.!32a. Yes (13( -1 Industrial Chemicals -2 IRECORD AT TOP OF APPROPRIATE COLUMN: Defoliants Or Herbicides -3 Insecticides . Or Pesticides Degreasing Chemicals -f- -5 -4 i IFCF. EACH SUBSTANCE | iCODED, ASK Q.132b. 1 132b. In what years were you in contact with (SUBSTANCE) on those jobs? 1 FROM > ! ! FROM ! 1 i 1 1 1 (14) (15) TO I 1 1 1 ! (14) (15> TO 1 I I l ! l ! ! FROM 1 1 1 1 ! 1 ( 4 ' (15) 1) TO 1 1 1 1 1 ! (16) (17) 1 ( 16) ( 17) 132c. How many days altogether would you say you came in contact with (SUBSTANCE) on these jobs? I32c. On those days you came irt contact with 'SUBSTANCE) how often did /ou wash to remove the (SUBSTANCE), or use protective clothing or gear — all the time, some of the time, or never? IttAND RESPONDENT CARD "H" i T32e. Which of the following did you use? 1 MULTIPLE RECORD IF | ! NECESSARY I (16) (17) DAYS i 1 1 1 1 1 1 I <18)( 19><20) DAYS 1 1 1 1 1 1 ! ! ( 18) (19X20) All of the time.( 21( Some of the -1 All of the time.(2l( Some of the -1 FROM 1 1 FROM 1 1 i i (14) (15) TO I I I ! ! I DAYS 1 1 1 1 1 ! (18)(19)(20) -1 ! DAYS 1 1 1 1 1 1 1 1 (18)(19X20) All of the time.(2l( Some of the l 1 FROM ! I l i (14) (15) TO 1 1 i 1 1 1 (16) (17) (16) (17) 1 All of the time.(2l( Some of- the ! I -1 DAYS i I (li) (!5) TO 1 1 1 (16) (17> ! i • ! i 1 1 1 I (18X19X20) i 1 ! i l l All of the time.(2l( Some of the ' I -1 DAYS : i 1 i (l8)(l9X20) All of the time.(21( Some of the -i _*i <t. Never -3 (IF "NEVER," GO TO NEXT SUBSTANCE REPORTED) Air Filter. ( 22( -1 Goggles (~23T -1 Face Shield( 24( -I Special Clothing.. ( 5 -1 2( Washing Facilities(26( -1 (GO TO NEXT SUBSTANCE REPORTED) 01 79-80 Never -3 (IF "NEVER," GO TO NEXT SUBSTANCE REPORTED) Air Filter. (22( -1 Goggles ("231 -1 Face Shield(24( -1 Special Clothing.. ( 5 2 ( -1 Washing Facilities(26( -1 (GO TO NEXT SUBSTANCE REPORTED) 02 79-80 Never -3 (IF "NEVER," GO TO NEXT SUBSTANCE REPORTED) Air Filter. (22( -1 Goggles (23( -1 Face Shield(24( -1 Special Clothing.. ( 5 2 ( -1 Washing Facilities(26( -1 (GO TO NEXT SUBSTANCE REPORTED) 03 79-80 Never -3 (IF "NEVER," GO TO NEXT SUBSTANCE REPORTED Air Filter. (22( -1 Goggles (23( -I Face Shield(24( -1 Special Clothing.. (25( -1 Washing Facilities(26( -1 (GO TO NEXT SUBSTANCE REPORTED) 04 79-80 Never -3 (IF "NEVER," GO TO NEXT SUBSTANCE REPORTED) -3 Air Filter. (22( -1 Air Filter. (22( — 1 Goggles. . . . (23( -1 Goggles (23( -1 Face Shield(24( -1 Face ShieldC-.( -1 Specia 1 Special Clothing. .(25( -1 Clothing. .(25( -1 Washing Washing Facilities(2d( -1 Facilities(26( -1 (GO TO NEXT SUBSTANCE REPORTED) ° A 7 u~bO 05 79-80 �CARD 133 ENTER ALL COUNTRIES FROM S.R.B. - PAGE 2 AT THE TOP OF THE COUNTRY COLUMNS ON PAGES 75 AND 76. IF NO COUNTRIES ENTERED IN S.R.B. PAGE 2 , SKIP TO Q.134 IF MORE THAN SIX COUNTRIES OK S.R.B. PAGE 2, RECORD ADDITIONAL COUNTRIES AT TOP OF COLUMNS ON PAGES 46 AND 47 IN S.R.B. 133. Next, I'd like you to think again about your active duty assignments. SECOND COUNTRY FIRST COUNTRY 12-13 (14-15) [HAND RESPONDENT CARD "G"| 133a. In your job assignments while stationed in (COUNTRY), ( t h a t t i m e ) ( d o / d i d ) you come in contact with any of the following substances? lASK Q.133b FOR EACH SUBSTANCE CCOED IN Q.133a ! 133b. In general, how many days a monn ( d o / d i d ) you come in contact wit*1. (SUBSTANCE)? Ol..(16( -1 O5.(20( -1 O2..(17( -1 06. (277 -1 03..(IFT~-1 - 07.C22T -1 O4..(l9( -1 (IF "07," SKIP TO NEXT COUNTRY) 1 01.. 1 1 I ! I I 04.. | (23) (24) l ! l ! (25) (26) THIRD COUNTRY (14-15) 01..U6( 02.. (FTl -1 O5.(20( -1 -1 O6.(21( -1 O7.(22l -1 -1 (IF "07," SKIP TO NEXT COUNTRY) Ol..(16( -1 O5.(20( -1 Q2..(17( -1 06.(21 ( -L 03.. (Is!-1 07. (ill-1 (IF "07," O4..(19( -1 SKIP TO NEXT COUNTRY) 03.. (nn -i O4..(19( 1 1 1 1 1 1 1 04.. I I | 01.. 1 1 (23)(24) (25) (26) 01..I ( 3 ( 4) 2)2 (25 )(26 ) I I I 02.. 1 I | ( 2 7 ) (28) 1 1 1 I I ! 52. .1 I 1 05..! I I (27 )(28) (29) (30) I I I I I 02..I I I 05..I I (29 ) GO ) (277(28) I I I 03.. 1 I I (31) (32) i l F ANY SUBSTANCE CODED IN Q.13?r. ASK Q.133cl 133c. Did you wash to remove tr.e (SUBSTANCE) or did you use protective clothing or eear when stationed in (COUNTRY) — = i. of the time some of the time, or never? I I I 05.. 1 1 1 (29) (30) I I I 06.. 1 I 1 (33) (34) 1 1 1 I I I 03. .1 ! ! 06.. 1 i 1 (31) (32) (33) (34) 03..I I 06..I 03 (31 )(32 ) T -lT(ASK Q. f 133d) -2J All the time.( 35 ( Some of -l](ASK Q. f 133d) -2 *J All the time. (35 ( Some of the time Never iHAND RESPONDENT CARD "H" ! 133d. Which of the f o l l o w i n e c:-- vou use on t h a t iob? (MULTIPLE RECORD :• NECESSARY I All the tiroe(35( Some of the time -3 (GO TO NEXT COUNTRY ) Never -3 (GO TO NEXT COUNTRY ) Never. Air f i l t e r (36 ( Goggles (37 ( Face shield (33 ( Special clothing. .(39 ( Washing f acilit ies(40 ( -l] -1 (GO TO -1 NEXT -1 COUNTRY) -1 01 79-80 Air filter Goggles ....(36( (J?l flu ( Special clothing. .(39 ( Washing f a c i l i t iesCip ( -?) -1 (CO TO -1,'NKVT -1 C O U N T R Y ) -ll 02 79-80 T )(34 ) -ll(ASK Q. V 133d) -2j _-3 (GO TO NEXT COUNTRY Air filter 06 ( Goggles (3_7 Face shield.......(• Special clothing.. (39 ( Washing facilitiesCiO �FOURTH COUNTRY .12-13 SIXTH COUNTRY FIFTH COUNTRY (14-15) (14-15) (14-15) IHAND RESPONDENT CARD "G"| I 3 3 a . In your job assignments w h i l e s t a t i o n e d in ( C O U N T R Y ) , ( t h a t t ime) ( d o / d i d ) you come in c o n t a c t with any of the following substances? l A S K 0.133b FOR EACH SUBSTANCE CODED IN Q.133a. 1 133b. In g e n e r a l , how many days a month ( d o / d i d ) you come in contact with (SUBSTANCE)? O2..(17( 03.. (T^ O4..(19( Q5.(20( -1 O6.(21( -1 07.(~22T -1 (IF "07," SKIP TO NEXT COUNTRY) -1 -1 -1 Ol..(16( Q2..(17( -1 -1 04..(T5( -1 O5.(20( -1 O6.(21( -1 07. ("221 -1 (IF "07," SKIP TO NEXT COUNTRY) :i..(16( ^..(nC ~3..(18( :i..<19( O 5 . ( 2 0 ( -1 06. ( 2 H -1 0 7 . ( 2 2 ( -1 (IF " 0 7 , " SKIP TO NEXT COUNTRY) I I I 01..! 1 I (23)(24) 1 1 1 04.. I 1 1 ( 25) (26) 1 1 1 1 1 1 1 1 01. .1 1 1 04.. I (23) (24) (25) (26) i 01. . i 1 ! 1 02. .1 1 1 ( 2 7 ) (28) 1 1 1 05. .1 t 1 ( 29) ( 30) 1 1 1 1 1 1 02.. 1 1 1 05. .1 1 1 ( 2 7 ) (28) ( 29) ( 30) i i 1 I I I 02.. 1 i 1 05..! ! i ( 2 7 ) (28) ( 2 9 ) (30) I I I 03.. 1 ! 1 (31) ( 3 2 ) i l F ANY SUBSTANCE CODED IN Q.133b, ASK Q.133cl 133c. Did you wash to remove the (SUBSTANCE) ^•c did you use protective clothing or gear when stationed in (COUNTRY) — all of the time, some of the time, or never? 1 I I 06.. I 1 1 ( 33) ( 34) I I I 1 1 1 03..! 1 1 06. .1 I 1 (31M32) (33M34) i l l I I I 03. .1 ! ! 06.. 1 1 1 (31)(32) (33)(34) T All the tiae.(35( -1HASK Q. Some of \f 133d) the tine -2J All the time( 35( Some of -1 (ASK Q. 133d) ! Never -3 (GO TO NEXT COUNTRY ) All the time.(35( Some of -l) (ASK Q. > 133d) -2j -3 (CO TO NEXT COUNTRY) l ! -1 -I -1 -1 l 1 1 04..! <.23)(24) Never ! ! i i (25) ( 2 6 ) -3 (GO TO NEXT COUNTRY IHAND RESPONDENT CARD "H"| 133d. Which of the f o l l o w i n g did you use Air f i Iter ( 36( Face shield ( 38( Special clothing.. ( 39( Washing f a c i l i t i e s ( 40( -1 -1 (GO TO -1 NEXT -1 COUNTRY) -1 04 79-80 . . . ( 36( Face shield . ..(38( Special c l o t h i n g . .( 39( Washing facilities(~50f -l) Air f i l t e r -1 NEXT -1 COUNTRY) -1 Face shie'.d ( 38( Special cSathing. .( 39( Washing f a:ilit ies( 40( 05 79-80 ( 36( -1 (RECORD -1 T10NAL -1 COUN-1 TRIES IN S.R.B. PAGE 46) 06 79-SO �CARD 812039 lHAND RESPONDENT CARD "G" 1 134a. Have your hobbies and spores a c t i v i t i e s ever brought you in contact w i t h any of the ( a l l o w i n g s u b s t a n c e s ? X-Ray Or Nuclear Radiation Asbestos 0.134a. Yes I F O R EACH SUBSTANCE I i C O D E D , ASK Q.134b. 1 134b. I n what month a n d y e a r did y o u r hobbies and s p o r t s a c t i v i t i e s f i r s t b r i n g you in cont a c t w i t h (SUBSTANCE)? 134c. For how many years did you c o n t i n u e to come in c o n t a c t w i t h (SUBSTANCE)? !34d. In g e n e r a l , how many days per year did you come i n c o n t a c t w i t h (SUBSTANCE)? !34e. On the days you came in c o n t a c t w i t h (SUBSTANCE) how o f t e n did t h i n g or gear or wash to remove (SUBSTANCE) — all of the t i n e , some of the t i m e , or never? i H A N D RESPONDENT CARD "H" I n i f . Which of the • o l l o w i t i e did vou use? ^ i ' J . ' l IPl.r RECORD IF 1 Sr.i i-SSASV ! ( 13( -1 -2 MOSTH YEAR 1 i 1 I 1 1 1 : 1 - 1 1 1 ( 1 4 ) (15) (16) (17) YEARS ! I ! 1 I 1 (18)(,9) MONTH 1 I 1 l (14) YEAR 1 l (15) 1 - 1 I 1 l l (16) (17) -i Never -2 (!? "NEVER," GO TO NEXT SUBSTANCE REPORTED) Air F i l t e r . (24( -1 Goggles ( 25( -1 Face 3 h i e l d ( 2 6 ( -1 Scecial Clothing.. (27( -1 Washing F a c i l i t i e s ( 2 8 ( -i <V> *'.'• N E X T SUBSTASC£ REPORTED) 01 79-30 All of the time.(23( Some of the MONTH YEAR ! 1 I I 1 1 I l l I l l (14) (15) (16) (17) YEARS 1 1 1 1 1 1 U8M19) DAYS 1 1 1 1 1 1 1 1 (20)(21)(22) -i (IF "NEVER," GO TO NEXT SUBSTANCE REPORTED) Air F i l t e r . (24( -1 Goggles (25( -1 Face S h i e l d ( 2 6 ( -] Special C l o t h i n g . . ( 2 7 ( -1 Washing Facilities(28( _j (CO TO NEXT SUBSTANCE REPORTED) 02 79-80 -1 MONTH YEAR 1 1 1 1 1 1 1 1 l-l 1 I (14) (15) (16) ( 1 7 ) Degreasing C h e m i c a l s -6 MONTH YEAR 1 1 1 1 1 1 I l l I l l ( 1 4 ) . (15) (16) ( 1 7 ) YEARS 1 1 1 1 1 1 (18)(19) DAYS 1 1 1 I 1 1 1 I (20)(21)(22) All of the t i m e . ( 2 3 ( Some of the Insecticides Or Pesticides -5 YEARS ' 1 1 1 1 1 1 (18)(19) DAYS 1 1 1 1 1 1 1 1 (20H21K22) All of the time.(23( Some of the time. .. Defoliants Or Herbicides -4 -3 YEARS 1 1 1 1 1 1 (18)(19) DAYS i i I I ' I I I (20) ( 2 1 ) ( 2 2 ) All of the t i m e . ( 2 3 ( Soae of the Industrial Chemicals (RECORD AT TOP OF APPROPRIATE COLUMN] -1 DAYS i 1 1 1 1 1 I I (20)(21)(22) All of the t i m e . ( ' J ( Some of the -1 MONTH YEAR i 1 1 1 i I 1 1 i-l 1 1 (14) ( 1 5 ) (16) ( 1 7 ) YEARS 1 1 1 1 ! 1 (18)(19) DAYS I I I ! 1 i 1. i (20) ( 2 1 ) ( 2 2 ) All of the tiir.e.(- J ( Some of the -1 -2 Never -2 (IF "NEVER," GO TO NEXT SUBSTANCE REPORTED) Air F i l t e r . (24( -1 Goggles.... ( 2 5 ( '-1 Face S h i e l d ( 2 6 ( -1 Special Clothing. .(27( -1 Washing 28 F a c i l i t i e s ( ( -1 (GO TO NKXT SUBSTANCE REPORTED) 03 79-80 Never -2 (IF "NEVER," GO TO NEXT SUBSTANCE REPORTED) Air F i l t e r . ( 2 4 ( -1 Goggles ( 2 5 ( -1 Face S h i e l d ( 2 6 ( -i Special Clothing. .(27( -1 Washing F a c i l i t i e s ( 2 8 ( -1 (CO TO NEXT Sl'BSTANCE R E P O R T E D ) 04 79-80 Neve** —1 (IF "NEVER," GO TO NEXT SUBSTANCE REPORTED) Air F i l t e r . ( 2 4 ( -1 G o g g l e s . . . . ( 2 5 ( -1 Face Shieldt 2 "?! -1 i Special Clothing. .(27( -1 Washing F a c i l i t i e s ( ' a ( -1 ; (CO TO Ni£XT SUBSTANCE REPORTED) i 05 79-80 Air F i l t e r . ( 2 - » i Goggles. . . . ( - 5 ( Face S h i e l d ( - ° ( Specia 1 Clothing. .C-7l Washing Fjcilicies(~Jt -1 -I -i -1 -1 i..., "••->- �CARD 135 812039 Now I have some questions about your income. I HAND RESPONDENT CARD "l"I 1 3 5 ^ P I e a s e tell me wh ich letter on this card best represents the total household income in 1980 before taxes or other deductions for all people in your household, not including roomers. This amount should include wages, net income from business, itltSteSt, dividends, pensions, and any other money income. Tell me the letter that comes closest. A. H. t, p F.. F. r, $?0,000-$2'i 999 -1 -2 "•' 1 ~-l\ $25,000-$29,999 $30,000-$34,999 $35,000-$39,999 -6 _-^ $5,000-$9,999 (12 ( $10,000-$ 14 ,999 lit 1 *. ,110(1 $ 1 'I , ''' !)) H. T ,7. K. 1.. M. N. 0. P. 0 R. R. T. "-fi $45,000-$49,999 $50 ,000-$54 ,999 $55,000-$59,999..(13( $60,000-$64,999 $65,000-369,999 $70,000-$74,999 $75,000-$79,999 $80,000-$84,999 $85,000-189,999 $90,000-$94 ,999 $95,000-499,999 $100,000 or more..... 136a. Did you earn any income from any job during 1980? retirement plans or pensions. Yes. _-l No.. -2 "-0 -1 -? -ft -5 _1 ~-n -9 -0 Do not include income from (ASK Q.136b) (SKIP TO Q.137a) I CONTINUE WITH CARD *JI_"J 136b. In vhich of these groups did your earnings from jobs in 1980 fall that is, before taxes or other deductions? Tell me the letter that conies closest. A. B. C. D. E. F. G. H. I. J. K. L. M. N. 0. P. Q. R. S. T. j5,000-*9,999----(15 ( tlO,000-tlA,999 ...... $15,000-$19,999......" $20,000~$24,999 ...... ~ $25,000-$29,999......" $30,000-$34,999......" *35,000-*39,999......" " -1 I"6 -7 I"8 -9 $45,000-$49,999 _ $50,000-$54,999 _ $55,000-$59,999..(16 ( ' " ' -1 "" $60,000-$64,999 -2 $65,000-$69,999 ^-3 $70,000-$74 ,999 -4 $75,000-$79,999 -5 $80,000-$84 ,999 -6 $85,000-$8P,999 -7 $90,000-$94 ,999 -8 *95,000-$99,999 -9 $100,000 or more -0 �CARD 135 812039 137a. We would like your consent for the doctors and medical facilities you mentioned during this interview to provide medical records to the Air Force Health Survey. These records will help us obtain more detailed information about the health services you talked about. TURN TO S.R.B. PAGES JT-// . ENTER NAMES OF MEDICAL PROVIDERS ON APPROPRIATE PERMISSION FORMS AND ASK RESPONDENT TO SIGN EACH FORM. IFOR EACH SIGNED FORM, ASK:I 137b. What is the current address of (DOCTOR/FACILITY)? 138a. To obtain the most complete and useful information that we can, ve are asking participants to have a physical examination. The examination will be conducted by a private medical facility and will take place over a five-day period that is convenient for you. (IF DISCHARGED OR RETIRED SAY: "You will also receive a stipend of tlOO.OO a day. ) The United States Air Force will pay for all travel and per-diem expenses so that participants can go to a nationally recognized medical facility. If you were asked, would you be willing to have a physical examination at a time convenient for you? Yes. No.. 138b. (RECORD IN S.R.B. PAGE 13 AND SKIP TO Q.139) (RECORD IN S.R.B. PAGE 13 AND ASK Q.138B) What is your reason for not wanting to have the examination? 1 RECORD IN S.R.B.L PAGE 13| 138c. Under what conditions would you be willing to have an examination? [RECORD IN S.R.B., PACE 131 [CHECK S.R.B. PAGE 2. IF ANY DIVORCED OR SEPARATED WIVES, READ:I 139. It i'E very important lor the success of this survey that we also conduct a brief interview with former-wives of respondents. This will provide a more complete and accurate picture of the health of the families of Air Force personnel. We would like to send this letter signed by you Co (each of your former/your former) (wife/wives). ENTER NAME OF EACH FORMER WIFE ON A LETTER. RECORD CURRENT ADDRESS FOR EACH WIFE ON LETTER. HAVE RESPONDENT SIGN EACH LETTER. ICHECK S.R.B. PAGE 2.IF LIVING WITH WIFE, SAY:! 1AO. I would like to speak to your wife briefly. Is she available now? IF WIFE IS NOT LIVING AT SAME ADDRESS, RECORD CURRENT ADDRESS ON STUDY SUBJECT NAME ASSIGNMENT SHEET. . Thank you for participating in the Air Force Health Study. TIME INTERVIEW ENDED: (am/pro) �FOR OFFICE USE ONLY: LOUIS HARRIS AND ASSOCIATES, INC. 630 Fifth Avenue New York, New York 10111 Case No. Study No. 812039 O.M.8. NUMBER 0701-0033 Approval Expires 11/30/82 Respondent CONFIDENTIAL AIR FORCE HEALTH SURVEY SUPPLEMENTAL RECORDING BOOK �812039 Q.2. Where born: City: State: Q.8b. Employers 1st job: 2nd job: 3rd job: 4th job: 5th job: 6th job: 7th job: 8th job; 9th job: 10th job: llth job: 12th job: Q.8e Main Duties �812039 Q.Hi: Countries Served In: _______ 1. 2 .__________ _ _ 9. 10. 4. 11. 12. Q. 20-22 and J_52-JVi Marital History b. Wife's Current Full Name First/ only wife Second wife Third wife Fourth wi fe Fifth wife Sixth wife Wife's Maiden Name Living With Wife Or Divorced/ Separated/ Widowed �^.:«-T2 and 157 - 165 CHILDREN CHILD Q.29 FIRST First Last SECOND d. BIRTHDATE .-:;:•:::-: First Place C/S Last C/S MONTH 1 I I 1 l-l DAY 1 1 Place First Place I! l-l YEAR 1 i Place i ! i i C/S Last C/S MONTH DAY Place First Place C/S Last C/S i ! I I l-l 1 YEAR 1 1 i i l-l C/S Place i < i 1 C'S DAY i 1 Place 1 i :-S First Place 1 ! l-l YEAR : •' Place i MONTH ! M 1 l-l C/S Last C/S MONTH I ! ! ! !-! DAY 1 I Place : ' C S First Place II I-! YEAR ! ' Place i ! C/S Last C/S MONTH DAY Place First Place C/S Last C/S First !.a = r x- DEATH RECORDS Place First Last h. MOTHER'S FULL NAME Place i | C'S ; FT-TH CURRENT MEDICAL RECORDS YEAR 1 i First Last g. 1 1 )-! i FOL'PTH BIRTH RECORDS MONTH DAY 1 1 1 1 | |-| | First L3«=t f. i ! First Last THIRD NAME ! i I 1 l-l 1 1 YEAR 1 1 l-l • 1 Place ! ! C S �<". 2'H-32 and CHILD 157 - 165 CHILDREN q.29 NAME d. f. BIRTH RECORDS BIRTHDATE RECORDS SEVENTH First Last EIGHTH First Last NI NTH TESTH 1 1 F irst ENTH ast 1 1 'r.\~ ' F^Hr irst Last DAY I M !-l YEAR Place i l 1 1 c/s ! 1 I M l-l YEAR Place i l 1 1 C/S MONTH 1 I 1 1- -DAY i I I YEAR Place 1 . i ! l-l ! 1- DAY i 1 M l-l l 1 MONTH 1 1 1 1- DAY 1 i 1 1 l-l YEAR 1 1 MONTH 1 1 F irst Last DAY 1 1 MONTH I 1 First Last MONTH 1 1 1 1- MONTH i 1 DAY i M l 1 1- I 1 ! 1 1 1 !! 1- ! 1 1 ! c/s YEAR Place ! 1 c/s Place 1 1 YEAR l-l C/S Place NAME Place First Place C/S Last C/S Place First Place C/S Last C/S Place First Place C/S Last C/S Place First Place C/S Last C/S Place First Place C/S Last C/S Place First Place C/S Last C/S 1 1 1 c/s �CARD 136 Q.35 Medical Providers — Pneumonia jst Time a. 2nd Timr Monthe/years had that time. MONTH 812039 YEAR T a. 3rd Time Months/years had that time. MONTH a. Months/years had that time. YEAR YKAR MONTH F I (28) (29) (20) (21 ) (22) (23) TO MONTH YEAR 112') (13) (14) (15) TO MONTH YEAR MONTH h. (25) (18) (19) Doctor/facility who made diagnosis. b. (32) (33) (2h) (27) Doctor/facility who made diagnosis. YEAR i—i—ri—i r ~T (16) (17) (30) (31) TO b. (34) (35 ) Doctor/facility who made diagnosis. Name Name Name Address_ Address Address C/S C/S C/S e. Name of h o s p i t a l • e. Name of h o s p i t a l . e. Name of h o s p i t a l . Name Nnmr Naim- Address Address^ Address_ C/S C/S C/S Q.36. Medical Providers — Cancer c. Month/year first diagnosed 1 1 d. c. Month/year f i r s t diagnosed MONTH YEAR 1 I I 1 1 1 l-l 1 1 ( 36) ( 3?) &8 ) ( 39) D o c t o r / f a c i l i t y where f i r s t diagnosis made: Part 3 Part 2 Part 1 1 I d. c. Month/year first diagnosed MONTH YEAR 1 I I 1 1 l l I l l U4) ( 4 5 ) (46) ( 4 7 ) D o c t o r / f a c i l i t y where f i r s t d i a g n o s i s made; MONTH YEAR i i i r i i i (52) i (53)1-1 ( 5 4 )i ( 5 5 ) i d. D o c t o r / f a c i l i t y where f i r s t d i a g n o s i s made: Name Name Name Address Address Address C/S C/S C/S e. D o c t o r / f a c i l i t y last consulted. e. D o c t o r / f a c i l i t y last consulted. e. D o c t o r / f a c i l i t y last consulted. Name Name Nnmo Address Address Address C/S C/S C/S f. Mnnth/ypar last c o n s u 1 1 od . 1 1 MONTH YEAR 1 I I 1 1 1 l-l 1 1 ( 4 0 ) (41 ) (42 ) (43 ) f. Month/year last consu 1 t rd . 1 I MONTH 1 l l (4 a ) ftq ) I I - YEAR 1 l rI ( >() OH ) f. Mcinrh/yr.ir I n s t f O I I K I l l I I'll . F l T I MONTH l 1 YEAK T1 1 ' T l - I l l 06 ) ( 57) (58 ) ( 59) �CARD Q.3d. MONTH I 1 1 1 l-l ( 60) ( 61) d. Part 6 Part 5 Month/year first diagnosed 1 1 812039 M e d i c a l Providers --~ Cancer Part 4 c. 136-137 c. YEAR ' 1 1 1 1 ((,1) (63) Doctor/facility where first diagnosis made: c. Month/year first diagnosed MONTH T I l 1 YEAR I I 1 l - I l (68) (69) d. Month/year first diagnosed 1 l 1 I ( 70) (71) Doctor/facility where first diagnosis made: d. MONTH YEAR 1 I I 1 1 l l I l l (12) (13) (14) (15) Doctor/facility where first diagnosis made: Name Name Name Address Address Address C/S c/s e. . e. Doctor/facility last consulted. ... , C/S e. Doctor/facility last consulted. Doctor/facility last consulted. Name Name Name Address Address Address C/S C/S C/S f . Month/year last consulted . f. f. Month/year last consulted. M6NTH ! 1 MONTH I l l YEAR 1 1 1 1 l-l (64 ) US) 1 1 1 1 ( 66) ( 6>) 1 I (72 ) (7.3) Q.37 Medical Providers — Leukemia b. Month/year f ir_et diagnosed MONTH YEAR 1 ! 1 1 1 1 l-l 1 1 (20 ') ( 21)' (22 ) ( 23) 1 1 c. Doctor/facility where first diagnosis made: Name Address C/S f. Doctor/facility last consulted. Name Address C/S g. Month/year last consulted . MONTH 1 1 1 J YEAR I I ! 1 l-l 1 - . 1 1 1 - YEAR 1 I l 1 l (ft,} (75) Month/year last consulted. i 1 MONTH f 1 (16) (17) ri L-l YEAR i 1 (18) (19) i 1 �CARD 137 M f d i c . - i l P r o v i d e r s — OTHER MEDICAL CONDITIONS o.'lH. THYROID DIABETES h. b. First toldh a d : 1 I e. 812039 First told had: MONTH YEAR 1 I I 1 1 l l I l l ( 28) < 29) ( 30) ( 3l) D o c t o r / f a c i l i t y where d i a g n o s i s made: ANEMIA MONTH 1 1 1 1 b. YEAR 1 1 II H MONTH (60) ( 6 1 ) D o c t o r / f a c i l i t y where d i a g n o s i s made: YEAR ( 6 2 ) (63) i i i r i i 1 1 1 - 1 1 1 1 1 (44) (45) (46) (47) e. First t o l dh a d : e. D o c t o r / f a c i l i t y where diagnosis made: NjiriP Name Name Ad d re s s Address Address C/S C/S C/S f>. Doctor last c o n s u l t e d : 1 I h. g. Doctor last consulted: MONTH YEAR 1 I I 1 1 l l I l l <32> (J3> (j4> <3b> 1 1 h. D o c t o r / F a c i l i t y last consulted. MONTH 1 1 YEAR 1 1 1 1 l-l Us) (49) (;>o) (51) g. Doctor last c o n s u l t e d : 1 1 D o c t o r / F a c i l i t y last consulted. 1 I h. MONTH YEAR 1 1 1 1 1 l l I l l (64) (65) ( 6 b ) .(67) D o c t o r / F a c i l i t y last consulted. Name Name Name Address Address Address C/S C/S C/S ENLARGED LIVER HEART CONDITION b. b. First told had: 1 I e. First told h a d : b. First told h a d : MONTH YKAR 1. 1 1 1 1 1 1 1 l-l 1 1 (52) (53) (54) (55) MONTH YKAR 1 1 1 1 1 l l I l l (36 ) ( 37) ( 38) ( 39) e. D o c t o r / f a c i l i t y where diagnosis made: JAUNDICE D o c t o r / f a c i l i t y where diagnosis made: 1 1 e. MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (68) (69) ( 7 0 ) (71) D o c t o r / f a c i l i t y where diagnosis made: Name Name Name Address Address Address C/S C/S C/S g. Doctor last c o n s u l t e d : T I MONTH l ! < 4(j» < 41> h. YEAR F 1 1 l - I l g. 1 l Doctor last consulted: I I I h. I - <56 ) (57 ) <M ) < 4 3 > D o c t o r / F a c i l i t y last consulted. MONTH I l l YEAR I I l g. I l Doctor last consulted: 1 1 ( 58) ( 59) D o c t o r / F a c i l i t y last consulted. h. MONTH YEAR 1 I I I 1 1 l - l I I ( 72) ( V3) • ( 74) ( 75) Doctor/Facility consulted . Name Name Name Address Address Address C/S C/S C/S 'J2 last �CARD 1 38 i.i. J K . M e d i c a l 1 ' r u v i d e r s -- OTHER M E D I C A L C O N D I T I O N S ( C O N T I N U E D ) C I R R H O S I S OK THE L I V E R MKI'ATITIK h. h. 1- i r s l l o l d h a d : MONTH r "i I l l " 0 ^ 5 (1 i) P. 812039 YEAH i i i - First told had: MONTH 1). First told had: YEAR T i ri r i 1 ( 2 8 1) (29 ) - (30) (31 ) 1 1 1 1 i I l l (I/, j (lrO e. H o r r o r / f a c i l i t y where d i aj'nos i fl made : INTESTINAL PARASITES D o c t o r / f a c i l i t y where diagnosis made: 1 I e. MONTH 1 I I (V, ) ( 4 5 ) I t - YEAR 1 I I (46 ) ( /|7) 1 I D o c t o r / f a c i l i t y where diagnosis made: Name Name Name Address Address Address c/s C/S C/S H. i i . i t i n r last consulted: MONTH I I I I ( l < > ) (17) li. g. Doctor last YEAR I I l-l I I ! I 1 1 ( 1 8 ) (l9l Doctor/Facility last (•(insulted. h. consulted: £. Doctor last c o n s u l t e d : MONTH YEAR 1 I I 1 1 1 l-l 1 1 ( ti ( 3 3 ) (3-4) (35) D o c t o r / F a c i l i t y last consulted. MONTH YEAR 1 I I 1 1 I I I I I ( 4 8 ) (49 ) (50) (51 > 1 I h. Doctor/Facility consulted. Name Name Name Address Address Address C/S C/S last C/S CALL BLADDER h. Fi rst told h a d : I I e. h. MONTH YEAR . I I I 1 1 l l I l l (20) (;>]} (?/>5 ( 2 3 ) D o c t o r / f a c i l i t y where diagnosis made: OTHER RESPIRATORY OTHER LIVER CONDITION First told had: MONTH b. F i r s t told h a d : YEAR in i i i . r- i i i i ii (36) (37) e. 1 1 (52) (53) (38> (39) D o c t o r / f a c i l i t y where d i a g n o s i s made: MONTH 1 1 e. II l-l YEAR 1 1 1 1 ( 5 4 ) (55) D o c t o r / f a c i l i t y where diagnosis made: Name Name Name Address Address Address C/S C/S C/S t>. Doctor l a s t c o n s u l t e d : 1 1 h. p. MONTH YF.AR 1 I I 1 1 1 l-l 1 1 <24> <25> <2b) (27> Doctor/Facility last Doctor l a s t c o n s u l t e d : MONTH 1 1 h. consulted. g. MONTH YEAR 1 1 1 1 l-l (56) ( 5 7 ) (58) ( 5 9 ) YEAR 1 1 ( ! 1 1 l - l I I (40) (/,i) ( 4 2 ) (^3) Doctor/Facility last consulted. Doctor last c o n s u l t e d : h. D o c t o r / F a c i l i t y last consulted. Name Name Name Address Address Address <:/s C/S C/S �. Q.'iH. Medical Providers — CARD 138 OTHER MEDICAL CONDITIONS (CONTINUED) OTHER MAJOR CONDITIONS b. First told had: MONTH YEAR (60) (61) (62) (63) e. Doctor/facility where diagnosis made: Name AdJress_ C/S g. Doctor last consulted: MONTH YEAR ~(6A) (65) (66) (67) h. Doctor/Facility last consulted . Name Address Q.M. Medical Providers — Acne b. Last consulted doctor TI MONTH l-l YEAR Ir T n Ii I (68) (69) fro ) (71) c. Doctor/facility last consulted: Name Address_ C/S 812039 �CARD Q.4? — McdiCii] Providers B. EASIER BRUISING OF SKIN PATCHES OF SKIN CHANCE COLOR K. 812039 Doctor/facility whore diagnosis made: e. Doctor/facility where diagnosis made: C. SKIN EXTRA SENSITIVE e. Doctor/facility where diagnosis made: Name* Name Name Address Address Address c/s C/S C/S f. f. Month/ypar diagnosis made: MONTH 1 1 1 1 K. YEAR 1 1 I I l-t Month/year diagnosis made: MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (TB ) ( 2'i1 ( 30* ( 3p 1 1 g. Doctor/Facility last consulted. f. Doctor/Facility last consulted. Month/year diagnosis made: 1 I g. MONTH I II l l ^4 ) ( 4bJ YEAR 1 1 I l l ( ltd ( 4?' Doctor/Facility last consulted. Name Name Name Address Address Address C/S C/S C/S h. h. Month/year last consulted: Month/year last consulted: h. TihV'n"?)" OiO (I'J) MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (32) (33) (34) (35) D. RASH ON BACK E. EXCESSIVE HAIR GROWTH 1 I e. MONTH 1 l l I I - YEAR 1 I l 1 l e. Doctor/facility where diagnosis made: Doctor/facility where diagnosis made: Name Name Address Address C/S C/S f. Month/year diagnosis made: f. MONTH YEAR '1 '1 1 '1 " 1 1 l-l 1 1 (20) (21) (22) (23) g. Doctor/Facility last consul ted . Month/year diagnosis made : 1 1 h. MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (36) (37) (38) (39) Doctor/Facility last consulted. Name Name Address Address C/S C/S h. Month/year last consul red: MONTH 1 M. Month/year last consul ted : YEAR l-l J MONTH 1 I l l YEAR - I l l Month/year last consulted: 1 I MONTH YEAR 1 1 1 1 1 l l I l l (48) (49) (50) (51) �CARD 139-140 812039 Q.43 — Medical Providers A. NUMBNESS IN LIMBS h. B. C. TINGLING IN :LIMBS >•? D o c t o r / f a c i l i t y where diagnosis made: h. BURNING IN LIMBS D o c t o r / f a c i l i t y where diagnosis made: h. D o c t o r / f a c i l i t y where d i a g n o s i s made: Name Name Name Address Address Address C/S C/S C/S i. Month/year diagnosis made: MONTH I I i. l 1 Month/year diagnosis made: YEAR 1 I ! l - I l ( 52) ( 53) i. ( 54) ( 55) 1 l D o c t o r / F a c i l i t y last consulted. 1 I i. i. MONTH YEAR 1 II 1 1 l l I l l (68) (69) (70) (71) D o c t o r / F a c i l i t y last consulted. Month/year d i a g n o s i s made: 1 I j. MONTH YEAR 1 1 1 1 1 l l I l l (16) ( 1 7 ) (18) ( 1 9 ) D o c t o r / F a c i l i t y last consulted. Name Name Name Address Address Address C/S C/S C/S k. Month/year last consulted: I I k. MONTH YEAR 1 1 1 1 1 l l I l l ( 56) ( 5 7 ) 68 )( 59) 1 I D. PERSISTENT ACHES IN LIMBS h. Month/year last consulted: D o c t o r / f a c i l i t y where diagnosis made: k. MONTH YEAR 1 1 1 1 1 l l I l l (73(71) (74) (75) E. REDUCTION IN CRIP STRENGTH h. D o c t o r / f a c i l i t y where diagnosis made: Name Name Address Address C/S C/S i. Month/yeflr d i a g n o s i s made: 1 I j. i. MONTH YEAR 1 I I I I l l I l l ( 60 (61 ) (62 )( 6.3) Doctor/Facility last consulted. Month/yea! diagnosis made : 1 MONTH 1 1 1 , YEAR 1 1 i (76 ) i (77 ) i (- 7tS>i (79 ) i i j. Doctor/Facility last CO' j u l t e d . Name Name Address Address C/S C/S k. Month/year last consulted: 1 I MONTH YEAR T M 1 1 l l I l l r (64) (6 >) (66) ( 6 7 ) k. Month/year last consulted: 1 I MONTH 1 l l (12) (13) II - YEAR 1 I l (14) (15) 1 l Month/year last consulted: 1 1 MONTH YEAR 1 1 I 1 1 1 . 1-1 1 1 f20) ( 2 1 ) (22)123) �CARD 812039 Respondent answer to Q.60 was . . . Yes No „ — _ — — — — ——— . — ——• — » . . —— — — — ™ _ _ _ _ _ _ - . — • • - « — . — _—.^ ^ . . — — w —._ __ *«•,_—T . ^ — __«._«.^__ _ ^ W ^ w __ H M_ **_ M M _ K H K. , • fl(MM • ' » « . —.^ Record "Yes" or "5" 62a. Did you lose your appetite 63a. Check Box If Any Recording In Quest ion group Did you lose weight without trying to — as much as ten pounds altogether , 64a. Did your eating increase so much that you gained ten pounds altogether , fi'ja. Did you have trouble falling asleep, staying asleep, or waking up too early «. °"\ : I 66a. Were you sleeping too much? I 67a. Did you feel tired out all the time? 68a. Did you have to he moving all the time — that is, you couldn't sit still and paced up and down?..... F Did you talk or move more slowly than is usual for you? 69a. T ____ 70a. Was your interest in sex a lot less than usual? 71a. Did your thoughts come much slower than usual or seem mixed up? 73. Did you feel worthless, sinful, or guilty? 74. Did you think a lot about death — either your own, someone else's, or death in general? *j Did you have a lot more trouble concentrating than is usual for you?. 72a. r 75. Did you feel like you wanted to die? 76. Did you fee] so low you thought of committing suicide?._ 77. Did you attempt suicide? I r T : /" 97 ^ T T 1 �CARD 812030 138a. If you were asked, would you be willing to have a physical examination at a time convenient for you? Yes. No.. -<59( -1 ~-2 (ASK Q.138b) 138b. What is your reason for not wanting to have the examination? 138c. Under what conditions would you be willing to have an examination? (60 (62( (63j (641 (65( (66( (67( (8 6( (70( �CARD OOf. 141. Additional Civilian Training Programs (0-6) 4 Hi Prog rum h. 812039 5th Program For what k i n d of work was ybufr Illix'" civilian training program preparing you? f. For what kind of work was your next c i v i l i a n training program preparing you? For what kind of work was your next c i v i l i a n training program preparing you? (15 ( 05 ( 05J_ (16 ( (16 ( (10 ( (17 ( (17 ( Jill (1H( (]«( HILL CH U9(. U9( c . In what month and year did you start this training? MONTH 1 1 I 1 ! l-l 1 1 1 I h. MONTH YEAR 1 II 1 1 1 hi 1 1 C., ) (25 ) (26 ) (27 ) '' Have you participated in any other c i v i l i a n iob training program that prepared you for a major change in your occupnt ion? Yes . ( 2H No. -1 (ASK Q. l/.lf) -2 (RETURN TO Q.7) O/. MONTH 1 l l I I - (20 ) (21 ) (22 ) (23 ) In what month and year H i d you complete t h i s training? 1 1 e. 1 1 YEAR 1 I l MONTH 1 II 1 l-l (24 ) (25 ) MONTH T (20) (21 ) 1,. (22 ) (23 ) In what month and year did you complete t h i s training? 2JONTH_ YEAR 1 J ( 26) (27 ) Have you p a r t i c i p a t e d in any other c i v i l i a n job t r a i n i n g program that prepared you lor a major change in your occupa t i on? Yes.GM^ -1 (ASK Q. l./ilj) No ' ~'i (RETUKN TO 0.7) _ 79"-80" YFAK T I ( 22) (23 ) YEAR 1 In what month and year did you start this training T 1 l In what month and year did you complete this training? 1 1 i. k. In what month and year did you start, this training? YEAR I (20 ) (21 ) »l. g. I j -J I i (2/4 )" (25 ) (26 ) (27 ) Have you p a r t i c i p a t e d i n a n y other c i v i l i a n job t r a i n i n g progratn that prepared you tor a major change in your occupa t ion? Yes.CH ( -1 (RETURN TO Q.7) No -2 _ 79-80 �CARD 007 Arid i I i on,-i I M i l i t a r y Jnh T r a i n i n g Programs (<l.7) !-. 6 t h Program Mil I'l-ogi-aiii •'tflt 1'rogr.im Kui wh.it kind o( work \ j , \ ' - your no.xt m i l l t n r v I r.'i i n i nu progr.-iin pi • • • - g. Kor w h a t k i n d ot work was your n e x t m i l i t a r y t r a i n i n g program preparing you? PHI' ing y»u? 1.. For what kind ot work w.is your n e x t m i l i t a r y t r a i n i n g program preparing you? <15< <16< (16( 17< (17( (17( (18( Q8( (19( i s the AFSC job? (16< IiSL_ What that ( 15( ( c. (15( (19( (19( for h. What i s the AFSC that joh? fen ( d. T n wh.it month and year did ynu > ! t a r t t h i s training? MONTH I 1 I I l-l (21 ) (22 ) e. 1 1 i. In what month and year did you s t a r t t h i s training? MONTH 1 1 1 I (23 ) ( 24) In what month and year did you complete this j. f. MONTH YEAR 1 I I 1 1 l l I l l (25) (26) (27) ( 2 8 ) Have you p a r t i c i p a t e d in any other m i l i t a r y job training program that prepared you for a major change in your occupation? Yes.(M( -1 (ASK Q. I42j.) What i s the AFSC that joh? for (20( l l (21 ) <22 ) I I - MONTH 1 I k. 1 MONTH I l (23 ) (24 ) 1 l 1 1 l (26) I I - 1 I (27) l (28) -1 -2 o. I I l-l 1 1 ( 23) ( 24) In what month and yeardid you complete this MONTH 1 I YEAR 1 l I I - l (25 ) (26 ) p. 1 I 05 79-80 JO I) l 1 l (27 ) (28 ) Have you participated in any other military job training program that prepared you for a major change in your occupation? Yes.(29( -1 fRpTMHN TO Q . R 1 Q.8) 04 79-80 1 1 training? 1 l (ASK Q.142]) (RETURN TO Q.8) YKAR 1 1 Gl ) C2 ) Have you p a r t i c i p a t e d in any other military joh training program that prepared you for a major change in your occupation? es.<29( 0 In what month and year did you s t a r t this training? YEAR 1 l (25) n. YEAR 1 In what month and year did you complete this training? training? 1 I m. (20< YRAR 1 1 for 06 79-80 �CARD QQ8 812039 143-145 A d d i t i o n a l jobs (Q.8-13) Seventh Job Eighth Job Ninth Job 143a.In what month and year did you start your next job that lasted three months or longer? 144a.In what month and year did you start your next job that lasted three months or longer? 145a.In what month and year did you start your next job that lasted three months or longer? I 1 MONTH t 1 M l-l (i5) (16 ) YEAR i 1 i I 1 I (17 ) (18 ) MONTH YEAR 1 I I 1 1 l l I l l (15 ) (16 ) (17 > OS ) 1 I MONTH YEAR 1 1 1 1 1 l l I l l 05 ) ( 16) ( 17) ( 18) b. What (is/was) the name of your employer? 1 RECORD IN S.R.B. - PC 1 I b. What (is/was) the name of your employer? b. What (is/was) the name of your employer? c. (Is/Was) the job fulltime or part-time? c. (Is/Was) the job f u l l time or part-time? c. (Is/Was) the job fulltime or part-time? Full time.. (19 ( IRECORD IN S.R.B. - PG i i -1 Full time. . 1 ( (9 IRECORD IN S.R.B. - PG i I Full time. .(19 ( -1 -1 d. What kind of business is thaf — what (do/ did) they make or do there? d. What kind of business is that — what (do/ did) they make or do there? d. What kind of business is that — what (do/ did) they make or do there? e. What ally what your 1 RECORD e. What ally what your e. What ally what your (do/did) you actudo on the job — (are/were) some of main duties? IN S.R.B. - PC 1 I (do/did) you actudo on the job — (are/were) some of main duties? IKECORD IN S.R.B. - PG 1 | lHAND RESPONDENT CARD "B"] f. Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. IHAND RESPONDENT CARD "B"! (WRITE IN NUMBER) (WRITE IN NUMBER) I 1 (20) ! | 1 I (21> g. In what month and year did this job end? I 1 MONTH YEAR 1 1 [ 1 1 1 .1-1 1 1 (23) (24) C5 ) (26) Current job..(27< -1 (RETURN TO Q.14a) h. What was the main reason you stopped working on your job? f. Please .look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. 1 1 (20 ) ( U 2 g. In what month and year did this job end? 1 I MONTH YEAR 1 I I I I l l I l l (23) (2/0 (25) (26) Current job..(27( -1 (RETURN TO Q.lAa) h. What was the main reason you stopped working on your job? IHAND RESPONDENT CARD "B"| f. Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. (WRITE IN NUMBER) 1 1 I | I I (20 ; (21) g. In what month and year did this job end? 1 I MONTH YEAR 1 1 1 1 .1 l l I l l (23) (24) (25) (26) Current job..(27( (RETURN TO -1 Q.14a) h. What was the main reason you stopped working on your job? (28( (29( ( 29( 07 7<j"-ft() IRECORD IN S.R.B. - PC i 1 (28( (2H( (ASK Q.144a) (do/did) you actudo on the job — (are/were) some of main duties? (ASK Q.145a) (29( (ASK __0_8 7 '"«) )"( Q. 146a) 09 T'i-'HO �CARD 008 H12039 A d d i t i o n a l Jobs (8-13) Cont'd. Trnth Job Eleventh Job 146a.ln what month and year did you start your next job that lasted three months or longer 1 1 Twelfth Job 147a.In what month and year did you start your next job that lasted three months or longer 148a.In what month and year did you start your next job that lasted three months or longer? MONTH YEAR 1 1 1 1 i 1 l-l 1 1 0'> ) 07. ) ( 17) ( IH) 1 I MONTH YEAR 1 II 1 1 l l I l l < 15) ( 16) ( 17) ( IB) MONTH ! I l I YEAR l l 45 ) (16) - l I l I I l (17) (18) h. What (in /was) the name of your employer? iHK.COHli !M S.R.R. - IT. 1 b. What (is/was) the name of your employer? b. What (is/was) the name of your employer? IRECORD IN s.u.u. - PC; i IRECORD IN S.R.R. - PG i i c. (Is/Wits) thi- job l u l l lime or part-time? c. (Ifi/Wns) the job f u l l time or part-time? c. (Is/Was) thf job l u l l time or part-limp? F u l l t i me , . ( ' '\ Full time..( l«( -1 Full time..( I9( -1 -1 d. '.'hiit k i n d of business is that -- what (do/ did) they make or do there? d. What kind of business is that — what (do/ did) they make or do there? d. What kind of business is that — what (do/ did) they make or do there? e. What ally what your ! RECORD e. What ally what your (RECORD e. What ally what your (do/did) you actudo on Che job — (are/were) some of main duties? IN S.R.B. - PG 1 1 (do/did) you actudo on the job — (are/were) some of main duties? IN S.R.B. - PG 1 1 ilUvND RESPONDENT CARD "B" f. Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. IHAND RESPONDENT CARD "B"| (WRITE I N NUMBER) (WRITE IN NUMBER) 1 1 (20) 1 1 I | (21) g. In what month and year did this job end? 1 I MONTH YEAR 1 1 1 1 1 l l I l l (2J) <24) (25) (26) Current job..(22J -1 (RETURN TO Q.lAa) h. What was the main reason you stopped working on your job? f. Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. I | | I | | (20) ( 21) g. In what month and year did this job end? 1 I MONTH YEAR 1 1 1 1 1 l l I l l (23) (24) (25) (26) Curre--; job..(27( -1 (RETURN TO Q.14a) h. What was the main reason you stopped working on your job? (28( (ASK Q.147n) Id <7¥< (28( (do/did) you actudo on the job — (are/were) some of main duties? IRECORD IN S.R.B. - PG i 1 IHAND RESPONDENT CARD "B"| f. Please look at this card and tell me the number which best describes the kind of industry you (work/ worked) in. (WRITE IN NUMBER) I i I I 1 1 (20 ) (21 ) g. In what month and year did this job end? 1 I MONTH 1 II l l (23 ) (24 ) Current job..(27( -1 YEAR 1 1 I l l (25 ) ( 26) (RETURN TO Q.14a) h. What was the main reason you stopped working on your jot)? (28 ( (29 ( (29 ( (ASK <).148a) 11 (RETURN TO Q.14a) 1.' �CARD 0 1 f> A d d i t i o n a l P e r i o d s ol M i l i t a r y S e r v i c i ' \ l i . l f i ) l i i w h M m o n t h ,'iiut y e a r did you next enter thr Armed f o r c e s ? MONTI! 1. MONTH YEAR i—r—r T <. In what montli ami y.';ir did you next rnt.rr tlur Armed Forces? YEAR I ! TTeTTTTf I Air Force.(j^jj Navy Army Marines Coast G u a r d . . . What branch of the m i l i tary war, that? -1 -2 -3 -4 Air Force. ( j _j Navy Army Marines Coast Guard... -5 Were you discharged or separated from the (BRANCH OF S E R V I C E ) ? "T I ' l-l h. !. . -1 -2 -3 -k -5 Were you discharged or separated from the (BRANCH OF SERVICE)? Wliat branch of the m i l . t.-iry was that? Air Force. (Ulj Navy Array.. . Marines Coast Guard... m. -1 (ASKQ.i) Discharged/ separated.( 1'X Still in (MILITARY) -2 (RETURN Still in (MILITARY) d. S t i l l in (MILITARY) TO Q.16) MONTH MONTH YEAR "(20) (21) e. (22 ) (23 ) Following your separation j. or discharge in (DATE IN "d"), did you reenter the Armed Forces? Yes..(24( -1 (ASK Q.149f) No __-2 (RETURN TO Q.16) _0_4 "79- 80 MONTH YEAR i r i i" (20 i—n—ii ) (23 ) I (20 )I (21) ) (21 ) (22 T Following your separation o. or discharge in (DATE IN "i"), did you reenter the Armed Forces? Yes..(24( -1 (ASK Q.149k) No __-2 (RETURN TO Q.16) __05_ 79-80 11)3 -/i _c -1 (ASK (J.n; _-2 V RETURN TO Q . 1 6 i In what month and year were you discharged/ separated from the (BRANCH OF MILITARY)? In w h a t m o n t h and year were yon d i s c h a r g e d / s e p a r a t e d from the (BRANCH OF M I L I T A R Y ) ? In what month and year were you discharged/ separated from the (BRANCH OF MILITARY)? -1 -2 -3 Were you discharged or separated from the (BRANCH OF SERVICE)? Discharged/ Discharged/ separated.(19 ( separated. <lj)_(___-l (ASKQ.d) -2 (RETURN TO 0-16 YEAR T "06)077 What branch of the m i l i tary was that? c. In what m o n t h nnd ye'.'ir did vou lU'xl rnter thu Armed Forces? YEAR Following your srparai ion or discharge in (l)ATh I N "n"), did you reentor theArmed Forces? Y e s . . ( 2 A ( -1 ( R E T U R N TO No -2 Q.16) _-. 7'J-80 �CARP OK, 150. Additional Count ri PS Stationed (Q.16/17) Srvrn t h Count ry a. 813030 Eighth Country Wli.it was the next count r t h d l you were stationed in for more than 10 days w h i l e on a c t i v e duty? f.. Ninth Country What was the next countr that you were stationed in for more than 90 days while on active duty? m. What was the next country that you were stationed in for more than 90 days while on a c t i v e duty? ( (14-15) (14-15) J4--.m (RECOUP COUNTRY HERE AN!) IN (RECORD COUNTRY HERE AND IN (RECORD COUNTRY HERE AND IN, S.H.It. I'd '} AND CONTINUE) S.R.U. PC ? AND CONTINUE) S.R.B. PC 2 AND CONTINUE) No o t h e r s . ( l f i ( -1 (RETURN TO Q.18) h. No others. (16 ( -1 (RETURN TO Q.18) In what month and vear h. did you begin and end act i v e duty in (COUNTRY)? No others. (16 ( In what month and year n. did yon hegin and end act i v e duty in (COUNTRY)? -1 ( RKTUKN TO Q.18) In what month and year did you b e g i n and end active duty in (COUNTRY)? KEG IN MONTH YEAR 1 1 M 1 1 I l l I l l < 1 7 ) (IK) (19) (20) BEGIN MONTH YEAR 1 1 1 [ 1 1 I l l I l l (17) (18) (19) (20) BEGIN MONTH YEAR 1 ! I I 1 1 I l l I l l (17 ) (.18) (19 ) (20 ) END MONTH YEAR 1 1 I I 1 1 I l l I l l (21) (22) (23) (24) END MONTH YEAR 1 1 1 1 I I I l l I l l (21 ) (22) (23) (24) END MONTH YEAR 1 1 1 1 1 1 I l l I l l (21) (22) (23) (24) Current .. .(25( Current ...(25 ( Current . . . (25 ( -1 -1 -1 c. What specific job assign- i. ments (do/did) you have in (COUNTRY)? Can you f>ivi> me the AFSC? What specific job assign- o. ments (do/did) you have in (COUNTRY)? Can you give me the AFSC? What specific job assignments (do/did) you havp in (COUNTRY)? Can you give me the AFSC? 1. (26-28) 1. (26-28) 1. (26-28) 2. (29-31) 2. (29-31) 2. 3. (32- 34) 3. (32-34) 3. (32-34) d. (Do/Did) your duties in j. (COUNTRY) include flying? (Do/Did) your duties in p. (COUNTRY) include flying? (Do/Did) your duties in (COUNTRY) include flying? Yes.(35( No p. Yes.(35( No -1 -2 How many flight hours d i d you 1 og while in (COUNTRY)? 1 I I 1 I ! Hours (36) (37) (38) I I How many flight hours did you log whi le in (COUNTRY)? 1 1 1 1 .(39( What specific letter and numerical designat ion(-s) did each aircraft have? L. q. 1 I Hours (36) (37) (38) -1 -2 How many flight hours did you log while in (COUNTRY)? 1 1 1 I 1 1 1 1 1 1 Hours (36) (37) (38) Other (SPECIFY) .(39( -1 (29-31) Yes.(35( 1o -1 -2 Other (SPECIFY) Other (SPECIFY) f. k. - .(39( -1 What specific letter and numerical designation(s) did each aircraft have? r. -1 What specific letter and numerical designat ion(s) did each a i r c r a f t have? 1. (40-43) 1. (40-43) 1. (40-43) 2. (44-47) ?. (44-47) 2. (44-47) 3. (48-51) 3. (48-51) 3. (48-51) l'i.>-y>) t>. A. ( 'i.'-V,) ,. 07 7'} -Ml . C->2-'j':) (ASK Q. 1r >l.-i) (ASK 0. 110m) (ASK <>. 1 M)r 1 08 7~')"-HU 04 79-"«0 �CARD |S|. El cvfilth Count ry S.K.B. PC 2 AND CONTINUE) Nu' oi he i :, .(lft( I n w h a t iniiiil h and y e a r li. did you begin and end act i v e d u t y in ( C O U N T R Y ) ? (RETURN TO 1,1. 18) lu u h a t m o u t l i and y e a r d i d you b e t f i n and end ac I i v c rintv i n ( C O U N T R Y ) ? J YEAIi ..-...,-. MONTH T 1 J I 1-1 I [ (1?") ( 1 8 ) (I9)~(~2<) In w h a t inontli cnul y e a r did vou he-gin and end act ive Viui y i n (COUNTRY )? ] 1 T-~] " 1 [-1 L [ I I I'.'l ) (2(1) T~~ I ( 2 3 ) (2 h ) ' (21 i -1 (2.' ) (17 ) (18 ) YEAR j [. J -1 (32-34: Current. .. (25 ( Wliat s p e i i f i c job assignmerits ( d o / d i d ) you have in (COUNTRY)? Can you K i v e me the AEKC? -1 What s p e c i f i c job assignments (do/did) you have in (COUNTRY)? Can you give me the AFSC? 1. (26-28) J^iliL 3. YEAR r ! r i—T T I (21 )I (22 )l-l (2'f )I (24 ) I 1 (2n-28) 2. (14 ) (20 ) MONTH (23 ) (24 ) Current. . . (25 ( What s p e c i f i c job a s s i g n - l i. ments ( d o / d i d ) you have ! in (COUNTRY)" Can you \ give me the AI'SC? YEAR 1' I" ' '] """T l-l I I END MONTH - T .. 1 r , L_ J_ J-1 L-J J_ J , . .Current .. .(25 ( I!_EC 1 N MONTH KN1' T (21 ) (22) c. u. YEAH (17 ) ( 1 8 ) F..N1) ___ YEAK MONTH T J -I (RETURN . I'd Q. 1H) 11 EC:: 1 N JJW: IN ] (I4-15J f RECORD COUNTRY' TiiVKiT" AND i N No , , | | , , T S . < l f ' < _ _ . - I (RETURN MONTH .. . j Wlial was (lie next country t h a t vou were stationed i n for more than 90 days w h i I e on a c t i ve duty? _ TO o . 1 8 ) 1..-..., in. (KF.COKll COUNTRY HERE AND IN S . R . I ) . PC; 2 ANII CONTINUE) S.K.l'.. PC :> AND CONTINUE) o t h l ' l :, . ( ll,( Twelft h Countiy What w a s t h e n e x t c o u n t r that you were s t a t i o n e d in for mo r i- I him 1'0 ilnys wh i 1 e on ,'ic t i ve d vit y ? What W.IK the next c o u n t r y ( Inn von were M at inned in I'M more t h a n ('U day; wh i 1 r mi .'ic t i ve d u t y ? T'KKCOKD" couN'tWiTERE'ANJ') IN' lio HI 2039 A.l'h t innvt1 I'oiiiit r i i ' S SI ,11 loiifd ( 0 . 1 ( , / i 7 ) T i ' i i t h COIIII_M-V ;i. Old d. (Do/Did) your d u t i e s in (COUNTRY) i n c l u d e t l y i n g ? (29-31) 3. 3. (32-34) j. ( D o / D i d ) your ilurips in p. (COUNTRY) include 1 l y i n R ? (Do/Did) your d u t i e s in (COUNTRY) i n c l u d e flyinfi? Yes.(35J -1 No ' 7 - ' " - ~ z e. k. Yes. (35 ( No Yos.(35( How many f l i p t i t hours d id you .1 op. wh i 1 e in (COUNTRY)? T —r I 1 how many f l i g h t liours d i d you lop w h i l e in (COUNTRY)? q. I I How many flight hours d id you log w h i l e i n ( COUNTRY ) ? 1 1 1 Hours (3d ) (37 ) (38 ) I Hours Ou") (37 roe" Other ( S P E C I F Y ) (SPECIFY) Other . (_3«,._ - 1 f . -1 ~-2 i—T —r—r J 1 Hours (3d ) (37 ) ( 3 8 ) Other -2 Ni Wh a t :; pec i I i r 1 e 11 e t and I.. numerical desi f>nar i on( s ) d i d each a i re i n 11 h.we1.' (SPECIFY) . (J9(__-1 What s p e c i f i c l e t t e r a n d nunier i ca 1 de.s i ^nat i on( s ) < l i d each a i r c r a f t have? What s p e c i f i c l e t t e r and numerical d e s i p n a l i o n ( s ) d i d each ai re r a I t have? j£°~i!L (44-47) 2. C,8-r)l) (A;;K Q.I s i p ) (4 8- 'in 3. (>2-jiii (ASK Q. I S I m ) (RETURN TO Q. 18) i_o 79-80 '79-80 �CARD 020 152-1W. A d d i t i o n a l Marriages (Q.18-22) FOURTH MARRIAGE FIFTH MARRIAGE MONTH YEAH 1 1 l l I l l (] 5 ) (] (, ) (] 7 ) ( ] fl) b. What ( i s / w a s ) her rent f u l 1 name? MONTH 1 1 1 l-l <15 ) ( 1 6 ) cur- c. What was her m a i d e n name? 154a.In what month and year did you Ret married the sixth time? YEAR 1 1 1 1 ( 1 7 ) (18) b. What ( i s / w a s ) her current f u l l name? IRECORD IN S . R . B . PC 2 1 IRECORD IN S . K . B . PC 2 I c. SIXTH MARRIAGE I S T a . I n what- m o n t h a n d y e a r d i d y o u get m a r r i e d t h e f i f t h time? 1523. In w h a t m o n t h a n d y e a r did you get married t h e f o u r t h time? I 812039 What was her maiden name? MONTH 1 YEAR 1 l-l 1 • 1 ( 15) ( 16) . ( 17) ( 1 8 ) h. What ( i s / w a s ) her current l u l l name? IRECORD IN S . R . B . .PC 2 1 c. What was her m a i d e n name? IRECOKD IN S.R..K. PC 2 1 IRECORD IN S.R.B. PG ? 1 IRECORD IN S.R^B. PC 2 1 d. What is her date oi hirth? d . W h a t i s h e r d a t e oi birth? d. What is her date of hirth? •1 I MONTH MONTH YEAR 1 1 1. 1 1 l l I l l ( 1?) ( 20) ( 21) ( 22) 1 1 ( 1 9 ) (20) f. YEAR 1 1 1 1 (21) ( 2 2 ) 1 I Yes...(23( No -1 -2 Yes...(23( No -1 -2 -l -2 f. Did your w i f e ever have any p r e g n a n c i e s hy you which ended in a miscarriage? f . D i d your w i f e ever have any p r e g n a n c i e s by you w h i c h ended in a m i s c a r riage? D i d your w i f e ever have any p r e g n a n c i e s by you w h i c h ended in a m i s c a r riage? MONTH YEAR 1 I I I 1 l l I l l (19) (20) (21) ( 2 2 ) e. Have you ever had any c h i l d r e n by ( y o u r / t h i s ) wife? e. Have you ever had any c h i l d r e n by ( y o u r / t h i s ) wife? e. Have you ever had any children by ( y o u r / t h i s ) wi fe? Yes...(23( No 1 1 l-l Yes...(24( -1 (ASK Q.g) Yes...(24( -i (ASK Q . g ) Yes...(24( -1 (ASK Q . g ) No -2\ I ( S K I P TO -3 1 Q . 1 5 2 L ) No -2") V (SKIP TO -3 1 Q.153L) No -2") L ( S K I P TO -3 f Q.154L) J D o n ' t know g. When was t h a t ? Any o t h e r s ? ) 1st " 1 I i 1 2nd " ' 3 1 d 1 4th 1 1 1 1 1 : 1 l-l 1 1 1 2nd ( 31) ( 32) YEAR 1 I I 1 1 I l l I l 3 l r (33) (34) (35) (36) MONTH YEAR 1 I I V I 1 l-l 1 1 4th ( 3 7 ) ( 38) (39) ( 40) (GO TO Q . 1 5 2 h ) J g. When was t h a t ? Any o t h e r s ? ) (PROBE: MONTH YEAR 1 1 1 1 1 l l I l l 1st ( 2 5 ) ( 26) ( 2 7 ) (28) MONTH YKAR ( 29) ( 30) MONTH r D o n ' t know MONTH "" 1 l l (25) ( 2 6 ) MONTH 1 1 1 1 (29) (30) MONTH 1 1 d I l ~ < j j ) (34) MONTH 1 1 1 1 1 I g. When was t h a t ? Any o t h e r s ? ) (PROBE: YEAR 1 1 I l l ( 2 7 ) (28) YEAK I I I 1 l-l 1 1 (31) (32) YEAR 1 1 . 1 1 l I l (35) (36) YEAR II 1 1 l-l. 1 1 1 1 - ( 3 7 ) OH) (39) ( 4 0 ) (GO TO Q. I53h) .100 D o n ' t know 1st 2 n d 3 rl d 4th (PROBE: MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (25) (26) (27) (2B) MONTH YEAR 1 1 1 1 1 1 1 - 1 1 J (29) (iO) (31) (32) MONTH YEAR 1 I I 1 1 1 l - l I . I (33) ( 3 4 ) (T5) (36) MONTH YEAR 1 1 1 1 1 1 l-l ,,.[....^1 ( 3 7 ) (38) (39) (40) (GO TO Q. 1*541)) �CARD 14:1 ADDITIONAI, RECORDING SHKKT_j'OR_ N(>N-j. 1 VK-JUK'[MS_ FOURj'll MARRIAGE M 1 ST,SRta~jVC ElT:-~CJ7 ) 5 2 h F 1 FTII M A R R I A G E MLsCARRtATTES'' - Q'.* 1 53h 1 2nd l__ I_ |__ I | Months ~<1~2) (13) W a s n ' t t r y i n g (j_4( -1 D o n ' t know -2 ." (V?y (fir Wasn't trying (\l<( __ -] Don' t know ........ -3 J L ' Months (I1))' (If,)" W a s n ' t t r y i n g (J_7_J_ - I D o n ' t know ? 1 2nd j 'In) J I | Months j (]',') ( I f , ) " | W a s n ' t t r y i n g ( 17J -1 ' D o n ' t know -2 1 I T 1 T I | | Months " ( 1 2 )( 1 3 ) W a s n ' t trying (14( -1 Don 1 1 know -2 ~~T I 3rd SIXTH MARRIAGE GES - Q.~154h r~~T ?nd I Months 812039 3rd 1 I I I Months W a s 1 ' t t r y i n g (l_?i n Don I k n o w . . . . . . . . -I -2 1-------T Months "(181 ( 1 9 ) W a s n ' t I r v i n g ( 2 (K______- I D o n ' t know ........ ._.'..' 2 ' ttl ' [_ "_ "I Mo n th s 4th Wasn't trying C 2 0 ( _ _ -1 D o n ' t know ........ _ -2 (CO TO Q . l r > ? i ) 4th L I Months (18) ( 1 9 ) W a s n ' t t r y i n g (20( -1 D o n ' t know -2 (CO TO Q . I 5 4 i ) (CO Til Q. I', ii) TAFTER Q. ij> I A S K F O R KACH Were p i tlu'r (i^f you u s i n g b i r t h c o n t r o l at t h o t i m e she hecnme p r e g n a n t ? IFOR ANY "YES'r ASKlJ Please look at this card and cell mo all the numbers that apply to the types of birth control you used. 1st: _(21- • 2 2 ) 1st: _(21-22) 1st: J21-22) 2nd: _ < 2 3 - 2 4 ) 2nd : J23-24) 2nd: _(23-24) _(2 >-26) 3rd: _(25-26) (27-28) 4th : 3rd: r J25- 26) 3 r d : (27- 28) 4 t h : ftth: (GO TO 0.20L) (CO TO Q.21I.) F1FTH MAKRlAOK ST1LLIURTMS - Q. 153n KOUKTH MARRI ACK STILLBlrRTHS - Q. " 2ml (29) (30) W a s n ' t t r y i n g (31( D o n ' t know 3rd ] 3rd 4l.h -1 -2 J r [_ _ I M o n t h s ( 2 9 ) ("30)"" W a s n ' t t r y i n g ; (31( -1 D o n ' t know -2 [ Months Wasn't tryinp, (3lj_ Don't know ........ SIXTH MARRIAGE STILLBIRTHS - Q.154n i—i—r T T 2nd (27-28) (CO TO Q.22O J 2nd I r 3rd _[_ I Months I Wasn't trying (34_(_ Dor,' t know ........ I I Months (32) (33)" 1 Wasn t trying (W* _ -1 Don' t" k now -/ -1 -7 i 4th J __ I __ I M o n t h s OS) ( 6 ) 3"" Wasn't t r y i n g ( 3_7( __ -1 Don' I know ........ __ •-- I Months -1 -2 I __ I __ I Months (32F(33)~ W a s n ' t t r y i n g ( 3 4 ( -1 ^,1,'t k n o w ........ _ -2 T r—i 4th i I_ (35) I I Months HfiT" r I I Months (3ST (3(>r Wasn' t trv iny. (37( -1 Don' t kiinw -? Wasn't trying (3?( -1 D o n ' t know ........ _ -2 (U) TO (J. 1 VJo) (GO TO O..154o) (CIO TO Q.l!)2o) . - .. Were ( M t h i - r o j ym| MS i n g b i r t h c o n t r o l at t h c > r i m e she hoc nine p r c g n . m t ? |FOR"ANY""YKS""ASK : I I'leasi' look at t h i s card and t e l l nn- all thr numbers tliat a p p l y to t h e t y p e s nf b i r t h control von used. . ( 1H-.)')) I (40-41 ) 2nd: (4 .' 4 ' i ) 4th: r ( 4 4 - 4 >) 4 t h : (44-45) CO Til (;. 1 S J Ji)7 4th: (44-45) (GO TO ?)Tl54qT" �A D D I T I O N A L KKCORD1 NO SIIKK'I KOK NON-LI VK-lll KTIIS I M ..... I.:, l-.'.isn't t r y i n g DIMI' I know _.____.___ :inl 1 " I -I -? I " MOIII.IIS W . - i s n ' t t r y inc. ( S Don ' r k n o w 1 nlli Vn.l I I ' i" ' ' I I I Mnnllis T/,fiT~'a77' W . - i s n ' t I r v i n j ! (liK(_ Don 1 (• k n o w 3rd I I I I T I j' Mimllirt W a s n ' t t r y i n f ! (f>4( -I D o n ' t know . . . . . . . . _ - 2 'it. ti T I .S1XJJI M A K K I A i . K A BOUT i'ON:;"-" ~j. r>'i- (). ] 'i Is _-1 -? I I Months W a s n ' t t r y inc. 'IlU __ "1 Don't know.. ...... _~'2 T ~_ AIIOKTIONS" "T I Hr.'il.f' i - 1 ri'ii MAivi;i_A(;i; KOl'KTH MARKIAOI-: A H O K T I O N S ' " " < ; . 1 M'.-i I CARD I'i3 j I ?ml I I ' T/76T"Wi Y I I J I [ Months W a s n ' t t r y i n g (51( Don't know I I 1 I (GO TO iJ. -1 -2 -1 ~-2 I I~ Monllis ('> -1 -2 I 3rd I Mon D I N W a s n ' t I ry i nf, Don' I know I ! Miniili: W a s n ' t t r y i n f , (/i8( Don' t k n o w r~T I i I Wasn't trying ( Vi( __ -1 Don1 1 know ........ ___ -1' (GO TO Q . l W t ) (GO TO Q. [AFTER (j.l52t/l')3t/15At ASK FOB F.ACU A'SORTlpNlT Were either of you using birth control at the time she became pregnant? IFOR ANY "YES" ASK:I THANH' RESPONDENT CARD "c" I P l o a s e look a t t h i s card a n d C e l l m o a l l control you used. i t h e numbers t h a t a p p l y t o t h e t y p e s o f b i r t h 1st : _C>5-S6) 1st: 2nd: 1st : (57- r .8) 2nd : :trd: (61-62) il.li: (co TO y.iTfu Qtt 79-80 (61-62) ( G O T O O T l 53 .05 74-80 lllo 4th: (c;o To qr 06 79-80 �; . . . FIFTH MARRIAGE FOURTH MARRIAGE l ' i 2 h . H o w many m o n t h s d i d i t t n k p your w i f e t o hrcomf p r e g n a n t t h i s 1 inif? 1 1 I"" 1 1 I Months ('•1 ) ( 43 ' W a s n ' t t r y i n g (43 ( . -1 . 1 1 1 1 Months (41 ) ( 42) W a s n ' t t r y i n g (43 ( -1 1 I 1 I I 1 1 1 1 1 1 I I Weeks Yes.(52( -1 (ASK Q.k) -2 ( S K I P TO Q.152L) N o . . . . . . 1 I Weeks 1 | 1 I Weeks (50 ) (51 ) j. Did a doctor t e l l you why t h i s (1st, etc.) miscarriage might have occurred? -1 (ASK Q.k) Yes. (52 ( -2 ( S K I P TO Q.153L) No k. What did the doctor say caused the miscarriage? k . W h a t d i d t h e doctor s a y caused the miscarriage? I: I 1 I 4th j. Did a doctor t e l l you why this ( 1 s t , etc.) miscarriage might have occurred? ). Did a doctor tell you why t h i s .(1st , etc. ) m i s c a r r i a g e might have occurred? 1 I Weeks (48 ) ( 49) 1 1 1 1 1 1 Weeks 150) (51> 4th 1 | 1 3 r d| p.8 ) (4V ) I 1 Weeks I I Weeks (44 ) (iti ) 1 | 2nd 1 1 (SO) ( 5 l ) Yes.(52( No i— i—r 1st (46 ) (4? ) T 3rd (48) (49) 4th l I Weeks (46) ( 4 7 ) . 1 Weeks 1 1 •744 HAS ) (4b) (47) 3rd i . How mnny weekis had y o u r w i f e been p r e g n a n t when the ( 1 s t , e t c . ) miscarriage occurred? I Weeks 2nd i 1 1 1 1 Months ( 41) ( 42) W a s n ' t t r y i n g (43 ( -1 D o n ' t know -2 I I 1st 1 Weeks l I 1 1 Weeks I 1 T '.'ml 1 1 i . How many weeks linil your w i f e been p r e g n a n t when the ( l e t , e t c . ) miscarriage occurred? _ 1 812039 SIXTH MARRIAGE 154h.How many m o n t h s d i d i t take your w i f e t o become p r e g n a n t t h i s time? 1 1 I J_ . CARD 0 2 0 153h.How many m o n t h s did it take your wi fe to become pregnant t h i s time? i , How ninny weeks hud your w i f e been pregnant when HIP. ( 1 s t , etc. ) m i s r n r i i «K<- o c c u r red? . .— .. .. 1st . -1 (ASK Q.k) -2 (SKIP TO Q.154U k. What did the doctor say caused the miscarriage? 1st 1st 1st 2nd 2nd 2nd 3rd 3rd 3rd 4th 4th Yes...(53( -1 (ASK Q.m) -2 (SKIP TO Q. No ' " 152q) in. When was t h a t ? Any o t h e r s ? ) 1st 2nd 3id MONTH YEAR 1 1 1 1 1 1 l-l I 1 1 s ( 5 4 ) (55l (56) (57) MONTH YEAR 1 I I I I I I 1 l-l 1 1 2nd (58) (59) (60^ (61) MONTH YEAR 1 1 I I i 1 I l l I l l 3rd ( 6 2 ) (63) (64) (65) MONTH YEAR 1 1 T 1 1 1 1 1 1 l-l 1 1 (rib) ( 6 7 ) (68) ( 6 9 ) (CIO TO (}. 152n) 04 79^80" th 1 t 1 1 h. Did your w i f e have any pregnancies by ( you which ended in a s t i l l b i r t h ? -] (ASK Q.m) Y e s . . . ( 5 3 ( -2 (SKIP TO Q. No 153q) m. When was t h a t ? Any others?) (PROBE: 1 iih 4th . L. Did your w i f e have any pregnancies by you which ended in a s t i l l b i r t h ? L. D i d your w i f e have any p r e g n a n c i e s hy yon w h i c h ended in a s t i l l b i r t h ? Yes...(53( No . . m. When was t h a t ? Any others?) (PROBE: MONTH MONTH YEAR 1 I I 1 1 I l l I l 1st l (54) (55) ($6) (37 ) MONTH YEAR ! 1 I I l-l 1 I -1 (ASK Q.m) -2 (SKIP TO Q. 154q) 1 I 1 1 | I 2nd (58) (59) (60) (61) MONTH YEAR f i 1 1 1 1 I | |-| | | 3 f62) (63) (64) (65) MONTH YEAR 1 1 1 1 1 1 1 1 l-l I 1 4 (66) ( 6 7 ) (68) (69) (GO TO Q. 153n)o5 79-80 r i dI I 1 (PROBE: YEAR I (54) (b5) MONTH I l-l 1 I 1 1 (bb) (57T YEAR 1 I I 1 1 I |-| | | (58) (59) (6d) (61 J MONTH YEAR i ir i i l l I l l (62) (63) ( 6 4 ) (bi>) MONTH YEAR 1 1 1 T 1 1 t h I l l I l (66) ( 6 7 ) (68) (69^ 06 (GO TO Q.154n) 7T^8CT l �CARD 220 FIFTH MARRIAGE FOURTH MARRIAGE 812039 SIXTH MARRIAGE 152n. How many months did it take your w i f e to become pregnant this time? 1 1 1 1 I 1 Months ( 12) ( 13) Wasn 1 1 trying (14 ( -1 153n.How many months did it take your w i f e to become pregnant this time? 1 1 1 1 1 1 Months ( 12) ( 13) W a s n ' t t r y i n g ( I4( -1 154n.llow many months did it take your w i f e to become pregnant this time? 1 1 1 1 1 1 Months (12) (13) W a s n ' t t r y i n g ( l*( -1 o. Did a doctor tell you why t h i s s t i l l b i r t h might have occurred? o. Did a doctor tell you why this s t i l l b i r t h might have occurred? o. Did a doctor t e l l you why this s t i l l b i r t h might have occurred? Yes. (15 ( No -1 (ASK Q . p ) -2 (SKIP TO Q.152q Yes. (15 ( No p. What did the doctor say caused the (1st, etc.) stillbirth? -1 (ASK Q.p) -2 (SKIP TO Q.153q Yes. 05 ( No p. What did the doctor say caused the (1st, etc.) stillbirth? -1 (ASK Q . p ) -2 (SKIP TO Q.154q) p. What did the doctor say caused the (1st, etc.) stillbirth? 1st 1st 1st 2nd 2nd 2nd 3rd 3rd Ath Ath q. Did your w i f e ever have any pregnancies by you which ended in abortion' Yes...(16( No -1 (ASK Q.r) -2 (SKIP TO Q 152u) r. When was that? Any others? ) .. . Ath q. Did your w i f e ever have any pregnancies by you which ended in abortion Yes...(16( No (PROBE: -1 (ASK Q.r) -2 (SKIP TO Q 153u) r. When was that? Any others?) MONTH YEAR MONTH YEAR 1 1 1 1 1 1 1 1 1 1st 1 l-l 1 1st 1 1 l-l 1 (17) ( I B ) (19) (20) (17) (18) (19) (20) MONTH YEAR MONTH YEAR 1 1 I I 1 1 1 1 II 1 2 m lI l l I l l 2nd j 1 l-l | 1 (21) ( 2 2 ) (23) (24) (21) ( 2 2 ) (23) ( 2 4 ) MONTH YEAR MONTH YEAR 1 1 1 1 1 1 1 1 1 1 1 1 3rd I 1 I-J I I 3 r dI l l I l l (25) (26) (27) (28) (25) (26) ( 2 7 ) (28) MONTH YEAR MONTH YEAR 1 1 II 1 1 A t hI l l I l l 4 t hI l l I l l (29) (30) (31) (32) (29) (30) (31) (32) i rr i i s. How many months did it take your w i f e to become pregnant t h i s time? 1 1 1 1 -1 t. What was the main reason for the (1st, etc.) abortion? 1 1 1 1 3rd rd s. How many months did it take your w i f e to become pregnant t h i s - time? 1 I 1 I Months t. What was thp (pain reason for t h e O u t , * t p > ) abortion? Et . nd rd th (GO TO Q . 1 5 2 v ) MONTH YEAR 1 II 1 1 1st 1 l-l 1 1 (17 ) (18) (19 ) «0) MONTH YEAR 1 1 1 1 1 1 2 n dI l l I l l ( 2 1 ) (22) (23) (24) MONTH YEAR 1 1 II 1 1 3 r dI l l I l l (25) (26) ( 2 7 ) (28) MONTH YEAR 1 1 I I 1 1 4 t hI l l I l l ( 2 9 ) (30) (31) (32) (33) (34j W a s n ' t trying (ffjL_-» -1 Et nd th (CO TO Q.153v) 110 (PROBE: 1 1 1 I t. What was the main reason for the (1st, e t c . ) abortion? 2nd -1 (ASK Q.r) -2 (SKIP TO Q. 15Au) r. When was that? Any others?) 1 1 Months (33) ( 3 4 ) W a s n ' t trying (35( 1st Ath Yes...»6( No s. How many months did it take your wife to become pregnant this time? 1 1 Months (33) (34) W a s n ' t t r y i n g ( 35( q. Did your w i f e ever have any pregnancies by you which ended in abortion? (PROBE: 1 1 i 3rd (GO TO Q . l S A v ) �CARD 220 812039 SIXTH MARRIAGE FIFTH MARRIAGE FOURTH MARRIAGE V, 'liSHIf'*' Jook at this card and t e l l me a l l the numbers that apply to the types of birth control you or your wife normally used. 06.(/>2( Ol.(37( ISAii. (IF ANY CONCEPTIONS -CHILD, MISCARRIAGE, STILLBIRTH, OR ABORTION: SKIP TO Q.w ALL OTHERS: ASK Q.u) Did either you or your wife use birth control techniques regularly? Yes. . (36( -l (ASK Q.v) No -2 (ASK Q.x) THAND RESPONDENT CARP ^c VPH..( '>>( -1 (ASK O.v) Nn......._J-2 (ASK Q.x) I53u. (IF ANY CONCEPTIONS -CHILD, MISCARRIAGE, STILLBIRTH, OK ABORTION: SKIP TO Q.w ALL OTHEKS: ASK Q.u) Did either you or your wife use birth control techniques regularly? Yes..(36( -1 (ASK Q.v) No ~-2 (ASK Q.x) IM'li. ( 1 K ANY CONCEPTIONS -CHILD, MISCARRIAGE, KTII.I.III HTM, (U( AllOKII OK: SKIP TO Q.w AI.I. OTHERS: ASK l).u) l i i i l rithtr you or your w i f e u so b i r t h control techniques r e g u l a r l y ? I HAND RESPONDENT CARD "C"I v. Please look at this card and tell me all the numbers that apply to the types of birth control you or your wife normally used. v. Please look at this card and t e l l me a l l the numbers t h a t a p p l y to the types of b i r t h control you or your w i f e normally used. Ol.(37( 02.(^ -1 -1 06.(42( 07.( -J Ol.(37( -1 Q2.(38( 03.(39< -1 O6.(«f( 07.(5TT 08. ( 09.( 1Q.(46( 11. (4T( -1 -1 11.(571 -1 -1 .<48< (SKIP TO oTx) w, D i d any of .these pregn a n c i e s occur w h i l e o i t h e r you or your w i f e were p r a c t i c i n g b i r t h control? (49( Yes I I I J_ I I Times ( 50f T3TT (SKIP TO Q.aa/bh) y. How many months did you live apart the (first/ next) time? i —i 1st _. 2nd I_ r I | (50) x. During this marriage, how many times were you living apart from your wife for more^ than three months? Times Times (50) -1 (SKIP TO Q.aa/bb) I I I I Months 2nd I I I j Months ( 53) ( 54) 1 J I [_ (51) Never..(52( -1 ! let I I I (53) I J (54) 1 -1 -T — -| - I I Months I Months 3rd I I I " I" Months 3rd I I I I I I Months ( 57) ( 58) TTTTTlur [ Months 4th Ath ]_ Minirhs I I 5th [ Months 4th I Months ( 39) ( 6 0 ) i MJi | | I Months 5th I Months 6th I ( fil) 1 ~T 'T filh I I I Months (r,o TO r>. l •>?!•.') Months 1 ^ I 1 [ Months ( 55) ( 56)'" 3rd (SKIP TO Q.aa/bb) y. How many months did you live apart the ( f i r s t / n e x t ) time? y. How many months did you live apart the (first/ next) time? I J -2 No (51) Never. .(J>2(_ 1st Mont: 7 53Y TW 1 _ I hs I __ L -2 During this marriage, how mnny I imps were you living apart from your wife for more than three months? x. During (his mari iape, how nirtny 1'iim'H were you l i v i n g apart from your wife for more than three months? .-1 Did any of these pregnancies occur while either you or your wife were practicing birth control? _ No -? Never. .( 52( (SKir TO q.x) (SKIP TO Q.x) Did any of these pregnancies occur while either you or your wife were practicing birth control? -1 No -1 -1 12 (SPECIFY) 12 (SPECIFY) 12 (SPECIFY) -1 -1 i I Months ( 6 1 ) (62T 1 6tli — I I -{rrrrTwr (GO TO O.l r >3z) 111 r j [ I I ~cm TBTT \ |_ Months (GO TO Q . l S A z ) �CARD FOURTH MARRIAGE Yes No (5 ( 6 -1 153z.As a result of (this/ those) separations, did you and your wife have fewer children than you wanted to have? -1 -2 IIF ONLY MARRIAGE 1 aa.Are yon currently married an<l living with your wife, or are you divorced, widowed, or separated? living with wife... (66 ( SIXTH MARRIAGE FIFTH MARRIAGE 152z.As a result of (this/ these) separations, did you and your wife have fewer children than you wanted to have? Yes No (65 ( Divorced -2! (SKIP TO Separated. . .~-3f Q.cc) . 154z.As a result of (this/ these) separations, did you and your wife have fewer children than you wanted to have? Yes No -1 -2 IIF LAST MARRIAGE I aa.Are you currently married and living with your wife, or are you divorced, widowed, or separated? (RETURN Living with TO Q.23) wife... ( 6 ( -i 6 812039 220 (5 ( 6 -1 -2 IIF LAST MARRIAGE 1 aa.Are you currently married and living with your wife, or are you divorced, widowed, or separated? Living with (RETURN wife. . ( 6 ( .6 TO Q.23) Divorced -1 (RETURN TO Q.23) -2~)(SKIP TO Divorced -2\(SKIP TO Separated. .. . - f Q.cc) ~3 —•»• ij Separated. . . . -3V Q.cc) Widowed -4 I (RECORD IN S.R.B. PG 2 I (RECORD IN S.R.B. PG 2 1 IIF OTHER MARRIAGES! IIF OTHER MARRIAGES 1 (IF OTHER MARRIAGES! bb.How did that marriage end — were you divorced or were you widowed? bb.How did that marriage end — were you divorced or were you widowed? bb.How did that marriage end — were you divorced or were you widowed? — j (RECORD IN S.R.B. PG 2 | Divorced(67( Widowed. ... -1\(ASK -21 Q.cc) (RECORD IN S.R.B. PG 2 [ cc.In what month and year were yon (divorced/ widowed/ separated)? MONTH YEAR 1 1 I 1 1 l l I l l (68) (69) (70) ('n ) (IF A FIFTH MARRIAGE GO TO Q.153a) 1 I Divorced(67 ( -fltASK Q.cc) Widowed . . -21 •—~~ J (RECORD IN S.R.B. PG 2 MONTH I 1 —— cc.In what month and year were you (divorced/ widowed/ separated)? YEAR I I .1-1 1 1 i 1 (68) (69) (70), (71) (IF A SIXTH MARRIAGE GO TO Q.lSAa) 05 79-80 04 79-80 11 j J (RECORD IN S.R.B. PG 2 L cc.ln what month and year were you (divorced/ widowed/ separated)? I 1 Divorced(6_7j__-l\(ASK Q.cc) Widowed . . . . -2i 1 1 MONTH YEAR 1 I I I 1 1 l-.l I 1 (68) (69 ) (70) (71 ) (RETURN TO Q.23a) 06 79-80 �CARD 155. 023 812039 Additional Partners (Q PARTNER FOURTH PARTNER onth and year e g in l i v i n g r t ner time? a. In what month and ypar did you bpp, in livmj.1. wi tli a partner the lourth time? i i MONTH | YEAR MONTH T—r I -I I 1 I (in'mry' (15) (16) "(21) 1 YEAR i (22)' (23) 1 Current. .(25( -(21) (22) MONTH i YEAR i i 1 I Current. .(25 ( n) Yes.(26( No 1st 1 1 u) MONTH YEAR 1 1 1 1 (29) (30) YEAR I 1 I-I (33) 1 (34) 2nd M i 1 I-I 1 (31) (32) l 1 (33) (34) t. What was the outcome of the outcome of that pregnancy? (What nancy? (What was the outcome of the utcoroe of the second pregnancy?) egnancy? ) First 1 Second irst 1 Second 1 35 5( -l|(36( _! Live birth. ( ( -ll(36( -1 -21 -2 ."--21 -2 Miscarriage... -31 -3 • 17-31 ~-3 Stillbirth -4| -A _5| -5 nancy? No. 1 1 I-I (27) (28) r your partner control reguavoid preg- (GO -1 1 (c: , MONTH 1 1 YEAR (31) "(32) O4) TO NEXT PARTNER Q.h) 1 1 MONTH -1 (ASK Q. s) -2 (SKIP TO Q. s. When was that? M I-I r i" i i -1 r. Did this partner ever become pregnant by you? YEAR 1 1 1 1 1 (19 ) (20) MONTH YEAR 1 1 1 ! 1 l l I l l (21) (22) (23) (24) (29) (30) g. Did you or your partner use birth control regularly to avoid pregnancy? ..7773-2 V J ! 1 -1 MONTH I 1st I 2nd (33) 1 ACE) 1 . q. In what month and year did this relationship end? L. When was that? f. What was the outcome of that pregnancy? (What was the outcome of the second pregnancy?) First I Second | Live birth.(35(.__-l | (3_6(__Miscarriage... -2 I " Stillbirth Stillbirth -3| Abortion -4 I -4 Abortion. Not sure. Not sure -5 I - Yes .(37( No (23) (24) ASK Q. L) SKI P TO Q g) 2nd (31) (32) (WRITE IN partner ever egnant by you? YEAR r 1 I Current. .(25( 5( ist i (27) i ^28) iwrr^T-i-1.. i i i 1 1 YEAR 1 -1 -1 ii 1 1 MONTH i e. When was t h a t ? MONTH MONTH YEAR 1 I I 1 1 l l I l l (15) (16) (17) (18) p. How old was she at that time? >nth and year re 1st ionship d. Did this partner ever become pregnant by you? i r 1 I (19 ) (20) (24) YP«.(26( -I (ASK Q. e) No __-2 (SKIP TO Q. YEAR ! 1 1 1 (17) (18) 1 (WRITE IN AGE) 1 J [ (19) (20) c. In what month and year did this relationship end? MONTH 1 -1 o. In what month and year did you begin living wi th a partner the s i x t h rime? is she at that time ? b. How old was she at that time? (WRITE IN AGE) J SIXTH PARTNER (CO TO NEXT PARTNER Q.o) 05. 79-80 04 115 u. Did you or your partner use birth control regularly to avoid pregnancy? Yes(37( No -ll(Rt'TURN TO Q.25a) -2J 06_ �CARD 141 156. b. Additional Pregnancies (Q.25) d. When was that? FIRST MONTH ! 1 I 1 1 1 c. What was the outcome of that pregnancy? L i v e b i r t h , ( 1?( Miscarriage.... Stillbirth Were there any others? (IF YES, ASK Q.156d) -1 -2 -3 f . When was that? When was that? SECOND MONTH YEAR 1 I I 1 l l I l YEAR I (ft) l (14) - (15) (16) l l I l PROBE: 812039 1 I (13) (14) 1 l (15) (16) THIRD MONTH YEAR I I I I 1 1 I l l I l l (13) (14) ( 1 5 ) (16) e. What was the outcome of that p r e g n a n c y ? g. What was the outcome of that pregnancy? Live b i r t h . U 7 ( Miscarriage.... Stillbirth Live b i r t h . ( 1 7 ( Miscarriage. ... Stillbirth PROBE: Were there any others? (IF YES, ASK Q.156f) 114 -1 -2 -3 PROBE: -1 -2 -3 Were t h e r e any others? (RETURN TO Q . 2 6 a ) �.CARD 028 l:. V - ! r , o . Miiunii c i i i i . M FIFTH C H I L I ) SIXTH C H I L D NAME : NAME : I W a . l l o u o l d i s ( C H I L D ) now? 158a.llow old is (CHILD) now? NAME: T ~i r 812039 A d d i t i o n a l (hilclron (Q.30-32) 1 1 1 1 Ape 1 I 1 I 159a.How old is (CHILD) now? 1 1 Ape (15 ) ( 11,1 . 1, i 1 . 1 1 i . ,! . 1 1 1 h. (Is/K.-i.O ( C H I L D ) m a l e or f t'm.i 1 e? Male Femal o (|K -1 -1 Male Female POUNDS 1 I -1 -2 Male Female l l ( 19) ( 2 0 ) I I - POUNDS I I l ( 2 1 ) ( 22) D o n ' t know. ..( 23( -' 1 l 1 I -1 MONTH MONTH DAY M l-l 1 1 YEAR 1 1 1 1 1 1 l A L S O RECORD I N s'.R.B.-l'C 3 i (• . Was t h e c h i l d p r e m a t u r e , f u l l term, or overdue? P r e m a t u r e . ( '3(X FuM term Overdue (CO TO 0. 1 S7f ) l l 1 I 1 l l (21) (22) -1 d. What is ( C H I L D ) 's b i r t h date? YEAR MONTH DAY YEAR 1 1 1 1 1 1 T f 1 1 1 l-l 1 l-l 1 1 ( 2H) ( 2~9) ( 24) (25 ) ( 26) ( 27) ( 28) ( 29) ( 24) ( 2 5 ) ( 26) ( 27) (28) ( 2 9 ) 1 1 l-l ( 24)" (^5)' ( 2 6 ) "(21) DAY -1 -2 OUNCES M - D o n ' t know. ..( 23( d. What is ( C H I L D ) 's b i r t h date? 1 ! 1 ( 19) (20) d . What i s ( C H I L D ) ' s b i r t h date? 1 1 ( 18( c . H o w much d i d ( C H I L D ) weigh a t b i r t h ? OUNCES 1 1 b. ( I s / W a s ) (CHILD) male or f e m a l e ? c. How much did ( C H I L D ) wei gh at bi r t h ? POUNDS OUNCF.S 1 I I 1 1 l-l 1 ( 19) ( 20) ( 21) ( 22) A m ' t know. . . ( 23( ( 1« Age C l i i 1,1 d i r d . .( !/_(___ h. (Is/Was) (CHILD) male or f e m a l e ? ,-. How much r!id ( C H I L.D) w f i ph a t h i r t h ? 1 1 \ i In I.I .1 i i - . l . , ( | f. 1 1 1 1 1 ( 1 5) T 16) -1 -2 -3 I 1 M l-l 1 1 1 1 l-l 1 1 1 1 lALSO RECORD IN S . R . B . - P G 3| e. Was the c h i l d premature, f u l l t e r m , or overdue? P r e m a t u r e . ( 30( F u l l term (CO TO 0. I S S f ) -1 -2 lALSO RECORD I N S . R . B . - P G 3 1 e. Was the c h i l d premature, f u l l t e r m , o r overdue? P r e m a t u r e . ( 30( F u l l term (GO TO Q. 159f ) -] -2 �812039 CARD 028 FOURTH CHILD SIXTH CHILD FIFTH CHILD 157f. Where are (CHILD) 's hirth registration records located? Tn what city and state is that? IRECORD IN S.R.B. PG 3 1 158f- Where are (CHILD) 's birth registration records located? In what city and state is that? IRECORD IN S.R.B. PG 3 l g. Where are (CHILD)'s rtirront medical records located? In what city and state is that? 159* -Where are (CHILD) 's birth registration records located? In what city and state is that? IRECORD IN S.R.B. PG 3 1 g. Where are (CHILD) 's current medical records located? In what city and state is that? g. Where are (CHILD)'s current medical records located? In what city and state is that? IRECORD IN S.R.B. PG 3 I IRECORD IN S.R.B. PG 3 1 IRECORD IN S.R.B. PG 3 1 li. What was (CHILD) 's mother's full name? h. What was (CHILD) 's mother's full name? h. What was (CHII.D)'s mother's full name? IRECORD IN S.R.B. PC 3 1 IRECORD IN S.R.B. PG 3 1 IRECORD IN S.R.B. PG 3 1 i. How old was the mother when (CHILD) was born? 1 1 1 1 1 1 1 1 Age (•11) (-12) No 1 1 1 1 Age (31 ) (32 ) 1 1 Yes.(33( -l (ASK Q.k) .. -2 (SKIP TO Q.L) lHAND RESPONDENT CARD "C" I k. Please look at this card and tell me all ol the numbers that apply to the types of birth control you or your partner were practicing? Ol.(34( -1 06.(39( -1 02.(35( 03. ( 5 3( (M. ( 7 3T ()5.(38( " -1 -1 -l -1 O7.(40( -1 08. (Al(" -1 09. ( 2 71 -i I()-(5^J_... •' 1 1 , 1 /T'i'f ~ " i -1 (ASK No -2 (SKIP TO Q.L) ,U5( (CO TO Q.157L) -1 Age Yes.(33( Q.k) (HAND RESPONDENT CARD "C"l k. Please look at this card and t e l l me all oj the numbers that apply to the types of birth control you or your partner were practicing? O1.(34( -1 Oh.(39( -1 02.(35( 03. (551 o/i. (T71 05. (751 -1 -1 -] -1 O7.(40( 08.(5T( 09. (571 10. ( 7 71 i i ('I'M -1 -i -i -1 i -1 (ASK Q.k) No -2 (SKIP TO Q.L) lHAND RESPONDENT CARD "C"l k. Please look at this card and tell me a l l of the numbers that apply to the types of hirth control you or your partner were practicing? Ol.(34( 02.(35( 03. (361 04. (571 O6.(3y( O7.(40( 08. ( T "* 09. (W( -1 -1 -I -1 .<45( -1 -1 -1 -l -1 o5.(jBT ""-i lo.t/rri -ii " i i i •' '"( I'J (SI'fciMFY) [2 (SPKCIKY) K> (HI'lf-ClKY) 1 1 1 1 (31 ) ( ) M j. Were either of you using birth control at the time she became pregnant with (CHILD)? j. Were either of you usinf birth control at the time she became pregnant with (CHILD)? j. Were either of you using birth control at the time she became pregnant with (CHILD)? Yes.(33( i. How old was the mother when (CHILD) was born? i. How old was the mother when (CHILD) was born? .(45< (GO TO Q.158L) .110 -1. (CO TO Q.159O �C ARC 028 FOURTH CHILD FIFTH CHILD 15 71,. How many m o u t h s did i t t f i k o h e r Co become preg n a n t u'ith t h i s c h i l d ? 1 1 1 1 SIXTH CHILD 158L.How many m o n t h s d i d i t 159L.How many m o n t h s did it t a k e her to become pregtake her to become pregnant with this child? nant w i t h this c h i l d ? 1 I Months 1 1 1 1 1 1 Months 1 1 ( /if)) ( <W) ( 40) ( 477 1 <-nv l l u n 812039 1 i i N n l l l i • 1 /i IH 1 I.BB n m. Did ( C H I L D ) have any b i r t h defects? 1 lirtil 1 1 1 (46) In. M i l l i . ( '.III I 1 1 Months (47) l.i' M b l l u i n 1 nutn t l i . ( 48( Not t r y i n f i in. Did ( C H I L D ) have any birth defects? ~| -2 m . D i d ( C H I L D ) have a n y birth defects? Y P S . ( 4'X -1 (ASK Q . n ) Y e s . ( 49( -l (ASK Q . n ) Y e s . ( 49( -1 (ASK Q . n ) No -2 ( S K I P TO Q . o ) No -2 ( S K I P TO Q . o ) No -2 ( S K I P TO Q . o ) n. W h a t kind of b i r t h def e c t s d i d (s)he have? Any o t h e r s ? n. What k i n d of b i r t h def e c t s d i d ( s ) h e have? Any o t h e r s ? n. W h a t k i n d of b i r t h def e c t s did ( s ) h e have? Any others? o. Was ( C H I L D ) ever d i a g nosed as h a v i n g cancer? o. Was ( C H I L D ) ever diagnosed as h a v i n g cancer? o. Was ( C H I L D ) ever diagnosed as having cancer? Y e s . ( 5(X -1 ( A S K Q . p ) Y e s . ( 5<X -1 (ASK Q . p ) Y e s . ( 50( -i (ASK Q . p ) No. . -2 (SKIP TO Q . r ) N0 -? ( S K I P TO Q . r ) No -2 (SKIP TO Q . r ) p. In what month and y e a r was the d i a g n o s i s made? MONTH ! I l 1 Gl ) G2 i 63 )( 5/0 In what month and year was the d i a g n o s i s made? MONTH YEAR 1 I 1 l - I l p. 1 l q. W h a t k i n d of cancer was diagnosed? YEAR i i ii r i 1 (51 1) £521 ) - (531 ) (1 54) 1 (CO TO Q . 1 5 7 r ) -1 1 I In w h a t month and year was the diagnosis made? MONTH YEAR 1 M 1 1 l l I l l ( 5 1 ) (52) (53) (54) q. What k i n d of cancer was diagnosed? q. What kind of cancer was diagnosed? (55-56 (55-56 ( 5.'i-56 Not s u r e . . ( S 7 ( p. Not sure. .(57 ( -1 (GO TO Q.158r) 117 Not sure. .(57 ( (CO TO Q . 1 5 9 r ) -1 �CARD FOUKTII CHILD 028 FIFTH CHILD 157r.(Oopa/Did)(CHILD) have 15Sr.(Does/Did)(CHILD) have diagnosed learning, d i s diagnosed learning d i s.ihi 1 i ry? ahility? 812039 SIXTH CHILD 159r.(Does/Did)(CHILD) hov« diagnosed learning d i g ahi 1 i tv? Yps.(5H( -1 (ASK Q.s) Yes. (58 ( -1 (ASK Q.s) Yes.t 5(< -1 (ASK O.s) No -2 (SKTP TO 0. t) No.... No -2 (SKIP TO Q t) -'1 (SKI)' TO Q t) .-,. What kind of learning d i s a b i l i t y (does/did) (s)lu' have? s. What k i n d of learning d i s a b i l i t y (does/did) (s)he luive? s. What, kind of learning d i s a b i l i t y (does/did) (s)he have? t. (Does/Did) (CHILD) have any physical, mental, o motor impairments? l. (Does/Did)(CHILD) have any physical, mental, o motor impairments? t. (Does/Did)(CHILD) have any physical, mental, motor impairments? Yes.(59( -1 (ASK Q.u) Yes.(59( -1 (ASK Q.u) Yes. (59 ( -] (ASK Q.u) No -2 (SKIP TO Q.v) No -2 (SKIP TO Q.v) No u. What kind of impairment (does/did) (s)he have? CONTINUE SKIP TO NEXT CHILD u. What kind of impairment (does/did) (s)he have? -2 (SKIP TO Q.v) u. What kind of impairment (does/did) (s)he have? IF CHILD IS DEAD: IF CHILD IS DEAD: IF CHILD IS DEAD: OTHERWISE: OTHERWISE: OTHERWISE: v. On what date did (CHILD) die? MONTH . DAY YEAR 1 1 1 I I 1 1 1 1 l-l 1 l-l 1 1 [60) (61) (62) (63) 76/0 (65) CONTINUE SKIP TO NEXT CHILD v. On what date did (CHILD) die? CONTINUE SKIP TO NEXT CHILD v. On what date did (CHILD) die? MONTH DAY YEAR MONTH DAY YEAR 1 I I 1 I I 1 1 1 I I 1 I I 1 1 l-l 1 l-l 1 1 l-l 1 l-l 1 1 60) (61) (62) (637 (64) (65) 60) (61) 162) (63) (64) (65 w. What was the cause of death? w. What was the cause of death? w. What was the cause of death? x. Where is (CHILD) 's •.Ifntli r"I»i --re re '' ! " ! x. Where is (CHILD) 's x. Where is (CHILD) 's Uiat '.' RECORD IN"~S". R .'» ". "PC 3T (GO TO NEXT CHILD Q.158a) 'I^n E'h t P T1 ' •• f i- : r tJ"* T n ',--!, 1 I*• i •.'•f' '-1 '.: '•>':" * . ' '=' ' t li a t: '.' Chat? WcolD~i NT. TtTiiT "Pc~3 T RECORD ADDITIONAL CHILDREN IN S.R.B. - PC 32-35) (CO TO NEXT CHILD Q.15')a) 05 79-80 " 04 79 -HO ' lib "RECORD IN sTR'.F." pc;"3'"f 06 79-8 �CARD 160-162. 812039 Addition.il C h i l d r e n ( Q . 30-32 KCONT 1 11) SEVENTH CII1U) Klf.HTH C H I L D NINTH CHILD NAME : NAME : KiOii.How old i s ( C H I L D ) now? I f c l a . H o w o l d is ( C H I L D ) now? 1 1 r (i > ) (if. 5 1 1. i 1 .1 ,1 i i : i l ' 1 1 M Male Femal e NAME: 1 1 Ape 1 1 1 1 1 6 2 a . H o w o l d i s ( C H I L D ) now? 1 1 Ape (15 ) ( 16) i i. ; i .1 .1 ; c , i 1 h. ( I s / W a s ) (CHILD) male or f e m a l e ? (Hi ( Male Female D o n ' t know. . .(23 ( i | ;i i 1 1, 1 1 .! ,1 1 p . l ( 18( -1 d . Whnt i s ( C H I L D ) ' s b i r t h date? 1 I 1 1 Age Male Female | ( 18( -1 -2 c. H o w much d i d ( C H I L D ) weigh a t b i r t h ? POUNDS OUNCES 1 I I 1 1 l l I l l <19 ) (20 ) ( 21) ( 2 2 ) D o n ' t know. ..(23 ( 1 | ,<< b. ( I s / W a s ) (CHILD) male or f e m a l e ? -1 -2 c. How much d i d ( C H I L D ) weigh a t b i r t h ? POUNDS OUNCES I I I 1 1 l l I l l (19 ) <20 ) (21 > (22 > 1 1 0 5 ) (16) b. ( I s / W a s ) ( C H I L D ) m a l e or f e m a l e ? -1 -2 c . . H o w much d i d ( C H I L D ) weigh at b i r t h ? 1 I 028 -1 d. What is ( C H I L D ) 's b i r t h date? 1 I POUNDS OUNCES 1 I I 1 1 l l I l l ( 19) feO ) (21) (22) Don' t know. . .( 23( -1 d. What is (CHILD) 's b i r t h date? MONTH DAY YEAR MONTH DAY YEAR • MONTH DAY YEAR 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l-l 1 l-l 1 1 1 1 l-l 1 l-l 1 1 1 1 l-l 1 l-l 1 1 (24 ) (25 ) (26 ) (27 ) (28 ) (29 ) GO (25 ) ( 26) ( 27) ( 28) ( 29) (24 ) (25 ) ( 26) ( 27) ( 28) ( 29) lALSO RECORD IN S.R.B.-PG 3l e. Was the c h i l d p r e m a t u r e , f u l 1 t e r m , or o v e r d u e ? lALSO RECORD IN S . R . B . - P G 3| o. W a s t h e c h i l d p r e m a t u r e , f u l l t e r m , o r overdue? lALSO RECORD IN S . R . B . - P G 3| e. Was the c h i l d premature, f u l l t e r m , o r overdue? l ' r e t n a t u r e . ( 3(X -1 P r e m a t u r e . ( 30( F u l l term -1 -2 P r e m a L u r e . ( 3U( -1 Not 8ure -4 Not sure -4 Not s u r e -4 (CO TO Q . 1 6 0 f ) (GO TO Q . 1 6 1 f ) 11J (GO TO Q . 1 6 2 C ) �CARD 02B SEVENTH CHILD EIGHTH CHILD 1601". Where are (CHILD) 's hirth repistration records located? In what city and state is that? 1 RECORD IN S.R.B. PC, 3 [ p. Whore arc ( CHILD) ' s current m e d i c a l records located? In what c i t y and state IE that? 1 RECORD IN S.R.B. PC 3 1 Hill'.Where nrr (CHILD) 's birth registration records located? In what city and state is that? IRECORD IN S.R.B. PG 31 h. What was (CHILD) 's mother's f u l l name? IRECORD IN S.R.B. PC 3 [ IRECORD IN S.R.B; PG 3 1 i. How old was the mother when (CHILD) was born? 1 1 Age Yes.O3( -2 (SKIP TO Q.1601^ No IHAND RESPONDENT CARD "C" | k. Please look at this card and t e l l me all of the numbers that apply to the types of birth control you or your partner were practicing? O6.(3y( 07.UO( OR. UK -1 OA.dTT '-i (iy.(T2J - Or).(j8< lf>.<43( 1 < i HIM' i: II 1 V ) " <W •</,!>< (HO TO Q.160L) 1 1 Age 1 1 1 1 Age (31) (32) (32) j. Were either of you using birth control at the time she became pregnant with (CHILD)? Yes.(33( -1 (ASK Q.k) -2 (SKIP TO Q.1611) No -1 (ASK Q.k) -2 (SKIP TO 0.162L) IHAND RESPONDENT CARD "C"l IHAND RESPONDENT CARD "cllT k. k. Please look at this card and tell me a l l of the numbers that apply to the types of .birth control you or your partner were practici ng? Ol.(3'i( 02.(:)5( 03.(36( o/i .(37T 01. ("TB? "" -1 -1 -1 -i --i 1 ' v ril'r i i f < i -1 1 1 i. How old was the mother when (CHILD) was born? j. Were either of you using birth control at the time she became pregnant with (CHILD)? No -1 1 1 (31) -1 (ASK Q.k) -1 -1 -1 g. Where are (CHILD) 's current medical records located? In what city and state is that? IRECORD IN S.R.B. PC 3 1 Yes.( 33( 01. (M 02. (3M 03.(36( IRECORD IN S.R.B. PC 3 1 IRECORD IN S.R.B. PC 3 1 j. Were either of you using hirth control at the t i m e she became pregnant with (CHILD)? . 162.* -Where are (CHILD) 's b i r t h registration records located? In what city and state is that? g. Where are (CHILD) 's current m e d i c a l records located? In what c i t y and state is that? i. How old was the mother when (CHILD) was born? 1 1 1 1 (31) (32) NINTH CHILD h. What was (CHILD) 's mother's f u l l name? h. Whnt was (CHILD) 's mother's f u l l name? IRECORD IN S.R.B. PC 3 1 1 1 81203') 06,. (39( 07.UO< OS. UK -1 -1 -1 oq.(T2T -i m.(/,3< -i i '.'II'. "•' Please look at t h i s card and tel 1 me n i l of the numbers t h a t a p p l y to tho types of b i r t h control you or your partner were p r a c t i c ing? 01.Ci4( 0?.(j5( O3.(3b( OA,(17T -1 -1 -1 -1 o5.(3aT -i Oh. O<M 07. CiO( OB.('«1( 09.1A7T -1 -1 -1 -1 io.(/TTf -i " I:,V. 1 I .' * r:l'r i l l ' , ' .(/,;>( (CO TO ( . 1611.) ) -1 . lVi( (CO TO 0. 1621.) -1 �CARD SEVENTH CHILI) EIGHTH CHILD 1601. .How many m o n t h s did it t a k e her to become preg nant with this child? 1 1 I 1 1 1 Months 1 1 NINTH CHILD I 1 1 1 Months 1 1 -1 ! I V(PI|'. •} -L Less t h a n 1 m o n t h . ( 4 8 ( Mi-, 1 i v IIIB -7 Mo, m. Did ( C H I L D ) have any birth defects? -\ (ASK Q . n ) No -? ( S K I P TO Q.o) 1 I Months I.PSS than 1 m o n t h , ( 4fl( m. Did (QIILD) have any hirth defects? Y e s . ( lA 1 1 fc6 ) ( 47) 0,6 ) (47 ) l.pss Minn 1 m o n t h . ( 4 B ( 812039 1611. .How many months did it 162L.How many months did it t a k e her to become pregtake her to become pregnant w i t h t h i s c h i l d ? nant w i t h t h i s c h i l d ? <46 ) <47 > fl'il 028 -j 1 rv ! i»ti . /' m. Did ( C H I L D ) have any birth defects? Y e s . (49 ( -1 (ASK Q.n) No -2 ( S K I P TO Q . o ) Yes.(4y( -] (ASK Q . n ) No -2 ( S K I P TO Q . o ) n. What kind of birth def e c t s did (s)he have? Any others? n. What kind of birth defects did (s)he have? Any others? n. What kind of birth defects did (s)he have? Any others? o. Was (CHILD) ever diagnosed as having cancer? o. Was (CHILD) ever diagnosed as having cancer? o. Was (CHILD) ever diagnosed as having cancer? Yes.(50( No -l (ASK Q . p ) — p. T r Yes.( -2 (SKIP TO Q.160r) : In what month and year was the d i a g n o s i s made? MONTH 1 i f 1 1-1 (51) < 5 2 > YEAR 1 i I i (53) (54) q. What k i n d of cancer was diagnosed? -1 (ASK Q.p) 50 v NO — p. Yes.(50( -2 (SKIP TO No Q.161r) In what month and year was the d i a g n o s i s made? MONTH 1 I 1 l l (51) (52) i I (GO TO Q.160r) -1 1 l l (53) (54) q. What k i n d of cancer was diagnosed? p. 1 1 -2 (SKIP TO Q.162r) In what month and year was the d i a g n o s i s made? MONTH 1 1 1 1 l-l (51 ) (52) YEAR [ 1 Not sure.. (57 ( (GO TO Q . 1 6 1 r ) -1 1 1 (53) (54 f q. What kind of cancer was diagnosed? (55-56) (55-56) (55-56) Not sure. .(57{ — YEAR M - -1 (ASK Q.p) Not sure. .(57 ( (GO TO Q . 1 6 2 r ) -1 �CARD 02« SEVENTH CHILD EIGHTH CHILD 812039 NINTH CHILD 160r.(Does/Did)(CHlLD) have a 161r.(Does/Did)(CHILD) have diagnosed learning disdiagnosed learning disabil ity? ability? 162r.(Does/Did)(CtllLD) have a diagnosed learning disabil ity? Ves.(58( -1 (ASK Yes. (58 ( -l (ASK Q.s) No -2 (SKIP TO Q.t) Q.s) Yes.(58( -1 (ASK Q.s) No -1 (SKIP TO Q.t) No -2 (SKIP TO Q.t) s. What kind of learning d i s a b i l i t y (does/did) (s)he have? s. What kind of learning disability (does/did) (s)he have? s. What kind of learning disability (does/did) (s)he have? t. (Does/DidHCMILD) have any physical, mental, or motor impairments? t. (Does/DidMCHILD) have any physical, mental, or motor impairments? t. (Does/Uid)(CHlLD) have any physical , mental , or motor impairments? Yes.(5y( -1 (ASK Q.u) Yes.(59( -1 (ASK Q.u) No -2 (SKIP TO Q.v) No -2 (SKIP TO Q.v) Yes.(59( -1 (ASK No -2 (SKIP TO Q.v) u. What kind of impairment (does/did) (s)he have? u. What kind of impairment (does/did) (s)he have? u. What kind of impairment (does/did) (s)he have? Q.u) IF CHILD IS DEAD: IF CHILD IS DEAD: IF CHILD IS DEAD: OTHERWISE: OTHERWISE: OTHERWISE: CONTINUE SKIP TO NEXT CHILD CONTINUE SKIP TO NEXT CHILD v. On what date did (CHILD) die? v. On what date did (CHILD) die? v. On what date did (CHILD) die? CONTINUE SKIP TO NEXT CHILD ) MONTH DAY YEAR 1 1 1 1 1 I I 1 l-l 1 1I I 60) (61) (62) (63) (64) (65) (60) (611 (62) (63 ) (64) (65) 60) (61) (62) (63) 6<i) (65) MONTH 1 1 1 1 l-l DAY 1 1 I I l-l YEAR 1 1 1 1 1 MONTH 1 1 1 1 l-l DAY 1 1 1 1 l-l 1 YEAR 1 1 1 1 1 w. What was the cause of death? w. What was th-e cause of death? w. What was the cause of death? X. Where is (CHILD) 's death registered? In what city and state is that? 1 l x. Whe .- is (CHILD) 's death registered? In what city and state is I ha I V x. Where is (CHILD) 's death registered? In what city and state is Mini' Rfii'Olifi THfi.R'.fl.K 1 | rKEcM1)TN"T!';v:B7 'T S'*y'T (GO TO NEXT CHILD Q.161a) jRfcofiY IN s.K.ii. f-c 3 I (GO TO NEXT CHILD Q.162a) RECORD ADDITIONAL CHILDREN IN S.R.B. - PC 36-39) 08 79-80 07 79-80 122 09 79-80 �CARD 163-165. Additional Children (Q.30-32)(CONT'D) TENTH CHILD TWELFTH CHILD ELEVENTH CHILD NAME :• NAME: 163a.How old is (CHILD) now? 164a.How old is (CHILD) now? ! I 1 1 <I5> 'IB) 1 1 Age 1 1 ( 18( Male D o n ' t know. . . ( 23( 1 I Male Female -1 YEAR POUNDS OUNCES 1 II 1 1 l l I l l (19 ) ( 20) ( 21) ( 22) MONTH 1 1 1 1 l-l DAY 1 1 1 I -1 -2 YEAR 1 1 1 1 I ALSO RECORD IN S.R.B.-PG 3 1 POUNDS OUNCES 1 I I 1 1 l l I l l 0.9 ) (20) (21) (22) D o n ' t know...(23( -1 1 i i r i i I I i i i 1-1 i 1-1 i i lALSO RECORD IN S.R.B.-PG 3| 1 1 l-l (18( c. How much did (CHILD) weigh at birth? 1 1 -1 d. What is (CHILD) 's birthdate? d. What is (CHILD) 's birthdate? d. What is (CHILD) 's birthdate? DAY ( 18( D o n ' t k n o w . . . ( 23< -1 Age b . ( I s / W a s ) ( C H I L D ) male or female? c. How much did (CHILD) weigh at birth? POUNDS OUNCES 1 I I 1 1 l l I l l (19 ) ( 20) (21 ) ( 2 2 ) 1 1 C h i l d d i e d , . <JL 7 J... ~ ' b. ( I s / W a s ) (CHILD) male or female? -1 -2 1 1 f5 ) (16) Chi I d d i e d . .( 17( -1 -1 c. How much did (CHILD) weigh at b i r t h ? MONTH 1 1 165a.How old is (CHILD) now? &5 ) ?16) b . ( I s / W a s ) (CHILD) m a l e or female? Male Female NAME : 1 1 .„ L ... 1 Age C ' h i I d d i e / I . . < |/ 1 I 812039 Q28 i i MONTH r i i i 1-1 DAY i i • M 1-1 YEAR i i i i <2O (25 ) ( 26) ( 27) ( 29 (29 ) ( 24) ( 25) ( 26) ( 27) (28 ) 09 ) ( 24) (25 ) (26 ) ( 27) ( 28) ( 29) e. Was the c h i l d premature, f u l l term, or overdue? Premature. 00 ( (GO TO Q.163f) -1 e. Was the child p r e m a t u r e , f u l l term, or overdue? Premature. ( 30( Full term (GO TO Q.164f) -1 -2 lALSO RECORD IN S.R.B.-PG 31 e. Was the child premature, f u l l term, or overdue? Premature. (30( Full term (GO TO Q.165f) -1 -2 �CARP TENTH CHILD JRECORD IN S.R.E. PG 3 I TWELFTH CHILD 164 f. Where are (CHILD) 's birth registration records located? In what city and state is that? IRECORD IN S.R.B. PG 3 g. Where are (CHILD) 's current medical records located? In what city and state is that? 1 RECORD IN S.R.B. PC 3 1 1651. Where are (CHILD) 's birth registration records located? In what city and state is that? IRECORD IN S.R.B. PG 3 | g. Where are (CHILD) 's current medical records located? In what city and state is that? IRECORD IN S.R.B. PC 3 h. What was (CHILD) 's mother's full name? IRECORD IN S.R.B. PG 3 1 g. Where are (CHILD) 's current medical records located? In what city and state is that? 1 RECORD IN E.R.R. PC 3 1 h. What was (CHILD) 's mother's full name? !>. What was (CHILD) 's mother's full name? IRECORD IN S.R.B. PG 3 1 IRECORD IN S.R.B. PG 3 i. How old was the mother when (CHILD) was born? I 1 1 1 (31) (32) 812039 ELEVENTH CHILD lf>3f. Where are (CHILD) 's birth registration records located? In what city and state is that? I 1 Q2B i. How old was the mother when (CHILD) was born? I 1 Age j. Were either of you using hirth control at the time she became pregnant with (CHILD)? l l 1 1 (31) (32) i. How old was the mother when (CHILD) was born? 1 1 Age 1 1 (31) (32)" j. Were either of you using birth control at the time she became pregnant with (CHILD)? j. Were either of you using birth control at the time she became pregnant with (CHILD)? Yes.(j3(:. -1 (ASK Q.k) Yes.(33( -1 (ASK Q.k-) Yes.(33( No No -2 (SKIP TO No ~ -2 (SKIP TO Q.163L) 01. ( j 3( 02. (3M 03.(36( (4(7 ).3( 0.(8 '.3( -1 -1 -l -l -1 Of).(39( 07. ( 6 71 08. UK 09. (42f in.(/.3( -1 -1 -1 -1 -1 1 1 . u/iT ' -i 1? (SPECIFY) lHAND RESPONDENT CARD "c" 1 k. Please look at this card and tell me a*l 1 ol the numbers thai apply to the types of hi rth control you or your partner werp practicing? Ol.(34( 02.(35( 03.(36( OA.(37( -1 -1 -1 -1 06.(39( O7.(40< 08. (41( 09.<42( -1 -1 -1 -1 OS.O«( I I0.(/i3( 1 1 .(7,'f, ( -1 ~-\ (CO TO 0.163L) -1 lHAND RESPONDENT CARD "C" 1 k. Please look at this card and t e l l me all of the numbers that apply to the types of birth control you or your partner were practicing? Ol.(34( 02. ( 3 51 03.(J6( 0/..(37( 0-).C>R( -1 -1 -1 -1 -1 Q6.O9( 07.(4<j( O8.(41< 09.(42( IO.(Vl( ll.(Vi7 17 (SI'K.CIKY) 12 (SPKC1KY) .''( (.> -1 (ASK Q.k) -2 (SKIP TO ~ Q.165L) Q.164L) (HAND RESPONDENT CARD "C" I k. Please look at this card and tell me all of the numbers that apply to the types of birth control you or your partner were practicing? 1 1 Age ,''( (.) (CO TO Q.164L) -1 .('•'>( (CO TO Q.16S1.) �CARD TENTH CHILD 028 81Z039 ELEVENTH CHILD TWELFTH CHILD 1631. .How many m o n t h s d i d i t 1 ML. How many months did it 165L.How many m o n t h s did it t a k e her to become pregtake her to become pregtake her to become pregnant w i t h this c h i l d ? nant w i t h t h i s child? n a n t with t h i s child? 1 1 1 ) 1 1 Months 1 1 Illflll | 1 I Months 1 1 U K I I l l l l . ( /|H( 1 \.pna lluui 1 lnnril h. ( AH( -I Nol trying m. Did (CHILD) have any birth defects? 1 1 1 I Months (46) (47) W !> ( 2 7 ) ?46) U7) liPBH 1 1 -2 Iri-rtn Hum | i m m f h . ( AH( Not t r y i n g m. Did (CHILD) have any birth defects? -| -2 m. Did (CHILD) have any birth defects? Yes.( 4;X -1 (ASK Q . n ) Y e s . ( 49( -1 ( A S K Q . n ) Yes.<49( -1 (ASK Q . n ) No -2 (SKIP TO Q . o ) No -2 ( S K I P TO Q . o ) No -2 ( S K I P TO Q . o ) n. What kind of b i r t h def e c t s did (s)he have? Any others? n. What kind of birth defects did (s)he have? Any others? n. What kind t>{ birth def e c t s did (s)he have? Any others? o. Was (CHILD) ever diagnosed as having cancer? o. Was (CHILD) ever diagnosed as having cancer? o. Was (CHILD) ever diagnosed as having cancer? Yes.( 50( -1 (ASK Q.p) Yes.(50( -1 (ASK Q.p) Yes.(50( -J (ASK Q . p ) No -2 (SKIP TO Q.163r) No -2 (SKIP TO Q.164r) No -2 (SKIP TO Q.165r) ... p. 1 I In what month and year was che diagnosis made? MONTH YEAR 1 1 1 I 1 l l I l l fcl ) (52 ) ( 53) ( 54) q. What kind of cancer was diagnosed? p. 1 I In what month and year was the diagnosis made? MONTH YEAR 1 1 1 1 1 l l I l l (51) ( 5 2 ) (53) (54) q. What kind of cancer was diagnosed? (GO TO Q.163r) -1 1 1 In what month and year was the diagnosis made? MONTH YEAR 1 I I 1 1 1 l-l 1 1 (51) ( 5 2 ) (53) (54!) q. What kind of cancer was diagnosed? (55-56) (55-56) ( 55-56) Not sure. .67 ( p. Not sure..(57( -1 (GO TO Q.164r) 125 Not sure..(57( (GO TO Q.165r) -1 �CARD 028 TENTH CHILD ELEVENTH CHILD 812039 TWELFTH CHILD l63r.(Uoes/Uid)(CHII.D) have a 164r. (Does/DidHCHILD) have a 165r.(Does/Did)(CHILD) have diagnosed learning disdiagnosed learning disdiagnosed learning dis ahiliry? ability? ability? Yes.(f>H( -1 (ASK Q.s) Yes.(58( -1 (ASK Q.s) No -2 (SKIP TO Q.t) No -•) fSKFP TO n.iO Yes.(58( -1 (ASK Q.s) No -2 (SKIP TO Q.t) s. What kind of learning disability (does/did) (s)ho have? s. What kind of learning disability (does/did) (s)he have? s. What kind of learning disability (does/did) (s)he have? t . (l)iies/iiLd)(Clii.l.l>) haviauv physical , m e n t a l , o mi- Lor i mpa i rim-nt ••? t. (Does:/Did)(CIIlLD) have any p h y s i c a l , mental, o motoi i mpa inm?n t s? t. (Does/rnd)(CllILD) have any physical, mental, t motor impairments? Yes.< W( -1 (ASK Q.u) Yes.CM -1 (ASK Q.u) Yes. ( 1)9 ( -1 (ASK No --2 (SKIP TO Q.v) No -2 (SKIP TO Q.v) No u. What kind of impairment (does/did) (s)he have? CONTINUE SKIP TO NEXT CHILD u. What kind of impairment (does/did) (s)he have? IF CHILD IS DEAD: IF CHILD IS DEAD: OTHERWISE: OTHERWISE: v. 0: what date did (CHILD) die? CONTINUE SKIP TO NEXT CHILD v. On what date did (CHILD) die? Q.u) -2 (SKIP TO Q.v) u. What kind of impairment (does/did) (s)he have? IF CHILD IS DEAD: CONTINUE OTHERWISE: KliTURN TO.Q.33. v. On what date did (CHILD) die? MONTH DAY YEAR MONTH DAY YEAR MONTH DAY YEAR 1 I 1 I 1 I I I 1 I I 1 1 I I 1 I I 1 1 l-l 1 11 l-l 1 l-l 1 1 l-l 1 l-l 1 60) (61) (62) (63) (64) (f-5) (60) (bl) 752) (63) M) (55) (faO) (61 ). (62) (63) (64) (65 w. What wns the cause of death? w. What was the cause of death? w. What was the cause of death? x. Where is (CHILD) 's death registered? In what r i t y and slate i ti Unit? x. Where is (CHILD) 's 'Jeath registered? In what c i t y and state is lluil ! x. Where is (CHILD)'s death registered? In what r i t y and state i i; (hut '1 IHKCOKIJ IN ti.lt. 11. i'C '! 1 RECORD IN S..R.B. i-c 3 1 (GO TO NEXT CHILD Q.lMa) IRKCORI") IN S.R.S. i'C :J 1 (CO TO NEXT CHILD Q.165n) 10 7 4- HO (RETURN TO Q.33) II /"J-HO .126 '-' y«7-~HD �CARD 03b 812039 > Part 4 Part 5 Part 6 166c In what month and year 1661 In what month and year 166o. In what month and year was cancer of the (BODY was cancer of the (BODY was cancer of the (BODY PART) first diagnosed? PART) first diagnosed? PART) first diagnosed? iRecf^ jTh fl»IUB. PG 6 1 IRECORD IN S.R.B. PG 6 1 IRECORD IN S.R.B. PG 6 1 d. What is the full name of the doctor or the fnedicfll facility where the diagnosis was made? IRECORD IN S.R.B. PG 6 1 e. j. What is the full name of the doctor or the medical facility where the diagnosis was made? IRECORD IN S.R.B. PC 6 1 What is the full name of the doctor or the medical facility you last consulted about cancer of the (BODY PART)? IRECORD IN S.RJB. PC 6 1 k. What is the full name of the doctor or the medical facility you last consulted about cancer of the (BODY PART)? IRECORD IN S.R.B. PG 6 1 p. What is the full name of the doctor or the medical facility where the diagnosis was made? IRECORD IN S.R.B. PC 6 1 q. What is the full name of the doctor or the medical facility you last consulted about cancer of the (BODY PART)? IRECORD IN S.R.B. PG 6 I f. During what month and year did you last consult (NAME FROM Q.iO L. During what month and year did you last consult (NAME FROM Q.k) IRECORD IN S.R.B. PG 6 | r. During what month and year did you last consult (NAME FROM Q.ci) g. What treatments or medicines did you take for cancer of the (BODY PART)? 1 MULTIPLE RECORD BELOW] m. What treatments or medicines did you take for cancer of the (BODY PART)? (MULTIPLE RECORD BELOW | s. What treatments or medicines did you take for cancer of the (BODY PART)? IRECORD IN S.R.B. PG 6 I Chemotherapy. '.( 16( .' -1 Other (SPECIFY) Chemotherapy... (lb( -1 Other (SPECIFY) IRECORD IN S.R.B. PC 6 I [MULTIPLE RECORD BELOW] Chemotherapy . . . (16( Surgery (TR -1 Other (SPECIFY) .(18( h. During what month and year did you first receive (EACH TREATMENT CODED IN Q.g) fur cancer of the (BODY PART)? MONTH YEAR RadiaI I I I 1 1 tion.... I I l-l 1 1 (19) (20) (21) (22) MONTH YEAR Chemo1 1 1 1 1 1 therapy. I 1 l-l 1 1 (23) (24) (25) (26) MONTH YEAR Surgery.. I i 1-1 1 1 T^7) (28) (29) (30) MONTH YEAR I t I I 1 1 Other.... I l l - I l l (31) (32) (33) (34) (GO TO NEXT BODY PART) n. During what month and year did you first receive (EACH TREATMENT CODED IN Q.m) for cancer of the (BODY PART)? t. l -1 During what month and year did you first receive (EACH TREATMENT CODED IN Q.s) for cancer of the (BODY PART)? MONTH YEAR RadiaI I I I tion.... | | I I (19) (20 (21) (22) MONTH YEAR Chemo1 1 I I therapy. I I ' I I (23) (24) (25) (26) MONTH YEAR I I 1 1 Surgery.. 1 1 1 1 Surgery.. 1 1 l-[ .1 (27) (28 J 29) UU) T27) (28) (29) (iff MONTH YEAR MONTH YEAR 1 1 I I I I I I 1 1 1 1 Other.... I I - I I I Other (3D (32T (33) (34) (3l) (32) (33) (34) (RETURN TO Q.37a) (GO TO NEXT BODY PART) MONTH YEAR Radia] 1 I I tion.... 1 1 l-l 1 1 (19) (20) <21) (22) MONTH YEAR Chemo1 T 1 1 I therapy. 1 1 ~l 1 1 (n) (24) (23) (26) MONTH YEAR i i ri i 04 /y-80 7 y-80 127 lid 79-80 �CARD 142 167. 812039 Additional Cigarette Periods (Q.45-47) 167c. After that, about how many packs per week did you smoke? i r •. I I packs per week d. Until what month and year did you continue to smoke (NUMBER) packs per week on a regular basis? MONTH YEAft 1 ~ 1 T 1 T T ( I F DATE I S T H E SAME A S Q.45c: RETURN T O Q.47a. I I j-l I I ALL OTHERS: CONTINUE) (14) (15) (16) (17) e. After that, about how many packs per week did you smoke? I J 1 \ | [_ packs per week (18) (19) f. Until what month and year did you continue to smoke (NUMBER) packs per week on a regular basis? MONTH YEAR T I T T f ~ " T (RETURN T O Q.47a) I I l-l I I . . '(20) (21) (22) (23) ' 168. 168c. Additional Pipe Periods (Q.48-50) After that, about how many pipefuls per week did you smoke? T I I I I I pipefuls per week (25) d. Until what month and year did you continue to smoke (NUMBER) pipefuls per week on a regular basis? T MONTH j.__ YEAR ' j. I -j | (ip DATE ig TIIE SAME AS Q>48c. RETURN T0 Q 5 Q_ j I |-j j | ALL OTHERS: CONTINUE) (26) (27) (28) (29) • e. After that, about how many pipefuls per week did you smoke? T J i! r . [ pipefuls per week (JO) (31) f. Until what month and year did you continue to smoke (NUMBER) pipefuls per week on a regular basis? .MONTH YEAR 1 1 Tl \ T (RETURN TO Q.SOa) I • I l-l I I (32) (33) (34) (35) 128 �CARD H2 169, 169c, 812039 Additional Cigar Periods (Q.51-53) After that, about how many cigars per week did you smoke? 1 I I J_ I I cigars per week (36) (37) d. Until what month and year did you continue to smoke (NUMBER) cigars per week on .1 regular basis? MONTH YEAR T I T T I T (IF DATE IS THE SAME AS Q.Slc: RETURN TO Q.53a. I ! I- 1 I I AU. OTHERS: CONT1NIIK) e. After that, about how many cigars per week did you smoke? I J I I I |_ cigars per U2) (A3) week f. Until what month and year did you continue to smoke (NUMBER) cigars per week on a regular basis? MONTH YEAR 1 I FT I T (RETURN TO Q.53a) I . I l-l I I (44) (45) (46) (47) 129 �CARD 142 ; 170. 170c. 812039 Additional Drinking Periods (Q.54-56) After that, about how many drinks per week did you have? I I I drinks per t* ) ( 49) week d. Until what month and year did you continue to drink (NUMBER) drinks per week on a regular basis? MONTH YEAR 1 T T1 I T ( I F DATE I S T H E SAME A S Q.54c: RETURN T O Q.56a. I I l-l I I ALL OTHERS: CONTINUE) C$0) C51 ) (52 ) ( 53 6' After that, about how many drinks per week did you have? I J I I | [_ drinks per week 64 ) ( 55) f. Until what month and regular basis? MONTH 1 I I I 66 ) 67 171, 171c. year did you continue to drink (NUMBER) drinks per week on a YEAR T 1 \ T (RETURN T O Q.56a) l-l I I ) 68 ) 69 ) Additional Marihuana Periods (Q.57-59) After that, about how many joints per week did you smoke? I J I T | [ joints per week fcO ) fcl ) d. Until what month and year did you continue to smoke (NUMBER) joints per week on a regular basis? MONTH YEAR 1 I TT I T (IF DATE IS THE SAME AS Q.57c: RETURN TO Q,59a. I I J - | j J ALL OTHERS: CONTINUE) 62 )( 63) ( 64) ( 65) e. After that, about how many joints per week did you smoke? I joints per week (66) (67) f. Until what month and year did you continue to smoke (NUMBER) joints per week on a regular basis? MONTH YEAR 1 I TT I T (RETURN TO Q.59a) I I l-l I I '«S8 ) (69) ( 70) (71 ) 130 �CARD 0.172. 812039 130 Additional Jobs and Toxic Substances (Q.13Q) NINTH JOB EIGHTH JOB SEVENTH JOB THAND RESPONDENT CARD "G"| 172a. W h i l e w o r k i n g at (EMPLOYER) as (DUTIES), d o / d i d you come in contact with any of the subs tances on t h i s c a r d ? By c o n t a c t , I mean that you inha l e d , t a s t e d , had skin contact with, or were radiated by any of these s u b s t a n c e s ? iMULTIPLE RECORD! ]ASK Q.172b FOR EACH SUBSTANCE CODED IN 0.172a. 1 172b. In g e n e r a l , how many days a month ( d o / d i d ) you come in contact with ( SUBSTANCE ) ? 01. .( 12( O2..(l3( O3..(ia( O4..(l5( 1 01. .1 1 1 -1 -1 -1 -1 ! ! ! 04. .1 (19)(20) Ol..(12( O2..(13( 03..(14( 04..(15( O5.(16< -1 06.(17( -1 O 7 . ( l 8 ( -1 < I F "07," SKIP TO NEXT JOB) I ! 1 I2l)(22) -1 -1 -1 -1 Ol..(12( O2..(13 O5.(16( -1 O6.(17( -1 O 7 . ( l 8 ( -1 (IF "07," SKIP TO NEXT JOB) -1 O5.(16( O6.(17( -1 -1 Q7.(18( -1 (IF "07," SKIP TO NEXT JOB) i l l I I ! 01.. 1 1 1 04..! 1 t (19X20> (21X22) 01..1 (19) (20) (21X22) i l l 02. .1 I 1 <23)(2A) I I I I I I 02.. | I 1 05.. 1 1 ! (23X24) (25X26) 1 1 1 1 1 1 1 ! 02. .1 1 ! 05. .1 (25) (26) (23) (24) T i l 03.. 1 I 1 (27)(28) !lF ANY SUBSTANCE CODED IN Q.172b, ASK Q.172cl 172c. W h i l e you were on that job, how o f t e n di d you wash to remove the (SUBSTANCES) or use pro- 1 1 05. .1 t 1 (25X26) 1 1 1 06.. 1 1 1 (29><30) I I I I I I 03.. | I I 06-.. 1 1 1 (27)(28) (29X30) 03..I I (27) (28) All the time(31( Some of. -l](ASK Q. | 172d) Never -3 (GO TO NEXT JOB) T (29) (30) All the time. (31 ( Some of -ll(ASK Q. V 172d) All the time. (31 ( Some of Che time -1~1(ASK Q. f 172d) -2j Never -3 (GO TO NEXT JOB) Never. -3 (GO TO NEXT JOB) J soine of the t i m e , or never? lHAND RESPONDENT CARD "H" | 172d. Which of the following did you use on t h a t iob 7 1 MULTIPLE RECORD IF NECESSARY! I Air f i l t e r (32( -?\ • Goggles (33( -i (-GO TO Face shield (34l -1 NEXT JOB) Special clothing. .(35( -1 Washing f a c i l i t i e s ( 3 6 ( ~~1 07 79-80 Air f i l t e r (32( -1} Goggles (537 -1 ( G O T O Face shield (34( -1 ' NEXT Special clothing. .(35( -1 JOB) Washing facilities(36C ~-l 08 79-80 Air filter (2 3 ( -1 Goggles (33T -1 (GO TO Face shield (34 ( -1•NEXT JOB) Special clothing..(35( -1 Washing facilities(36( -1 09 79-80 �CARD Q.172. IHAND RESPONDENT CARD "G" 1 172a. W h i l e working at (EMPLOYER) as ( D U T I E S ) , d o / d i d you come in c o n t a c t w i t h any of the subs tances or t h i s c a r d ? By c o n t a c t , I mean t h a t you inha led, t a s t e d , had s k i n c o n t a c t w i t h , or were r a d i a t e d by any of these s u b s t a n c e s ? [MULTIPLE RECORDl lASK Q.172b FOR EACH SUBSTANCE CODED IN Q.172a. 1 1 7 2 b . In g e n e r a l , how many days a month ( d o / d i d ) you come in contact with (SUBSTANCE)? Ol..(12( 02..(13< O3..(14( O4..(15( -1 -1 -1 -1 Ol..(12{ O2..(13~t 03..U4( O4..(15( 0 5 . U 6 C -1 06. (ITT -1 O 7 . ( l 8 ( -1 (IF "07," SKIP TO NEXT JOB) -1 -1 -1 -1 1 1 05.. 1 1 (25X26) I I I 06. .1 1 1 (29X30) ! ! i (21M22) All the t i m e ( 3 l ( Some of Never 1 ! 04.. 1 .. -l"\(ASK Q. f 172d) -3 (GO TO NEXT ~~ JOB) Air f i l t e r (32( Goggles (33( Face shield (34( Special clothing. .(35( Washing f a c i l i t i e s ( 3 6 ( -1 -lUGO TO -1|NEXT JOB) -1 ~lj 10 79-80 -1 I 01.. I 1 1 1 I I I 03. .1 1 ! 06. .1 1 1 ( 2 7 ) (28) (29) (30) I -1 O4..(l5l I I I 1 1 1 02. .1 1 1 05. .1 1 i (23X24) (25X26) I I I 03. .1 1 1 (27)(28) 0 2 . . (IT? I I I 1 1 1 01.. 1 1 1 04.. | I I (19X20) (21X22) t i l 02.. 1 1 1 (23X24) [HAND RESPONDENT CARD "H" ! 172d. Which of the following did you use on t h a t job? IMULTIPLE RECORD IF NECESSARY! O5.(16( O5.(16( -1 0 6 . ( 1 7 ( -1 O7.(18( -1 (IF "07," SKIP TO NEXT JOB) I I 01.. 1 (19X20) I I F ANY SUBSTANCE CODED IN Q.172b, ASK 0^172c! 172c. W h i l e you were on t h a t job, how o f t e n d' d you wash to remove the (SUBSTANCES) or use prot e c t i v e gear — w o u l d you say all of the t i m e , some of the t i m e , or never? TWELFTH JOB ELEVENTH JOB TENTH JOB Additional Jobs and Toxic Substances (Q.130) (CONT'D) 812039 130 I I I I I T T T 1 02..I T T i (23) (24) i r 03..I I I (27)(28) All the time.(31( Some of the t ime Never -3 (GO TO NEXT JOB) Never. 11 79-80 I 04..J (21 )(22 ) -l)(ASK Q. f 172d) -l\ -l}(GO TO -ITNEXT -1 JOB) -1J O7.(18( (IF "07," SKIP TO NEXT JOB) (19)UO) All the time. (31 ( Some of Air f i l t e r (32( Goggles.. (33 ( Face s h i e l d (34( Special clothing. .(35( Washing f acilities(36( - 06. (rn - (25) (26) I I T I (29) (30) -l) (ASK 0. 172d) _-3 (RETURN TO Q.131) Air filter (32 ( -1 Goggle* (33 ( -1 RETURN TO Face shield . 3 ( -1 Q.131) (4 Special clothing..(35( -1 Washing facilities(36( -1 12 79-80 �CARD 173. 133 812039 Additional Countries and Toxic Substances (Q.133) lASK Q.173b FOR EACH SUBSTANCE CODED IN Q.173a. 1 173b. In g e n e r a l , how many days a month ( d o / d i d ) you come in c o n t a c t with (SUBSTANCE)? Ol..(1h( 02. . ( ) 7 ( 03..qtT O4..(,q( 1 1 1 01. .1 1 1 (23) (24) ! 02. .i ! 1 -1 -1 -1 -1 I I ! OA..I 1 1 (25) (26) 1 I I 05. .1 ( 2 7 ) (28) i l l 03. .1 1 ! (31)(32) I ! ! 1 (29) (30) I I I 06..! I 1 (33)(34) I I F ANY SUBSTANCE CODED IN Q.173b, ASK Q.173c| 173c. Did vou wash to remove the (SUBSTANCE) or did you use protective clothing or gear when stationed in (COUNTRY) — all of the time J some of the time, or never? Never IHAND RESPONDENT CARD "H"| 173d. Which of the following did you use on that job? I MULTIPLE RECORD IF NECESSARY 1 O5.(20( -1 06. (HI -1 07.(22T -1 (IF "07," SKIP TO NEXT COUNTRY) Air f i l t e r (:te( Goggles.(-}7< Face shield (-^( Special clothing- .(:rt< Washing facilities(/|i-i( « ASK Q. 173d) -3 (GO TO NEXT COUNTRY) "U -ll(GO TO -ifNEXT -IICOUNTRY) -11 ri7 (14-15) (14-15) (14-15) IHAND RESPONDENT CARD "G" 1 173a. In your job assignments while stationed in (COUNTRY), ( t h a t t ime) ( d o / d i d ) you come in contact w i t h any of the following s u b s t a n c e s ? NINTH COUNTRY EIGHTH COUNTRY SEVENTH COUNTRY Ol..(lb( O2..(17( O3..(l8l O4..(19( 1 -1 -1 -1 -1 I -1 O2..(17( -1 Q3..(18( -1 Oi..(19j__-l 05-(20( -1 O 6 . ( 2 l ( -1 07. (22? -1 (IF "07," SKIP TO NEXT COUNTRY) I T 1 ! I 04..II 01..I 01.. 1 1 I 04.. [ 1 1 (23)(Z4) (Z5)(Zb) (25X26) 1 1 ! I I ! 02.. 1 1 ! 05.. 1 I 1 (27)(28) (29)(30) J 02 27)(28) 1 1 1 1 I I I 03.. 1 1 1 06.. 1 1 1 (31X32) (33)(34) -IHASK Q. T 173d) -2J T Never -3 (CO TO NEXT COUNTRY) (33X34) -ll(ASK Q. 173d) All the Some of the time Never. Air f i l t e r ($(,( -1 Goggles ( i y ( -1 (GO TO Face shield (^ a ( -1 NEXT Special clothing. .(-jq( -1 COUNTRY) Washing facilities(/| 0 ( ~-l [ 05.. I _ I I (29)(30) (31)(32) All the t i m e . ( 3 5 ( Same of the time 08 O5.(20( -1 O6.(21( -1 Q7.(22( -1 (IF "07," SKIP TO NEXT COUNTRY) _-3 (GO TO NEXT COUNTRY Air filter (6 3( Goggles (37 ( Face shield (8 3( Special clothing..(39( Washing facilitie?(4Q( 09 -1 -1 /GO TO -1 NEXT -1 COUNTRY) ~1 �CARD 173. 812039 133 A d d i t i o n a l C?-j-:ries and Toxic Substances (0-133) (CONT'D) ELEVENTH COUNTRY TENTH COUNTRY lASK Q.173b FOR EACH SUBSTANCE CODED IN Q.173a. ! 173b. In general, how many dayt a month (do/did) you come in contact with (SUBSTANCE)? 1 O5.(20( -1 02..(17( -1 O6.(2l( -1 03.. ( s l ^ -1 Q . ; ( -1 7(; 04. . 19( -1 ( (IF "07," SKIP TO NEXT ' COUNTRY ) 1 01. .1 i ! i 1 ! 04. .1 i 1 ! 03. .1 i 1 (31X32) IIF ANY SUBSTANCE CCi-D IN Q.173b, ASK Q.173c! 173c. Did you was'- : : remove the (SUBSTANCE) i ! < 25X26) (23) (24) ! i ! 02..! i ! (27) (28) ! 1 ! ! I 05. ( 29) ( 30) 1 i 1 06. .! 1 I (33) (34) -A (ASK 0- Never when stationed in CIVN'TRY) — some of the time, cr -ever? All the tine(35( Some o f 01..(16( -1 O2..<17( -1 03..(T8T~~-1 O5.(20( -1 O6.(2l( -1 07.C22T~-1 (IF "07," SKIP TO NEXT COUNTRY ) -,) " ' " Air filter ( 36( -1 Cosgles ( 37( -1 (GO TO ^ace shield.. , . . 3^( -1. NEXT ..( Special clothing. . 39( -1 COUNTRY) ( Washing facilit ies( 4Q( '0 *'" •• -1 -1 -1 O5.(20( -1 O6.(21< -1 07.(22( -'(IF "07," SKIP TO NEXT COUNTRY ) ! 1 1 1 I 1 1 ! 01.. 1 ! 1 04 ..! (23X24) (25X26) 1 ! I .! 1 1 02. .1 i ! 05. .1 i I (29) (30) (27) (28) i i ! 02.. 1 1 1 (27X28) 1 1 1 05 ..! i 1 (29X30) 1 I I I 1 i 03.. I I 1 06..! I 1 (33X34) (31X32) i 1 1 03..! 1 1 (31X32) 1 1 ! 1 ! 06 ..! (33X34) -fl (ASK Q. V 173d) . -2J All the tir.e.(35( Some o f the time -1)(ASK Q. V 173d) -2 -3 (GO TO NEXT COUNTRY ) All the time.( 35( Some of IHAND RESPONDENT CA=I "H"! 173d. Which of tre f:llowing did you use on that job? iMTLTiri. RECORD IF NECESSARY! O2..(r7( O3..(l8( 04..C19C 1 1 I 1 i ! I 04..! i 1 01. .1 1 (25X26) (23)(24) -3 (GO TO NEXT COUNTRY) all of the time (14-15) (14-15) (14-15) iHAND RESPONDENT CARI "G" 1 173a. In your job assignments while stationed in (COUNTRY), (that ti-« 'do/did) you come in contact with any of ~~e following substances? TWELFTH COUNTRY Never -3 (RETURN TO Q.134a) Air filter. . . . (36( (37( Goggles ( 1ST Special clotning. . ( 39i Washing faeilities( 4p( 11 79-80 Air filter ( 6 _-! (RETURN 3( -1 Goggles (37( _-1 TO Q. -1 (GO TO -1 NEXT Face shield (8 3 ( -1 1 TAa 1 -1 COUNTRY) Special clothing. .(39( Washing f acilities(40( -1 ( 12 79-80 �DEPARTMENT OF THE AIR FORCE WASHINGTON DC 20330 OFFICE OF THE SECRETARY James W. Doe 1215 Middle Grove Norfork, MD 23456 Dear i Mr Doe • ' The Air Force will soon begin conducting a very comprehensive health assessment of certain Air Force members who served our Nation in the Vietnam conflict. This health assessment is part of a medical study designed to help determine if you or your fellow Vietnam veterans may have had any compromise to your health as a result of exposure to the complex environment of Southeast Asia. Scientists at the USAF School of Aerospace Medicine have been given the responsibility for conducting this important project. The Air Force Surgeon General will contact you soon with more details and ask for your voluntary participation. . . A major focus of the President's program for veterans is the resolution of health issues raised by them. The Air Force and I are committed to doing our part in resolving these issues. I ask that you help us and all Vietnam veterans by voluntarily participating in this major study. Sincerely, Verne Orr Secretary of the Air Force 135 �DEPARTMENT OF THE AIR FORCE HEADQUARTERS UNITED STATES AIR FORCE BOLUNG AFB DC 2O332 James W. Doe 1215 Middle Grove Norfork, MD 23456 Dear Mr Doe The Air Force is conducting a very comprehensive health assessment of certain Air Force members who served our Nation in the Vietnam conflict. The USAF School of Aerospace Medicine has been given the responsibility for conducting this study. The.purpose of the study is to determine whether there may be any causal relationship between health problems and exposure to the complex and unique environment of the war in Southeast Asia. Simply stated, we do not know if such health effects exist. You are being asked to voluntarily participate in this study because of your unique Southeast Asia experience. Your participation is critical to the success of this study. However, you should not view this invitation to participate as a cause for alarm nor as an implication that you are at risk for any known disease. To insure the scientific validity of the study, both an in-depth interview and a detailed physical examination will be conducted. The administration of the interview will begin soon under the direction of a nationally recognized health survey organization. You will be contacted by phone or letter to arrange a convenient time for an in-home interview which will take from two to three hours. Shortly after the interview you will again be contacted to schedule a physical examination at a nationally recognized civilian medical facility. The physical examination will take approximately four days. Every effort will be made to minimize disruption of your normal activities and to facilitate your participation in the study. Travel and per diem will be paid by the Air Force. For those not precluded by law, a stipend of $100 per day will be paid as a partial compensation for your time. Our intent is to maintain all individual health data in strictest confidence. In case outside parties attempt to gain access to the data, the Air Force and the Department of Justice are committed to protect this individual confidentiality. Only in the event of an adverse final court decision, or in the highly unlikely instance where serious medical deficiencies must be shared with appropriate medical authorities to protect public health and safety, will any personal health data be revealed. You are referred to the Fact Sheet for further information regarding this matter. 136. �This is perhaps one of the most important health studies undertaken by the Air Force. Your voluntary participation is critical to its success. Although you may feel healthy, numerous Vietnam veterans believe that they have illnesses which may be attributable to service in Southeast Asia. The only way we can get clarification of these difficult questions is through your cooperation and participation. Sincerely PAUL W. MYERS Lieutenant General, USAF, MC Surgeon General 1 Atch Fact Sheet 137 �FACT SHEET INTRODUCTION - The USAF School of Aerospace Medicine, Brooks AFB, Texas, is conducting the study. You are being invited to participate in this study because of your specific duties and period of assignment in Southeast Asia. PURPOSE To determine whether there is a causal relationship between adverse health effects and exposure to the complex environment of Southeast Asia. METHODS - An in-depth health questionnaire will be administered to you by a member of a health evaluation team from Louis Harris and Associates, Inc. - A complete profile of your current health will be obtained by a physical examination which will be conducted by a nationally recognized outpatient clinic. - Follow-up abbreviated health questionnaires and physical examinations will be conducted at years 3, 5, 10, 15, and 20 of the study. - Travel expenses (including board and lodging) for the physical examination will be paid by the Air Force. - Stipend of $100 per day will be paid to study participants who are not on active duty, Government employed or otherwise precluded by law from receiving such a stipend. t - Confidentiality is to be maintained except in two cases: - A judicial order to release personal medical data following an Air Force and Justice Department defended lawsuit. Serious medical findings which impact public health and safety. Two examples of situations in which public health and safety would raise the questions of disclosure are: a participatnt has typhoid fever, a participant who directly impacts the safety of others either in his profession, or as a volunteer, is found to have a serious nerve, heart or mental disorder. In this instance a committee composed of a physician (whose specialty is the area of the identified problem), a physician of your choice, a flight surgeon, a judge advocate (lawyer) and a representative from your field of expertise w i l l be convened to review the medical findings. Before any disclosure is made to medical authorities, the committee must determine that the findings jeopardize the public health and safety. 138 �BENEFITS TO YOU - You will receive a complete health review and physical examination of top level executive calibre at no cost to yourself. - You will be completely informed of all examination results. The information from this study will be provided to a physician of your choice if you so request. -. Questions concerning the study may be referred to the USAF School of Aerospace Medicine, Epidemiology Division, Brooks Air Force Base, Texas 78235, or by calling collect AC 512 536-3309. - If you have recently changed your address or have an unlisted phone number, please advise the USAF School of Aerospace Medicine at the above address and phone number so that your records may be properly updated. 139. �LOUIS HARRIS AND ASSOCIATES, INC. 630 FIFTH AVENUE NEW YORK, NEW YORK 10111 Dear Mr. Doe Louis Harris and Associates has been asked by conduct interviews for a health study of Air served during the Vietnam conflict. The U.S. Medicine is undertaking this study in order to health effects of having served in Vietnam. the United States Air Force to Force pilots and servicemen who Air Force School of Aerospace answer questions about possible We need your cooperation in this study. The validity of the results of the study depends on the willingness of veterans like yourself, who have been selected for the survey, to participate. Reliable information will enable us to reach sound conclusions of vital relevance to all Vietnam veterans. One of our interviewers will be calling you in the next two weeks to arrange an appointment with you. The interview will cover many aspects of your military experience, occupational experience, family history, health history and health care utilization. Since the interview may take one or two hours to complete, we will schedule the interview at your convenience. Thank you for your cooperation. tant project. I hope that you will join us in this impor- Sincerely, LOUIS HARRIS 140 �;iS HARRIS .AND ASSOCIATES. INC. i_OdS HAPRIS INTEPNATION A _ '.. !<,. M A R » . * - rc.A-.CE - ;>PiNiOW P»ES£APCH i.'' RL £ v . ^ . ^ E N ^ ^ E 7SOCS PAP*- re TEL »6fc^ Or-260 CtNTRE -*^ *£LBEC^ 5T. »AN^£ TE..C*: » O N O O N WIM 0AB tNOlpA'Jt ffrn^r, r . "EL Oi - -66 - * > » ! TtLtJt' »- - PRIVACY ACT STATEMENT - EPIDEMIOLOGIC STUDY AUTHORITY: Section 133, 1071-87, 301.2, 5031 and 8012, Title 10, United States Code and Executive Order .9397. PRINCIPAL AND PURPOSE(S): The purpose of requesting personal information is to assist medical/technical personnel in developing records relative to your participation in an approved epidemiologic investigation. The Social Security Number (SSN) and Armed Forces Service Number (AFSN) are necessary to identify the person and records. ROUTINE USES: This information will be used to initiate, coordinate, and conduct the investigation. It will be used to compile statistical data, but information allowing identification of the individual volunteer will not be included. Data and results from this investigation may be used to supplement other approved research studies conducted at the USAF School of Aerospace Medicine or at other Federal agencies engaged in the: conduct of similar studies. WHETHER DISCLOSURE IS MANDATORY OR VOLUNTARY, AND EFFECT ON INDIVIDUAL FOR NOT PROVIDING INFORMATION: Disclosure or requested information is voluntary. If the information is not furnished, acceptance as a subject is not possible. This is an all-inclusive Privacy Act Statement which will apply to all requests for personal information made by medical/technical personnel during the time you are a volunteer subject. A copy of this form will be placed in your investigation subject folder as evidence of tnis notification. Your signature merely acknowledges that you have been advised of the foregoing. If requested, a copy of this form will be furnished to you. Signature of Volunteer SSN 141 ' Date !f �LlhEEVENTS CHART U.S. Air Force Survey YOUR AGE THEN DEATH SCHOOLS MILITARY EXPERIENCE OTHER JOBS MARRIAGE 1/19 CHILDREN FAMILY MAJOR ILLNESS OTHER SPECIAL EVENTS �SHOW CARD "A1 STUDY NO. 812039 High School Diploma High School Equivalency Diploma Associate of Arts (A.A.) Bachelor of Arts (B.A.) or Bachelor of . Science (B.S.) Masters (M.A. or M.S.) I Doctorate (Ph.D., M.D., Ed.D., Sc.D.) Others 143 �SHOW CARD "B1 STUDY NO. 512039 01 Aerospace 02 Aircraft^ 03 Agriculture 04 Automotive 05 Chemical 06 Electronic 07 Mining 08 Pest Control 09 Petroleum 10 Textile 11 None Apply 144 �SHOW CARD "C" STUDY NO. 8120: 01 Pill 02 Douche 03 Foam 04 Jelly, Cream, Suppository 05 IUD 06 Condom, Rubber 07 Diaphragm 08 Diaphragm and Jelly 09 Rhythm - Calendar 10 Rhythm - Temperature 11 Withdrawal 12 Other 145 �SHOW CARD "D( STUDY NO. 812039 a. Sterility due to surgery b. Known sterility due to injury, accident, or illness c. Sterility due to unknown causes d. Impotence e. Other known medical/physical conditions f. Some other reason 146 �SHOW •''" STUDY:^or 812039*'"' * 147 �STUDY NO. 81203C SHOW CARD "F1 Very often Fairly often Sometimes Almost never Never 148 �SHOW CARD "G1 STUDY NO. 81 01 Asbestos 02 X-ray or nuclear radiation 03 Industrial chemicals 04 Defoliants or herbicides 05 Insecticides or pesticides 06 Degreasing chemicals 07 None of these 149 �SHOW CARD "H STUDY NO. 81203- Air filter Goggles Face shield Special clothing Washing facilities 150, �SHOW CARD "I1 STUDY NO. 81 A. $5,000-$9,999 B. $10,000-$14,999 C. $15,000-$19,999 D. $20,000-$24,999 E. $25,000-$29,999 F. $30,000-$34,999 G. $35,000-$39,999 H. $40,OQO-$44,999 I'. $45,000-$49,999 J. $50,000-$54,999 K. $55,000-$59,999 L. $60,000-$64,999 H. $65,000-$69,999 N. $70,000-$74,999 0. $75,000-$79,999 P. $SO,000-$84,999 Q. $85,000-$89,999 R. $90,000-$94,999 S. $95,000-$99,999 T. $100,000 or more 151 �CARD 135 Case No._ SEI.K-ADMINtSTI-HF.I) SHKKT These next questions are about liow you have felt du_ri_nj> the last three months. for each question, please circle a number for the one answer that comes closest to the way you have been feeling during the last three months. 1. How often did you become nervous or jumpy when faced with excitement or unexpected situations during the past 3 months? Always 1 (17) Very often..... 2 Fairly often 3 Sometimes k Almost never 5 Never. 6 2. How much of the t i m e , during the past 3 months, did you feel relaxed and free of tension? All of the time 1 Most of the time 2 A good deal of the time... 3 Some of the time 4 A little of the time 5 None of the time (18) 6 3. Huring the past 3 months, how much of the time have you felt restless, fidgety, or impatient? A l l of the time Most of the i iuie 2 A good d e a l of the t i m e . . . 3 Home of tht' t i m e 'i A I iule of the time 5 None of the time l<, 1 (ly) 6 During the past 3 months, have you been anxious or worried? Yes, extremely so, to the point of being :. ick or almost sick 1 Yes, very much so 2 Yes, q u i t e a l o t . 3 Yes, some, enough to bother me 4 Yes, a l i t t l e b i t 5 No, not at a l l 6 I'l.KASE TUK.N OVKR 152 (-'") �n. .w o i i u i v , o i i r i n K tin- p a s t .) m o n t h s , have you lieen wak i'n(> up f eel id)- f resli an<j l.'Stfll? " '. ' . - ' " . " "','• • • - A l w t i V i > • evv.i y d a y ' ........ ' / ' , 1 A l m o s t e v e r y (lay . . . . . . . . . . " . ;, ' . ' ' • ••'-''•' ••'<'. '(21) 2 Mont <l;iys ....... . . . . ...... 3 Some. d a y s , ))nl u s u a l l y not 'i H a r d l y i.'Virr 5 ........ . ....... Never wake u p ( o e l i n g u - s t i ' d . . ---- ...... ........ .6 '). l u i r i i i ) ' , t h e p.'iHt '< iiiontlit. , how o l t i ' n i l l . I yoiii luiiuls shake when you t r i e d lo do . iu.)int.'Ui in^? A l w a y s , eve -y <l.iy ......... V e r y ol I #.n .......... . ..... 2 Most d a y s ........... . ..... 3 Some d a y s , hut u s u a l l y not A Hardly ever.... 7. 1 5 ........... (*-) How much of the t i m e d u r i n g the uatil 3 months have you f e l t c a l m and p e a c e f u l ? A l l of the t i m e . . . . ....... 1 (2!1) Most of the time .......... 2 A good deal of the time... 3 Some of the t ime ..... . . .. . (t A l i t t l e of the time...... 5 None of the time .......... 6 ti. During the past 3 months, how oltcjn did you put rattled, upset, or confused? Always.................... 1 Very ofti.-n ................ (24) 2 Ka i r 1 y u ( t *.' n ............. . 1 Some t ilia1 s ........... . . . A ... Almost never. ........... .. 5 Never.....................0 'i. How much have you been bothered hy nervousness, or your "nerves," during the past 3 months? Extremely so, to the point where I could not take care of things .......... 1 Very much bothered. .......... ,.....»•• Bothered ijnite a lot by nerves. ........ 2 3 Bothered some, enough to notice ........ 4 Bothered .jus I a 1 i 1 1 1 e. by nerves ...... 5 Not bothered al all hy nerves....... ... 6 you Tiavy ccimpleteTTtcma' 1 - j ' r e t u r n t f i j i h e f C 153 ( *-->) . ; �Louis HARRIS AND ASSOCIATES, INC. 63O NEW TEL FIFTH Y O R K . NEW AVENUE YORK I2iej 9 7 5 - I 6 0 O T E L E X I O I I I 148363 LOUIS H A R R I S I N T E R N A T I O N A L . INC OPINION R E S E A R C H C E N T R E LOUIS H A R R I S FRANCE 21 RUE VIV1ENNE 3 O W E L B E C K ST 7SOO2 PARIS. FRANCE LONDON WIM 6AB E N G L A N D TEL. oi-aoo -0034 TELEX: 200001 F TEL O I - i a e - 5 I S I TELEX UNITED STATES AIR FORCE HEALTH STUDY Name of Medical Provider/Medical Facility Street Address City State Zip 1 Phone Number Dear Doctor or Administrator: I am participating in a survey conducted for the United States Air Force to gather information on the health of current and former Air Force personnel. As part of this survey, medical providers who have delivered health care services to me are being asked to supplement the information that I have already provided to the study. By this statement or a pfiotocopy of it; 1 heret>y authorize and request you to furnish the United States Air Force Health Study with any medical information in your records concerning the health services received by me, These services were provided during the period to Thank you very much. Sincerely, Resp. # Signature of Patient FOR OFFICE USTOMLT: Date I I 2-a-lO3 �HARRIS AND ASSOCIATES. 'NC , £ Vv. "• C1 ° «•..' N E .'. T. Z ? $•>* •,"-'• '6CV '!.,£« 3 0 *e^8ECK ST ..CNCON W)M 6A6 TEL c .*<)£• SIB< TC. Oi- zfto -ttibJ Tluik*: zooeci Dear Louis Harris and Associates has been asked by the United States Air Force to conduct a study &f the health of Air Force pilots and servicemen who served during the Vietnam conflict. The U.S. Air Force School of Aerospace Medicine is undertaking this study in order to answer questions about possible effects of having served in Vietnam. I have just completed an interview with Louis Harris and Associates on the United States Air Force Health Study, As part of this study, they would like to interview the former wive^s of study participants. You will be asked to provide information on health and health care services. It is essential to the accuracy and completeness of the study that all selected participants and their families participate in the study. Reliable information will help produce sound conclusions of vital relevance to all Vietnam veterans and their families. I would appreciate it very much if you also would grant a representative of Louis Harris and Associates an interview. Shortly after receiving this letter, you will be called on by an interviewer from Louis Harris and Associates who, at your convenience, will either conduct the interview or set up an appointment. The interviewer will answer any questions you may have about the study. Thank you. Sincerely, (SIGNATURE OF STUDY RESPONDENT; (PRINTED NAME OF STUDY RESPONDENT) 155 �FOR OFFICE USE ONLY: LOUIS HARRIS AND ASSOCIATES, INC 630 Fi fth Avenue New York, New York 10111 Case * # 812039 Air Force Health Survey Respondent INTERVIEW EVALUATION INTERVIEWER: "COMPLETE THE FOLLOWING IN PRIVATE" IMMEDIATELY YOUR BEST JUDGMENT TO ANSWER FACH ITEM. _ AFTER THE INTERVIEW, USING] _J • lace of respondent: Black.... Nonblack. 2a. Did the respondent want to terminate the interview before it was finished? No (SKIP TO Q3a) Yes (ANSWER 2b AND ?c) 2b. At what question number or during what question series? '(. l-'hat was the reason? la i'ore there any (other) significant problems uurmg Uie interview? No. Yes (SKIP TO Q4a) (ANSWKR 3b) 3b. Describe the problems. 4a. Did respondent refer to records during the interview? No Yes (SKIP TO Q5a) (ANSWER 4b) 4b. What records did the respondent use? : 5a. Was anyone else present at any time during the interview? No Yes ^ (SKIP TO Q6) (ANSWER 5b and 5c) 5b. Who was present? | RECORD RELATIONSHIP | 5c. During which section(s)? _ 6. Length of interview: minutes 156 �l', MAR'ftl'.. '-VU) Av.''iCI.AU$",' INC, " '" Sjiydy « 812039 630. Fifth Avenue ... New. York, New York luli'J' AIR FORCE HEALTH SURVEY TO: New York Office. Louis Harris'- and Associates FROM: . . ; • ' . ' _ Int.erviewer Nairn? - PYease Print _ " „_...*_ ..... ' _____ " Ihi-, pai'kdiji.' iiontrt in , the1 Col lowing mdterial for Study*"*>uhject Respondent" Number W r i l e iii NUMW..K of'- card item being sent; on the l,ine_jit the rijjht —- "' .STUDY SUUJDCT INTERVIEW . ' Study Subject Name 'Assignment Sheet. " Study Subject Prlvvicy Act Statement (Signed), Study Subject Questionnaire Study Subject Supplemental Recording Book Study Subject Self Administered Form. Study Subject Medical Consent Form Study Subject. Former Wife Consent Letter..... Study Subject Interviewer Evaluation Form.... I PRFSF.NT Will: INTFRVII-W P'n vacy Ac t S ta"teiiK!h t. (Si gned). Spouse Questionnaire Spouse. Supplemental Recording [iook. Spouse Medical Consent Form Spouse Interviewer -L valuation form. •FfiKHl'R Win' I oniu.'r W i f e Nome As'1, iijimient Sheet.. Privricy Act St.ateim.Mit (Signed) Spouse Questionnaire. Spnir.e ',up|)lenierUal Recording Book. Spouse Medicdl Corisi.Mit Form Spouse Interviewer I'.yal'uation Form. PROXY INTERVIEW Proxy Name Assignment Sheet Privacy Act Statement (Signed) ,••..,..,..,. .m J w _ . . /•'. -._. . ' Proxy Questionnaire. •..„...;..'....'., Proxy Supplemental Recording book... ..,0.,..?.,,. Prox.y Medical Consent Proxy Interviewer; Evaluation ,-,.. ,.. •_„_„,_ Received: • ""liaTi? . ' '". ""• . I ' ,,,. ,^,_ >....( . : Ihecked in by: „ ' ' i 157 ' �CHAPTER II SPOUSE QUESTIONNAIRE The following Spouse/Partner Questionnaire was used to collect baseline data for the Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicide Orange. This data was collected during 1981-1982. All available spouse/partners, both present and former, were included in this data collection effort. The questionnaire and supplemental recording book are the actual field instruments. They have been photocopyed and reduced for the purpose of this report. Additional field documents, such as show cards, are included as attachments to the questionnaire. Additional attachments include: Introductory Letter, Privacy Act Statement, Medical Permission Form, Interview Evaluation, and Mailing Transmittal Form. The Spouse/Partner Questionnaire, as used in the field, follows. 158 �LOUIS HARRIS AND ASSOCIATES, INC. 630 Fifth Avenue O.M.B. NUMBER New York, New York 10111 0701-0033 Approval Expires 11/30/82 FOR OFFICE USE ONLY: Case No. 12-17 Study No. 812039 Respondent #: SPOUSE QUESTIONNAIRE 5-8 CONFIDENTIAL Present wife Former wife ( -1 This study is being conducted to collect information on the health of current and former Air Force personnel and their families. I will be asking background questions and questions about health. First, I have a few background questions to ask you. 1. ;Whflt is your date of birth? i (WRITE IN DATE) MONTH I DAY T 1 (19) (20) (21) (22) YEAR T I T (23) 159 I (24) �CARD 001 812039 2. How many children have you had — that is, of how many children are you the natural mother? Please include children who live w i t h you, those who live elsewhere, and those who may no longer be living. I I (WKITE IN NUMBER) I children (ASK Q.3) (25) (26) No children ....(2U -1 (SKIP TO Q.8) 3. Starting w i t h your first child, what is the first and last name of the child as it appears on the birth certificate? RECORD FIRST AND LAST NAMES OF ALL CHILDREN IN S.R.B. - PAGE 1. NAME ONLY AT THE TOP OF THE APPROPRIATE COLUMN(S). SKCOND CHILD FIRST CHILD How old is (CHILD) now? 1 1 1 1 (28) 1 1 Age Male Female (31( 1 1 5b. How much did (CHILD) weigh at birth? Male Female 5c. How old is (CHILD) now? 1 1 Age -1 (31( 6b. 6c. OUNCES I I l-l 1 1 (33) (34) (35) (32) Don' t know.. . 3 ( (6 -1 What is (CHILD) 's birth- 5d. date? -1 (Is/Was) (CHILD) male or female? (31( POUNDS OUNCES 1 1 1 1 1 1 ' l-l 1 1 (32) (33) (34) (35) Don't know. . ( 6 .3( -1 What is (CHILD) 's birth- 6d. date? -1 -2 How much did (CHILD) weigh at birth? 1 1 I Age (29) Male Female -1 -2 How much did (CHILD) weigh at birth? 1 1 1 1 Child died..(30( POUNDS POUNDS OUNCES 1 M 1 1 1 > l-l 1 1 (32) (33) (34) (35) 1 1 (28) (Is/Was) (CHILD) male or female? -1 -2 6a. (29) Child died..(30( -1 Don't know...(36( d. 1 1 (28) (Is/Was) (CHILD) male or female? 1 1 How old is (CHILD) now? (29) Jhild died..(30( 4C. 5a. NAME: ! 1 4h. THIRD CHILD NAME: MAME: ^ . WRITE IN THE FIRST -1 What is (CHILD)'s birthdate? MONTH DAY YEAR MONTH DAY YEAR MONTH DAY YEAR I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I I - I l l - I l l 1 1 l-l 1 l-l 1 1 •1 l-l 1 l-l 1 1 I (42) 37) (38) (39) (40) (41) (42) (37) (38) (39) (40) (41) (42) (37) (38) (39) (40) (41) ALSO RECORD IN S.R.B. -PC l l e. (ALSO RECORD IN S.R.B.-PG ll Was the child premature, 5e. full term, or overdue? lALSO RECORD IN S.R.B.-PG ll Was the child premature, 6e. full term, or overdue? Was the child premature, full term, or overdue? remature.(43( -A(ASK Premature. (43( Overdue. .... -1\(ASK -2j Q.5f) Premature. ( 3 4( Overdue -l^ASK -2( Q.6f) ull term.... -3^ (SKIP TO Full term -3 ((SKIP TO Full term -3*1(SKIP TO ~ f. X How many weeks (overdue/ 5f. premature) was (CHILD)? 1 1 1 1 (44) . 1 1 weeks (4S) (GO TO Q.4f;) How many weeks (overdue/ 6f. premature) was (CHILD)? 1 1 1 1 (44) 1 1 weeks (45) (GO TO Q.5g) 160 How many weeks (overdue/ premature) was (CHILD)? 1 1 1 1 (44) "1 1 weeks (45) (GO TO Q.6g) �CARD 001 FIRST CHILD 812039 SECOND CHILD THIRD CHILD 4g. Where are (CHILD)'s birth registration records located? In what city and state is thqt? (RECORD IN S.R.B. PG 1 I 5g. Where are (CHILD)'s birth registration records located? In what city and state is that? 6g. Where are (CHILD)'s birth registration records located? In what city and state is that? 4h. Where are (CHILD)'s current medical records Jlocated? .In what city and state is that? 5h. Where are (CHILD)'s current medical records located? In what city and state is that? 6h. Where are (CHILD)'s current medical records located? In what city and state is that? ii. 6i. What was (CHILD)'s f a ther's f ul 1 name? I RECORD IN'S.R.B. PC i I 4i. What was (CHILD)'s father's full name? IRECORD IN S.R.B. PG 1"I IRECORD IN S.R.B. PC i I IRECORD IN S.R.B. PC i I IRECORD IN S.R.B. PG i I How old were you when (CHILD) was born? 4j. How old were you when (CHILD) was born? I I I (46) (47) What was (CHILD)'s f a t h e r ' s f u l l name? IRECORD IN S.R.B. PG i I' IRECORD IN S.R'.B. PG~ i I (RECORD IN S.R.B. PG i I How old were you when (CHILD) was born? Age I (46) (47) Age Age (46) (47) 4k' Were either of you using 5k. Were either of you using 6k, Were either of you using birth control at the birth control at the birth control at the time you became pregnant time you became pregnant time you became pregnant with (CHILD)? with (CHILD)? with (CHILD)? Ye8.(4_8(__-l (ASK Q.4L) Yes.(48( No...... -2 (SKIP TO Q.4m) No THAND RESPONDENT CARD "c'T 4L. -1 -1 -1 -1 -1 06.(54( -1 __-2 (SKIP TO Q.5m) Ol.(49( -1 06.(54( -1 Q8.(56( Q9.(J7( 1Q.(58( ll.(59( -1 -1 -1 -1 No...... -2 (SKIP TO Q.6m) .(60( .6( (0 (GO TO Q.5m) 161 Ol.(49( O2.(50( O3.(51( 04.(5_2( O5.(53(. -1 -1 -1 Q6.(54( _ -1 Q7.(55( -1 Q8.(56( -1 Q9.(57( -1 1Q.(58( -1 11.(59( -1 12 (SPECIFY) 12 (SPECIFY) (00 TO Q.4m) -1 (ASK Q.6L) I HAND RESPONDENT CARD 1rCTT Please look at this 6L. ' Please look at this card and tell me all of card and tell me all of the numbers that apply the numbers that apply to the types of birth to the types of birth control you or your control you or your partner were practicpartner were practicing? ing? Q8.(56( -1 Q9.(57( '_, -1 1Q.(58( -1 11. (59C .-1 12 (SPECIFY) Yes.(48( THAND RESPONDENT CARD "c'T Please look at this 5L. card and tell me all of the numbers that apply to the types of birth cortrol you or your partner were practicing? Ol.(49( O2.(50( O3.(51( Q4.(S2( 05.(5J( -1 (ASK Q.5L) .6( (0 (GO TO Q.6m) �812039 CARD 001 SECOND CHILD FIRST CHILD 4m. How many months did it 5m. take you to become pregnant with this child? 1 1 1 I 1 1 1 I Months Less than 1 month, (63( -1 Less than 1 month. (63( Did (CHILD) have any birth defects? 5n. How many months did it take you to become pregnant with this child? 1 I "(61) (62) (61) (62) 4n. How many months did it 6m. take you to become pregnant with this child? I 1 1 1 Months THIRD CHILD -1 Did (CHILD) have any birth defects? 1 1 I | Months (61) (62^ Less than 1 month. (63(_._ -1 6n. Did (CHILD) have any birth defects? Yes.(64( -1 (ASK Q.4o) Yes.(64( -1 (ASK Q.5o) Yes.(64( -1 (ASK Q.6o) No -2 (SKIP TO Q.4p) No -2 (SKIP TO Q.5p) No -2 (SKIP TO Q.6p) 4o. What kind of birth defects did (s)he have? Any others? 5o. What kind of birth defects did (s)he have? Any others? 60. What kind of birth defects did (s)he have? Any others? 4p. Was (CHILD) ever diagnosed as having cancer? 5p. Was (CHILD) ever diagnosed as having cancer? 6p. Was (CHILD) ever diagnosed as having cancer'- Yes.(65( -1 (ASK Q.Aq) Yes.(65( -1 (ASK Q.5q) Yes.(65( No -2 (SKIP TO Q.4s) No -2 (SKIP TO Q. 5s) No. 4q. In what month and year 5q. was the diagnosis made? 1 1 4r. MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (66) (67) (68) (69) What kind of cancer was diagnosed? Not sure..(70( (GO TO Q.4s) 01 79-80 -1 5r. . . -2 (SKIP TO Q.6s) In what month and year 6q. was the diagnosis made? 1 I MONTH YEAR 1 I I 1 1 l l I l l (66) (67) (68) (69) What kind of cancer was diagnosed? Not sure..(70( -1 (CO TO Q.5s) 02 79-80 162 -1 (ASK Q.6q) In what month and year was the diagnosis made? MONTH 1 1 YEAR 1 1 1 1 (66) (67) 6r. (68) (69) What kind of cancer was diagnosed? Not sure. . 7 ( (0 (GO TO Q.fes) 03 79-fo -1 �812039 CARD 004 FIRST CHILD t;s. SECOND CHILD (Does/Did)(CIIILD) have & diagnosed learning disability? 5s. THIRD CHILD (Does/Did)(CIIILD) have a diagnosed learning disahil ity? 6s. (Does/DidMCHILD) have a diagnosed learning disability? Yes.(12( -1 (ASK Q.4t) Yes.(12( -1 (ASK Q.5t) Yes.(12( -1 (ASK No -2 (SKIP TO Q.4u) No -2 (SKIP TO No -2 (SKIP TO Q.6u) 4t. What kind of learning disability (does/did) (s,)he have? 4u. (Does/Did) (CHILD) have 5u. any physical, mental, or motor impairments? Yes.(13( -1 (ASK No -2 (SKIP TO Q.4w) 4v. Q.4v) What kind of impairment (does/did) (s)he have? 5t. What kind of learning disability (does/did) (s)he have? 6t. 6u. (Does/Did)(CHILD) have any physical, mental, 01 motor impairments? Yes.(13( -1 (ASK No -2 (SKIP TO 5v. 0.5u) Q.5v) Q.6t) What kind of learning disability (does/did) (s)he have? (Does/DidMCHILD) have any physical, mental, motor impairments? Yes.(13( -1 (ASK Q.6v) Q.5w) What kind of impairment (does/did) (s)he have? 6v. 'What kind of impairment (does/did) (s)he have? IF CHILD IS DEAD: IF CHILD IS DEAD: IF CHILD IS DEAD: OTHERWISE: OTHERWISE: OTHERWISE: iw. CONTINUE SKIP TO Q.4z On what date did (CHILD) die? MONTH DAY 5w. YEAR CONTINUE SKIP TO Q.5z On what date did (CHILD) die? 6w. CONTINUE SKIP TO Q.6z On what date did (CHILD) die? i i i i 11 r f i 1-1 i 1-1 i i 14) (155 (16) (17) (18) (19) MONTH DAY YEAR MONTH DAY YEAR 1 1 1 I I I 1 1 I I I 1 11 1 1 l-l 1 1-1 1 1 1 l-l 1 l-l 1 (14) (15) Tl6) (17) (18) (19) ( ) (15) (16) (17) (18) (19 H x. 5x. What was the cause of death? y. Where is (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. PG 1 I (CO TO Q.4z) What was the cause of death? 5y. Where is (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. PC 1 1 (GO TO Q.5z) 6x. What was the cause of death? "' 6y. Where is (CHILD) 's death registered? In what city and state is that? IRECORD IN S.R.B. PG 1 1 (GO TO Q.6z) �CARD 004 FIRST CHILD 4z. 812039 SECOND CHILD Did you smoke on a fairly regular basis during this pregnancy? 5z. THIRD CHILD Did you smoke on a fairly regular basis during this pregnancy? 6z. Did you smoke on a fairly regular basis during this pregnancy? Yes.(20( -1 (ASK Q.4aa) Yes.(20(_ -1 (ASK Q.5aa) Yes.(20( -1 (ASK Q.6aa) No -2 (SKIP TO NEXT CHILD) No No -2 (SKIP TO NEXT CHILD) 4aa. When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. I I I -2 (SKIP TO NEXT CHILD) 5aa. When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. T I I Packs Less than one pack.(23( 4bb. Less than one paek.(23( Did you drink alco5bb. holic beverages (beer, wine, or hard liquor) on a regular basis during this pregnancy? Yes..(24( No 4cc. (21) -1 -1 -2 (ASK Q.4cc) (GO TO NEXT CHILD) About how many drinks a week would you say that you had during this pregnancy? (25) I (26) -1 -1 -2 (ASK Q.Scc) (GO TO NEXT CHILD) About how many drinks a week would you say that you had during this pregnancy? T I drinks I Packs (21) (22) Less than one pack.(23( Did you drink alco6bb. holic beverages (beer, wine, or hard liquor) on a regular basis during this pregnancy? Yes..(2A( No 5cc. T T T I Packs (22) I (21) (22) baa. When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. Did you drink alcoholic beverages (beer, wine, or hard liquor) on a regular basis during this pregnancy? Yes..(24( No 6cc. T I drinks (25) (26) -1 -2 (ASK Q.6cc) (GO TO NEXT CHILD) About how many drinks a week would you say that you had during this pregnancy? I I drinks (25) (26) 02 03 79-80 (GO TO NEXT CHILD) 79-80 (GO TO NEXT CHILD) (RECORD ADDITIONAL CHILDREN IN S.R.B. PAGE 4) CARD 007 IF ANY CHILDREN: ASK Q . 7 . IF NO CHILDREN: SKIP TO Q.8. 7. Did you and (STUDY RESPONDENT; have the number of children you planned on? No. 8. ..... (12( (SKIP TO Q.9) Did you and (STUDY RESPONDENT) plan to have children? Yes No -1 (13( -1 "-2 104 �CARD 007 812039 9a. Did you and (STUDY RESPONDENT) ever try for a period of a year or more to conceive a child without heing able to? -1 -2 No 9b. (ASK Q.9b) (SKIP TO Q.ll) For how many periods of one year or more did this happen? \ I T I I I periods (15) (16) Not sure FIRST PERIOD lOa. i ! 1 In what month and year did the f i r s t period begin? And in what month and .year d i d it end? FROM MONTH YEAR 1 1 1 1 I I I l l I l l (18) (19) (20) (21) TO MONTH YEAR I I I 1 I I I 1 (22) (23) (24) (25) IjOb. • ' How old were you in (BEGINNING DATE OF PERIOD)? -1 ..(17( SECOND PERIOD lOd. 1 i THIRD PERIOD In what month flnd year did the second period begin? And in what month and year did it end? FROM MONTH YEAR 1 (18) 1 T ' 1 In what month and year d i d the t h i r d period begin? And in what month and year did it end? FROM MONTH YEAR 1 1 I I I 1 I l l I l l (18) (i'9) (20) (21) TO MONTH YEAR I I I I I I i 1 l-l 1 1 (22) (23) (24) (25) M i . i - i i i (19) lOg. (20) (21) i TO MONTH YEAR I I I I 1 1 I l l I l l (22) (23) (24) ( 2 5 ) lOe. How old were you in (BEGINNING DATE OF PERIOD)? lOh. How old were you in (BEGINNING DATE OF PERIOD)? • 1 1 LOc. fes (o 1 1 ! 1 AGE (26) ( 2 7 ) During t h i s period did either of you see a doctor to discuss any d i f f i c u l t i e s in conceiving children? (2R( -1 -2 1 1 lOf. , Yes I I 1 1 (26) ( 2 7 ) 1 1 AGE During t h i s period d i d e i t h e r of you see a doctor to discuss any d i f f i c u l t i e s in conceiving c h i l d r e n ? (28( -1 No Yes AGE During t h i s period did either of you see a doctor to discuss any d i f f i c u l t i e s in conceiving c h i l d r e n ? ..(28( -1 -2 02 79-80 (GO TO NEXT PERIOD) lOi. I I 1 1 (26) ( 2 7 ) 03 79-80 (GO TO NEXT PERIOD) (RECORD ADDITIONAL PERIODS IN S . R . B . PAGE 19) 11. Did you ever have difficulties in conceiving a child with any other husband or partner? Yes. No.. .(29( No other husband/ partner . 01 79-80 1.65 -1 | �.CARD 012 812039 I ASK EVERYONE I I HAND RESPONDENT CARD "D-T 12a. There are many reasons that some couples find it difficult or impossible to conceive a child. Please read this card and tell me the letter for each reason which ever applied to (STUDY RESPONDENT). Any other reason? I ASK Q.12b AND Q.12c FOR EACH REASON IN Q.I2a.I 12b. Did reason (LETTER) apply to you or your spouse? 12c. [MULTIPLE RECORD BELOW I In what year did this occur or become known to you? Q.12b Q.12a q.!2c Year A. Spouse respondent. . 18( ( 1 1 -1 1 1 1 1 (39) Study respondent. . ( 19( (31) 1 -1 ! 1 1 1 1 (3.2) Spouse respondent . . ( 0 2( B. -1 1 1 1 1 Sterility due to injury, accident, or illness (SPECIFY) .(13( (33) 1 1 (34) -1 Study respondent. .(21 ( -1 (35) ! I 1 ! (36) I 1 (37) C. Spouse respondent . . ( 2 2( Sterility due to unknown -1 1 1 1 1 (38) Study respondent.. (23( -1 Spouse respondent .. ( 4 2( -1 1 1 (39) 1 1 1 1 1 1 •(40) (4H D. Lack of interest in sex (15( -1 1 1 1 1 (42) Study ' respondent. . (25( (431 1 1 -1 1 1 1 1 (44) 1 1 (451 E. Spouse respondent. . (26( Other known medical or physical conditions (SPECIFY) I 1 -1 (46) .(16( -1 Study respondent. .(27( (47) 1 1 -1 1 1 I 1 1 1 (48) ! 1 (49) F. Spouse respondent .. (28( Some other reason (SPECIFY) .(17< -1 1" I 1 1 1 1 (50) (51) Spouse respondent . (29( . -1 1 -1 1 (52) 1 (53) �CARD 013 812039 13. Now I'd like to know about any othrr pregnancies you had that did not end in live births — that is, any pregnancies that ended in miscarriage, stillbirth, or abortion. Did you ever have a pregnancy that ended in miscarriage, stillbirth, or abortion? Yes. (ASK Q.H) No Not sure. .. -A • •' -31 (SKIP TO Q.lSa) 14. How many such pregnancies did you have? T Number (13) (14) PREGNANCY 2 PREGNANCY 1 ISa. In what month and year did the first such pregnancy end? MONTH PREGNANCY 3 16a. In what month and year did the next such pregnancy end? 17a, In what month and year did the next such pregnancy end? YEAR MONTH YEAR MONTH T (17) (18) (15) (16) (17) (18) (15) (16) tr Tr YEAR I Tl5) (16) I (17) (18) I 17b. Did this pregnancy end 16b. Did this pregnancy end 15b. Did this pregnancy end in a miscarriage, stillin a miscarriage, still' in a miscarriage, still' birth, or abortion? birth, or abortion? birth, or abortion? Miscarriage..(19( Stillbirth Abortion -1 -2 -3 15c. After how many weeks did the pregnancy end? Weeks (20) (21) Miscarriage..(19( Stillbirth Abortion 16c. I -1 -2 -3 Miscarriage..(19( Stillbirth.. Abortion After how many weeks did the pregnancy end? I (WRITE IN AGE) I I (WRITE IN AGE) I 17d. How old were you at that time? Was (STUDY RESPONDENT) 16e. your partner in this pregnancy? Were either of you using birth control at the time you became pregnant? I I (WRITE IN AGE) (22) (23) Yes.(24( -1 No. -2 15f. | Weeks (20) (21) 16d. How old were you at that time? 15e. (22) (23) Was (STUDY RESPONDENT) 17e. your partner in this pregnancy? Were either of you using birth control at the time you became pregnant? Was (STUDY RESPONDENT) your partner in this pregnancy? Yes.(24( -1 -2 No. Yes.(24( -1 No -2 16f. After how many weeks did the pregnancy end? I I I Weeks (20) (21) 15d. How old were you at that time? I I I I (22) (23) 17c. -1 -2 -3 17f. Were either of you using birth control at the time you became pregnant? Ves.(2_5 (___-! (ASK Q.15g) Yes.(_25( -1 (ASK Q.16g) Yes.(25( -1 (ASK Q.17g) No _-2 (SKIP TO Q.15h) No _-2 (SKIP TO Q.16h) No _-2 (SKIP TO Q.17h) (GO TO Q.15 g/h) (GO TO Q.16g/h) 11)7 (GO TO Q.17g/h) �CARD 013 PREGNANCY 3 PREGNANCY 2 PREGNANCY 1 812039 lHANI) RESPONDENT CARD "C" 1 (HAND RESPONDENT CARD "C" I 1 HAND RESPONDENT CARD "C" 1 15g. Please look at this card 16g. Please look at this card 17g. Please look at this card and tell me all the numand tell me all the numand tell me all the numbers that apply to the bers that apply to the bers that apply to the types of birth control types of birth control types of birth control you or your partner were you or your partner were you or your partner were using. using. using. Ol.(26( 02.(27( 03.(28( O4.(21( O5.(30( -1 -1 -1 -1 -1 06. (311 07.(32( 08.(33( ('. 34( )(( Ht.(35( ll.(3'f>( -1 -1 -1 -1 -1 -1 12 (SPECIFY) 1 I (38) O6.(31( 07.(32( 08.(33( 09.(34( 10.(35( ll.(36( -1 -1 -1 -1 -1 -1 ! 1 1 Months 1 -1 -1 -1 -1 -1 -1 O6.(31( 07.(32( 08.(33( 09.(34( 1<).'.35( ll.(36( -1 -1 -1 -1 -1 -1 .(37( -1 How many months did it take you to become pregnant this time? | (38) (39) Less than 1 month. U ( 0 16h. Ol.(26( 02.(27( 03.(28( 04.(29( O5.(30( -1 12 (SPECIFY) .(37( -1 How many months did it take you to become pregnant this time? 1 I -1 -1 -1 -1 -1 12 (SPECIFY) .(37( 15h. Ol.(26( 02.(27( 03.(28( ()4.(29< O5.(30( 1 How many months did it take you to become pregnant this time? I . I I 1 1 1 Months (38) (39) | | Months (39) Less than 1 month. ( 0 4( 17h. -1 Less than 1 month. U0( -1 17i. (IF MISCARRIAGE OR 16i. (IF MISCARRIAGE OR 15i. (IF MISCARRIAGE OR STILLBIRTH IN Q.17b, STILLBIRTH IN Q.15b, STILLBIRTH IN Q.16b, ASK Q.17i. IF ABORTION ASK Q.16i. IF ABORTION ASK Q.15i . IF ABORTION IN Q.17b, SKIP TO i). 17m) IN Q.16b, SKIP TO Q.lttm) IN Q.15b, SKIP TO Q.ISm) . Yes. UK No ... Did a doctor tell you why this (miscarriage/ stillbirth) might have occurred? Did a doctor tell you why this (miscarriage/ stillbirth) might have occurred? Did a doctor tell you why this (miscarriage/ s t i l l b i r t h ) might have occurred? Yes. UK -1 (ASK Q.15J) -2 (SKIP TO Q.15n) No.... Yes. UK -1 (ASK Q.16J) -2 (SKIP TO Q.16n) lo 7j . What did the doctor say caused the (miscarriage/stillbirth)? 15j. What did the doctor say 16j. caused the (miscarriage/stillbirth)? What did the doctor say caused the (miscarriage/stillbirth)? 15k. What is the name of the 16k. doctor or me.dical facility that you consulted about this? What is the name of the 7k. doctor or medical facili t v that you consulted about this? I RECORD IN S.R.B. - PG 3 1 15L. I RECORD IN S.R.B. - PG 3 1 In what month and year was that? .6L. In what month and year was that? (RECORD IN S.R.B. - PG 3 1 [RECORD IN S.K.B. - PG 3 1 (SKIP TO Q.5n) (SKIP TO Q.l>n) -1 (ASK Q.17J) -2 (SKIP TO Q.17n) What is the name of ths doctor or medical facil ity that you consulted about this? IRECORD IN S.R.B. - PG 3 1 7L. In what month and year was that? IRECORD IN S.R.B. - PG 3 1 (SKIP TO Q.7n) (GO TO NEXT PAGE) 168 �812039 CARD 013 PREGNANCY 1 PREGNANCY 3 PREGNANCY 2 15m. What was the main reason for the abortion? 16m. What was the main reason for the abortion? 17m. 15n. Did you smoke cigarettes on a fairly regular basis during this pregnancy? 16n. Did you smoke cigarettes on a fairly regular basis during this pregnancy? 17n. Did you smoke cigarettes on a fairly regular basis during this pregnancy? Yes.(42( -1 (ASK Q.lSo) Yes.(42( -1 (ASK Q,16o) Yes,(42( "1 (ASK Q.17o) No. -2 (SKIP TO Q.ISp) No. -2 (SKIP TO Q.l6p) No. "•2 (SKIP TO Q.l7p) 15o, When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? Dy pack we mean 20 cigarettes. 16o, When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we menn 20 cigarettes. Packs (43) ( 4 4) Less than one pack.(45( 15p. 15q. -1 Less than one pack.(45( -1 16p. Did you drink alcoholic beverages (beer, wine, or hard liquor) on a regular basis during this pregnancy? Yes..(A6(-1 No -2 (ASK Q.15q) (GO TO NEXT PREGNANCY) About how many drinks a week would you say that you had during this pregnancy? I I I drinks (47) (48) 17o. When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. ! I I | I I Packs (43) ( 4 4) T T | I Packs (43) ( 4 4) T Less than one pack.(45( -1 17p. Did you drink alcoDid you drink alcoholic beverages (beer, holic beverages (beer, wine, or hard liquor) wine, or hard liquor) on a regular basis duron a regular basis during this pregnancy? ing this pregnancy? Yes..(A6( _-l -2 No 16q. What was the main reason for the abortion? (ASK Q.16q) (GO TO NEXT PREGNANCY) About how many drinks a week would you say that you had during this pregnancy? T \ I I | drinks (47) (48) Yes..(46( -1 No......._-2 17q. (ASK Q.17q) (GO TO NEXT PREGNANCY) About how many drinks a week would you say that you had during this pregnancy? T T I drinks 747) (48) 01 79-80 (GO TO NEXT PREGNANCY) (GO TO NEXT PREGNANCY) 109 (RECORD ADDITIONAL PREGNANCIES IN S.R.B. PAGE 20) �CARD 812039 18a. We would like your consent for the doctors and medical facilities you mentioned during this interview to provide medical records to the Air Force Health Survey. These records will help us obtain more detailed information about the health services you talked about. TURN TO S.R.B. PG 3. ENTER NAMES OF MEDICAL PROVIDERS ON APPROPRIATE PERMISSION FORMS AND ASK RESPONDENT TO SIGN EACH FORM. IFOR EACH SIGNED'FORM, ASKTT 18b. What is the current address of (DOCTOR/FACILITY)? Thank you for participating in the Air Force Health Study! TIME INTERVIEW ENDED: (am/pro) 170 �FOR OFFICE USE ONLY: LOUIS HARRIS AND ASSOCIATES, INC. 630 Fifth Avenue New York, New York 10111 Case No. Study No. H12039 O.M.B. NUMBER 0701-0033 Approval Expires 11/30/82 Respondent t: CONFIDENTIAL AIR FORCE HEALTH SURVEY SUPPLEMENTAL RECORDING BOOK SPOUSE QUESTIONNAIRE 171 �CARD Q.4a-6a and 19-21 CHILDREN CHILD Q.29 FIRST First NAME 1 1 1 1 DAY 1 1 M I-l 1 1 MONTH 1 I I 1 I-l DAY 1 1 YEAR I I ! I-l 1 ) 1 • . 1 1 MONTH 1 II 1 I-l DAY I 1 YEAR I I I I-l 1 ! 1 I 1 MONTH 1 I I 1 I-l DAY 1 1 1 1 I-l YEAR I 1 1 1 First Last SIXTH MONTH 1 1 1 1 I-l First Last FIFTH MONTH DAY 1 1 1 1 ! ! 1 1 First Last FOURTH First Last g. MONTH 1 II 1 I-l First Last THIRD BIRTHDATE 1 1 Last SECOND <J. DAY 1 1 BIRTH RECORDS i. FATHER '« FULL NAME y. DEATH RECORDS First Place C/S Last C/S Place Place First Place ffc C/S Last C/S Place First Place C/S Last C/S Place First Place C/S Last C/S Place First .Place C/S Last C/S Place First Place C/S Last C/S YEAR Place i l ! 1 C/S . 1 1 YEAR 1 I CURRENT MEDICAL RECORDS Place M I-l | | h. 1 1 YEAR Place i l I 1 C/S Place' C/S Place C/S Place C/S �CARD Q.4a-6a and 22-24 CHILDREN CHILD Q.29 NAME d. SEVENTH First BIRTHDATE g. BIRTH RECORDS h. CURRENT MEDICAL RECORDS i . FATHER ' S FULL MAffi y. DEATH RECORDS EIGHTH C/S Last C/S MONTH 1 I 1 DAY 1 YEAR i Place Place First Place 1 1 1 1 l-l C/S Last C/S DAY 1 Place First Place 1 1 ! c/s C/S Last C/S Place First Place C/S Last C/S Place First Place C/S Last C/S Place First Place C/S Last C/S 1 1 1 First l-l MONTH 1 1 1 ! l-i 1 l-l I YEAR 1 1 I c/s Place 1 1 1 MONTH DAY i 1 1 1 1 1 - 1 1 I I l-l YEAR 1 1 Place 1 1 C/S 1 1 MONTH 1 1 1 1 l-l DAY 1 ! 1 1 l-l YEAR 1 i 1 1 1 Last ELEVENTH Place First Last TENTH Place 1 1 C/S First Last NINTH YEAR 1 1 First I I 1-1 1 Last MONTH DAY 1 I I 1 1 1 - 1 1 Place ! 1 MONTH 1 1 1 DAY 1 1 1 YEAR 1 1 First Last TWELFTH First Last 1 1 1 - 1 1 l-t 1 Place C/S Place 1 c/s �CAKU Ol/i Q.l."ik-17k. Medical Providers — Miscarriages/Stillbirth Pregnancy 1 lf>k. 812039 Pregnancy 2 Doctor/facility consul ted:- 16k. Pregnancy 3 17k. Doctor/faci1ity consul tod: Doctor/facility consul ted: Name Name Name Address Address Address C/S C/S C/S 15L. 17L. 161.. MONTH YEAR 1 M "1 1 I l l I l l ( 12) ( i:i) ( 14) ( 15) ( I MONTI YEAR 1 1 I 11 I ( 16) ( 7). ( 18) ( 19) MONTH 1 I 1 l l CO ) ( 21) YEAR M - 1 I l 1 l (22) (23) Q.29k-31k. Medical Providers -- Miscarriages/Stillbirth 29k. Doctor/facility consulted: 30k. Doctor/facility consul ted: 31k. Doctor/facility consulted: Name Name Name Address Address Address C/S C/S C/S i j 31L. 301.. 29L. MONTH "T24TT255 MONTH 1 YEAH r ri i i i i - i l l (26) ( 2 7 ) I YEAR -r— f l l l ( 2 4 ) (25) (26) ( 2 7 ) 174 MONTH l I YEAR l I (24) (25) (26) ( 2 7 ) r i I �Ft] 2039 CART) 00 I Q.19-21 Additional Children FOURTH CHILD FIFTH CHILD SIXTH CHILD NAME : NAME : 19a. How old is (CHILD) now? 20a. How old is (CHILD) now? 1 1 1 1 1 Age (.28) (29) Chi Id died. ,OO( (31 ( 1 1 (2fi) (29) Male Female -1 -2 Don 1 t know. . . (36( 1 1 1 1 Age ' (28) (29) -1 Chi Id died. .(30( I 11( POUNDS 1 1 Male 1 T l-l 1 1 (32) (33) -1 POUNDS OUNCES 1 1 1 1 1 1 l-l 1 1 (32) (33) (34) (35) 1 1 1 1 -1 Don't know. , ( 6 .3( (34) (35) Don't know. . ( 6 .3( -1 (31( 21c. How much did (CHILD) weigh at birth? OUNCES 1 I -1 21b. (Is/Was) (CHILD) male or femnle? -1 -2 20c. How much did (CHILD) weigh at birth? POUNDS OUNCES 1 1 1 1 1 l l I l l (32) (33) (34) (35) 21a. How old is (CHILD) now? Age 20b. (Is/Was) (CHILD) male or t oma 1 c? 19c. How much did (CHILD) weigh at birth? 1 I 1 1 Child died. .(30( -1 191,. (Is/Was) (CHILD) male or female? Male Female NAME : -1 19d. What is (CHILD) 's birth- 20d. What: is (CHILD) 's birth- 21d. What is (CHILD) 's birthdate'? date? date? I 1 MONTH DAY YEAR I I I I I I 1 1 l-l 1 . l-l 1 1 1 1 1 MONTH 1 II 1 l-l DAY 1 1 1 1 l-l YEAR 1 1 1 1 1 1 MONTH 1 1 1 1 l-l DAY 1 1 1 1 l-l YEAR 1 1 1 1 (37) 4) (37) (38) (39) (40) (41) (42) (37) (38) (39) (40) (41) (42) (38) (39) ( 0 (41) (42) lALSO RECORD IN S.R.B.-PG l| (ALSO RECORD IN S.R.B.-PG l l lALSO RECORD IN S.R.B.-PG l| 19e. Was the child premature, 20e. Was the child premature, 2le. Was the child premature, full term, or overdue? full term, or overdue? full term, or overdue? Premature. ( 3 4( Overdue . . -lt(ASK ~2\ Q.19f) Premature. (4 3( Overdue -ll(ASK -2j Q.20f) Premature. ( 3 4( -1\(ASK Full term Not sure -3l(SKIP TO ~4J Q.19g) Full term . . .. -3)(SKIP TO Full term -3*1 (SKIP TO . 19f. How many weeks (overdue/ 20f. How many weeks (overdue/ 21 f. How many weeks (overdue/ premature) was (CHILD)? premature) was (CHILD)? premature) was (CHILD)? 1 I 1 ! 1 1 weeks 1 1 1 1 1 1 weeks (44) (45) (GO TO Q.19g) (CO TO Q.20g) 175 1 1 1 1 1 1 weeks (44) (45) (GO TO Q.21g) �CARD 001 FOURTH CHILD FIFTH CHILD 19g. Where are (CHILD) 's birth registration records located? In what city and state is that? [RECORD IN S.R.B. PG 1 1 19h. Where are (CHILD) 's current medical records located? In what city and state is that? [RECORD IN S.R.B. PG 1 1 19i. What was (CHILD) 's father's full name? (RECORD IN S.R.B. PG i 1 19 j. How old were you when (CHILD) was born? 1 1 (46) 1 1 812039 SIXTH CHILD 20g. Where are (CHILD) 's birth registration records located? In what city and state is that? TRECORD IN S.R.B. PG i 1 21g. Where are (CHILD) 's birth registration records located? In what city and state is that? 1 RECORD IN S.R.B. PG 1 20h. Where are (CHILD) 's current medical records located? In what city and state is that? 1 RECORD IN S.R.B. PG 1 1 21h. Where are (CHILD) 's current medical records located? In what city and state is that? 20i. What was (CHILD) 's father's full name? 21i. What was (CHILD) 's father's full name? (RECORD IN S.R.B. PG i 1 20j . How old were you when (CHILD) was born? 1 ! Age I 1 (46) (47) 1 1 [RECORD IN S.R.B, PG 1 IRECORD IN S.R.B. PG 1 21j. How old were you when (CHILD) was born? 1 1 Age (47) 1 1 (46) 1 1 Age (47) 19k. Were either of you using 20k. Were either of you using 21k. Were either of you using birth control at the birth control at the birth control at the time you became pregnant time you became pregnant time you became pregnant with (CHILD)? with (CHILD)? with (CHILD)? Yes.(48( Yes.(48( -1 (ASK Q.19L) No -2 (SKIP TO Q.19m) No Yes,(48( -1 (ASK Q.20L) -2 (SKIP TO Q.20m) No -1 (ASK Q.21L) -2 (SKIP TO Q.21m) IHAND RESPONDENT CARD "C"| IHAND RESPONDENT CARD "c"i IHAND RESPONDENT CARD "c"l 19L. Please look at this 20L. Please look at this 21L. Please look at this card and tell me all of card and tell me all of card and tell me all of the numbers that apply the numbers that apply the numbers that apply to the types of birth to the types of birth to the types of birth control you or your control you or your control you or your partner were practicpartner were practicpartner were practicing? ing? ing? Ol.(49( O2.(50( O3.(51( 04.C52T 05.(53( -1 -1 -) -1 -1 12 (SPECIFY) 06.(54( -1 07.<55( -1 08.(56( -1 O9.(57ij -1 I0.(?8( -1 ll.(59( -1 ' ' .6( (0 (GO TO Q.19m) Ol.(49( O2.(50( 03. (5U O4.(52l 05.(53( -1 -1 -1 -1 -1 06.(54( 07.(55( 08.(56( 09.(T77 10.(58( ll.(59( -1 -1 -1 -1 -1 -1 2 (SPECIFY) Ol.(49( O2.(50( O3.(51( O4.(52l 05-(53( -1 -1 -1 -1 -1 06.(54( -1 07.(55( ... -1 08.(56( -1 09. (TJI -1 10.(58( -1 ll.(59( -1 12 (SPECIFY) -1 .(60( (GO TO Q.20m) 17o -1 .6( (0 (GO TO Q.21m) -1 �CARD 001 FOURTH CHILD 812039 FIFTH CHILD SIXTH CHILD 19m. How many months did it 20m. How many months did it 21m. How many months did it take you to become pregtake you to become pregtake you to become pregnant with this child? nant with this child? nant with this child? 1 I ! 1 ( I Months 1 1 (615^ ( 2 6) Less than 1 month. ( 3 6( 1 1 1 I Months (1 (2 6) 6) 1 I 1 1 | | Months (1 (2 6) 6) -1 Less than 1 month. ( 3 6( "I Less than 1 month. ( 3 6( Wasn't trying , . . -2 19n. Did (CHILD) have any birth defects? 20n. Did (CHILD) have any birth defects? 21n. Did (CHILD) have any birth defects? Yes.(6A( -1 (ASK Q.19o) Yes.(64( -1 (ASK Q.20o) Yes.(64( -1 (ASK Q.21o) No -2 (SKIP TO Q.19p) No -1 .. -2 (SKIP TO Q.20p) 19o. What kind of birth defects did (s)he have? Any others? 20o. What kind of birth defects did (s)he have? Any others? 21o. What kind of birth defects did (s)he have? Any others? 19p. Was (CHILD) ever diagnosed as having cancer? 20p. Was (CHILD) ever diagnosed as having cancer? 21p. Was (CHILD) ever diagnosed as having cancer? Yes.(65( -l.(ASK Q.19q) Yes.(65(. -1 (ASK Q.20q) Yes.(65( —7 ( <1KT P TO n 1 Qs} No 19q. 1 I -2 (SKIP TO Q.20s) No No In what month and year 20q. iwas the diagnosis made? MONTH YEAR 1 M I I l l I l l ( 6 (67) ( 8 ( 9 6) 6) 6) -1 (SKIP TO Q. 21s) In what month and year 21q. In what month and year was the diagnosis made? was the diagnosis made? MONTH 1 1 -1 (ASK Q.21q) YEAR 1 1 1 1 1 1 H 1 1 ( 6 (67) (68) (69) 6) I 1 MONTH YEAR I I I 1 I 1 l-l 1 1 (66) (673 ( 8 (69) 6) 19r. What kind of cancer was diagnosed? 20r. What kind of cancer was diagnosed? 21r. What kind of cancer was diagnosed? Not sure..(70( Not sure..(70( Not sure..(70( (GO TO Q.As) 04 T^0 9$ -1 -1 (GO TO Q.5s) 05 Jtto 177 (GO TO Q.6s) 06 79^80 -1 �CARD 004 FOURTH CHILD 812039 FIFTH CHILD SIXTH CHILD 19s. (Does/Did)(CHILD) have a diagnosed Iparning disabil ity? 2()s. (Does/Did) (CHILD) have a diagnosed learning disabi 1 ity? 21s. (Does/Did)(CHILD) have a diagnosed learning disability? Yes.(12( -1 (ASK Q.19t) Yes.(l2( -1 (ASK Q,?(U) Yes.(12( -1 (ASK Q.21t) No -2 (SKIP TO () 19u) No , -2 (SKIP TO Q,20u) No -2 (SKIP TO Q.21u 19t. What kind of learning disability (does/did) Cjjhe have? 20t. What kind of learning disability (does/did) (s)he have? 2H. What kind of learning disability (does/did) (s)he have? 21u. (Does/Did)(CHILD) have 20u. (Does/Did)(CHILD) have 19u. (Does/Pid)(CHIU» have any physical, mental, any physical, mental, or any physical, mental, or motor impairments? motor impairments? motor impairments? Yes.(13( Ves.(13( -1 (ASK Q.19v) No -2 (SKIP TO Q.19w) No .. . 19v. What kind of impairment (does/did) (s)he have? -1 (ASK Q.20v) Yes.(13( -2 (SKIP TO Q.20w) No 20v. Wliat kind of impairment (does/did) (s)he have? -1 (ASK Q.21v) -2 (SKIP TO Q.21w 21v. What kind of impairmen (does/did) (s)he have? IF CHILD IS DEAD: IF CHILD IS DEAD: IF CHILD IS DEAD: OTHERWISE: OTHERWISE: OTHERWISE: CONTINUE SKIP TO Q. 19z 19w. On what date did (CHILD) die? CONTINUE SKIP TO Q.20z 20w. On what date did (CHILD) die? CONTINUE SKIP TO Q.21z 21w. On what date did (CHILD) die? MONTH DAY YEAR MONTH DAY YEAR MONTH DAY YEAR 1 I I 1 I I 1 1 1 1 1 I I 1 1 1 I I 1 II 1 1 1 1 l-l 1 l-l 1 1 i-l 1 l-l 1 1 1 1 l-l 1 l-l 1 1 14) (15) (16) (17) (18) (19) (14) (15) (16) (17) (18) (19) (14) (15) (16) (17) (18) (19 9x. What was the cause of death? 20x. What was the cause of death? 21x. Wliat was the cause of death? 9y. Where is (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. PG 1 I 20y. Where is (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. PG 1 I 21y. Where is (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. PG 1 J (GO TO Q.19z) (GO TO Q.20z) 178 (GO TO Q.21z) �CARD 004 FOUKTH CHILD 812039 FIFTH CHILD SIXTH CHILD 19z. Did you smoke on a fairly regular basis during this prpgnancy? 20?,. Did you smoke on a fairly regular basis during this pregnancy? 21z. Did you smoke on a fairly regular basis during this pregnancy? Yes.(20( _ -1 (ASK Q.19aa) Yes.(20( -1 (ASK Q.20aa) Yes.(20( -1 (ASK Q.21aa) No...... No -2 (SKIP TO Q.20bb) No -2 (SKIP TO Q.21bb) "2 (SKIP TO "*" Q.l'Jbb) 19aa.Wh.en you were smoking cigarettes on a fairly regular basis during this pregnancy, on Che average, how many packs per week did you smoke? By pack we mean 20 cigarettes. I I I I Packs (21) (22) Less than one pack.(23( -1 20aa.When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. 21aa.When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. I I I ' I Packs \ I (21) (22) Less than one pack.(2_3( I I I I (21) -I Packs (22) Less than one paek.(23( -1 20bb. Did you drink alco21bb. Did you drink alco19bb. Did you drink a.lcoholic beveragfes (beer, holic beverages (beer, holic beverages (beer, wipe, or hard liquor) wine, or hard liquor) wine, or hard liquor) on a regular basis duron a regular basis duron a regular basis during this pregnancy? ing this pregnancy? ing this pregnancy? Yes..(24(T -1 No -2 ~~ (ASK Q.19cc) (GO TO NEXT CHILD) 19cc. About how many drinks a week would you say that you had during this pregnancy? I I Yes..(24( -1 No __-2 (ASK Q.20cc) (GO TO NEXT CHILD) 20cc. About how many drinks a week would you say that you had during this pregnancy? T I drinks . . .1 (2$) (26) I I drinks (25) (26) Yes..(24( -1 N . . . . "-2 o... (ASK Q.21cc) (GO TO NEXT CHILD) 21cc. About how many drinks a week would you say that you had during this pregnancy? I drinks I (25) (26) 05 79-80 (GO TO NEXT CHILD) (GO TO NEXT CHILD) (GO TO NEXT CHILD) �812039 CARD 001 Q. 22-24 Additional Children 1 EIGHTH CHILD SEVENTH CHILD NINTH CHILD NAME: NAMF. : NAMF: 22a. How old is (CHILD) now? 23a. How old is (CHILD) now? 1 1 1 1 I 1 (28) 1 1 Age (28) (29) C h i l d died..(30( (31( 1 I (31( -1 24b. (Is/Was) (CHILD) male or female? -1 -2 Male Female (31( -1 -2 24c. How much did (CHILD) weigh at birth? POUNDS OUNCES 1' 1 1 1 1 l l I l l (32) (33) (34) (35) Don't know. . ( 6 .3( -1 1 1 Age Child died..(30( -1 23c. How much did (CHILD) weigh at birth? POUNDS OUNCES 1 1 1 i 1 l l I l l (32) (33) (34) (35) 1 1 (28) (29) (29) Male Female -1 -2 Don' t know. . . 3 ( (6 1 1 Age 23b. (Is/Was) (CHILD) male or female? 22c. How much did (CHILD) weigh at birth? I I 1 1 Child died..(30( -1 22h. (Is/Was) (CHILD) male or female? Male. Fema le 1 1 24a. How old is (CHILD) now? 1 1 POUNDS 1 1 1 1 l-l (32) (33) OUNCES 1 1 1 1 (34) (35) Don' t know. ..(36( -1 22d. What is (CHILD) 's birth- 23d. What is (CHILD) 's birth- 24d. date? date? -1 What is (CHILD) 's birthdate? MONTH DAY YEAR 1 1 1 1 1 1 1 1 1 1 1 l-l 1 l-l 1 1 (37) (38) (39!) (405 (41) (42) (37) (38) (39) (40) (41) (42) (37) (38) (39) (40) (41) (42) I 1 MONTH 1 I I 1 l-l DAY YEAR I I I 1 1 l-l 1 1 1 1 1 MONTH 1 II 1 l-l DAY 1 1 I I l-l YEAR I 1 IALSO RECORD IN S.R.B.-PG 2| lALSO RECORD IN S.R.B.-PG 2| 1 1 lALSO RECORD IN S.R.B.-PG 2| 22e. Was the child premature, 23e. Was the child premature, 24e . Was the child premature, full term, or overdue? full term, or overdue? full term, or overdue? Premature. (43( Overdue Full term Not sure . • . Premature. (4 3( Overdue -3((SKIP TO -4 [ Q.22g) ——~~ J -l)(ASK Premature. (43( -2] Q.23f) Overdue -1\(ASK -2) Q.24f) Full term -ll(ASK -2| Q.22O -3l(';KlP TO Full term .... -3*1 (SKIP TO . 22f. How many weeks (overdue/ 23f. How many weeks (overdue/ 24f. How many weeks (overdue/ premature) was (CHILD)? premature) was (CHILD)? premature) was (CHILD)? I I 1 1 1 1 weeks (44) (45) 1 I 1 I 1 I weeks (GO TO Q.22g) (GO TO Q.23g) " 4 (45) ,) i 180 I I l I l 1 weeks (44) (45) (GO TO Q.24g) �CARP 001 EIGHTH CHILD SEVKNTH CHILD NINTH CHILD 22s. When' arc (CHILD) 's birth registration records located? In what city and state is that? 1 RECORD IN S.R.B. PG 2 1 23g. Wherp are (CHILD 's hirth registration records located? In what city and state is that? 22h. Where are (CHILD) 's current medical records located? In what city and state is that? I RECORD IN S.R.B. PC 2 \ 23h. Where are (CHILD) 's current medical records located? In what city and state is that? 22i. What was (CHILD) 's father's f u l l nnmr? IHKCORD IN S.K.It. l'<; :> F 23i. What was (CHILD) 's father" s fill 1 mum;? 22j, How old were you when (CHILD) was born? 23 j. How old were you when (CHILD) was born? T . 1 I IRECORD IN S.R.B. PC 2 i IKHCOKH IN S.R.B. ]>c. <\ I I Ag<- (46) 24j;. Where are (CHILD) 's birth registration records located? In what city an'd state is that? IRECORD IN S.R.B. PG 2 I 1 1 812039 1 1 (47) 1 1 (46) IRECORD IN S.R.B. PG 2 24h. Where are (CHILD) 's current medical records located? In what city and state is that? IRECORD IN S.R.B. PG 2 24i. What was (CHILD) 's father's f u l l name? IKKCORD IN S.R.B. PC :'. 1 24 j. How old were you when (CHILD) was born? 1 T 1 1 1 1 Age 1 .1 (47) (46) (47) 22k. Were either of you using 23k. Were either of you using birth control at the birth control at the time you became pregnant ' time you became pregnant with (CHILD)? with (CHILD)? Age 24k. Were either of you using birth control at the time you became pregnant with (CHILD)? Yes.(48( -1 (ASK Q.22L) Yes.(48( -1 (ASK Q.23L) Yes.(48( -1 (ASK Q.24L) No -2 (SKIP TO Q.22m) No -2 (SKIP TO Q.23m) No -2 (SKIP TO Q 24m) IHANU RESPONDENT CARD "C" 1 22L. Please look at this card and tell me a l l of the numbers that apply to the types of birth control you or your partner were practicing? Ol.(4i9( O2.(50( O3.(5,l( 04. ( 2 ?l 05.(53( -1 -1 -1 -1 -1 06.(54( 07.(55( 08.(S6( 09. ( 7 17 10.(58( ll.<59( -1 -1 -1 -1 -1 -1 IHAND RESPONDENT CARD "c"l 3L. l.(49( 2.(50( 3.(51( 4(2 .5l 5.(53( -1 -1 -1 -1 -1 06.(54( 07.(55( 08.(56( 09. (Til 10.(58( ll.(59( -1 -1 -1 -1 -1 -1 2 (SPECIFY) 12 (SPECIFY) .6( (0 (CO TO Q.22m) IHAND RESPONDENT CARD "c"l Pluase look at this 24L. card and tell me all of the numbers that apply to the types of birth control you or your partner were practicing? Please look at this card and tell me all of the numbers that apply to the types of birth control you or your partner were practicing? Ol.(49( O2.(50( O3.(5l( 04. ( 7 5? 05.(53( -1 -1 -1 -1 -1 06.(54( 07. ( F I? 08.(56( 09. (TR 10.(58( ll.(59( -1 -1 -1 -I -1 -1 .6( (0 -1 12 (SPECIFY) .6( (0 -1 (GO TO Q.23m) ibl -1 (GO TO Q.24m) �CARD 001 SEVENTH CHILD 1 1 24m. How many months did it 23m. How many months did it take you to become pregtake you to become pregnnnt with this child? nant with this child? 1 1 Months 1 1 (61) (62) 1 I (61) Less than 1 month. (63( NINTH CHILD EIGHTH CHILD 22m. How many monLtis did it take you to become preg riant with this child? 1 1 812039 -1 1 1 Months 1 1 1 I (61) (62) Less than 1 month. (63( -1 1 1 Months (62) Less than 1 month. (63( 22n, Did (CHILD) have any birth defects? 23n. Did (CHILD) have any birth defects? 24n. Did (CHILD) have any birth defects? Yes.(64( -1 (ASK Q.22o) Yes.(64( -1 (ASK Q.23o) Yes.(64( No -2 (SKIP TO Q.22p) No .... -2 (SKIP TO Q 23p) No -1 -1 (ASK Q.24o) -2 (SKIP TO Q.24p) 22o. What kind of birth defects did (s)he have? Any others? 23o. What kind of birth defects did (s)he have? Any others? 24o. What kind of birth defects did (s)he have? Any others? 22p. Was (CHILD) ever diagnosed as having cancer? 23p. Was (CHILD) ever diag. nosed as having cancer? 24p. Was (CHILD) ever diagnosed as having cancer? Yes.(65( -1 (ASK Q.22q) • Yes.(65( Yes.(65( No -2 (SKIP TO Q.22s) No 22q. 1 I -2 (SKIP TO Q 23s) No. In what month and year 23q. was the diagnosis made? MONTH YEAR 1 I I 1 1 l l I l l (66) (67) (68) (691) -1 (ASK Q.23q) 1 1 . -2 (SKIP TO Q.24s) In what month and year 24q. was the diagnosis made? MONTH YEAR 1 I I 1 1 1 l-l 1 1 (66) (67) (68) (69) -1 (ASK Q.24q) 1 1 In what month and year was the diagnosis made MONTH YEAR 1 II 1 1 1 l-l 1 1 (66) (67) (68) (69) 22r. What kind of cancer was diagnosed? 23r. What kind of cancer was diagnosed? 24r. What kind of cancer was diagnosed? Not sure..(70( Not sure. . 7 ( (0 Not sure..(70( (GO TO Q.4s) 07 Two -1 (GO TO Q.5s) 08 TFfo 182 -1 (GO TO Q.6s) 09 79-fiO -1 �812039 CARD 004 SEVENTH CHILD NINTH CHILD EIGHTH CHILD 22s. (Does/Did) (CHILD) have a diagnosed learning disability? 23s. (Does/Did)(CHILD) have a diagnosed learning disability? 24s. (Does/Did) (CHILI)) have a diagnosed learning disability? Vee.tlj!(....-l (ASK Q.22t) Yes.(12( -1 (ASK Q.23t) Yes.(12( No No -2 (SKIP TO Q.23u) No -2 (SKIP TO Q 22u) 22t. What kind of learning disability (does/did) (s)he have? 23t. What kind of learning disability (does/did) (s)he have? -1 (ASK Q.24t) -2 (SKIP TO Q.24u 24t. What kind of learning disability (does/did) (»)he have? 22u. (Does/Did) (CHILD) have 23u. (Does/Did)(CHILD) have 24u. (Does/Did)(CHILD) have any physical, mental, or any physical, mental, or any physical, mental, motor impairments? motor impairments? motor impairments? Yes.(13( -1 (ASK Q.22v) IF CHILD IS DEAD: OTHERWISE: CONTINUE SKIP TO Q.22z !2w. On what date did . (CHILD) die? -1 (ASK Q.23v) No 22v. What kind of impairment (does/did) (s)lie have? Yes.(13( -2 (SKIP TO Q.23w) No 23v. What kind of impairment (does/did) (s)he have? IF CHILD IS DEAD: CONTINUE OTHERWISE: SKIP TO Q.23z 23w. On what date did (CHILD) die? Yes.(13( -1 (ASK Q.24v) -2 (SKIP TO Q.24w 24v. What kind of impairment (does/did) (s)he have? IF CHILD IS DEAD: CONTINUE OTHERWISE: SKIP TO Q.24z 24w. On what date did (CHILD) die? MONTH DAY YEAR MONTH DAY YEAR MONTH DAY YEAR I I I 1 I I 1 1 1 1 1 I 1 1 1 1 1 I I I 1 I I 1 1 l-l 1 1 I - ! 1 1 - 1 1 1 1 1 l-l 1 l-l 1 l-l 1 1 (14) (15) (16) (17) (18) (19 14) (15) (16) (17) (18) (19) (14) (15) (16) (17) (IB) (19) 2x. What was the cause of death? 23x. What was the cause of death? 24x. What was the cause of death? 2y. Where is (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. PG 2 | 23y. Where is (CHILD) 's death registered? In what city and state is that? RECORD .IN S.R.B.1 PC T[ 24y. Where is (CHILD) 's death registered? In what city and state is that? (GO TO Q.22z) (GO TO Q.23z) 183 (RECORD IN S.R.B. PC ^ i (GO TO Q.24z) �CARD 004 812039 EIGHTH CHILD SEVENTH CHILD NINTH CHILD 22z. Did you smoke on a fairly regular basis during this pregnancy? 23z. Did you smoke on a fairly regular basis during this pregnancy? 24 z. Did you smoke on a fairly regular basis during this pregnancy? Yes.(20( Yes.(20( Yes.(20( No -1 (ASK Q.22aa) __-2 (SKIP TO No. No _-2 (SKIP TO Q.23aa) Q.22aa) 22qa.When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. -1 (ASK Q.23aa) 23aa.When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. -1 (ASK'Q.24aa) _-2 (SKIP TO Q.24aa) 24aa.Wlien you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. I I I Packs I Packs Packs (21) (22) ( 1 T 2 2 T 2 T Less than one pack.(23( -1 Lees than one pack.(23( (21) -1 (22) Less than one pack.(23( -1 24bb. Did you drink alco23bb. Did you drink alco22bb. Did you drink alcoholic beverages (beer, holic beverages (beer, holic beverages (beer, wine, or hard liquor) wine, or hard liquor) wine, or hard liquor) on a regular basis duron a regular basis duron a regular basis during this pregnancy? ing this pregnancy? ing this pregnancy? Yes..(2A( -1 No __-2 (ASK Q.22cc) (GO TO NEXT CHILD) 22cc. About how many drinks a week would you say that you had during this pregnancy? I I I _l (25) (26) Yes..(24( No (ASK Q.23cc) (GO TO NEXT CHILD) 23cc. About how many drinks a week would you say 'that you had during this pregnancy? I drinks -1 -2 I I I I drinks (25) (26) Yes..(24( No -1 -2 (ASK Q.24cc) (GO TO NEXT CHILD) 24cc. About how many drinks a week would you say that you had during this pregnancy? I I I I I I drinks (25) (26) 07 (GO TO NEXT CHILD) (GO TO NEXT CHILD) 184 (GO TO NEXT CHILD) �CARD 001 812039 Q.25-27 Additional Children TENTH .CHILD ELEVENTH CHILD TWELFTH CHILD NAME: NAME : 25«. How old is (CHILD) now? 26a. How old is (CHILD) now? 1 I * I 1 Age ''I t (28) 1 1 1 I ARE (28) -1 1 1 Aae (28) (29) (29) Child died..(30( 26b. (Is/Was) (CHILb) male or female? 27b. (Is/Was) (CHILD) male or female? Male -1 ..(31( -1 26c. How much did (CHILD) weigh at birth? 25c. How much did (CHILD) weigh at birth? POUNDS OUNCES 1 1 ( 1 1 1 l-l 1 1 (32) (33) (34) (35) Don't know. . .(36( 1 1 Child died..(30( -1 25b. (Is/Was) (CHILD) male or female? Male, . . . K ..O 27a. Hov old is (CHILD) now? 1 1 (29) Child djed..(30( 1 1 NAME: POUNDS 1 1 -1 Male OK Female. . . . . ... " I 1 ' 1 1 l-l 1 1 (32) (33) (34l (35) Don't know. . ( 6 .3( -1 -1 -2 27c. How much did (CHILD) weigh at birth? OUNCES I 1 -1 1 1 POUNDS OUNCES 1 1 1 1 1 1 hi 1 1 (32) (33) (34)^(35) Don't know.. . 3 ( (6 -1 25d. What is (CHILD) 's birth- 26d. What is (CHILD) 's birth- 27d. What is (CHILD) 's birthdate? date? ~ date? MONTH DAY YEAR MONTH DAY , YEAR MONTH DAV YEAR I II 1 I I 1 1 1 I I 1 I I I I I 1 1 II 1 I I 1 1 1 1 l-l 1 l-l 1 .1 1 l-l 1 .H 1 1 1 1 1 l-l 1 l-l 1 1 1 4) (37) (38) (39) (40) (41) <42) (37) (38) (39) ( 0 (41) (42) 137) (38) (39) (40) T41) (42) [ALSO RECORD IN S.R.B.-PG2 1 ALSO RECORD IN S.fe.B,-PG2 1 [ALSO RECORD IN S.R.B.-PG2 1 25e. Was the child premature, 26e. Was the child premature, 27e. Was the child premature, full term, or overdue? full term, dr overdue? full term, or overdue? Premature. (43( Overdue .. Full term Not sure i Premature. (43( -ll(ASK -2j Q.25f) Overdue -ll(ASK Premature. (43( -2j Q.26f) Overdue -3")(SKIP TO Full term ,. -3l(SKIP TO Full term .. -4i 0 25g) *—"*~~ J -l"\(ASK -2j Q.27O -3! (SKIP TO 25f . How many weeks (overdue/ 26f. How many weeks (overdue/ 27f. How many weeks (overdue/ premature) was (CHILD)? ' premature) was (CHILD)? premature) was (CHILD)? I 1 I 1 | I weeks (44) (45) (GO TO Q.25g) 1 1 1 1 (44) 1 1 weeks (45) (GO TO Q.26g) 185 I 1 l 1 (44) l 1 weeks (45) (GO TO Q.27g) �CARD 001 TENTH CHILD 812039 ELEVENTH CHILD TWELFTH CHILD 25g. Where are (CHILD)'s birth registration records located? In what c i t y nnd state is that? 26g. Where are (CHILD)'s hirth registration records located? In what c i t y nnd stnte is that? IRECORD IN S.R.B. PC TT 27g. Where are (CHILD)'s h i r t h registration records located? In what city and state in that? IRECORD IN S.R.B. 1^2~T IRECORD IN S.R.B. PG_Ji'l 25h. Where are (CHILD)'s current medical records located? In what city and state is that? IRECORD IN S.R.B. vc 2 \ 26h. Where are (CHILD)'s current medical records located? In what city and state is that? IRECORD IN S.R.B. PC 2 I 27h. Where are (CHILD)'s current medical records located? In what city and state is that? IRECORD IN S.R.B. PC 2 I 25i. What was (CHILD)'s father's full name? JRECORD IN S.R.B. PC 2 I 26i. What was ( C H I L D ) ' s f a t h e r ' s f u l l name? 27i. What was (CHILD)'s father's full name? 25j. 26j. How old were you 27j. How old were you when (CHILD) was born? i1 i1 1 IRECORD IN S.R.B. PC 2 I when (CHILD) was born? Were either of you using 26k. birth control at the time you became pregnant with (CHILD)? Were either of you using 27k. birth control at the time you became pregnant with (CHILD)? Yes.(48( -2 (SKIP TO Q.25m) Nb I HAND RESPONDENT CARD' "C" I 25L. 1 I (46) (47) (46) (47) Yes.(A8(._._-! (ASK Q.25L) No -I (ASK Q.26L) -2 (SKIP TO Q.26m) No IHAND RESPONDENT CARD "CVM Q6.(54( Q7.(5S( Q8.(56( Q9.(57( 1Q.(58( ll.(59( 12 (SPECIFY) -1 -1 -1 -1 -1 -1 Ol.(49( 02.(M>< 03 ( - JLI 0.5 4(^ 05.(52( _ 06.(54( O7.(5_l( 08. (5j> 09. (V7 10.(58 ll.(59( (GO TO Q.25m) -1 (ASK Q.27L) -2 (SKIP TO Q.27m) IHAND RESPONDENT CARD "c1 I Please look at this card and tell me all of the numbers that apply to the types of birth control you or your partner were practicing? Ol.(49( O2.(50( O3.(5l( Q4.(52( Q5.(53( -1 '-1 -1 -1 -I -1 06.(54( 0.| 7(_ 0.^ 8(6 09. (JiT. 1Q.(58( ll.(59( 12 (SPECIFY) 12 (SPECIFY) .6( (0 Age Were either of you using birth control at the time you became pregnant with (CHILD)? Yes.(48( 27L. 26L. Please look at Chis Please look at this card and tell me all of card and tell me all of the numbers that apply the numbers that apply to the types of birth to the types of birth control you or your control you or your partner were practicpartner were practicing? ing? Q2.(50T -1 O3.(51( -1 Q4.(52( ~-l Q5.(53( -1 How old were you when (CHILD) was born? Age 1 Age (46) (47) 25k. IRECORD IN S.R.B. PC.2 I .6( (0 (GO TO Q.26tn) .(60( (GO TO Q.27m) �CARD 001 TENTH CHILD 25m. ELEVENTH CHILD How many months did it take you to become preg nant with this c h i l d ? f j 1 I TWELFTH CHILD 26m. How many months did it 27m. How many months did it take you to become pregtake you to become pregnant with this child? nant with this child? 1 I Months 1 1 U t J (62) 1 1 T I Months 1 I (61) (621 Less than 1 month. (63( 25n. 812035 -1 Did (CHILD) have any birth defects? 1 I 1 | Months (61) (62) Less than 1 month. (63( -1 Less than 1 month. (63( 26n. Did (CHILD) have any b i r t h defects? 27n. Did (CHILD) have any birth defects? Yes.(64( -1 Yes.(64( -1 (ASK Q.25o) Yes.(64( -1 (ASK Q.26o) No...... -2 (SKIP TO Q.25p) No -1 (ASK Q.27o) -2 (SKIP TO Q.26p) 25o. What kind of birth defects did (s)he have? Any others? 26o. What kind of bi.rth defects did (s)he have? Any others? 27o. What kind of birth defects did (s)he have? Any others? 25p. Was (CHILD) ever diagnosed as having cancer? 26p. Was (CHILD) ever diagnosed as having cancer? 27p. Was (CHILD) ever dragnosed as having cancer? Yes.'(65( -1 (ASK Q.25q) Yes.(65( -1 (ASK Q.26q) Yes.(65( -1 (ASK Q.27q) No ..... -2 (SKIP TO Q.25s) No -2 (SKIP TO Q.26s) No -2 (SKIP TO Q.27e) 25q. 1 I In what month and year 26q. was the diagnosis made? MONTH 1 l l (66) (67) II - YEAR 1 I l (68) (69) 1 l In what month and year 27q. was the diagnosis made? 1 I MONTH YEAR 1 I I I 1 l l I l l (66) (67) (68) (69) In what month and year was the diagnosis made? 1 1 MONTH YEAR . I I I I I 1 l-l 1 1 (66) (67) (68) (69) 25r. What kind of cancer was diagnosed? 26r. What kind of cancer was diagnosed? 27r. What kind of cancer was diagnosed? Not s u r e . . ( 7 0 ( Not sure..(70( Not sure..(70( (GO TO Q.4s) 10 79-80 -1 -1 (GO TO Q.5s) 11 TF50" 1S7 (GO TO Q.6s) 12 79^50 -1 �CARD 004 TENTH CHILD 812039 ELEVENTH CHILD TWELFTH CHILD 25s. (Does/Did) (CHILD) have a diagnosed learning disability? 26s. (Does/Did)(ClllLD) have a diagnosed learning disability? 27s. (Does/Did)(CHILD) have a diagnosed learning disability? Ves.(12( -1 (ASK Q.25t) Yes.(l2( Yes.(l2( No -1 (SKIP TO Q.25u) No 25t. What kind of learning disability (does/did) (s)he have? -1 (ASK Q.26t) -2 (SKIP TO Q.26u) No 26t. What kind of learning disability (does/did) (s)he have? 26u. (Does/Did)(CHILD) have 25u. (Does/DidMCHILD) have any physical, mental, or any physical, mental, o motor impairments? motor impairments? Yes.(l3( Yes.(13( -1 (ASK Q.25v) No -2 (SKIP TO Q 25w) No 25v. What kind of impairment (does/did) (s)he have? -1 (ASK Q.26v) -2 (SKIP TO Q.27u 27t. What kind of learning disability (does/did) (s)he have? 27u. (Does/Did)(CHILD) have any physical, mental, motor impairments? Yes.(13( . . -2 (SKIP TO Q.26w) No .. 26v. What kind of impairment (does/did) (s)he have? -1 (ASK Q.27t) -1 (ASK Q.27v) -2 (SKIP TO Q.27w 27v. What kind of impairmen (does/did) (s)he have? IF CHILD IS DEAD: IF CHILD IS DEAD: IF CHILD IS DEAD: OTHERWISE: OTHERWISE: OTHERWISE: CONTINUE SKIP TO Q.25z >5w. On what date did (CHILD) die? CONTINUE SKIP TO Q.26z 26w. On what date did (CHILD) die? CONTINUE SKIP TO Q.27z 27w. On what date did (CHILD) die? MONTH DAY YEAR MONTH DAY YEAR MONTH DAY YEAR 1 I I i l l 1 1 1 1 1 1 I I 1 1 1 I I 1 I I 1 1 l-l 1 l-l 1 1 1 1 l-l 1 l-l 1 1 l-l 1 l-l 1 1 14) (15) (16) (177 U8) (19) (14) (15) (16) (17) (18) (19) (14) (15) (16) (17) (18) (19 5x. What was the cause of death? 26x. What was the cause of death? 27x. What was the cause of death? 5y. Where is (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. PG 2 1 26y. Where is (CHILD) 's death registered? In what city and state is that? 1 RECORD IN S.R.B. PG 2 1 27y. Where is (CHILD)'s death registered? In what city and state is that? RECORD IN S.R.B. PG 2 1 (GO TO Q.25z> (GO TO Q.26z) 188 (GO TO Q.27z) �CARD 007 Q.28 Additional Periods of Infertility FIFTH PERIOD FOURTH PEKIOD 28a. In what month and year did the fourth period begin? And in what month and year did it end? FROM MONTH YEAR T 28d. In what month and y e a r d i d t h e f i f t h period h e g i n ? And in what month and y e a r did it end? FROM MONTH YEAR 1 1 i 1 1 I I l l I l l (18) ( 1 9 ) ( 2 0 ) (21) ' TO MONTH YEAR 1 1 1 1 1 1 I l l I l l (22) (23) ( 2 4 ) (25) I (187(195 (20) (21) TO MONTH YEAR I I I (22) (23) 28h. I (24) (25) How old were you in (BEGINNING DATE OF PERIOD)? 28e. (26) (27) Yes. No.. During this period did either of you see a doctor to discuss any difficulties in conceiving children? .2( (8 How old were you in (BEGINNING DATE OF PERIOD)? 1 1 AGE 28c. 812039 28f. 1 1 1 1 (26) ( 2 7 ) 2fig. 1 I 77T -2 02 79-80 (GO TO NKXT PERIOD) M - l 1 I l 1 l ( 2 0 ) (21) MONTH 1 I 1 l l YEAR M - (22) ( 2 3 ) 28h. 28i. Yes No 1 l TO 1 1 During this period did either of you see a doctor to discuss any d i f f i c u l t i e s in conceiving children? In what month and year did the sixth period h e g i n ? And in what month and year did it end? FROM MONTH YEAR fl8) (19) AGE -1 (GO TO NEXT PERIOD) SIXTH PERIOD 1 I l 1 l (24) (25) How old were you in (BEGINNING DATE OF PERIOD)? 1 1 1 1 (26) ( 2 7 ) AGE During this period did either of you see a doctor to discuss any d i f f i c u l t i e s in conceiving children? OR( No -1 -2 03 79-80 (RETUKN TO Q.].2s) �;:CARP 004. TENTH CHILD 812039 ELEVENTH CHILD TWELFTH CHILD 25z. Did ybu smoke on a fairly regular basis during this pregnancy? 26z. Did you smoke on a fairly regular basis during this pregnancy? 27z. Did you smoke on a fairly regular basis during this pregnancy? Yps.(20( Yes.(20( -1 (ASK Q.26aa) Yes.(20( -1 (ASK Q.27aa) No -2 (SKIP TO 26aa) No -2 (RETURN TO Q.7) No -1 (ASK Q.25aa) ^-2 (SKIP TO Q.25aa) 25aa.When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack wi; mean 20 c igarettes. — 26aa.When you were smoking cigarettes oh i fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. I I I 1 1 Packs I 27aa.When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. 7 1 (21) (22) <21) (22) Less than frne pack.(23( -1 I I I Packs Less than one pack. (23( Packs (21) (22) -I Less than one pack.(23( -1 25bb. Dj.d you drink alco26bb. Did you drink alco27bb. Did you drink alco-, holic beverages (beer, holic beverages (beer, holic beverages (beer, wine, Or hard liquor) wine, or hard liquor) wine, or hard liquor) on a regular basis duron a Regular basis duron a regular basis during this pregriancy? ing this pregnancy? ing ttiis pregnancy? Y..s..(24( -1 (ASK Q.25cc) Nn -2 (GO TQ NKXT ~~ CHILD) 25cc. About how many drinks a week Would you say that y6u had during this pregnancy? Yes..(24( ; -I (ASK.Q.26cc) No....... -2 (GO *0 NEXT' CHILD) Yes..(2A( -1 No _-2 26cc. About how many drinks a week would you say that you had during this pregnancy? 27cc, About how many drinks a week would you say that you had during this pregnancy? 1 (ASK Q.27cc) (GO TO NEXT CHILD) 1 drinks 1 (25) (25) 1 drinks 1 (26) to (GO tO NEXT CHILD) 1 drinks 1 (2'S) (26) 12 79-80 (GO TO NEXT CHILD) (RETURN TO Q.7) �812039 CARD 013 <>-') I Add i t ion.-i I M i ' i i - . - i r r i . T R i ' s / S r i 11 h i r r h s / A b o r t i o n s PKKCNANCV t< I'UECNANCYJH I i i wh;it t i u n i l l i a n d vi'«'ir d i d l;h*' f i r s I s u r h pregnancy end? MONTH IIOn. In w l i a t montli and y e a r did the next such pregnancy end? MONTH YKAK T YtAK MONTH Mi scar i i.i^r. . (J^(_ Si i l l b i r l l i Abortion.. _. 2''c. iI -1 -2 "3 After how mnny weeks did I (17) (is) 2'ib. Did this pregnancy end in a mi Kearriage , s t i l l b i r t h , "r abortion? (15) Misoarriage. . ( I <U SI: i I I b i rth t Abort ion 30c. r I Weeks (?0) (21) (20) d . How old were yon at that limp? -1 -2 -3 31b. Did this pregnancy end in a miscarriage, s t i l l b i r t h , or abortion? Mi scarriage. . ( 1 *)( Stillbirth Abort ion '.ilc. Was (STUDY RESPONDENT) your partner in this pregnancy? I I Weeks (20) (21) (21) 31d. How o l d were you at t h a t time? 30d. How old were you at that time? No 2c)t'. 30e. I I Were e i t h e r of you . using b i r t h control at the time you became pregnant? Y e s . ( 2 5 ( -1 ( A S K Q . 2 9 g ) No _-2 ( S K I P TO O . J O h ) (CO TO Q . 2 Q g / h ) Was (STUDY RESPONDENT) your p a r t n e r in t h i s pregnancy? Ye s. ( 2/4 ( No -1 -2 30f. -1 -2 -3 After how many weeks did the pregnancy end? ( W R I T E IN AGE) (22) (23) 2')e. ( 1 7 ) (18) I I ! Weeks (WRITE IN ACE) I (WHITK I.N ACE) I I (15) ( 1 6 ) After how many weeks did the pregnancy end? I I I I ( 1 7 ) (18) (If.) 30b. Did this pregnancy end in a miscarriage, s t i l l h i r t h , or abort ion? the prugnancy end? iI YEAR T I <n) (is) 3 1 a . In what, month and year did rhe next sucli pregnancy end? Yes.(25( -1 (ASK Q.30g) Mo _-:> (SKIP TO I (CO TO ().30g/h) 1'Jl 31e. I Was (STUDY RESPONDENT) your partner in this pregnancy? -1 -2 Were either of you using birth control at the time you became pregnant? I ( 2 2 ) (23) Ves.(2'(( No. 31f. -1 -2 Were either of you using hirth control at the time you became pregnant? Yes.(25( T-I (ASK Q.31g) No -2 (SKIP TO Q . 3 1 h ) (CO TO Q . 3 1 g . / h ) �H I2 CARD OH Q.29-31 PREGNANCY 4 PREGNANCY (> PKKGNANCY 5 lHANI) RESPONDENT CAKU "C" \ IllAND RESPONDENT CARD "C" I IllANI) RESPONDENT CARD "C" 1 29g. Please look at tliis cart 30g. Please look at this car 31g. Please look at this card and tell rap a l l the num and tell me all the numand tell me .ill the numhers that apply to the bers that apply to the bers that apply to the types of birth control types of birth control types of birth control you or your partner wen you or your partner were you or your partner were using. us inn • usiniiOl.(2(>( 0?.(27< O3.(28r~ (K..(29( O5.(30( -1 -1 -1 -1 -1 Of,. OK -1 <17.(32( -1 08. (337" "-1 09.<34( -1 10.(35( -1 ll.(30( -1 12 (SPECIFY) 1 1 (38) Ofi.OK ()7.(32( OH.(33(~ 0)U( '.4 1().(35( ll.(36( -1 How many months did it take you Co become pregnant this time? 1 -1 -1 "-1 -1 -1 -1 -1 "-1 -1 -1 -1 12 (SPECIFY) .(37( 29h. Ol.(2h< 0?.(27< 03.(2H7~ 0/..(2<M 05.(3(J( (39) Less th.in 1 month. (A0( Wasn't trying 1 I I (38) -1 -'1 T 31h. (39) 1 1 (38) LI-MS Hum ! month. UiO( -1 (IF MISCARRIAGE OR STILLBIRTH IN Q.30h, ASK Q.30i . IF ABORTION IN Q.30b, SKIP TO Q.30m) Did a doctor t e l l you why this (miscarriage/ stillbirth) might have occurred? 29j . What did the doctor say caused the (miscarringe/st i 1 Ibirth) .' ?%-. 1 1 Months (39) -1 •)i i. (IF MISCARRIAGE OR STILLBIRTH >N Q.31b, ASK Q.3H. IF ABORTION IN Q.31b, SKIP TO Q.31m) Did a doctor tell you why this (miscarriage/ stillbirth) might have occurred? -1 (ASK Q.30J) Yes.UH -i (SKIP TO Q 30n) No -1 (ASK Q.31J) -2 (SKIP TO 0 31n) Oj. What did the doctor say caused the (miscarr iaRc/st i 1 I h i rt'li)? 1 j. Wlial did the doctor say caused the (miscarriani.'/si i 1 Ibirth)? Wh.-it is NIP nniiir of 1 In' Ilk. duct 01 or mini ioi 1 1 m- i 1i t y rh.il you consul ted about this? WIl.H i K Mil- namr of I'lir tlin- 1 (>t or nusl i i- H 1 1 m- i 1 i t v t h a t yon consulted .Tlr-ut this? : Ik. Wh.il is llu> ii.-iiiu- of the doctor ttr nu'd i en 1 l a c i l i t y that you consulted about this? TRECORD IN S.K.B. - PC 3 I 29L. es.(4K (, -1 Less than 1 month. ( 0 4( Did n doctor t e l l you why t h i s (miscarriage/ s t i l l b i r t h ) might have oc c u r re d ? -1 (ASK Q.29J) -1 -1 -1 -1 -1 -1 How many months did it take you to become pregnant this time? 1 1 1 Months 29i. (IF MISCARRIAGE OK 30 i. STILLBIRTH IN Q.29b, ASK Q.29i. IF ABORTION IN Q.29h, SKIP TO Q.29m) Yes.(41( Qft.OK 07.(3?( OH.(T3l 09.(34( 1().(35( ll.(36( .(37( -1 How many months did it take you to become pregnant this time? 1 1 1 Months -1 -1 '-1 -1 -1 12 (SPECIFY) .(37( 30h. Ol.(26( 02.(27( 03.(?H( 04. (29( ()5.(30( In what month and year was that? IRECORD IN S.R.B. - pn i 1 '.)[.. In what month and year wns t h a t ? IRECORD IN S.R.B. - PC; 3 1 1L. In what month and year was that? IRECORD IN S.R.B. - PC 3 1 IRECORD IN S,R.B. - PC 3 1 (SKIP TO ij.29n) (SKIP TO Q.30n) (SKIP TO Q.31n) (CO TO NEXT PACK.) (CO TO NKXT PACK) (CO TO NEXT I'AGE) IRECORD IN S.R.B. - PC 3 1 192 �CARD 013 812039 Q.29-31 PREGNANCY 4 PREGNANCY 5 PREGNANCY 6 29m. What was the main reason for the abortion? 30m. What was the main reason for the abortion? 3)m. What was the main reason for the abortion? 29n. Did you smoke cigarettes on n fairly regular basis (lurinp this pregnancy? ')0n. Did you smoke ciparettcs on a fairly regular basis during this pregnancy? 31n. Did you smoke cigarettes on a fairly regular basis during this pregnancy? Yes.(42(. -1 (ASK Q.29o) -2 (SKIP TO Q.29p) No. 29o, When you' were smoking cigarettes qn a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. Yes.(4_2j__-l (ASK Q.30o) Yes.(42( -1 (ASK Q.31o) No. No. -2 (RETURN TO Q.lBa) 30o I I ,\ Packs (43) (44) Less than one pack.(45(__-l (ASK Q.29q) (SKIP TO About how many drinks a week would you say that you had during t h i s pregnancy? T I Packs I 30q. -1 l.ess than one pack. (45 ( -1 Tip. Did you drink alcoholic beverages (beer, wine, or hard liquor) on a regular basis during this pregnancy? Yes..(46j__-l No __'-2 T I Packs (43) (44) Less than one pack.(45( Q.30a) 29q. 31o. When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. (43) (44) 30p. Did you drink alcoholic beverages (beer, wine, or hard liquor) on a regular basis during this pregnancy? Yes..(46( -1 No __-2 When you were smoking cigarettes on a fairly regular basis during this pregnancy, on the average, how many packs per week did you smoke? By pack we mean 20 cigarettes. I T 29p. -2 (SKIP TO Q.30p) (ASK Q.30q) (SKIP TO Q.31a) About how many drinks a Wi'eVi would you say that you had during this pregnancy? Did you drink alcoholic beverages (beer, winp, or hard liquor) on a regular basis during this pregnancy? Yes. . (46( -1 (ASK Q.31q) No __-2 (GO TO NEXT PREGNANCY) 31q. About how many drinks a week would you say that you had during this pregnancy? T—r I drinks I drinks (47) (48) (47) (48) 01 70-80 I I I drinks (47) (48) 02 79-80 (ASK Q.30a) (ASK Q.31a) (RETl'KN TO Q. IPa) �Louis HARRIS AND ASSOCIATES. INC. 63O NEW TEL FIFTH AVENUE Y O R K , NEW YORK (ZI2I 9 7 5 - I 6 O C T E L E X I O I I I 146383 LOUIS H A R R I S I N T E R N A T I O N A L . INC LOUIS H A R R I S P R A N C E OPINION R E S E A R C H 21 RUE VIVI6NNE SO W E L B E C K ST. 7SOO2 PARIS, FRANCE LONDON WIM BAB ENGLAND TEL. 01- zeo -ees*! TELEX: zooeo> TEL: O I - 4 8 6 - 5 I 5 I T E L E X . 21-«O3 Dear Louis Harris and Associates has been asked by the United States Air Force to conduct a study of the health of Air Force pilots and servicemen who served during the Vietnam conflict. The U.S. Air Force School of Aerospace Medicine is undertaking this study in order to answer questions about possible effects of having served in Vietnam. I have just completed an interview with Louis Harris and Associates on the United States Air Force Health Study. As part of this study, they would like to interview the former wives of study participants. You will be asked to provide information on health and health care services. It is essential to the accuracy and completeness of the study that all selected participants .and their families participate in the study. Reliable information will help produce sound conclusions of vital relevance to all Vietnam veterans and their families. I would appreciate it very much if you also would grant a representative of Louis Harris and Associates an interview. Shortly after receiving this letter, you will be called on by an interviewer from Louis Harris and Associates who, at your convenience, will either conduct the interview or set up an appointment. The interviewer will answer any questions you may have about the study. Thank you. Sincerely, TSIGNATURE OF STUDY RESPONDENT; (PRINTED NAME OF STUDY RESPONDENT) 1Q4 CENTRE �LOUIS HARRIS AND ASSOCIATES, INC. 630 FIFTH AVENUE NEW YORK, NEW YORK 10111 Dear Louis Harris and Associates has been asked by the United States Air Force to conduct a study of the health of former .and current AirForce., servicemen who served during the Vietnam conflict. The U.S,tAir Force 5 | 1 of Aeroc^ space Medicine is undertaking this study in order to answer questions about possible effects of having served in Vietnam. In order to complete the study, we need to interview both tKe Air Force personnel selected for this study and their wives. We have alrea^ly completed an interview with your former husband and now we need your cooperation in this endeavor. The interview is quite shor*t and should take no longer than twenty minutes to complete. The questionnaire focuses on the health of you and your family. The validity of the results depends on the willingness of women like yourself to participate in the study. Reliable information will enable the Air Force to reach sound conclusions of vital relevance to all Vietnam veterans and their families. A copy of the letter from the Surgeon General of the Air Force which was sent to your former husband is attached. It will explain the purpose of the study in detail. The fact sheet, which is part of this letter, includes a telephone number which you may call if you have additional questions. One of our interviewers will be contacting you in the next two weeks to arrange an appointment. We will schedule the interview at your convenience. Thank you for your cooperation. important project. I hope that you will join us in this Sincerely, Louis Harris 195 �Louis HARRIS AND.ASSQCIATES, INC. 63O NEW TEL FIFTH AVENUE Y O R K , NEW Y O R K '2121 9 7 5 - I 6 O O T E L E X I O I I I 118383 LOUIS HARRIS INTERNATIONAL. INC. OPINION RESEARCH CENTRE LOUIS H A R R I S FRANCE 21 RUE V I V I E N N E 3O WELBECK ST. 7SOOE PARIS, FRANCE LONDON WIM BAB ENGLAND TEL. oi-zeo -9OB4 TELEX: aooeoi r TEL: oi-«iee-6i6i TELEX: 24*03 PRIVACY ACT STATEMENT - EPIDEMIOLOGIC STUDY AUTHORITY: Section 133, 1071-87, 3012, 5031 and 8012, Title 10, United States Code and Executive Order 9397. PRINCIPAL AND PURPOSEJSJ: The purpose of requesting personal information is to assis~t medipal/technical personnel in developing records relative to your participation in an approved epidemiologic investigation. The Social Security Number (SSN) and Armed Forces Service Number (AFSN) are necessary to identify the person and records. ROUTINE USES: This information will be used to initiate, coordinate, and conduct the investigation. It will be used to compile statistical data, but information allowing identification of the individual volunteer will not be included. Data and results from this investigation may be used to supplement other approved research studies conducted at the USAF School of Aerospace Medicine or at other Federal agencies engaged in the conduct of similar studies. WHETHER DISCLOSURE IS MANDATORY OR VOLUNTARY, AND EFFECT ON INDIVIDUAL FOR NOT PROVIDING INFORMATION:" Disclosure or requested information is voluntary. If the information is not furnished, acceptance as a subject is not possible. This is an all-inclusive Privacy Act Statement which will apply to all requests for personal information made by medical/technical personnel during the time you are a volunteer subject. A copy of this form will be placed in your investigation subject folder as evidence of this notification. Your signature merely acknowledges that you have been advised of the foregoing. If requested, a copy of this form will be furnished to you. Signature of Volunteer SSN 196 ' Date �SHOW CARD "C1 NO. 812039 01 Pill 02 Douche 03 Foam 04 Jelly, Cream, Suppository 05 IUD 06 Condom, Rubber 07 Diaphragm 08 Diaphragm attd Jelly 09 Rhythm *- Calendar 10 Rhythm - Temperature 11 Withdrawal 12 Other 197 �SHOW CARD "D-l? STUDY NO. 812039 a. Sterility due to surgery b. Known sterility due to injury, accident, or illness c. Sterility due to unknown causes d. Lack of interest in sex e. Other known medical/physical conditions f. Some other reason J.98 �t.S. LQUIS MARHIS AND N E W Y O R K . N E W Y O' f t' * . I O I I I TEi. 2 I Z I 9 7 5 - I 6 Q O T£i,E» ii>93e? LOVIIS HAIR'S I N T E R N A T I O N A L . ' N C . LOUIS H A R R I S FRANCE 30 21 RUE VIVI6NNE 7SOO2 PARIS, TRANCE LONDON WIM 6A6 ENGLAND TCL. oi-aoo -oestf TELEX: iooeoi TEL: Oi-«»86-6isi TELEX'. £*- UNITED STATES AIR FORCE HEALTH STUDY Name of Medical Provider/Medical Facility Name of Place Street Address "State" ( 1 Phone Number Dear Doctor or Administrator: I am participating in a survey conducted for the United States Air Force to gather information on- the health of current and former Air Force personnel and their families. As part of this survey, medical providers who have delivered health care services to 'me are being asked to supplement the information that I have already provided to the study. By this statement 0r a photocopy of it, I hereby authorize and request you to furnish the United States Health Stu^y with any medical information in your records on the health services received by me, ^ in connection with a birth on • •/_, _ , '; _,_ ,___ _ . Related health care was provided during the period to .: Thank you very much. Sincerely, Resp. ? Signature cif ;Pati;enlf FOR OFFICE USE ONLY: Date MEDICAL PROVIDER PERMISSION FORM: SPOUSE �FOR OFFICE USE ONLY: LOUIi HARRIS AND ASSOCIATES, INC 630 Fi fth Avenue New York, New York 10111 c # 812039 Air Force Health Survey Respondent # INTERVIEW EVALUATION (NI'tRVlEWER; COMPLETE THE FOLLOWING IN PRIVATE IMMEDIATELY AFTER THE INTERVIEW, USING YOUR BEST JUDGMENT TO ANSWER F.ACH ITEM. I. .lace of respondent: Black.... Nonblack. ?.a. Did the respondent want to terminate the interview before it was finished? No (SKIP TO Q3a) Yes, (ANSWER 2b AND 2c) 2b. At what question number or during what question series? >r V'hat was the reason? la viere there any (other) significant problems during f.he interview? No Yes _ (SKIP TO Q4a) _H'J (ANSWKR 3b) 3b. Describe the problems. 4a. Did respondent refer to records during the interview? No Yes (SKIP TO Q5a) (ANSWER 4b) 4b. What records did the respondent use? _____ 5a. Was anyone else present at any time during the interview? No Yes (SKIP TO Q6) (ANSWER 5b and 5c) 5b. Who was present? rRTCQROELA~TIONSHIP | 5c. During which section(s)? 6. , Length of interview: minutes 200 �LOUIS flAKKls AND ASSOCIATES, INC, 630 I i H:ji Avenue New York, New York 10.111 Study # B12039 AIR FORCE 'HEALTH SURVEY k FQRM TU: New York Office. Lbuis Harris and Associates FROM; \________.„,_..... . . . . . . . __ "" "~~* TnteTviewer'Nanie "-" "PTease" "PrTnt" • __ ...... ' fliis pat:k,tge contain:, the following material for Study Subject Respondent" Numhei- I W r i t e in NtlMnr.R of eai:h J.teni.. bgtflS. s.pn.1;. PH. .*}!§.. It.'iV. at tjl Study Subject Name Assignment Sheet Study Subject. Pri vciey Act Statement (Signed)... , __ Study Subject Questionnaire......' Study Subject Supplemental Recording Book .., Study Subject Sel f Administered Form ; Study Subject Medical Consent Form. ,, Study Subject former Wi fe Consent Letter., , Study Subject Interviewer Evaluation Form. PRESENT.WlFt INTERVir.W Privacy Act StateriuVnt (Signed) , Spouse (jui'':>t l o n r i a i i c . Spou'ii-1 '.upMliMiiontril Recording Book. SIJOUSL* Medical Consent Form Spouse .Interviewer I valuation form. Former W^Tc Name Assignment Sheet.. l'rivm.y Act. Stotemeiil. (Si(jned) Spouse (jut", t Minna i reSpouse 'iup|'li-.'iiient.<i I Kecordituj Book. Spouse M u d i c d l Coii.i'iil I orm Spoust; I n t e r v i c w e i ' I v.iluiit. ion form. PK'iXY. i f i r i k V i r w I'coxy Name Assignment. Shi-nl Privacy A c t '.taleim-nt. '.(Signed) Proxy f)ui.'Stionnaifi'.. . ......... •. .• I'roxy Supp|i'iw;nt..i I Kei.-ordin.g Book Proxy Modji'.d I LOIIMTI! ........... I'roxy I rtl.ci v ii'wci I Viiluat i on l<ci.i.'i vi/d: Date l.ln.r.|.ei! ii: by: I ! I 201 ^___ .__ , , �CHAPTER III NEXT OF KIN (PROXY) QUESTIONNAIRE The following Next of Kin (Proxy) Questionnaire was used to collect baseline data for the Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicide Orange. This data was collected during 1981-1982. All available proxies were Included in this data collection effort. The questionnaire and supplemental recording book are the actual field instruments. They have been photocopyed and reduced for the purpose of this report. One show card, anatomical representation, is included as an attachment to demonstrate to the reader complete data collection methods. Additional attachments include the Privacy Act Statement, Life Events Chart, Medical Permission Form, Introductory Letters, Interview Evaluation, and Mailing Transmittal Form. The Next of Kin (Proxy) Questionnaire, as used in the field, follows. 202 �630 Fifth Avenue New York, New York O.M.B. NUMBER 0701-0033 Approval Expires 11/30/62 10111 Case No. Study No. 812039 Respondent t: PROXY QOKST10NNAJRE CONFIDENTIAL This study is being conducted to collect information on the health of current and former Air Force personnel and their families. Since I will be asking you questions about the health, career, and personal history of (STUDY RESPONDENT), we have prepared a Life Events Chart to help you remember when various events in his life occurred. The best way to use the Life Events Chart is to first record when he was born in Column, or how old he was in 1930, if he was born before 1930. Then, record his subsoquent 5-year intervals in the Age Column. Next, note the year he graduated hiflli school and/or college in the next column. You can enter the year he joined military in the next column. There are other columns to record any marriages or children he may have had, as well as other major events in his life. the Age age at from the 1 w i l l tie asking you questions about each of these areas during the interview. If you w i l l take a few moments to fill out the Life Events Chart now, it w i l l help you to recall d,it.es and ages during Che interview. First, 1 have a feu background questions to ask you. 1. What (is/was) (STUDY RESPONDENT'S) date of birth? (WRITE IN DATE) I MONTH I DAY YEAR I I I T~7iT 1 ) ( 2. I I In what city and state was (STUDY RESPONDENT) born? T RECORD IN SUP PIGMENTARY RECORDING BOOKTON PACE l] 3. What was his religious preference — was it Protestant, Catholic, Jewish, some other, religion, or no religion? -1 Protestant Catholic Jewish Other (SPECIFY) -A "-5 None. What van the highest grade or year in high school that he completed? Less than 1 year of H.S...C (_ 1st year H.S. (9th Grade) ^ 2nd year H.S. (10th Grade)....* 3rd year H.S. (llth Grade) " Ath year H.S. (12th Grade) " Z03 -1 _ -3 �CARD 812039 I HAND RESPONDENT CARD "A"I , 5a. Please look at this card and tell roe which of these regular academic school certificates, diplomas, or degrees (STUDY RESPONDENT) had obtained? I MULTIPLE RECORD BELOWl YEAR High school diploma (__J -1 ( ) ( YEAR High school equivalency diploma ( ( -1 Associate of Arts (A.A.) ( ( -1 YEAH YEAR Bachelor of Arts (B.A.) or Bachelor of Science (B.S.)... (_J -1 I )( ) 1 YEAR Masters ( ( -1 ( ) () YEAR Doctorate ( ( -1 Others (SPECIFY) YEAR (1) -1 YEAR (2) -1 YEAR (3) -1 No certificate, diploma, or degree (/olunteered)....(( -1 IFOR EACH DEGREE. PIPLOMA. OR CERTIFICATE, ASK Q.Sbj 5b. In what year did he receive his (CERTIFICATE/DIPLOMA/DEGREE)? IRECORP ABOVET �CARD 812039 6a. 1 ai" interested in train ing programs which prepared (STUDY RESPONDENT) for a major change in his occupation. Fi rst, I will ask about civilian job training programs. Resides the formal schooling you told me about, did he participate in any civilian job training programs that prepar ed him for a major change in his occupation? Yes...( ( -1 (ASK Q.6b) No -2 1st Program b. (SKIP TO Q.7a) 2nd Program For what kind of work was his first civilian . training program preparing him? 3rd Program f. For what kind of work was his next civilian training program preparing him? (( c. d. MONTH ,1 l l 1 1 - YEAR 1 I l MONTH 1 1 T 5 '( ')' I I l-l YEAR 1 I' (")(> ( MONTH 1 1 1 1 t. I l-l I YfiAR I 1 1 1 1 1 I I 1 ( MONTH I 1 I I hi YEAR [ 1 1 1 L. In what month and year did he complete this training? 1 1 m. MONTH YEAR 1 I I I I 1 l-l 1 1 < )( ) ( J( } Did he participate in any other civilian job training program that prepared him for a major change in bis occupation? -1 (ASK Q.6j) Yes.( -2 (SKIP TO Q.7a) ( Nb....k. 205 < In what month and year did he start this training? I 1 1 1 i. Did he participate in any other civilian job training program that prepared him for a major change in his occupation? -1 (ASK Q.6f) Yes.( -2 (SKIP TO Q.7a) jjo k. YEAR I l-l ( ( 1 1 In what month and year did he complete this training? MONTH 1 1 Did he participate in any other civilian job training program that prepared him for a major change in his occupation? Yes.( No g. In what month and year did he start this training? 1 l In what month and year did he complete this training? 1 1 e. (( In what month and year did he start this training? I I < ( ( (( j. For what kind of work was his next civilian training program preparing him? -1 (RECORD ADDITIONAL TRAINING PROGRAMS . IN S.R.B. ON PG. 13) -2 (GO TO Q.7a) �CARD 812039 7a. Now, let's calk about military technical and specialized training programs that prepared (STUDY RESPONDENT) for a major change in his occupation. Besides the formal schooling (and the job training programs) you've told me about, did he participate in any military technical or specialized training programs that prepared him for a major change in his occupation? Yes...( ( No -2 1st Program b. -1 (ASK Q.7b) (SKIP TO Q.8) 2nd Program For what kind of work was his first military training program preparing him? 3rd . Program g. For what kind of work was his next military training program preparing him? L. For what kind of work was his next military training program preparing him? ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( c. What was the AFSC for that job? ( ( d. ( ( In what month and year did he start this training? i. In what month and year did he start this training? MONTH YEAR 1 1 '[ I ' " 1 ' 1 I . I H I 1 ( ) ( ) (• ) ( ) e. m. What was the AFSC for that job? h. What was the AFSC for that job? In what month and year did he complete this training? MONTH 1 1 1 1 .1-1 ( ) ( ) YEAR 1 1 1 1 ( )( ) f. Did he participate in any other military job training program that prepared him for a major change in his occupation? Yes.( ( rl (ASK Q.7g) I 1 j. MONTH YEAR I I I 1 1 1 l-l 1. 1 ( )( ) ( )( ) In what month -and year did he complete this training? MONTH 1 I ( ( n. 1 . I 1 o. 1 1 1 1 1 1 l l I l l ( )( ) ( ) ( ) k. Did he participate in any other military job training program that prepared him for a major change in his occupation? MONTH YEAR I I I 1 1 1 1 - 1 1 1 ( ) ( ). ( ) ( ) In what month and year did he complete this training? YEAR 1 es.( ( No •••• In what month and year did he start this training? MONTH YEAR 1 i 1 I I 1 l-l 1 1 ( )( ) ( )( > p. Did he participate in any other military job training program that prepared him for a major change in his occupation? -1 (ASK Q.7L) Yes.( -2 (SKIP TO Q 8 ) No ( -1 (RECORD ADDITIONAL TRAINING PROGRAMS IN S.R.B. ON PG. 14) -2 (GO TO Q.8) �612039, 8. Now I have some questions about working. please tell me about all his jobs that lasted three months or ^longer since the first elate (STUDY RESPONDENT) stopped going to school lull time. Count changes of jobs for the Same employer as separate jobs. Do npt include jobs in the military. ' First Job Ha. Second Job In what month and year 9a. In what month and year did he start his did he Start his f i r s t job that lasted next job that lasted three months or longer? three months or longer? ; YEAR •MON'fH I I 1 I -l I TTT 8b. Third Job What was the name of J|i8 employer? MONTH 10s. In what month and year did he start his next job that lasted three months or longer? VEAR 1 1 MONTH ( ) ( y (. ) < T Be. Was the job foiltime or part-time? 9b. What was the name of his emp_lojrer? 10b. What was the name of his employer? 9c, Was the job fulltime or part-time? TRECQRn i N s. R ., B . - PC i T 10c. Was the job fulltime or part-time? TRECORD IN. S.R.B, - p<j 1 " 1 -1 "-2 $d. What kind of business was that — what did they make or do there? 9d. What kind of business was that -- what did they make or do there? TRECORD TNS.R. i\ . ~-"p"(3~f T Full time. i( r ( -1 Part tin*...... ~2 Full tirae..( Part time..,? YEAR I i 1 i f i i"(•' }i I .. 5 1-1 ( )T ) i, I 1 -' I 1 l-l . . . 1 „ J Full time..( r( -1 lOd. What kind of business was that — what did they make or do there? (.( .. (( Be. Whflt did he actually dp on the job — w h a t were Some of his main ( . ( . 9e. What did he actually do on the job — what were some of his main duties? lOe. What did he actually do on the job — what were some of his main duties? 9f. 10f. Please look at this card and .tell me the number which best describes the kind of industry he worked in. IRECORD IN S.R.B. - p'd i i IRECORD IN S.IUB. - PC i I . ', . ,/ IHAND RESPONDENT CARD "Blf IHAND RESp'oNDEtif .CA'RD "ts"l Please look 4't this card and tell me the number which best describes the kind of industry he worked in. I (WRITK tH NUMBER) In whdt month find year did this job end? HdHTtl _ T r "T i Current (WRITE IN NUMBER) 9g. (SKIf TO 8h. What was the main reason he stopped working at that job? 1 I ( (WRITE I N NUMBER) | [ | I 3 ( ) In what month and year did this J6b end? YKAR T "T I (T ( ) ( ) ( ) Please look at this card and tell me the number which best describes the kind of industry he worked in. 1 1 MONTH 1 1 '11 ' l-r YEAR I' "1 1 ,_J ( ) ( ) {• ) ( ) 1 I ) 1 I . I lOg. In what month and year did this job end? r MONTH 1 i i i 1 YEAR l-l i i 1 1 Current (SKIP TO job..(_T_-l Q.l/t) Current jOb..( 9h. What was Uhe main reason he stopped working at that job? lOh. What was the main reason he stopped working . at that job? (SKIP TO ( -1 Q.14) ( ( < (ASK Q.9af (ASK Q,}6a!> 101 < ( ( (ASK Q.lla) �CARD Fifth Job Sixth Job 12a. In what month and year did he start his next job that lasted three months or longer? 13a. In what month and year did he start his next job that lasted three months or longer? Fourth Job lla. In what month and year did he start his next job that lasted three months or longer? MONTI! I I 1 I I - ( )( ) 812039 YEAR I I I I I 1 I ( )( ) lib. What was the name of his employer? 12b. What was the name of his employer? 1 RECORD IN S.R.B, - PC 1 | lie. Was the job fulltime or part-time? Full time..( MONTH YEAR 1 II 1 1 I l - l 1 1 ( )( ) ( ) ( ) ( 12c. Was the job fulltime or part-time? lid. What kind of business was that — what (do/ did) they make or do there? MONTH YEAR 1 I I 1 1 l l I l l ( )( ) ( )( ) 13b. What was the name of his employer? (RECORD IN S.R.B. - PG 1 1 Full time..( ( Part time...... -1 1 I (RECORD IN S.R.B. - PG 1 | 13c. Was the job fulltime or part-time? Full time.,( -1 -2 12d. What kind of business was that — what (do/ did) they make or do there? ( -1 13d. What kind of business was that — what (do/ did) they make or do there? ( ( ( ( ( ( ( ( ( ( ( ( 12e. What did he actually do on the job — what were some of his main duties? 13e. What did he actually do on the job — what were some of his main duties? IHAND RESPONDENT CART) "B"l IRECORD IN S.R.B. - PC i 1 IRECORD IN S.R.B. - PC i 1 IHAND RESPONDENT CARD "B"| IHAND RESPONDENT CARD "B'M llf. Please look at this card and tell me the number which best describes the kind of industry he worked in. m. Please look at this card and tell me Che number which best describes the kind of industry he worked in. 13f. Please look at this card and tell me the number which best. describes the kind of industry he worked in. (WRITE IN NUMBER) (WRITE IN NUMBER) (WRITE IN NUMBER) lie. What did he actually do on the job — what were some of his main duties? IRECORD IN S.R.B. - PG i 1 1 1 ( i 1 1 1 ) () llg. In what month and year did this job end? MONTH I 1 ( | I I 1 ).( ) 12g. In what month and year did this job end? YEAR I I I 1 1 1 l-l 1 1 ) ( .) ( ) ( ) I I ( MONTH 1 1 ( ( 1 1 I 1 ^ (1 13g. In what month and year did this job end? YEAR 1 I I 1 1 1 l-l 1 1 )( ) ( ) ( ) 1 I 1 1 MONTH YEAR 1 I I 1 1 1 l-l 1 1 ( ) ( ) ( )( ) Current (SKIP TO job..( ( -1 Q.14) Current job..( llh. What was the main reason he stopped working at that job? 12h. What was the main reason he stopped working at that job? 13h. What was the main reason he stopped working at that job? ( ( ( ( ( ( ( ( ( ( (RECORD ADDITIONAL JOBS IN S.R.B. - PG 15 AND 16) ( -1 (SKIP TO Q.14) ( ( (ASK Q.12a) (ASK Q.13a) Current job..( ( -1 (SKIP TO Q.14) �812039 .CARD 14. Now 1 am going to ask you about (STUDY RESPONDENT'S) years in the military. a. In what month and year did he first enter the Armed Forces? MONTH t. YEAR i t t i. i i i c * i ( • )1-1 (• j i r > i . h. MONTH I I I 1 l-l 1 1 YEAR 1, ,.1, -1 Coast Guard. , . _ Air Force. '( J 1 , H Discharged/ separated. ( ; ( 1 , 1 ( ) ( 1 ( •")•( )'-• Following his separation or discharge in. (DATE IN "d"), did he reciter the Armed Forces? L. 1 1 l-l )( ) YEAR 1 1 (MILITARY) ' ' ••)' T T" ( Whaf branch of the military was that? -1 -2 rS Coast Guard, . . '^ -5 { -1 (ASK m. Was he discharged or separated from the (BRANCH OF SERVICE)? Discharged/ separated. (^ -1 (ASK """ i. In what -month and year was he discharged/ separated from the (BRANCH OF MltlTARY)? T 1 MONTH 1 ' 1 1 1 Still in ••*% (SKIP TO (MILITARY).,... -2 (SKIP TO Still in YKAR i "T. ( MONTH 1 1 Air Force. ( ( Navy. . . . . . .... (BRANCH OF se^idfe)? In what month and year was he dischargee!/ separated from the (BRANCH OF MILITARY)? MONTH _ 1 1 "1 h. Was he discharged or separated from the Discharged/ separated. ( ( Jl (ASK Q.14d) S t i l l in (MILITARY)..... -? (SKIP TO ~ Q.15) r~ r™T "i ' Coast Guard... -5 c. Was he discharged or separated from the (BRANClll OF SERVICE)? d. ( k. In what month and year did he next enter the Armed Forges? 1 1 What branch of the. mili- g. What branch dj the mi}itary wa's that? tary was that? A i r Force. ( ( e. In what month and year did he next enter the Armed Forces? J 1 l-l n. YEAR 1 ' .' 1 ..1 1 j. Following his separation or discharge in (DATE IN "i"), did he teenier tlie Armed Forces? Q.15) In what month and year was he discharged/ separated from the (BRANCH OF MILITARY)? 1 1 MONTH 1 1 1 1 l-l YEAR : l 1 I ' 1 o. Following his separation or discharge in (DATE if) "n!'), did he reenter the Armed Forces? Yes..( (_ -1 (ASK q.Hif) Ye 9..^ (: _-l (ASK Q. 14k) Ves..( (^ -1 ~~'"':~1''; No. . . . . . '. -2 (SKIP TO Q.15) . .Ti . "-2 (SKIP TO Q.J5) No. . No -2 (RECORD ADDITIONAL SERVICE PERIODS IN S.R.B. PG 17) (SKIP TO Q.15) �CARD 812,039 15. I would like to ask you the names of all the countries (STUDY RESPONDENT) wag stationed in while on active duty in the Armed Forces. a. Starting with induction, in what Country was he first stationed while on active duty? Include temporary duties of greater than 90 days. g. What was the next count r> ro. What was the next count, thit he was stationed that he was stationed in for more than 90 days in for more than 90 day: while on active duty? while on active duty? ..,('(. (RECORD COUNTRY HERE AND IN S.R.B. PC 2 AND CONTINUE) . . .: ( ( (RECORD COUNTRY HERE AND IN S.R.B. PC 2 AND CONTINUE) nECORD COUNW HERE" "AND' IN" ' S.R.B. PG 2 AND CONTINUE) No others.( No others. ( ( -1 (SKIP TO Q.17) b. h. ( ( In what month and year did he begin and end active duty in (COUNTRY)? In what month and year did he begin and end active duty in (COUNTRY)? BEGIN MONTH ' YEAR 1 1 I I 1 I l l I l ( )( ) ( )( END MONTH YEAR I I I I 1 I I , l-l . 1 BEGIN MONTH I I I I 1 I l-l YEAR 1 1 1 1 ( )( ) c >< ) END MONTH YEAR 1 1 I I I I .) 1 hi 1 1 ( ) ( ) ( ) ( ( -1 (SKIP TO Q.17) Current. . ( . ( In what month and year did he begin and end active duty in (COUNTRY)? BEGIN MONTI! YEAR 1 1 1 1 1 1 1 1 l-l 1 1 ( ) ( )"' ( ) ( ) " END MONTH YEAR 1 I I I 1 1 I I l - l 1 1 ••(')<) ( )( ) 1 l ) 1 1 ( )( > ( . V ( ) J Current . . . ( ( ^ -1 n. Current. . . ( ( -1 -1 c. What specific job assign- i. What specific job assign- o. What specific job assigi ments did he have ments did he have ments did he have in (COUNTRY)? Can you in (COUNTRY)? Can you in (COUNTRY)? Can you give me the AFSC? give me the AFSC? give me the AFSC? i. < . < .:" V. ' 2. .(. < ... 2.. 3. (( 3- d. . H " (( (,( 1. ( :(.,L 2. ,( ( ( ( . 3. '_ Did his duties in j. Did his duties in p. (COUNTRY) include flying? (COUNTRY) include flying? Yes.( No ( Yes.( ( No...... -1 -2 e. How many flight hours did he log while in (COUNTRY)? 1 1 1 1 1 1 I I Hours ( )(•')( ) Other (SPECIFY) .( K. How many flight hours did he log while in (COUNTRY)? 1 1 1 1 1 1 Hours ( >(")'<'> Other (SPECIFY) f. Wliat specific letter and numerical designat ion( s) did each aircraft have? 1 I -1 -2 1 I 1 1 1 | Hours ( ) ( M .) Other (SPECIFY) .( ( -1 .( ( -1 j. d q. How many flight hours did he log while in (COUNTRY)? 1 1 ( -1 Did his duties in (COUNTRY) include flyini Yes.( ( No .. -1 -2 -,.(....( What specific letter and r. What specific letter am numerical designationCs numerical designations) did each aircraft have? did each aircraft have? 1. (( 1. (( 1. ( ( 2. (( 2. ( ( 2. ( ( 3. (( 3. ( ( 3. ( ( (( <•-. ( ( 4. 4. (ASK Q.15g) (ASK Q.lSm) 210 (( (ASK Q.lba) �ft 13.6.3 9.'. -CARP, Question id ' Country a. sixth Country Fifth Cp.Mji$.t!3! What was the nexf country g. that ho was s t a t i o n e d in far mote than 90 days w h i l e On active duty? What was the neitt q o u n f r y m. that h« was sfaliiined •' in ior more tyft .90 d«y« while on actiyl: avty? p/ What was the next country that ne w>s stationed in for more than 90 day 6 whije pn active duty? , ,.,...,;. . HERS ;: AtfO •<;..' < IN (RECORb COjMTKf ' • •-Ar - " i ' (RECORD doWfRY 'Htttp 'AND fif), ' S.R.B. PG 2 AND CONTINUE) S.R.B. PC 2 AND cBrfttNtlB) ; (RECORD COUNTRY HERE AND IN S.R.B., PC 2 AMD CONTtNUE) No other*. ( Ho others. ( No others. ( ( -1 (SKIP TO Q.17) b. In what Hionth and year did he begin and end Active d u t y in (COUNTRY)? BEQIN 4i MONtfl , , YEAR h. ( j ( 1 1 1 ( ) .J( ' H( END ' YEAR 1 .•••".'} MONTH 1 1 ( ) : MONTH i. i. *) ( ^ ..'. ' • ( n. ' PAR i I • • ' • • • ! • • . - . ?• i • Current . . .( 2. , . ..':•. ( ( • ( . ' YEAR i ; :;i r i n I1 I i hi i 1 .tv) ; ( ) ( )-<' ) END YE.AR : r .; i ' 'i 1 >" • ()V( , ^^v;**! 1 ; i • i h ) L ( Current, ..( t l 1 H ( ) ,-1 What specific )bb assign- p. What specific job as«igh; ments di4 he have Bients did he <1a^e , in (COUNTRY)? Can you in (COtJHTRY)l, iC4n you giv« Me the AFSC? give we the Ai^Sci '.• What epecifie job assign- i. ments did he have in (COUNTRY)? (Jfliv yOu give me She AFS<5? . ,: ;...>,( .:( . . ( PCIN MONTH : ,«EAR i'(•')•.(' '> ~r ( 1"",T 1 ' -1 (SKIP TO MONTH END 1 : In What month and year di,d he begin and end ait* tiv«s duty in (COUNTRY)? BEGlil MONTH i. 1*.: '..,.„., 2.."...:' . ..,.,,..,(,.. • '., i. ,.-.-• . . ; . ' . . . . < c.., ,,,.J.'.r : 2. ... .....:.•:• , (.. I' .( ( 3.-' : . . . ' ,., ''../,,,,.,...;'( i - . , 3 . ' . . .. ; ' . . . . - . ( ( , .1 Did his d u t i e s in j. Did his duties in p. Did. his duties in (COUNTRY) include flying? (COUNTRY) includij flying? (cOONTrtV) include flying? I ' ' . 3. d. Yes.( ( u < ,j-l ; No. . . . J . ' -2 Ye*.( ( No.../,. e. k. Howrosilyf l i g h t hours did he! lofc w h i l e in (COUNTRY)? 1 1 1 " 1 ' 1 i How many f l i g h t hours did he log w h i l e in (COUNTRY)? 1 ..., ."' :"(,.:.(.} q. • . . . "( (AS'K Q,ilg) ' ( -1 How many f l i g h t hours did he lot; w h i l e in (.COUNTRY)? '1 .T • • ' • ! " I l.» 1 L ,,-.!- Hours ( ) ( ) ( ) Ofh«r (SPECIFY) . .. :..• . ,. .,.--(„.. x 'TV • . ..,. . :V . . . . ....'...•(. ,( -i What spe'iifie letref arid . r. numericat de8ignatioti(s) " ' . ..did ea'ch'''fli.rc'r*ft.. (ia.v^l' ' L. : •• " . • . " • • ' • r,:;:.,, .'...-';:. .' . • • . ,' '• '.' •.• • '• ,\ .' . ''. \. :. ,' •' : '!":.• '. *} ' ' ' .. .•.;•::,;•;.:.;;:•. .-*.••:$••." ...•; ...... (.,f ... 2. .',,..:'• ••/'; :;;.' r-..( .•(• : : .. ,,',.:.'( ',(,' .': ?. ' ./... •" •:'•;., :,:.' ':^ ( 2. 3..... 'r 'i: .'.'.' ' Yes.( • • • ( ' . Other (SPECIFY) ,. ... ,., .,..(, ,(,.:.,-! "-"""• '' '"" ?' ..<'?. f. What s p e c i f i c l e t t e r and fiumcrieal designat ion(e) did each a i r c r a f t ; have!? l. -1 -2 , I 1 t . •! Hours . ,(. - y (••".y-jvy .. ;. 1 . , ,4 •>. 1 Hours ( ')(.)'( ) Other (SPtfClJ'Y) 4. ' Q.17) c > 1 { , ) M (-^-4-( .:'y j i Current * . . ( J _,via rl c. . - ''I.". ) . ( ( -i (skip TP --^Q,17j In what month «ftd year did he begin jj'jiii end a,?* tive duty in (cb'BNTRY)1 : r i ." i ' T • [ . .. i i ...1- .1 . t. ' ( : A . . .'. •(.'(' (AS'tf Q.lta) What S p e c i f i c Ht^er and nuitierical designation(s) rfid!(Bach a i r c r a f t have? i.' ;:v.;.'0 •«.'::'.: '-. •;:•."(• ( '. 2. . : ::/-' . :: • . ( ( 3;.'.:"' V.; ' • • ' • : • : ' • . . • • • . ' ,•'( ( A. .'.'.'. ( ( ( RECOR& ADDt t iONAL COU^RI ES < IN S.R.B. PC 18 ANP 19) �CARD 812039 Now I would like to ask you about about (STUDY RESPONDENT'S) marital history. 17. Was he ever legally married? (ASK Q.18) 22) ... . - . . . , . . . IB. , .* • • • . - • . . • • . . , • . •• ^ '. v . '; How many times was he legally married? : . ,: .. (WRITE IN NUMBER) 1 1 ( 1 ' 1 | I times ) ( ) FIRST/ONLY, MARRIAGE 19a. In w h a t month and year did he get m a r r i e d (the f i r s t time)? 1 MONTH 2da. tn what month and year did he get married ( t h e second tilde? YEAK " 1 ""IT" I "T 19b. What (is/was) the current f u l l name ,of that w i f e (RECORD IN S.R.B. PG. 2 1 19c. THIRD MARRIAGE What was her f u l l maiden name? ]RECORD IN S.R.B, PG 2 | 1 1 MONTH 1 .1.1 1 '. H ( ) ( ) ( 21a. In what month and y e a r did he get m a r r i e d (the t h i r d time? YEAR 1 1 1 1 ) ( ) 1 1 MONTH 1 I I 1 l-l ( ) ( ) ( YEAR 1 1 1 1 ) ( ) 20b- What (is/was) the cur21b. What (is/was) the current f u l l name of rent full name of .•V : .^vths ; t,ijife;. ••,>. •• : '-'"' .•''•'' IfcECOKD .XN S.R.B. .PG,2: 1 (RECORD IN S.R.B. P(» 2 ( 20o . Wiat w4f. hftf full . ':•"•': ^.^'waiden';n«meY '•.'•'.•• .••".". •• 21c. (RECORD m st tR.fc. . PG 2 i What was her full maiden name? 1 RECORD IN S.R.B. PG 2 | 20d. During this marriage, 19d. During this Marriage, 21d. During this marriage, how many times was he how many times was he how many times was he living apart from his living apart from his living apart from hit w i f e (you) for more than wife (you) for more than wife (you) for mote than three months? three months? three months? 1 1 1 1 ( Never.. ( • 1 J Times ) ( -I I ,1 . 1 | Times I 1 I I 1 I Times ) ( I 1 (SKIP TO Q.19f) Never. .( ( '....-1. (SKIP TO / Q.20f) ' Never.. ( ( -1 (SKIP TO . Q^if) 19e. How many months did they 20e. How many months did they 21e. How many months did they (you) live apart the (you) live apart the (you) live apart the ( f i r s t / n e x t ) time? ( f i r s t / n e x t ) time? ( f i r s t / n e x t ) time? 1st 2nd '1 I ' 1 J ( ) ( I I I 1 Ath ( I 1 Months ) ( ) I ] ( I | 1st ) I I ( 3rd Months I 1 Months ) ( i ^ i I 3rd ( Months ( 2nd | I ' . ( . ), ( ) j ) I 1 Months Ath i 2nd ) i | Months 3rd j ( Months Ath 1 1 1 | I J ( 6th — 1 I } ( I | I | Months 5th J ) I I Months TTT~T (CO TO Q , 1 9 f ) i " " • " I """"."I | I Months ( ). ( ') i .: i 6th | T Months ' . • • ( • ) ( ) (GO TO Q . 2 0 I ) 5th 6th 1 I Months ) I 1 Months 'i j j Months ) ( .) I 1 ( ( • > ( ) . 5th ) ( I I | | . ( ' • ) ( ) _ 1 1 ( ...' 1 Months I | ) ( 1 I 1st ) ( I | Months ) i 1 ) I :1 | ()•(;>• Mpnths 1 1 I I I I Months (GO TO Q . 2 1 f ) �CARD FIRST/ONLY MARRIAGE TfF ONLY MARRlAGEl 19f. At the time he (died/ became incapacitated) was he divorced , widowed, separated, or w,ls lit married and living, with his wife? Uvjiif With (SKIP TO vifp.'-'^jL--1 Divorced. ... >i °-'22) -2 (SKIP TO rM!|>nratcd....~-3 " Q.l«h) 812039 SECOND MARRIAGE THIRD MARRIAGE IIP LAST MARRIAGE! 20f. At the time he (died/ became incapacitated) was he divorced, widowed, separated, or was he married and living with his wife? llF LAST MARRIAdEl 21f. At the time he (died/ became incapacitated) was he divorced, widowed, separated, or was he married and living with his wife? Living with wife...( _j_ -1 Living with (SKIP TO wife...( ( -1 Q.22) (SKIP TO Q.22) Divorced ... -2)(SK1P TO Separated -3| Q.20h) Divorced -2/(SKlP TO Separated....~-3f Q.21h) ~— (RECORD IN S.R.B.; PG 2 1 (Kia'.Oiil) IN S.R.B. PC 2 1 IIF OTHER MARRIAGES! [IF OTHER MARRIAGES I 19g. How did that marriage end --• was he divorced or was tie widowed? 20g. How did that marriage end -- was he divorced or was he widowed? Divorced( ( -lX(ASK Q.19h) Pivorced( Widowed Widowed . . -2J J [RECORb IN S.R.B. PG 2 1 |IF OTHER MARRIAGES I 21g. How did that marriage en(J -7 w*§ he divorced or was he widowed? ( -l\(ASK Q.20h) Divorced( -2j ( -1\(ASK Q.21h) [RECORD IN S.R.B. PG 2 1 IRECORD IN S.R.B. PG 2 1 IRECORD IN S.R.B. PC 2 1 19h. In what month and year was he (divorced/ widowed /separated)? 20h. In what month and year was lie (divorced/ widowed/ separated)? 21h. In what month and year was he (divorced/ widowed/separated)? I 1 MONTH I 1 I l-l I YEAR I 1 MONTH 1 1 ( ) < ) (: ) ( ) (IK A SKCOND MARRIAGE GO TO Q.20a) YEAR I I I I l-l 1 1 ( )( ) ( )( ) (IF A THIRD MARRIAGE GO TO Q.21a) 1 1 MONTH ' YEAR 1 1 I I I I 1 1 l-l 1 1 . ' ( ) ( ) ( )( ) (RECORD OTHER MARRIAGES IN S.R.B. PG 20 AND 21) �CARD 812039 22. How many children (has/did) (STUDY RESPONDENT) (had/have) — that is, of how many children was he the natural father? Please include all children, both those who are living and those who may no longer be living. i i r I I I children (WRITE IN NUMBER) Ho children. -1 (ASK Q.23) (SKIP TO Q.27a) 23. Starting with the oldest child, what is the first and last name of the child as it appears on the birth certificate? RECORD FIRST AND LAST NAMES OF ALL CHILDREN IN S.R.B. - PAGE 3-4. WRITE IN THE FIRST NAME ONLY AT THE TOP OF THE APPROPRIATE COLUMN,(S). FIRST CHILD THIRD CHILD SECOND CHILD NAME: NAME: NAME: . How old ie (CHILD) now? 25a. How old is (CHILD) now? 1 Child died..( ( -1 ( -1 (_ 24c. How much did (CHILD) weigh at birth? OUNCES POUNDS 1 1 1 Age Child died..( 24b. (Is/Was) (CHILD) male or female? Male Female 1 ( ( -1 ( -1 24d. What is j(CHILD)'s birth- 25d. What is (CHILD) 's birth date? date? MONTH DAY YEAR i—r T r~i—ri-ir~i—n i i i i „ rT~rri-ir~rr~y r 7T~fi MONTH DAY «• 1 (_ ~-2 POUNDS OUNCES Don't know. . . ( ( 26c. How much did (CHILD) weigh at birth? 1 Don't know...( ( -1 Male Female 25c. How much did (CHILD) weigh at birth? POUNDS ! 1 AKe 26b. (Is/Was) (CHILD) male or female? -1 ~-2 (_ 1 1 Child died..( ( -1 25b. (Is/Was) (CHILD) male or female? Male Female 26a. How old is (CHILD) now? YEAR lAl.SO RECORD IN S.R.B.-PG 3| OUNCES 1 Don't know. . . ( ( -1 26d. What is (CHILD) 's birthdate? MONTH DAY YEAR I ALSO RECORD IN S.R.B.-PG'TT 24e. Was the child premature, 25e. Vi... the child premature, 26e. Was the child premature, full term, or overdue? full term, or overdue? full term, or overdue? Premature. ( Full term Overdue Not sure (_ _ (CO TO Q.24f) _ ~ Premature.( Full term Overdue Not sure ( (GO TO Q.25f) -1 -2 -3 -A Premature.( ( -1 Full term -2 Overdue ____~3 Not sure...,.. -4 (GO TO Q.26£) �.812039 F|RST CHILD SECOND CHILD 24f. Where are (CHILD) 's birth registration records located? In what city and state is that? _ ' I RECORD IN S.R.B. p'c""3"t 25f. Where are (CHILD) '« birth registration records located? In what city and state is 24g. Where are (CHILD) 's current medical records located? In whnt city and stale is .that? TRECO>D IN a.fc.b. PC 3 T ~ 25g . Where are (CHILD) 's current medical records located? In what city and state is that? TRECORD IN S.R.B; PG 3 1 26g. Where are (CHILD) 's current medical records located? In what city and state is that? [feECORD IN S.R.B. PTT '3" I 24h. What was (CHILD) 's rfiother '.s ful 1 name? I RECORD IN S . PC 3 1 25h . Wliat was (CHILD) 's mother's full name? (RECORD IN S.R.B. PC 3 1 26h. What was (CHILD) 's mother's tul 1 name? 2Ai. How old was the mother when (CHILD) was born? 25i . How old was the mother when (CHILD) was born? 26i. How old was the mother when (CHILD) was born? IRECORD IN S.R.B. PC 3 1 1 1 T"rr~r I 1 I 1 26f. Where are (CHILD) 's birth registration records located? In what city and state is .that?_ ' . IRgCORD IN S.R.B. PGTT IRECQRD IN S.R.B. PC 3 1 1 Age Age 2/ij. Did (CHILD) have any b i r t h defects? 25j . Did (CHILD) have any bin.h defects? 26j. Did (CHILD) have any birth defects? Yes.C, ( Yes _-l (ASK Q.25k) Yes.( .j -1 (ASK Q,26k) No. _-2 (SKIP TO Q.25L) No No -1 (ASK Q . 2 4 k ) __-2 (SKIP TO Q.24L) 2Ak. What kind of birth defects did (s)he have? Any others? 25k . What kind of birth defeet s did (s)he have? Any others? __-2 (SKIP TO Q.26L) 26k. What kind of birth de^ fects did (s)he have? Any others? ( ( ( "( 24L, Was (CHILD) ever diagnosed as having cancer? 25L . Was (CHILD) ever diagnosed as having cancer? 26L. Was (CHILD) ever diagnosed as having cancer? Yes.( Yes ( Yes.( No 26m. ( -1 (ASK Q.24m) ( -1 (ASK Q.25m) -2 (SKIP TO Q.24o) No. _-2 (SKIP TO Q.25o) No ( -1 (ASK Q.26m) _-2 (SKIP TO Q.26o) In wlmt month and year 25m was the diagnosis made? In what month and year 26m. was the diagnosis made? In what month and year was the diagnosis made? MONTH MON Til MONTH YEAR 1 T I I YEAR 1 1 1 24n. W)iat k i n d of cancer was diagnosed? 25n. What kind of cancer was diagnosed? 1 1 (GO TO Q.24o) -1 Not sure. .( ( (CO TO Q.25o) -1 II l-l YEAR 1 1 26n. What kind of cancer was diagnosed? . .. Not stire..( ( 1 1 ''-r";''.. /...If; < Not sure.,( J (GO TO Q.Z6o) -I �CARD FIRST CHI 1.1) SECOND CHILD 2Ao. (DofS/Did)(CHII.n) have a diagnosed learning disability? Yes.( 812039 THIRD CHILD s/Did)(CinLD) have nosed learning disity? ( -1 (ASK Q.24p) -1 (ASK Q.2Sp) No...... -2 (SKIP TO Q.24q) No. 26o. (Does/DidXCHILD) have diagnosed learning disability? Yes.( _-2 (SKIP TO Q.25q) No kind of learning b i l i t y (does/did) «• have? ( ( Uliat k i n d o f l e a r n i n g d i s a b i l i t y (does/did) (s)h<> have? ( -1 (ASK Q.26p) -2 (SKIP TO Q.26q 26p. What kind of learning disability (docs/did) (s)he have? ( ( ( ( 24q. (Do<.'s/Did)(CHJU>) have any physical, mental, or motor impairments? Yes.( s/nid)(CHILD) have physical, mental, o r impairments? (__-! (ASK Q.2Ar) No ( ( 26q. (Does/DidHCHILD) have any physical, mental, motor impairments? Yee.C -1 (ASK Q.25r) _-2 (SKIP TO Q.25s) No __-2 (SKIP TO Q.24s) No. 2Ar. What kind of impairment (does/did) (s)he have? kind of impairment s/did) (s)he have? ( -1 (ASK Q.26r) ' -2 (SKIP TO Q.26s 26r. What kind of impairment (does/did) (s)he have? ( ( ( ( OTHERWISE: 24s. E: DAY CONTINUE SKIP TO NEXT CHILD IF CHILD IS DEAD:. OTHERWISE: tat date did LD) die? On what date did (CHILD) die? MONTH ( ( IS MAD: ONTINUE SKIP TO NI'.XT CHILD i—r~T r~~i ( ( YEAR 1 DAY 1 1 1 -1 ! 26s. l-l MONTH YEAR 1 I 1 I ( ) ( )( ) ( ) ( ) ( ) ( 24t. What was the cause at death? was the cause of i? death? CONTINUE SKIP TO NEXT CHILD On what date did (CHILD) die? MONTH 1 I I 1 l-l DAY 1 1 I I l-l YEAR 1 1 )( ) ( ) ( )( ) ( 26t. What was the cause of death? ( ( ( ( Where is (CHILD)'s death registered? In what city and state IE that? _ JJL S. R. B PC 3 J v (GO TO NEXT CHILD Q.25a) ( ( ( ( is (CHILD) 'B registered? In city and state is 26u. Where is (CHILD) 's death registered? In what city and state is that? that? S.R.B. PG 3 1 NEXT CHILD 26a) [RECORD IN S.R.B. PC 3 1 (RECORD ADDITIONAL CHILDREN IN S.R.B. - PG 22-30) �CARD 27a. 812039 Now let's talk about (STUDY SUBJECT'S) health. Did (STUDY SUBJECT) ever have pneumonia? Yes.( No 27b. ( , -1 _-2 (ASK Q.27b) (SKIP TO Q.29a) How many times did he have pneumonia? (WRITE IN NUMBER) times I ( )< T First Time 28a. Second Time- During what months and years did he have pneumonia (the first time)? 28f. During what months and years did he have pneumonia (the second 28k. 1 RECORD IN S.R.B. PC 5 1 IF IlliKORE 1961, SKIP TO Q.28f , RECORD IN S.R.B. PG 5 1 IF 'BEFORE 1961, SKIP TO Q.28k. 28g. What is the full name of the doctor who mode the diagnosis or the medical facility where the diagnosis was made? RECORD IN S.R.B. PG 5 1 IF BEFORE 1961, SKIP TO Q.29a. What is the full name 28L. of the doctor who made the diagnosis or the medical facility where the diagnosis was made? 28h. What is the full name of the doctor who made the diagnosis or the medical facility where the diagnosis was made? I RECORD IN S.R.B. PG S 1 (RECORD IN S.R.B. PC 5 1 8c. What prescribed medicine did he take for the pneumonia he had that time? During what months and years did he have pneumonia (the third time)? r i rue ) ? RECORD IN S.R.B. PC 5 1 !8b. Third Time 28m. What prescribed medicine did he take for . the pneumonia he had that time? What prescribed medicine did he Cake for the pneumonia he had that time? 1. t / ( ( 8d. t?s.( c, Be, Was he hospitalized for the pneumonia he had that time? ( -1 (ASK Q.28«) -2 (SKIP TO Q 28f) What was the full name of that hospital? RECORD IN S.R.B. pcf'5 ] ( ( 1. (. ( 2. ( ( 2. (( 3. (( 3. (( 28i. 28n. Was he hospitalized for the pneumonia he had that time? Yes.( ( -1 (ASK Q.28i) Yes.( No ... -2 (SKIP TO Q.28k) No, Was he hospitalized for the pneumonia he had that time? ( -1 (ASK Q.28o) -2 (SKIP TO Q.77a IN S.R.B. PG 31) 28 j . What was the full name 2 80. What was the full name of that hospital? of that hospital? 1 RECORD IN SiR.B^lp 5"! 1 RECORD IN S.R.B. PG 5 1 (RECORD, ADJD.ITION.AI. PERIODS IN S.R,B. PAGE 31) 1 -' 217 �CARD 29a. 812039 Did (STUDY RESPONDENT) ever have cancer? Yes..(_J (ASK Q.29b) No 29b. -1 -2 (SKIP TO Q. 30) In which parts of his body was cancer located? LIST EACH BODY PART BELOW. FOR ADDITIONAL PARTS. IF MORE THAN THREE BODY PARTS, USE S.R.B. - PAGE 32 Part 1 Part 3 Part 2 29c. In what month and year was cancer of the (BOD PART) first djagnoBed? I RECORD IN j R. B. P(T .6, I . 29i. In what month and year 29o. In what month and year was cancer of the (BODY was cancer of the (BODY PART) first diagnosed? PART) first diagnosed? I RECORD IN S.R.B. PC 6 I IRKCORD~TN S.R.B. PC 6 I 29d. What is the full name 29j. Wliat is the full name 29p. What is the full name of the doctor or the of the doctor or the of the doctor or the medical facility where medical facility where medical facility where the diagnosis was made7 the diagnosis was made? the diagnosis was made? I RECORD IN S.R.B., PC' f>_"\_ |RECORD~~IN S.R.B. PC 6 I I RECORD IN S._R.B. PC 6 I 29e. What is the full name of the doctor or the medical facility he last consulted about cancer of the (BODY PART)? {RECORD IN Sr.R.B.L PC 6 j 29k. Wliat is the full name of the doctor or the medical facility he last consulted about cancer of the (BODY PART)? 'IRECORD IN S.R.B. PG~ 6 j 29f. 29q. What is the full name of the doctor or the medical facility he last consulted about cancer of the (BODY PART)? I RECORD IN S.R.B. 'PC 6 I 29L. D u r i n g what month and year did he last cons u l t (NAffljJtOM Q.29e)? During what month and 29r. During what month and year did he last conyear did he last consult (NAME FROM Q.29q)? suit (NAME FROM Q.29k)? [RECORD IN S.R.B. PC 6 1 jRECORD IN S.R.B. PC 6 I TRECORD IN'S.R.B. PC (TT PART)? "IMULTIPLE RECORD BELOW I ' What treatments or 29e. What treatments or medicines did he take , medicines did he take for cancer of the (BODY for cancer of the (BODY PART)? FART)? IMUI.TI PLE RECpRD"BEi.OWI TMU'LTI PLE 'RECORD TF.IXJW I Radiation (_ ( -1 Chemotherapy...(~ ( -1 Surgery (_ ( -1 Other (SPECIFY) Radiation......( ( -1 Chemotherapy...( ( -1 Surgery ( ( -1 Other (SPECIFY) 29g. 29m. What treatments or medicines did he take for cancer of the (BODY ( .( 29h. DurinK what month and year did he first receive (EACH TREATMENT CODF.O IN Q.29g) for cancer of the (BODY PART)? MONTH Radiation... .( ( -1 9n. MONTH I I 29t. MONTH YEAR hemoT I T 1 I I j-| \ l-l I I therapy. I ( M ) ( M ) ( )( ) 7~7T~ I MONTH YEAR MONTH YEAR Surgery. . J \ I MONTH _ ~~TT I urgery. 1 I I I YEAR YEAR hemotherapy 1 MONTH I T [_ _H urgery. J ther... i i M I I l-l YEAR ther... (GO TO NEXT BODY PART) 1 YEAR ( )( T r MONTH r~T Other.... I !_ I-J J J < ) ( ) < ) ( )" (CO TO NEXT BODY PART) MONTH | j MONTH *adiaI tion.... I T) < ) ( ) ( ) MONTH YEAR Chemotherapy. -1 During what month and year did he first receive (EACH TREATMENT CODED IN Q.29s) for cancer of the (BODY PART)? YEAR i n r adiation .( ( -1 During what month and year did he first receive (EACH TREATMENT CODED IN Q.29m) for cancer of the (BODY FART)? YEAR Radiation (_ ( -1 Chemotherapy... ( ( -1 ~ Surgery ( ( -1 Other (SPECIFY) YEAR i r I I (GO TO NEXT BODY PART IN S.R.B. PAGE 32) �61203$ (JOT PREVIOUSLY MENTIQNte.P, ASKVj loinDid (STUDY RESPONDENT) ever have leukemia? Yes..( •..(.t . -1 No 30b. (ASK Q.30b) (SKIP TO Q.31a) -2 In what month and year Was his leukemia first diagnosed? I RECORD IN S . R | •, -, PC ?,.]. .., 30c. What is the. full name of the doctor or the medical facility where the diagnosis was made? J RECORD IN.; S,-K».B.-.- -v '$.. 7,. I 30d. b. What treatments or medicines did he take for leukemia? |REC_Q^D JBELpjIj hCPIClNE/TREATMEj4T |..yiRST..RECEIVED MONTH. .,..YEAR T I " ' I l'"\"-' | I I l-l I. I I. ) ( ) I ) ( ) 2. _ 3. ' , ,_ :i ( ..( ( 1 11 { iI.....H......L T i y ( <•>(j MONTH .YEAR. . \T Tl T T I ... I .... ...... . . T J{ ) ( )( ) 30e. During what month and year did he first receive (EACH TREATMEHT OR MtDICINE IN Q.30d>? J.RECORD. ABOVE |. 30f. What is the full name of the doctor or medical facility he last consulted about his leukemia? .1 RECORD IN .S..R..B. - PC .?.. |. 30g. During what month and year did he last consult (NAME IN Q.30O? FREGORD IN S.R.B. - PG 7 �CARD 812039 31a. I would like to ask you some questions about other medical conditions (STUDY RESPONDENT) may have had. 1. Did he ever have diabetes? Yes........( (_ No............._ 2. -1 -2 ("X" BOX ON PAGE 18) Did he ever have thyroid problems? Yes (SPECIFY) ____ . ( _ ( ___ -1 ("X" BOX ON PAGE 18) _ No.....". ~................._ -2 3- Hid he ever have anemia? Yes........( (_ No............._ A. -1 -2 ("X" BOX ON PAGE 18) Did he ever have a heart condition? Yes (SPECIFY) ______ . ( _ ( ___ -1 ("X" BOX ON PAGE 18) No----. 7 ................._ "7 _-2 5. Did he ever have an enlarged liver? Yes........( (_ No............._ 6. Yes........( j_ No......... . . _ .. ("X" BOX ON PAGE 19) -1 -2 ("X" BOX ON PAGE 19) -1 ("X" BOX ON PAGE 19) -2 Did he ever have gall bladder problems? Yes........( (_ No............._ 11. -1 -2 Did he ever have intestinal parasites? N 10. ("X" BOX ON PAGE 19) Did he ever have cirrhosis of the liver? Yes........(__<_ No............._ 9. -1 -2 Did he ever have hepatitis? Yes........( (_ No............._ 8. ("X" BOX ON PAGE 18) Did he ever have jaundice? Yes........( (_ No............._ 1. -1 -2 -1 -2 ("X" BOX ON PAGE 19) Did he ever have any other liver condition? Yes (SPECIFY) ._. _____ . ( (_ -1 ("X" BOX ON PAGE 20) No....... ................._ ' _-2 12. Did he ever have a respiratory condition other than pneumonia? Yes (SPECIFY) ___ . ( (_ No........................_ 13. -1 ("X" j-2 BOX ON PAGE 20) Did he ever have any other major condition? Yea (SPECIFY ALL OTHER CONDITIONS) <"X" BOX ON PAGE 20) No "-2 �CARD DIABETES 1/tSK (J.jl'b THROUGH Q.31e ifos E/ICW BOX "X"ED ON ipp. J£-20 1 ! 1 1 ! 1 I ! 1 I 1 THYROID PROBLEMS 1 ! 1 ! i t 1 ! 31-t?., Wfftm <Jid a doctor first e«lt him than he had eeeiJDITION)? 3i-c. ANEMIA I 1 SI2039 I 1 ! 1 ! 1 A HEART CONDITION 1 1 i 1 1 1 1 1 Whwt is- th« f u l l name of the doctor who made the rfi^gnosis or the medical facility where the diag- a doctor for (CONDITION)?. Jle. 1 1 1 1 [RECORD -IN S . R . B . 1 IPAGE 8 1 IRECORD IN S. R . B . I t IRECORD is S.R.B.I IPAGE 8 ! IRECORD is S.R.B.I . (RECORD IN S.R.B.] IPAGE 8 ! , IRECORD is S.R.B. ! IPAGE 8 1 iRECORD IK S.R.B. ! (PAGE 8 1 iRECORD IK S . R . B . I JRECORD IN S.R.B. I IPAGE 8 ! (RECORD IN S.R.B. I IPAGE 8 ' 1 IRECORD IN S.R.B. i IPAGE 8 1 IRECORD IN S.R.B.I IPAGE s 1 (PAGE 8 I RECORD IN S.R.B.! IPAGE 8 1 IRECORD IN S.R.B.! IPAGE s i [RECORD IN S.R.B. 1 IRECORD IN S.R.B.I IPAGE 8 i IRECORD iw S.R.B. 1 IPAGE 8 I (GO TO NEXT CONDITION "X"ED) (GO TO SEXT CONDITION 'X"ED) (GO TO NEXT COSDITIOW "X"ED) TO NEXT COROTTION "X"ED) ) IPAGE 8 (PAGE 8 IPAGE 8 I j 31 d. Vffifeti1 did he l»»t consult •~j •^i AK ENLARGED LIVER Wha-t is the full name of tlte doer tor o^r ined-ica.!; facility he last eonsfttlted about hi s> teOSDITIOU)? ....«.„. -.,...„ TO NEXT CONDITION "X"ED) IPAGE s 1 IRECORD IN S.R.B. 1 i IPAGE 8 IRECORB IN S.R.B. 1 IPAGE 8 1 IRECORD IN S.R.B. 1 1 �CARD JAUNDICE !ASK Q.31b THROUGH Q.3.1e [FOR EACH BOX "X"ED ON [PP. 18-20 31b. 31c. 1 1 1 1 1 i 1 1 1 I 1 812039 HEPATITIS i CIRRHOSIS OF THE LIVER ! I 1 1 1 1 1 1 1 ! 1 i 1 1 1 1 1 1 1 1 1 1 1 GALL BLADDER PROBLEMS i 1 ! ! When did a doctor first tell him that he had (CONDITION)? (RECORD IN S. R.B.I IPAGE 8 I [RECORD IN S.R.B. 1 IPAGE 9 1 IRECORD IN S.R.B.I i [RECORD IN S.R.B.I IPAGE 9 • 1 IRECORD IN S.R.B.I IPAGE 9 I What is the full name of the doctor who made the diagnosis or the medical facility where the diag- [RECORD IN S.R.B. I I [RECORD IN S. R.B.I IPAGE 9 1 [RECORD IN S.R.B. 1 JPAGE 9 1 IRECORD IN S.R.B.I IPAGE 9 1 IRECORD IN S.R.B. 1 iFAGE 9 1 .RECORD IN S.R.B. I IPAGE 8 1 [RECORD IN S.R.B. 1 IPAGE 9 1 [RECORD IN S.R.B. I IPAGE 9 1 IRECORD IN S.R.B. I IPAGE 9 ! i RECORD IN S.R.B. 1 IPAGE 9 1 [RECORD IN S.R.B. | JPAGE 8 1 (GO TO NEXT CONDITION "X"ED) [RECORD IN S.R.B.i IRECORD IN S.R.B. 1 IPAGE 9 .1 i RECORD IN S.R.B. I [RECORD IN S.R.B.I IPAGE 9 1 (GO TO NEXT CONDITION "X"ED) (GO TO NEXT CONDITION "X"ED) IPAGE e IPAGE 9 31d. When did he last consult a doctor for (CONDITION)?. 31e. INTESTINAL PARASITES What is the full name of the doctor or medical facility he last consulted about his (CONDITION)? IPAGE 9 1 (GO TO NEXT CONDITION "X"ED) IPAGE 9 1 (GO TO NEXT CONDITION "X"ED) ) �CARD [ASK Q.31b THROUGH Q.31e (FOR EACH BOX "X"ED ON IFF, 16-20 31b. 31c, 31d. I I I When did a doctor f i r s t tell him that he had (CONDITION ) ? What is the f u l l name of the doctor who made the diagnosis or the medical f a c i l i t y where the diag- Wben did he last consult a doctor for (CONDITION)?. What is the f u l l name of the doctor or medical facility he last consulted about his (CONDITION)? , 1 (RECORD IN S. R . B . I IPAGE 9 1 ! 1 1 (RECORD IN S. R.B.I IPAGE 9 1 ANY OTHER MAJOR CONDITION 1 1 i i 1 1 1 1 SECOND OTHER MAJOR CONDITION i ! 1 ! I 1 i 1 THIRD OTHER MAJOR CONDITION 1 [ 1 1 IRECORD IN S.R.B.I IPAGE 10 1 IRECORD IN S.R.B.I IPAGE 10 1 IRECORD IN S.R.B. 1 1 IPAGE 10 « IRECORD IN S.R.B. | IPAGE 9 ! IRECORD IN S.R.B. I IRECORB IN S.R.B. [ IPAGE 10 1 iRECORD IN S . R . B . 1 | 1 IRECORD IN S.R.B. 1 IPAGE 10 1 IRECORD IN s. R . B . I 1 IRECORD IN s. R . B . I IPAGE 9 1 [RECORD IN S . R . B . | 1 PAGE 10 1 IRECORD IN S . R . B . 1 IPAGE 10 1 IRECORD IN S.R.B. 1 IPAGE ib 1 (RECORD IN S.R.B. I IPAGE 9 1 IRECORD IK S.R.B. 1 IPAGE 9.. 1 (RECORD IN S . R . B . I (RECORD IN S.R.B. 1 IPAGE 10 1 iPAGE 9 31e. |A RESPIRATORY CONDITION I ! OTHER THAN PNEUMONIA 1 ! I I 1 ANY OTHER LIVER CONDITION 1 1 1 1 812039 (GO TO NEXT CONDITION "X"ED) I PAGE 9 (GO TO NEXT CONDITION "X"ED) IRECORD IN S.R.B.I iPAGE 10 I (GO TO NEXT CONDITION "X"ED) IPAGE 10 IPAGE 10 I (GO TO NEXT CONDITION "X"ED) �CARD 812039 32. Did (STUDY RESPONDENT) ever have acne on hie face? Yes..( ( -1 (ASK Q.33a) No.. 33a. -2 (SKIP TO Q.35a) As far as you know, during what year did he last have acne on his face? I Year J. (ASK Q.33b) T (WRITE IN YEAR) First Period YKAK MONTH r I 33c. Until when did that last? 33j. Think about the third time he had acne on his face — when did it start? MONTH YEAR I I MONTH YEAR —i—n—i— i r 33g. Until when did that last? MONTH YEAR I T T 1 T I H I I (SKIP TO Q.35a) Third Period 33f. Think about the second time he had acne on his 'face -- when did it start? L 1( 1 I -l (_ Second Period 33b. Think about the first time he had acne on his face -- when did it start? MONTH Before 1961.. ( l-l I 33k. Until when did that last? YEAR 1 I MONTH YEAR i i rr i r I I j I I 7 n l-li n r 7 n nnr 33d. Please show me on this diagram where the acne was located (the first time). 33h. Please show me on this diagram where the acne was located. 33L. Please show me on this diagram where the acne was located. THAND RESPONDENT CARD "E"| I MULTIPLE RECORD BELOW I I HAND RESPONDENT CARD "E"T IllAND RESPONDENT CARD "E" | I MULTIPLE RECORD BELOW| [MULTIPLE RECORD BELOW I Temples ."(_ Eyes or eyelids.( ( Ears ( Cheeks ( Nose ( Forehead ( Jaw, Chin, Other( -1 -1 Eyes or eye] ids. ( ( _-l Ears ( T~ -1 Cheeks ( ( Nose ( ( -1 ~-l Jaw, Chin, Other( ( -1 33e. Did he ever have another period of acne on his face? 33i. Did he ever have another period of acne on his face? Yes.( Yes.( No <__-l (ASK Q.33f) N . . 7 " -2 (SKIP TO Q.41a) o.7. Temples .( Eyes or eyelids.( Ears ( Cheeks ( Nose ( Forehead ( Jaw, Chin, Othert ( -1 ( -1 (""" -1 ( -1 ( -1 ( -1 ( -1 33m. Did he ever have another period of acne on his face? ( -1 (ASK Q.33J) Yes.( -2 (SKIP TO Q.34a) No ( -1 -2 IF ANY~nY"ES1' TO TEMPLE ,~fY*ES",EYELlUS7"OR1 IN Q.33d, ABOVE: ASK Q.34a. ALL OTHERS: SKIP TO Q.35a. 3Aa. Did he ever consult a doctor or medical facility about the acne on his (temples/eyes or eyelids/ears)? Yes. No Don't know.. -1 (ASK Q.34b) ~2 KSKIP TO Q.35a) -3J 34b. When did he last consult a doctor about the acne on his (temples/eyes or eyelids/ears)? [RECORD IN _S^R.j^._- PC 34c. What w.-is the name of the doctor or medical J_ facility he consulted at the time? �812039 Did (STUDY RESPONDENT) ever have (READ EACH COLUMN HEADING)? I IF "YES" TO AUY COLUMN HEADING,'ASK Q>35^-hFORTHAT COLUMNt A. ( p£ B,._'_ Patches hie. .stein change color? Easier bruising of the akin than usual? Yes..( No DK Yea..( ( -1 No. .. -2 DK. . . . -!i Yes..( ( -1 No -2 DK ...~3 ( -1 -2 ~~-3 b. On what part of his body did he have (CONDITION)? Any other part? b. On what part of his body did he have (CONDITION)? Any other part? b. On what part of his body d i d he have (CONDITION)? Any other part? •' C. Skin that was extra sensitive or seemed to hurt for no reason? ( ( < ( ( ( < ( ( ( ( ( c. Did he discuss (CONDITION) with a doctor? c. Did he discuss (CONDITION) with a doctor? c. Did he discuss (CONDITION) with a doctor? Yes.( Yes.( Yes.( No. DK ( -1 (ASK Q.35d) .. -2\(CO TO NEXT -3| CONDITION) *" — No DK ( -1 (ASK Q.35d) -2\(GO TO NEXT -3J CONDI TION) No DK ( -1 (ASK Q.35d) -2") (SKIP TO -3J Q.36a) iJ Wliat was the diagnosis? d. -What was the diagnosis? d. What was the diagnosis? ( ( ( ( ( ( ( ( d. ( ( ( ( e. What is the name of the e. What is the name of the e. What is the name of the doctor who made the diagdoctor who made the diag. doctor who ma do the diagnosis or the medical nosis or the medical nosis or the medical facility where the diagfacility where the diagfacility where the diagnosis was made? nosis was made? nosis was made? (RECORD.- IN S.R.B. - PG ill f. During what month and year was the diagnosis made? 1 RECORD IN S.K.B. - PO 111 f. During what month and year was the diagnosis made? IRECORD IN S.R.B. - PG ill f. During what month and year was the diagnosis, made? IRKCORD IN S.R.B. - PG li 1 I RECORD IN S.R.B. - PG 1 1 1 [RECORD IN S..R.B. - PC ill g. What is the name of the doctor or medical facility he last consulted about (CONDITION)? g. What is the name of the doctor or medical facility he last consulted about (CONDITION)? g. What is the name of the doctor or medical facility he last consulted about (CONDITION)? FRE'CORD IN S.R.B. - PC nl h. During what month and year did he last consult (NAME IN Q.35g)? 1-HECOUI) IN S.R.B. - PG 111 ] RECORD IN S.R.B. - PG 111 h. During what month and year did. he last consult (NAME IN Q.35g>? (RECORD IN S.R.B. - PG 111 -"•••: .:••••./• -.•:.; ,•'• 225 [RECORD IN S.R.B, - PG III h. During what month and year did he last consuit (NAME IN Q.35g)? IRECORD IN s j | , - ?& nl ..l �812039 CARD 36a. Aside from injury, (was there ever/has there ever been) a period of time when (STUDY RESPONDENT) had (READ EACH COLUMN HEADING)? IIF "YES" TO ANY COLUMN, HEADING, ASK Q.36b-j FOR THAT COLUMN! B. Persistent tingling sensations in any of his limbs? A. Persistent numbness in any of his limbs? Ycs..( No DK b. ( -1 -2 ' -3 When did he first notice (CONDITION)? MONTH C. Persistent deep burning sensations in any of his limbs? Ye s.. (__(__- 1 No. . . . . . . -2 DK -3 Yes..(. 4 -1 No -2 DK -3 b. YEAR b. When did he first notice (CONDITION)? ii ii • - r i i i i (>()()() c. . MONTH YEAR 1 1 1 1 1 1 I l l - I l l ( ) ( ) ( ) ( ) Which limbs or muscles were affected? (CONDITION)? Any other part? c. Which limbs or muscles were affected? (CONDITION)? Any other part? When did he first notice (CONDITION)? MONTH YEAR 1 1 I I 1 I I I I I ( ) ( ) ( ) ( ) c. Which limbs or muscles were affected? (CONDITION)? Any other part? ( ( e. ( ( ( ( d. ( ( ( ( ( ( d. During what period was the (CONDITION) most intense? FROM MONTH YEAR I I I I 1 1 r H 1 ( )( ) ( ) ( ) TO MONTH YEAR 1 1 1 1 1 1 1 1 l-j 1 1 ( ) ( ) ( ) () Did he see a doctor for (CONDITION)? Yes.( No DK ( -1 ((IF NO OR DK, -2< GO TO NFXT -3 (^CONDITION) During what period was the (CONDITION) most intense? FKOM MONTH YEAR 1 1 1 1( )( )( )( ) TO MONTH YEAR d. During what period was the (CONDITION) most intense? FROM MONTH YEAR 1 1 1 1( )( )( )( ) TO MONTH YEAR 1 1 1 1 1 1 1 1 l-l 1 1 ( } ( )( )( 1 e. Did he see a doctor for (CONDITION)? e. Did he see a doctor for (CONDITION)? Yes.( ( -1 (?IF NO OR DK. No ... -2J GO TO NEXT Yes.( ( -1 | F NO OR DK, U No . . . -2J CO TO NEXT . DK -3\ CONDITION) DK 1 1 I 1 What was the diagnosis? :. 1 1 1 1 ( V( ) ( ) ( ) -3 1 CONDITION) __«- f. M hi ^ What was the diagnosis? :. What was the diagnosis? ( ( ( ( ( ( ( ( ( ( ( ( g. What is the name of the doctor who made the diagnosis or the medical facility where the diagnosis was made? JRECPRD IN S.R.B. - PC 12j [RECORD IN S.R.B. - PC 12] [RECORD IN S.R.B. - PC 12| h. During what month and year wae the diagnosis made? |RECORD IN_ S.R.B. - PC 12J IRECORD IN S.R.B. - PC 12| IRECORD IN S.ft.B. - pc 121 i. What is the name of the doctor or medical facility he last consulted about (CONDITION)? TRECORD IN S.R.B. -~PG 12! IRECORD IN S.R.B. - PG 121 IRECORD IN; S.R.B: - 3' Pur ing what month and_year did he last consult (NAME IN Q.36g)? TRECORD IN S.R.B'. '-'PC 12]' JRECORD i N "s". R^ sT_ - PG^ 1 2! 220 I2 ___ IRECORD IN S.R.B. - PC 121 �36a. Aside Irom injury, (was there ever/has there ever been) a period of time when (STUDY RESPONDENT) had (READ EACH COLUMN HEADING)? TlF "YES" TO ANY COLUMN HEADING, ASK Q^3b-j FOR THAT COl.UMJj1 D. E. P e r s i s t e n t aches and paint in any of his l i m b s ? A reduction in grip strength? YL-S..< ( Yes..( -1 No DK b. • ' -3 [)(: b. When did he f i r s t nulice (CONDITION)? MONTH 1 1 -1 ! ~~~-3 When did he f i r s t notice- (CONDITION)? YEAR M ( No. . . . . . . -2 ....-? 1 I I I l l I l l. 7— TIT- r~rr~7 MONTH YEAR 1 M i l l l - I l l Which limbs or muscles were affected? (CONDITION)? Any other part? Which limbs or muscles were aftncted? (CONDITION)? Any olher part? ( ( During what period was the (CONDITION) most intense? FROM MONTH YEAR 1 1 1 1 I V ! l-l 1 d. During what period was the (CONDITION) most intense? FROM MONTH YEAR 1 1 I 1 T C ") ( 7 ( ) ( 5 I 1 I l-l ) ( TO ) ( TO MONTH 1 1 fJ YEAR 1 IH ) ( 1 ) ( 1 ) 1 ( e. ( I - 1 K 1 F NO ..OR DK , 1 ) ( YEAR 1- , 1 ( ) ( ) ( ) ( ) e. Did lie sec- a doctor for (CONDITION)? Yes . ( Mn MONTH T 1 ) ( 1 1 1 ) Did he see a doctor for (CONDITION)? Yes.( ( -1 No -2 IH' -2 J GO TO NKXT . . . . -3J CONDITION) DK -3 f. What was the d i a g n o s i s ? f. What was the diagnosis? ( ( ( ( g. ( ( ( ( What is the name of the doctor who made the diagnosis or the medical facility where the diagnosis was made? TRECORD IN s.R.B.^- PC nl [RECORD IN S.R.B. - PC 12! h._ During "hat month ami year wa» the dia^nosie madgl I RECORD IN S.R.B. - P C 12 1 I RECORD 1 N S, R ~,"i"f'-* W " '' ' ' " " • -- i. What is the name of the doctor or medical facility he last consulted about (CONDITION)? "> ! j. During what month and year did he lust consult (NAME IN Q.36g)? ___ „ __ -.-._..., __ ,..-.-_____K .r, TRKCORD IN S . K . B . - J7: ijT "i RECORD TN S . R . B . - pc_tiT 227 �CARD 812039 37a. Did (STUDY RESPONDENT) ever smoke cigarettes regularly for a period of ac least onemonth? \ Yes ..( ( -1 (ASK Q.37b) No 37b. -2 (SKIP TO Q.39a) In what month and year did he start smoking cigarettes on a fairly regular basis? MONTH _JEM_. I I ( n r ( ) () 37c. In what month and year did he jast smoke cigarettes on a fairly regular basis? MONTH YEAR T I I l-l I I 37d. Between (START DATE) and (END DATE), for about how many years altogether did (STUDY RESPONDENT) smoke cigarettes, not counting times when he stopped smoking? I ( I I Years )( ) 38. When (STUDY RESPONDENT) was smoking cigarettes on a fairly regular basis, about how many packs per week did he smoke? By "pack" we mean 20 cigarettes. I J I j I |_ packs per ~i n r week 39a. Did (STUDY RESPONDENT) ewer smoke a pipe regularly for a period of at least one month? Yes 39b. ( ( -1 (ASK Q.39b) No -2 (SKIP TO Q.Ala) * In what month and year did he start smoking a pipe on a fairly regular basis? MONTH YEAR i i n i r I I I"!.. I T~rr~r '"i > I(• r 39c. In what month and year did he last smoke a pipe on a fairly regular basis? MONTH YEAR <:n. r (i ( ) 3<ld. Between (START DATE) and (END DATE), for about how many years altogether did (STUDY RESPONDENT) smoke a pipe, not counting times when he stopped smoking? T—i—r Years I I I ( )( ) 40. When (STUDY RESPONDENT) was smoking a pipe on a fairly regular basis in (START DATE), about how many pipefuls per week did he smoke? I I I pipefuls per TTT T week �Ala. Did (STUDY RESPONDENT) ever smoke cigars regularly for a period of at least one month? Yes ( ( -1 (ASK Q.Alb) No Alb. ,,-2 (SKIP TO Q,A3a) In what month and year did he start smoking cigars on a fairly regular basis? MONTH YEAR 1 1 1 1 _ Ale. In what month and year did he last smoke cigars on a fairly regular basis? MONTH YEAR 1 1 11 1- 1 1 1 Aid. Between (START DATE) and (END DATE), for about how many years altogether did (STUDY RESPONDENT) smoke cigars, not counting times when he stopped smoking? 1—T—T I J , J Years ( V (' ) 42. When (STUDY RESPONDENT) was smoking cigars on a fairly regular basis in (START DATE), about how many cigars per week did he smoke? 1 I I I \ 1 cigars per week v ) (r 'IJt^STUDV.'R^spONbfeMf SMOKED' tiftXKgTt^. A' pip|^b£fcjeA|s7Tgg;T A3. In general, when he was smoking did he inhale the smoke? ( ( ,r -1 -2 Yes No AAa. Now let's talk about drinking alcoholic beverages, that is, beer, wine, or hard liquor. Did he ever drink alcoholic beverages on a fairly regular basis? Yes ( ( No ,:..-! .....2 .'.,- (ASK Q.AAb) (SKIP TO Q.46) AAb. When did he start drinking alcoholic beverages on a fairly regular basis? MONTH YEAR 1 T ' ~ I I ... AAc. When did he lajj drink on a fairly regular basis? MONTH YEAR I i T I" I I ( ) ( ) ( M ) A5. When (STUDY RESPONDENT) drank alcoholic beverages on a fairly regular basis in (START DATE), about how many drinks per week did he Usually have?. I ( I I drinks per ) ( T week �_____ CARD 812039 Now I'm going to ask you a feu questions about his recreation and leisure activities. 46. What are some of the hobbies and sports he participated in on a regular basis? Any others? (( 1. . 2,. . 3. „ (. ( ; (( (_ ( ____. ' 5. (( 6. 47. Did he p a r t i c i p a t e three or more times in (READ EACH 1TKM)? ITEM MKNT10NEU IN Q.46 AND DO NOT READ THAT ITEM) (CODE "YES" FOR ANY Yes 1. ( ( . ~1 2. Auto, boat, or motorcycle racing ( ( -1 3. Skydiving ( ( -1 -2 4. Mountain climbing ( ( -1 -2 5. Hang gliding ( ( -•! -2 ( ( ~1 ~2 6. Scuba diving Plane racing or plane acrobatics, not including flight training or any assignments for the Armed Forces.. �CAkD J 812039 TlK STUDY KKSl'QNDENT IS DECEASED, ASKQ.48-52; OTHERWISE CO TO (J.53.J Now T would like to know more about the circumstances surrounding (STUDV RESPONDENT'S) death. 48. What was the official cause of Ins death? H9. In what c i t y and state was (STUDY RESPONDENT) living at the time of his death? £i_,j! __________ , _____________________ , ____ ; ______ : - J -. Kcnte ( ( (IF OUTS IDK U. S . )_CounMrg , r (rj(. 50u. Wns lie in a hospital at the t imo of his death? Yes No.. 50b. (......(-...,,--1 -2 < ASK Q-50b) (SKIP TO Q.51) What was the name of the hospital? iRKCORj).' IN S.R.B. PAGE 33 i 51. What is the name of the primary physician who was responsible for his care at the timi1 of death? SK.B..PAGE S2. Was an autopsy performed? Yes ......... (_J ___ -1 No.............._ -2 S"i. Wu would like your const-lit for the doctors and medical facilities you mentioned .luring this interview lo p r o v i d e (STUDY UliSPONDIvNT'S) medical records to the Air Force H e a l t h Smvuy. This w i l l h e l p us to obi. lin uituu complete and detailed information about rhe h e a l t h services you talked about, Thank you loi participating in the Air Force Health Study! TIMI-: ItriKHVII-.W KIIOEI): ______________ ...... __ __ (um/|.i.,) �LOUIS HARRIS AND ASSOCIATES, INC. 630 Fifth Avenue New York, New York 10111 FOR OFFICE USE ONLY: Case No. Study No. 812039 O.M.B. NUMBER 0701-0033 Approval Expires 11/30/82 Respondent #: CONFIDENTIAL AIR FORCE HEALTH SURVEY SUPPLEMENTAL RECORDING BOOK PROXY 232 �812039 Q.2. Where born: City: State: ; Q.8b-13b, ! 1st job: 2nd job: 3rd job: 4th job: 5th job: 6th job: 7th job: 8th job: 9th job: 10th job: llth job: 12th job: Employers Q.8e -13e Main Puties �812039 Q.16: Countries Served In: 1. 7. 2. 8. 3. 9. A. 10. 5. 11. 6. 12. Q. 19-21 and (.';-(>? Marital History f/K b. Wife's Current Full Name First/ only wife Second wife Third wife Fourth wife Fifth wife Sixth wife Wife's Maiden Name Living W i t h W i f e Or Divorced/ Separated/ Widowed �i and 68- 7 b CHILDREN CHILD Q.29 FIRST NAME i . BI RTHDATE First . First ; : Last THIRD First ' Last FOURTH First First ; Last First La s t . MONTH C/S Place First Place C/S Last C/S YEAR Place i l 1 1 C/S Place First Place C/S Last C/S YEAR 1 ! Place First Place 1 1 l-l C/S Last C/S YEAR Place i l 1 1 C/S Place First Place M l-l C/S Last C/S YEAR I 1 Place First Place I I l-l C/S Last C/S DAY i I M l-l YEAR Place i l I 1 C/S MONTH 1 1 1 1 l-l DAY 1 1 M l-l DAY 1 i i l-l I-! . 1 1 I DAY II I-! 1 1 MONTH DAY 1 1 1 1 1 l-l 1 u.- DEATH RECORDS Last Place 1 1 C/S Place 1 1 C/S Place I 1 C/S Place h. MOTHER'S FULL NAME Place YEAR 1 1 i I g. . CURRENT MEDICAL RECORDS First II l-l MONTH ! SI XTH. DAY 1 1 1 Last FIFTH MONTH 1 1 1 1 l-l BIRTH RECORDS MONTH 1 1 1 Last SECOND f. ' C/S �Q.2A-26 and 63-76 CHILDREN CHILD Q.29 NAME d. MONTH SEVENTH First Last EIGHTH NI NTH ELEVENTH DAY 1 1 !-| MONTH DAY 1 1 1 1 I-l 1 ! YEAR 1 i YEAR 1 1 1 I I-l I I I-l 1 1 T 1 MONTH \ 1 1 1 I-l DAY i 1 YEAR 1 1 T 1 MONTH 1 I I 1 l-t DAY YEAR 1 I I i I I I 1 g. CURRENT MEDICAL RECORDS h. MOTHER'S FULL NAME u. DEATH RECORDS First Place C/S Last C/S Place First Place C/S Last C/S Place First Place C/S Last C/S Place First Place C/S Last C/S Place 1 1 C/S • Place First Place C/S Last C/S Place 1 1 r/s Place First Place C/S Last c/s • Place ' 1 1 c/s YEAR DAY 1 1 ! I I 1 BIRTH RECORDS Place YEAR I 1 1-1 MONTH 1 1 1 1 I-l 1 I TWELFTH First Last 1 1 1 ! i-l First Last DAY I ! I-! MONTH First Last 1 1 f. T 1 1 ! 1 I I I - ! ! First Last TENTH ! 1 First Last BIRTHDATE Place 1 1 c/s Place ! 1 C/S Place I I c/s �136 CARD Q.28 Medical Providers — Pneumonia 1st Tint' a. 1'nd Time M o n t h s / y e a r ; ; lint) th.it t i me . 1 L MONTH f YEAR 1 1 1 L . l-l (1.3 0 1 ) 1 n. MONTH 1 1 1 1 1 I 1 1 l - l YEAR ' 1 I t:1 •) CM ) CM > ) C'i ) <!•'. ) (TO a. M o n t h s / y e a r s had t h a t time. MONTH T I 1 I i I D o c t o r / f a c i l i r y who murie dingnosis. 1 I hi YEAR 1 l l (!•', ) I'O I I - 1 1 1 1 (10 ) (U ) TO MONTH YKAR YEAR M l-l C'H ) C.") ) TO i "' V ' i i i i i (| h 5 i ( 1.7 ) 1-1 d (j ) i <| o } i b. 3rd Time M o n t h r . / y o i i r s h.ld t h a t t imo , TO MONTH. 812039 1 I l (;>{, ) C-'?1) 1 I 1 l D o c t o r / f a c i l i t y who made diagnosis. b. MONTH YEAR 1 1 1 1 -I l l I l l ( 3 2 ) (3'3) (.14) ( i S ) D o c t o r / f a c i l i t y who made d i a g n o s i s. Name Name Name Address Address Address C/S C/S C/S e. Name of h o s p i t a l . e. Name of h o s p i t a l . e. Name of h o s p i t a l . Name Name Name Address Address Address C/S C/S C/S a. Months/years had that time. a. MONTH YEAR 1 I I 1 1 l l I l l 1 ) '( } ( ) ( ) TO MONTH YKAR I I I I I I I I I I ( ) ( ) ( ) ( ) D o c t o r / f a c i l i t y w h o made diagnosis. M o n t h s / y e a r s had that t ime. 1 I 1 I b. 6th Time 5th Time 4111 Time I I MONTH YEAR 1 1 1 1 1 l l I l l ( ) ( ) ( ) ( ) TO MONTH YEAR I i ( ) ( ) h. a. M l-l MONTH YEAR T 1 I I 1 ' i I l l - I l l ( ) ( ) ( ) ( ) TO MONTH 1 1 I I 1 1 l-l ( ) ( ) ( I I I .1 ( ) ( ) Doctor/ f a c i l i t y who made diagnosis. Months/years had that time. b. YEAR 1 1 1 .1 ) '( ) Doc tor/ f a c i l i l y who made diagnosis. Name Name Name Address Address Address C/S C/S C/S e. Name of h o s p i t a l . e. Name of h o s p i t a l . e. Name of hospital. Name Ma me Name Address Address Address C/S C/S •.""/• •'.' .' . ••'/:'. C/S . . �CARD Q.29. Part 2 Month/year f i r s t diagnosed 1 I d. 812039 Medical Providers -- Cancer Part 1 c. 1 (6-H7 c. Month/year f i r s t diagnosed MONTH YEAR 1 1 1 1 1 l l I l l ( ) ( > ) ('17") (J8> (•]«> Doctor/ f a c il i t y where f i r s t d i a g n o s i s made: Part 3 c. Month/year first diagnosed MONTH YEAR 1 1 I I 1 1 I l l I l l (/,.', ) (45 ) ( 4 ( > ) (47 ) <1. D o c t o r / f a c i l i t y where f i r s t d i a g n o s i s made: 1 1. d. MONTH YEAR 1 1 1 1 1 1 l-l 1 1 ( W ) (VJ) (54) (55) D o c t o r / f a c i l i t y where f i r s t diagnosis made: Name Name Name Address Address Address C/S C/S C/S e. D o c t o r / f a c i l i t y last consulted. e. D o c t o r / f a c i l i t y last consulted. e. Doctor/facility consulted. Name Name Name Address Address Address C/S C/S last C/S f. Month/year last consulted • MONTH 1 l l 1 I YEAR I I l 1 1 - (40) ( 4 1 ) f. f. M o n t h / y e a r last consulted . 1 I 1 l 0,2> 0 , j ) M o n t h / y e a r last consulted. MONTH YEAR 1 1 1 I 1 l l I l l ( 4 8 > O.y) ( 5 0 ) (51 ) Part it 1 I MONTH YEAR 1 I I 1 1 l l I l l ('>6> (57) (58 > < 5 9 > Part 5 Part 6 * c. * Month/year f i r s t diagnosed 1 1 MONTH 1 1 (do) <6i ) d. YEAR 1 1 I I l-l c. Month/year f i r s t diagnosed 1 I 1 I (6:') ( b j ) D o c t o r / f a c i l i t y where f i r s t d i a g n o s i s made: d. c. Month/year f i r s t diagnosed MONTH YEAR 1 I I 1 1 l l I l l (68) ( 6 4 ) (?()) ( ? l ) D o c t o r / f a c i l i t y where f i r s t d i a g n o s i s made: 1 I d. MONTH YEAR 1 I I 1 1 l l I l l (12) (13> <14> <15> D o c t o r / f a c i l i t y where f i r s t diagnosis made: Name Name Name Address Address Address C/S C/S e. D o c t o r / f a c i l i t y last consul ted . C/S 1 e. Doctor/ " .icil itv l a s t consu 1 ted . e. D o c t o r / f a c i l i t y last consul ted. Name Name Name Address Address Add res s C/S C/S C/S f. Mnnth/vfliir last r o i i K u l I .••!. MONTH I l (64) f. Munrh/yoai- lasl m i i K t i i t ctl . YKAR l I l l (63) (6(1) ( ( > ? ) 1 J MONTI! 1 .1 1 I i-l f. M o n t h / y e a r last I . . I I H I l l 1 ,-.!. YKAR 1 1 1 1 1 I MONTH 1 l l ( I I . ) (17 ) II fl.H YEAR 1 I l ) 0') ) 1 l �137 Q.30 Medical Providers — Leukemia b. Month/year first diagnosis! MONTH YKAR —i—n—r~ 1-1 c. Doctor/facility where first diagnosis made: Name Add ress_ c./s f. Doctor/facility last consulted. Name Address_ C/S g. Month/year last consulted. I I MONTH I I YEAR 8)2039 �CARD Q.31. Medical Providers — THYROID b. First Cold h a d : MONTI! 1 I l First told h a d : YEAR I I 1 l 1 - (28) (29) 812039 OTHER MEDICAL CONDITIONS DIABETES b. 137 I ! l 1 I l (IJO) (31) c. D o c t o r / f a c i l i t y where diagnosis made; c. ANEMIA b. First told h a d : MONTH YEAR 1 I I 1 1 l l I l l (44) (45) <4(>) (-'•?) Doctor/ f a c i l i t y where diagnosi s made: 1 1 MONTH YEAR 1 I I 1 1 1 l-l 1 1 (6()) (<>] ) (62) (63) D o c t o r / f a c i l i t y where d i a g n o s i s made: c< Name Name Name Address Address Address C/S C/S C/S d. Doctor last consulted: 1 I ,| Doctor last c o n s u l t e d : MONTH YEAR 1 1 1 1 1 l l I l l (32) ( 3 3 ) (34) (15) e. Doctor/Facility consulted. last 1 I e. Doctor last consulted: d MONTH YEAR 1 1 1 1 1 l l I l l (48) ( 4 9 ) ( 5 0 ) (51 ) Doctor/Facility last consulted. 1 1 e. MONTH YEAR 1 I I 1 1 1 l-l 1 1 (64) (65) (66) (67) Doctor/Facility last consulted. Name Name Name Address Address Address C/S C/S C/S ENLARGED LIVER HEART CONDITION b. First told had: 1 1 b. First t o l d h a d : MONTH YEAR 1 I I 1 1 1 ' l-l 1 1 (36) (37) (38) (39) c. Doctor/ f a c i l i t y where diagnosis made: 1 I c. JAUNDICE b. First told had: MONTH YEAR 1 . 1 1 1 1 l l I l l (52) (53) ( 5 4 ) (55^> D o c t o r / f a c i l i t y where diagnosis made: 1 I MONTH l (68) c. 1 YEAR I I I l - I l (69) T l (70) (71) D o c t o r / f a c i l i t y where diagnosis made: Name Name Name Address Address Address C/S C/S C/S d. Doctor last consulted: MONTH 1 I l (40) e. l (41) 1 1 - d. YKAR 1 1 I l l ( 4 2 ) (43) Doctor/Facility last consulted. Doctor j as t c o n s u l t e d : 1 I e. d. MONTH YKAR 1 1 1 1 1 l l I l l (56) ( 5 7 ) (58) (59) u o c t o r / F a c i l i t y last consulted. Doctor last consulted: 1 I e. MONTH YEAR 1 I I 1 1 l l I l l (72) ( 7 3 ) (74) (75) Doctor/Facility last consulted. Name Name Name Address Address Address C/S C/S C/S Mi) _ _ �CAKD Q.31. CIRRHOSIS OF THE L1VEK First told had: MONTH i l 1 b. YEAR 1 (12) (13) 1 b. MONTH YEAR 1 I I I 1 1 l-l 1 1 (28) (29) (.30) (31) 1 1 (14) (15) Doctor/facility where diagnosis made: INTESTINAL PARASITES First told had: 1 1 l I,.. .1 . .!-.!.. c. §12039 Medical Providers -- OTHEK MEDICAL CONDITIONS (CONTINUED) HEPATITIS b. I.JH c. Doctor/facility where diagnosis made: First told had; MONTH YEAR 1 I I 1 1 1 l-l 1 1 (44) (45) (46) (47) I '1 c. Doctor/facility where diagnosis made: Name Name Name Address Address Address C/S C/S C/S ^ Doctor last consulted: MONTH I 1 ' I 1 (1_ Doctor last consulted: MONTH 1 1 r <16) (17) (18) 1 1 I 1 1 l I! l - (32) (33) (19) Doctor last consulted: YEAR 1 YEAR I I . l-l. d. I 1 l l MONTH YEAR 1 I I 1 1 .1 l-l 1 | (48) (49) (50) (51) 1 1 (34) (35) c. Doctor/Facility Vast consulted. c. Name Name Name Address Address Address C/S C/S C/S Doctor/Facility last consulted. GALL BLADDER b. MONTH - l 1 1 h. i ( 20)r ( 21) i-i YEAR 1 i Doctor/Facility last consulted. OTHKR I.1VKK CONDITION First told had: l c. First told had: 1 1 1 i ( 22) ( 23) c. Doctor/facility where diagnosis made: c. OTHEft RESPIRATORY h. First told had: MONTH YEAR MONTH YEAR 1 .1 1 . 1 1 1 1 I I I I 1 l-l 1' 1 I l l I l l ( 36) ( 37) ( 3H> ( 39) 62 ) 63 ) ^54) (55) Doctor/facility where diagnosis made: c. Doctor/facility where diagnosis made: Address Name i 'Address Address C/S C/S C/S Name d- Doctor last consulted: MONTH 1 1 e. 1 1 („>/,) (Zljj d. Doctor last consulted: YEAK i 1 l-l 1 1 (26) (27) Doctor/Facility last consulted. Name MONTH 1 1 i ! : 1 1 i'. j ! 1 d< Doctor last consulted: YEAR 1 ! l-l ( 40) ( 41) 1 1 1 1 1 I (42) Uj> Doctor/Facility last consulted. e. MONTH YEAR 1 M 1 1 l l I l l < 56) ( 5?) < 58' '( 59' Doctor/Facility last consulted. Name Name Name Address l\d dress Address C/S :/s :/s J4J .. �138 CARD Q.31. Medical Providers -- OTHER MEDICAL CONDITIONS (CONTINUED) OTHER MAJOR CONDITIONS b. F i r s t told h a d : c. SECOND MAJOR CONDITIONS b. MONTH YEAR 1 1 ! I I l l I l l (60) (61) ( 6 2 ) (63) I I 812039 THIRD MAJOR CONDITIONS F i r s t told h a d : MONTH 1 l l 1 I I I - YEAR 1 I l b. 1 l F i r s t told h a d : : 1 I MONTH 1 l l i I l ^ Doctor/ f a c i l i t y where d i a g n o s i s made: D o c t o r / f a c i l i t y where diagnosis m a d e : YEAR I I l II - D o c t o r / f a c i l i t y where diagnosis made: Name Name Name Address Address Address C/S C/K C/S d. Doctor l a s t c o n s u l t e d : Doctor l a s t c o n s u l t e d : MONTH YEAR 1 I I I I l l I l l (64) (65) ?66) ( 6 7 ) 1 I e. Doctor/Facility consulted. 1 I ( MONTH 1 l l i ( !) I I - YEAR 1 I l ( ) ( ) 1 l Doctor l a s t c o n s u l t e d : 1 - D o c t o r / F a c i l i t y last consulted. last MONTH 1 1 1 I l l - ( ) ( ) Namp Name Address Address Address C/S C/S C/S Q. 34 Medical P r o v i d e r s — Ac First b. Last consulted doctor 1 I c. MONTH YEAR 1 1 1 1 1 l l I l l ( 6H> (6<l ) C 7 ( i » ( 7!) D o c t o r / f a c i l i t y last consulted: Name Address C/S 242 ( ) ( ) Doctor/Facility consulted. Name YEAR 1 I l last 1 l �CARD A. PATCHES OF S K I N CHANCE COLOR c. D o c t o r / f a c i l i t y where diagnosis made: 812039 C. SKIN EXTRA SENSITIVE B. EASIER BRUISING OF SKIN e. D o c t o r / f a c i l i t y where diagnosis made: e. D o c t o r / f a c i l i t y where diagnosis made: Name Name Name Address Address Address c/s C/S C/S f. Month/year diagnosis made: f I g. f. MONTH . YEAR "i I 1 1 1 l l I l l (11!) ( 1 ! ) ( I t , ) (!')) D o c t o r / F a c i l i t y last consulted. f. Month/year diagnosis made: 1 I MONTH 1 l l I I - (28) (L'y) g. YEAR 1 I l Month/year" diagnosis made: MONTH 1 1 l I (ill) ( i l ) l •! l ( 4 4 ) (45) g. D o c t o r / F a c i l i t y last consulted. YEAR II - I 1 l l 1 (46) (47) D o c t o r / F a c i l i t y last consulted. Name Name Name Address Address Address C/S C/S C/S h. Month/year last consulted: I 1 MONTH YEAR 1 1 1 1 1 1 1-1 1 1 ( ] f , ) ( 1 7 ) (18) ( 1 9 ) h. h. Month/year last consulted: ' MONTH 1 I 1 l (32) l (33) Month/year last consulted: YEAR I I - 1 I l ( 3 4 > (15) 1 l 1 1 MONTH YEAR I I I 1 i 1 l-l 1 I (48) (49) . (50) (51) �CARD 1'30-UO 812039 Q.36 -- M e d i c a l P r o v i d e r s >;. C. B U R N I N G I N LIMBS B. T I N K L I N G I N 1. 1 MBS A. NUMBNKSS I N I.IMBS g. K . Doctor/ f a c i l i t y where d i a g n o s i s made: D o c t o r / f a c i l i t y where d i a g n o s i s made: Doctor/ f a c i 1 i t y where d i a g n o s i s made: Name Name Name Address Address Address C./S C/S C/S li. 1 I i. h. h. M o n t h / y e a r d i a g n o s i s made: M o n t h / y e a r diagnosis made: MONTH MONTH YEAR 1 1 1 1 1 l l I l l <:>2> < 5 J > ( 5 4 ) tyj ! I ! Doctor/Facility last consulted. YEAR ! l ! l (68) Month/year diagnosis made: ! - (69) I ! l l 1 I : ( 7 0 ) (71.) i. i. D o c t o r / F a c i l i t y last consulted. MONTH YEAR 1 I I 1 1 l l I l l (16 ) ( 1 7 ) (18) (19) D o c t o r / F a c i l i t y last consulted. Name Name Name Address Address Address C/S C/S C/S j_ M o n t h / y e a r last consul t e d : 1 I MONTH YEAK 1 I I 1 1 l l I l l <56> (57) ^58) «59'" MONTH 1 I 1 l l 1 I l ( //.) ( 7 r >) 1 l E. REDUCTION IN C R I P STRENGTH g. D o c t o r / f a c i l i t y where d i a g n o s i s made: D o c t o r / f a c i l i t y where d i a g n o s i s made: Name Name Address Address C/S . h, C/S M o n t h / y e a r diagnosis made : MONTH YEAR r i i i i r I l l - I l l h. Month/year diagnosis made : = i 1 MONTH i 1 D o c t o r / F a c i l i t y las I consulted. Name Address C/S YEAR i i l-l ( 7<>) (/ / ) i. i 1 1 1 ' 7.s' ( ?')) i. Doctor/Facilitv last . c o n s u l ' . d. i i Namp 1 Address C/S i. Month/year last consulted: 1 I MONTH YEAR 1 I I "1 1 l l I l l (04) ((>'>) (6f>) ( ( > 7 ) i . Month/year last consul ted: 1 1 M o n t h / y e a r last consulted: YEAR M - ( 7 2 ) ( 7l) D. PERSISTENT ACHF.S IN L I M B S g. j. j . Month/year last consul ted : MONTH 1 II 1 l-l ( 12) ( J .1) YEAR 1 1 ( 1 ',) ( 1 5) 244 r. 1 I MONTH YEAR 1 1 1 1 1 l l I l l (20 ) ( 21) ( 22) ( 23) �CARD °06 Q.5A 812039 Additional Civilian Training Programs (Q.6) b. For what kind of work was his next civilian training program preparing him? ... , , 6th Program 5th Program 4th Program f. For what kind of work was his next civilian training program preparing him? ar>( j. For what kind of work was his next civilian training program preparing him? < \ h< nsc nisr C\f,t (\7( (17( . .. (17 C [l8-19[ |l8-19| ...... c. In what month and year did he start this training? 1 1 MONTH I I I 1 l-l (2<>) (:>i> YEAR 1 1 1 1 g. In what month and year did he start this training? 1 1 (22) (•>•}) d. In what month and year did he complete this training? MONTH YEAR 1 1 II 1 1 I . I l-l 1 1 (24) (25) (26) (27) n ^t MONTH YEAR 1 I I 1 1 1 l-l 1 ] (20) <!l ) ( ) (2:i) « h. In what month and year did he complete this training? MONTH 1 1 18-19 k. In what month and year did he start this training? MONTH 1 1 1 1 l-l (24) (25) 1 1 (26) (27) 1 1 1 1 (22) ( ' ) 2) L. In what month and year did he complete this training? MONTH 1 1 YEAR 1 1 l-l (20) (>l ) YEAR 1 1 1 1 1 1 YEAR 1 1 1 1 1 l-l 1 C24 > < 25) ( 26) ( 27) e. Did he participate in any other civilian job training program that prepared him for a major change in his occupation? i. Did he participate in any other civilian job training program that prepared him for a major change in his occupation? m. Did he participate in any other civilian job training program that prepared him for a major change in his occupation? Yes.(jp( Yes.(?8( No. • Yes.(28( -1 (ASK Q.f) " Q.7a) 04 79-80 -1 (ASK Q.j) -2 (RETURN TO Q.7a) 05 79-8"0 245 -1 (RETURN TO 0.6 75^80 �007 CARD Q.55. A d d i t i o n a l M i l i t a r y T r a i n i n g Progr.-ims ( Q . 7 ) 4th P r o g r a m b. 812039 6th Program 5t.li Program For w h a t k i n d of work was h i s next m i l i t a r y t r a i n i n g program prep a r i n g him? g. L, Kor what k i n d of work was h i s next m i l i t a r y t r a i n i n g program preparing him? For what k i n d of work was h i s next m i l i t a r y t r a i n i n g program prep a r i n g him? d')( (!')( (!'->( • (lb( (16( (16( (U ( (17( ' c. W h a t was the AFSC for t h a t job? < d. \ I MONTH l \ (21) e. l l - (22) \ \ ( I i. \ l l MONTH \ 1 l (21) ( 2 2 ) In what month and year did he complete t h i s training? j. m. l-l \ YKAR 1 \ What was the AFSC for t h a t job? ( ( In what month and year did lie s t a r t t h i s training? \ 1 (23) (24) ^B.2()| What was the AFSC for t h a t job? ( YEAR d7( [IH-2()|- h. 7n what month and year did he s t a r t t h i s training? ' \ 1 n. . In what month and year did he start this training? ! 1 : MONTH ! 1 (21) (23) (24) In what month and year did he c o m p l e t e t h i s training? ( o. YEAR 1 1 H (22) 1 1 1 1 (23) (24) In what month and year did he complete this training? • MONTH •1 I f. 1 YEAR II 1 1 MONTH YEAR 1 1 1 i 1 1 1 1 1 1 1 1 (2rO (2(0 (27)(2H) 1 l l I l l ( L ' r > ) (2(0 ( 2 7 ) (28) Did he participate in any other m i l i t a r y job t r a i n i n g program that prepared him for a major change in his occupation? Yes.Q'it No -1 (ASK Q.g) "-2 ( R K T U K N TO Q.8) k. Did he p a r t i c i p a t e i n a n y other m i l i t a r y job t r a i n i n g program that prepared him for a major change in his occupation? Yus. (;>•)< No -1 (ASK Q . L ) -2 ( R K T U K N TO Q.S) 1).'. ()'> 7<f-HO 7y"-'m> 240 1 1 : MONTH 1 1 1 r 1 1 - (J O (26) p. YEAR 1 1 1 1 1 ( 2 7 ) (28) D i d he p a r t i c i p a t e i n a n y other m i l i t a r y job t r a i n i n g program t h a t prepared him for a m a j o r change in his occupation? Yes.(:")( No -1 (RETURN TO -2 Q.8) 06 79-80" �CARD Q. iio-bl. , , MO.OTH ... 1 . I I 1 . (l'j) (id) I .1 (17) I 1 1 1 (18) b. b, Wh*f was th* name of his employer? IRECORD IN S.R.B. - PC I I MONTH YEAR 1 I I 1 1 1 l-l I I (15) (10) (17) (18) 1 1 What was the name of his employer? b. c. Was the job fulltime or part-time? c. Fu 1 1 t ime . . ( I <)( -1 MONTH YEAR 1 I I 1 1 1 l-l 1 1 (15) (16) (17) (18) What was the name of his employer? IRECORD IN S.R.B. - PC 1 i TRECORD IN S.R.B. - PG i \ Was the job fulltime or part-time? Full tim«. . I9( ( 58a. In what month and year did he start his next job that lasted three months or longer? 57a. In what month and year did he start his next job that lasted three months or longer? YEAR ! l-l Ninth Job Eighth Job 56a. In what month and year did he start his next job that lasted three months or longer? c. 812039 Additional Jobs (C .8-13) Seventh Job I 1 008 -1 Was the job fulltime or part-time? Full time..(!9( d. What kind of business was that — what did they make or do there? d. What kind of business was that — what did they make or do there? d. e. What did he actually do on the job -- what were some of his main duties? e. What did he actually do on the job — what were some of his main duties? -1 e. What kind of business was that — what did they make or do there? What did he actually do on the job — what were some of his main duties? IRECORD IN S.R.B. - PC 1 1 TRECORD IN S.R.B. - PC i I IRECORD IN S.R.B. - PG I 1 IHAND RESPONDENT CARD "B"| IHAND RESPONDENT CARD "j»"| IHAND RESPONDENT CARD "B"| n,, i f. Please look at this card and tell me the number which best describes the kind of industry he worked in. (WRITE IN NUMBER) g. 1 1 1 I I 1 (20) (2l) In what month .and year did this job end? I I f. -1 (RETURN TO Q.14) What was the main reason he stopped working .il that job? - " ' !, ' „ f. Please look at this card and tell me the number which best describes the kind of industry he worked in.' Please look at this card and tell me the number which best describes the kind of industry he worked in. (WRITE IN NUMBER) 1 I 1 1 (WRITE IN NUMBER) 1 1 | I I I .1 1 (20) (21) r-, Ell In what month and year did this job end? (20) (21) g. MONTH YEAR I I I 1 1 l l I l l (2H) (24) (25) (2ft) Current job..(27( h. r r.lr In what month and year did this job end? 1 1 MONTH 1 1 Current job..(27( h. I I l-l -1 YEAR 1 I . g. 1 1 1 I (RETURN TO Q.14) MONTH 1 II 1 l-l (2;)) (24) Current job,.(27( -1 YEAR 1 (25) 1 1 (26) (RETURN TO Q.14) What was the main rea- i h.. What was the main reason he stopped working son he stopped working at that job? at that job? (28 (28 i (28 (20 (29 i (29 (ASK Q.57a) no-Try (ASK Q.58a) 07 79-80 "Oo-'l /) OH 70-80 247 I (ASK Q.59a) Vi'o-fi') _o.9. 79-80 �CARP 59a. In what month and year did he start his next job that lasted three months or longer? 60a. In w h a t month and year MONTH 1 I (15) ( i n ) 61 a. did he s t a r t his next job t h a t lasted three months or longer? YEAR UTTOH) YEAR 1 M 812039 T w e l f t h Job Eleventh Job Tenth Job MONTH OQH In w h a t month and year did he s t a r t his next job t h a t l a s t e d three months or .longer? MONTH I I I l l I l l <15 > (16 ) (17 ) (18 ) I YEAR 1 II l l 1 - 1 I (15 ) (16 ) l l C17 ) U8 ) b. What was the name of his employer? [RECORD, IN S.R.B,. - PC f I b. What was the nnme of h i s employer? I RECORD IN S.R.B. - PG 1 I ! b. What was the name of his e m p l o y e r ? (RECORD IN S.R.B. - PC 1 c. Was the job fulltime or part-time? c. Was the job fulltime or part-time? c. Was the job fulltime or part-time? Fu 11 t ime..( Part Lime... Full time..(l9( Part time -1 -2 Full time..(l9(_ Part time -1 -2 d. What kind of business was that — what (do/ did) they make or do there? d. What kind of business was that — what (do/ did) they make or do there? d. What kind of business was that — what (do/ did) they make or do there? e. What did he actually do on the job — what were some of his main duties? e. What did he actually do' on the job — what were some of his main duties? e. What did he actually do on the job — what were some of his main duties? IRECORD IN S.R.B. - PG i 1 IHAND RESPONDENT CARD "B"| f. Please look at this card and tell me the number which best describes the kind- of industry he worked in. (WRITE IN NUMBER) g. 1 | (20) 1 1 1 1 H In what month and year MONTH YEAR I I I 1 1 l l I l l (23) ( 2 4 ) (25) ( 2 6 ) Current j o b . . ( 27( -1 (RETURN TO Q.14) h. What was the m a i n reason he stopped working at that job? IRECORD IN S.R.B. - PG j 1 i IHAND RESPONDENT CARD "Bir| f. Please look at this card and t e l l me the [ number which best describes the kind *of industry he worked in. • (WRITE IN NUMBER) 1 I 1 | 1 I IHAND RESPONDENT CARD "B"| f. Please look at t h i s card and tell me the number which best describes the kind of ind u s t r y he worked in. ; (WRITE IN i NUMBER) 1 I (20) ( 2 1 ) (21) did this job end? 1 I IRECORD IN S.R.B. - PG i 1 I (20) g. In what month and year did t h i s job end? 1 I I I MONTH YEAR 1 1 1 1 1 l l I l l r (23) (24) (2 >) (26) Current (RETURN TO job..(_27(__-l Q . 1 4 ) h. What was the m a i n reason he stopped working at t' ...t job? (21) HI ; g. In what month and year did t h i s job end? MONTH 1 I YEAR 1 l I I l - (23) (24) Current job..(27( -1 1 I (25) 1 l (26) (RETURN TO Q-14) h. What was the m a i n reason he stopped working at that job? (28( (28( (2B( (29( (29( (29 ( (ASK Q.60a) (ASK Q.61a) TJo"-7jy (RETURN TO Q.14) ~(To-T7j" 11 10 '7 9- HO O't-36) l TiTT-TJ")" 12 "79-80 7<J-~80 Cl/,-36) (34-36) 248 �015 CARD 812039^ Q.62. Additional Periods in Military (Q.U) a. In what month and year did he next enter the Armed Forces? 1 1 MONTH ' YEAR 1 I I 1 1 1 l-l 1 1 (14) (15) (16) (17) f. In what month and year did he next enter the Armed Forces? MONTH 1 1 k. In what month and year did he next enter the Armed Forces? YEAR 1 1 1 1 I I l-l (14) (15) 1 1 (16) (17) 1 1 MONTH YEAR 1 1 1 I I 1 l - l I I (14) (15) (16) (175 b. What branch of the mili- g. What branch of the mili- L. What branch of the military was that? tary was that? tary was that? Air Force. ( 8 1( -1 Army -3 Coast Guard... -5 c. Was he discharged or separated from the (BRANCH OF SERVICE)? Air Force. ( 8 1( -1 Air Force. (IB( Coast Guard... -5 Coast Guard. . . 'r h. Was he discharged or separated from the (BRANCH OF SERVICE)? -1 -5 m. Was he discharged or separated from the (BRANCH OF SERVICE)? Discharged/ Discharged/ Discharged/ separated. (ig( -1 (ASK separated. (19( -1 (ASK separated. ( ig( -1 (ASK Q.62d) Q.62n) Q.62i) Still in Still in Still in (MILITARY) -2 (RETURN (MILITARY) -2 (RETURN (MILITARY) -2 (RETURN TO Q.15 ~" TO Q.15 1 TO Q.15) d. In what month and year was he discharged/ separated from the (BRANCH OF MILITARY)? 1 I MONTH YEAR 1 I I 1 1 l l - I l l (20) (21) (22) (23) i. In what month and year was he discharged/ separated from the (BRANCH OF MILITARY)? 1 1 MONTH YEAR 1 1 I 1 1 1 l-l 1 1 (20) (21) (22) (23) n. In what month and year was he discharged/ separated from the (BRANCH OF MILITARY)? I I MONTH YEAR I I 1 I -1 1 (20) (2l) (22) (23) e. Following his separation j. Following his separation o. Following his separation or discharge in (DATE IN or discharge in (DATE IN or discharge in (DATE IN "i"), did he reenter the "n"), did he reenter the "d"), did he reenter the Armed Forces? Armed Forces? Armed Forces? Yes..(24( -1 (ASK Q.62f) Yes..(24.( -1 No -2 No....77T -2 (RETURN TO " Q.15) 04 79-80 249 Yes..(24( -I\;(ReTURN TO (ASK Q.62k) (RETURN TO No.... 771 -2 ( (JilS) Q.15) Q5 06 79-80 79-80 �CARD Q.63. 812039 016 Additional Countries (Q.15-16) Seventh Country Ninth Country Eighth Country What was the next country that he was stationed in for more than 90 days while on active duty? What was the next countr that he was stationed in for more than 90 days while on active duty? a. What was the next country that he was stationed in for more than 90 days while on active duty? (14 -15 (lit -15 (14 -15 (RECORD COUNTRY HERE AND IN (RECORD COUNTRY HERE AND IN (RECORD COUNTRY HERE AND IN S.R.B. PG 2 AND CONTINUE) S.R.B. PG 2 AND CONTINUE) (S.R.B. PG 2 AND CONTINUE) ! (No others.(l(,( -1 (RETURN No others.(j,6< -1 (RETURN No others.(lb( -1 (RETURN TO Q.17) TO Q.17) j TO Q.17) b. n. In what month and year did he begin and end active duty in (COUNTRY)? In what month and year |h. did he begin and end ac- | tive duty in (COUNTRY)? j In what month and year did he begin and end active duty in (COUNTRY)? BEGIN BEGIN MONTH YEAR 1 1 I I 1 1 I l l I l l (17) (18) (19) (20) BEGIN MONTH YEAR 1 1 1 1 1 1 I l l I l l (17) (18) (19) (20) END MONTH YEAR 1 1 II I 1 I l l I l l (21) (22) (23) (24) END MONTH YEAR 1 1 I I 1 1 I l l I l l (21) (22) (23) (24) END MONTH YEAR 1 1 I I 1 1 I l l I l l (21) (22) (23) (24) Current . . . (25( Current... (25( Current . . . (25( -1 1 I •-! MONTH 1 l l I I - (17) (18) YEAR I I l Tl9) (20) 1 l -1 c. What specific job assign- i. ments did he have in (COUNTRY)? Can you give me the AFSC? What specific job assign- o. ments did he have in (COUNTRY)? Can you give me the AFSC? What specific job assignments did he have in (COUNTRY)? Can you give me the AFSC? 1. (26 -28 1. (2f> -28 1. (?6-28 2. (29 -31 2. (29 -31 2. (29'-31 3. (32 -34 3. (32 -34 3. (3,2 -34, d. Did his duties in j. (COUNTRY) include flying? Did his duties in p. (COUNTRY) include f l y i n g ? Did his duties in (COUNTRY) include flying? Yes.<35( No e. -1 -2 Yes.(35( No How many flight hours did he log while in (COUNTRY)? 1 1 1 1 1 1 k. 1 1 Hours 1 1 -1 -2 How many f l i g h t hours did he log while in (COUNTRY)? 1 1 (38) -1 What specific letter and numerical deeignation(s) did each a i r c r a f t have? q. 1 1 Hours 1 1 1 1 (36) (37) 1 1 Hours (38) Other (SPECIFY) Other (SPECIFY) .(39( f. 1 1 (36) (37) Other (SPECIFY) Yes.(35< No How many f l i g h t hours did he log while in (COUNTRY)? 1 1 (36) (37) (38} -1 -2 .(IQ( •<32< -l L. What specific l e t t e r and numerical designation(s) did each aircraft have? r. -1 What specific letter and numerical designation(s) did each aircraft have? 1. (<W><-43 1- (4n(_4f 1. 2. (44 (-47 2. ( 4 4 ( - 4 7 2. j44(-47 3. (48(-'>l 3. (48(-. r >L 3. (48(-51 4. (56-59) (60-63) ((..'.- h 7) (56-59) (ISO- ft 3) (i.-'i • ( . / ) J il_M( 1 ?r, (40{-43 (52(-55 (ASK Q.64a) (ASK Q.m) (68-71) "(72-75) 07 . (52(-55 4. (52(-55 4. (ASK Q.g") „ (68-71) " (7?-7r>) on /'/-HO (56-59) "(60-r.-)) (d.'i- I./) (68-71) (72-75) D'J 79-80 �CARD 01b 812039 Q.64. Additional Countries (Q. 15-16) Tenth Country Eleventh Country a. What was the next country g. that he was stationed in for more than 90 days while on active duty? Twelfth Country What was the next country m. that he was stationed in for more than 90 days while on active duty? (i4(-u (14(-15 (RECORD COUNTRY HERE AND IN (RECORD COUNTRY HERE AND IN S.R.B. PG 2 AND CONTINUE) S.R.B. PG 2 AND CONTINUE) No others. (lb( b. -1 (RETURN No others. (]f>( TO Q.17) In what month and year did he begin and end active duty in (COUNTRY)? h. What was the next country that he was stationed in for more than 90 days while on active duty? (14(-15 (RECORD, COUNTRY HERE AND IN S.R.B. PG 2 AND CONTINUE) No other s ( 6 .!( -1 (RETURN TO Q.17) In what month and year did he begin find end active duty in (COUNTRY)? n. tn what month and year did he begin and end active duty in (COUNTRY)? BEGIN BEGIN MONTH YEAR 1 1 I I I I 1 1 .. l-l . 1 1 (17) (18) (19)"(20) MONTH 1 1 1 1 BEGIN YEAR 1 I I I l-l -1 (RETURN TO Q.17) MONTH 1 I 1 1 1. l YEAR 1 T 1 l - I l 1 l END MONTH YEAR 1 1 1 1 1 1 1 1 l-l 1 1 ( 2l5 ( 22V ( 23) ( 24) END MONTH YEAR 1 1 I I I 1 1 1 l-l 1 1 ( 21) ( 22) ( 23)" I 2&) ' END MONTH YEAR I I I I I 1 I l l I l l (21 ) ( 22) ( 23) ( 24) Current ... ( 25( . .. - 1 Current . . . ( 25( Current . . . C5 ( c. -1 -1 What specific job assign- i. ments did he have . in (COUNTRY)? Can you give me the AFSC? What specific job assign- o. ments did he have in (COUNTRY)? Ca.n you give me the AFSC? What specific job assignments did he have in (COUNTRY)? Can you give me the AFSC? 1. (26 (-28 1. ( 2d -28 1. (26 (-28 2. ( 2<X -31 2. ( 2 * -31 2. C29 (-31 3. ( 3 2 ( - 3 4 3. ( 3 * -34 3. (3X-34 d. Did his duties in j. (COUNTRY) include f l y i n g ? Did his duties in p. (COUNTRY) include flying? Did his duties in (COUNTRY) include flying? Yes.( 3* e. -1 Yes. 05 ( No . . How many f l i g h t hours did he log while in (COUNTRY)? I 1 l 1 l 1 1 1 Hours (36) ( 3 7 ) (38) Yes.(35( | Hr, How many f l i g h t hours did he log while in (COUNTRY)? 1 1 1 1 1 1 q. .<3?( L. 1 1 1 1 Hours Other (SPECIFY) -1 What specific l e t t e r and numerical designation(s) did each aircraft have? 1 1 (36) ( 3 7 ) (38) Other (SPECIFY) • <39< -1 -i "• How many f l i g h t hours did he log while in (COUNTRY)? 1 1 1 1 Hours ( 36) ( 37) ( 38) Other (SPECIFY) f. k. -1 -2 .<V* -1 What s p e c i f i c letter and numerical designat ion(s) did each a i r c r a f t have? r. -1 What specific letter, and numerical des,igna,<i:Q,0.(s): did each a i r c r a f t have.? 1. ( 4 0 ( - 4 3 1. ( i p ( - 4 ' l 1. (40(-43 2. ( 4 4 ( - 4 7 2. < 4 4 ( - 4 7 2. (44(-47 3. (48(-51 _3. Ufi(-5l 3. (48(-51 1 (52(-55 r 4, ( ) 2 ( - V j 4, (56-59.) (60-63) "(64-07). (6K-7I ), "(72-7'j) 10 T 7 .. . (5.6-59) "(60-63) ••"(64-67) (^(-S -, 4 . (68-71) ~ (72-7'i) I1 77-TTO 251 (56-5.9.) (60-63) '(64-,67.) (68-71.) "(72-75) 12 Tf-'fffl �CARD Q. 65-67. 020 A d d i t i o n a l Marriagiis FOURTH MARRIAGF. FIFTH MARRtACK b ^ f l . I n w h a t month am! yt*«r did he get married the fourth time? h f m . I n w h n i month nm1 year did hp get married the f i f t h time? 1 I MONTH YEAR 1 1 1 1 1 l l I l l <15) <16> ( l 7 > d « ) I I c. What was her f u l l maiden name? I RECORD IN S . R . B . PG 2 I What was her f u l l maiden name? 1 1 T I l l 1 1 I Times 1 1 Never.. (52( (SKIP TO Q.f) Never. .(52( e. ; What was her f u l l maiden name? (RECORD IN S.R.B. PG 2 I BeRin card 220 19-49 d. During this marriage, how many times was he living apart from his wife (you) for more than three months? 1 1 1 I 1 1 1 1 Months 1st 1 I 1 I 1 I 1 1 Months 2nd 1 1 1 I 1 I Months 3rd 1 I 1 1 1 Months -1 (SKIP TO Never. .(52( 1 Months t 61) ( 6T" 1 Months (63) (64) (GO TO Q.65f ) -i 1 I 1 I Months 1 1 1 I 4th 1 I 1 1 5th 1 I 1 1 e. How many months did they (you) live apart the ( f i r s t / n e x t ) time? 1st I I iI i1 I I 1 I Months (53) < 5 4 ) 1 I Months 2nd 1 1 I I . I I Months (55) (56> 1 I Months 3rd 1 I 1 I 1 I Months (57) (58) 1 1 Months 4th 1 1 1 I 1 I Months (59) (60^ 1 I Months 5th 1 1 1 1 1 1 Months (&l) (62) (61 > ( 6 2 ) 6th (SKIP TO Q.f) <59~5 ( 6()> 1 1 | Times (50) (51) ( 5 7 ) (58) ( 59> ( 6(i 1 I | (55 ) (56 ) ( 57) ( 5® 1 1 I 1 (53) (54) (55) (56) , I e. How many months did they] (you) live apart the ; ( f i r s t / n e x t ) time? ( 53) ( 54) 1 1 i Q.f) e. How many months did they (you) live apart the ( f i r s t / n e x t ) time? 6th < : (50) (51) -1 cur- IRECORD IN S.R.B. PC 2 I 1 Times (50) (51) 5th MONTH YEAR 1 1 1 1 1 l l I l l (15) (16) (17) (18) b. What ( i s / w a s ) the rent f u l l name of that w i f e Bet-in curd 220 19-49 : d. During this marriage, i how many times was he ! l i v i n g apart from his j w i f e (you) for more than three months? d. During this marriage, how many times was he living apart from his w i f e (you) for more than three months? 4th 1 I cur- IRECORD IN S.R.B. PG 2 1 19-49 3rd bla. In what month mij y e a r did he get married the sixth time? IRECORD IN S.R.B. PG 2 I c. 2nd STXTH MARR1AGK MONTH YEAR I I I I I l l I l l (15> dh> d?) < 1 8 > b. What ( i s / w a s ) the rent f u l l name of that wife b. What ( i s / w a s ) the current f u l l name of that w i f e 1 RECORD IN S.¥.B. PG 2 1 1st 812039 r1 Months (63) (64) (GO TO Q . 6 6 f ) 6th 1 1 1 1 1 1 Months (63) < 6 4 ) (GO TO Q . 6 7 f ) �020 CARD Q,65-67. 812039 Additional Marriages (CONTINUED) FOURTH MARRIAGE FIFTH MARRIAGE IIF ONLY MARRIAGE! SIXTH MARRIAGE 65f. At the time he (died/ became incapacitated) was he divorced , widowed, separated, or was he married and living with his wife? llF LAST MARRIAGEl 66f. At the time he (died/ became incapacitated) was he divorced, widowed, separated, or was he married and living with his wife? JIF LAST MARRIAGEl 67f. At the time he (died/ became incapacitated) was he divorced, widowed, separated, or was he married and living with his wife? Living with wife...((j6( Living with wife...((,6( Living with wife...(66( Divorced Separated Widowed (RETURN TO Q.22) -1 -?](SKM' TO .Z-^IQ.h) ~4j -1 (RETURN TO Q.22) nivorrprf -?1(SKIP TO S<"p/ir«l«-<1....~-3( Q.h) -1 (RETURN TO Q.22) i Separated Widowed J ~-3f Q- h ) -4 1 ""*"* J IRECORD IN S.R.B. PG 2 1 IRECORD IN S.R.B. PC 2 ,1 JRECORD iii S.R.B. PG 2 I |lF OTHER MARRlAGESl ll F OTHER MARRlAGESl IIF OTHER MARRlAGESl g. How did that marriage end —- was he divorced or was he widowed? Pivoreed(67( -1\(ASK Q.h) Widowed..T T - J .22 g. How did that marriage end — was he divorced or was he widowed? Divorced(67( Widowed -UUSK Q.h) -2\ IRECORD IN S.R.B. PG 2 ) IRECORD IN S.R.B. PG 2 I h. In what month and year was he (divorced/ widowed'/separated)? h. In what month and year was he (divorced/ widowed/separated)? MONTH 1 I i l YEAR l l (68) (69) 1 l - I l MONTH 1 l (70) (71) (IF A FIFTH MARRIAGE GO TO Q.66a) 1 I Divorced(67( -ll(ASK Q.h) Widowed -2) IRECORD. IN S.R.B. PG 2 I h. In what month and year was he (divorced/ widowed/separated)? MONTH YEAR 1 l g. How did that marriage end — was he divorced or was he widowed? 1 1 1 - I l l (68) ( , ) (9 1 l Oo) (71) (IF A SIXTH MARRIAGE GO TO Q.67a) 1 1 YEAR 1 1 1 1 l-l (68) (69) 1 I 1 I (70> (71-) (RETURN TO Q.22) (72-73) (72-73) (72-73) (74-75) (74-75) (74-75) (76-77) (76-77) (76-77) 79-80 �CARD 68-70. 02H 812039 Additional Children (Q.22-2 FOURTH CHILU FIFTH CHILD SIXTH CHILD NAME: NAME: 68a. How old is (CHILD) now? 69a. How old is (CHILD) now? I I I I I I I I I I I I .(l'5) (10) Age (15) (16) Child died..(U( Child died..( 17( -1 b. (Is/Was) (CHILD) male or female? Male Female. I OUNCES T b. (Is/Was) (CHILD) male or female? I (19) (') .() (21) (22) Don't know. , ,(_23( d. What is (CHILD)'s birthdate? ! POUNDS I (21) (22) I I OUNCES I (19) (2"0)(21) (22) Don't know...( 23( d. What is (CHILD)'s birthdate? -1 d. What is (CHILD)'s birthdate? YEAR MONTH DAY MONTH DAY YEAR 1 1 I I 1 I I 1 1 1 1 I I 1 1 1 1 1 1 1 l-l 1 l-l 1 1 1 1 l-l 1 l-l 1 1 28) ( 29) (24) (25) (26) (27) (28) (29) ( 24) ( 25) ( 26) ( 27) ( (2/0 (25) (26) (27) (28) (29) MONTH I I 1 1 II l-l DAY 1 1 1 1 l-l YEAR T I IT l-l Don't know.. .(23( -1 c. How much did (CHILD) weigh at birth? OUNCES POUNDS I I -1 Male. Female. c. How much did (CHILD) weigh at birth? T I I I Age (15) (16) Child died..(17( -1 Male. Female. c. How much did (CHILD) weigh at birth? (19) (20) 70a. How old is (CHILD) now? ! | Age b. (Is/Was) (CHILD) male or female? <18( POUNDS NAME: 1 1 1 1 [ALSO RECORD IN S.R.B.-PG 3| e. Was the child premature, full term, or overdue? Premature. ( jg( (CO TO Q.f) -1 lALSO RECORD IN S.R.B.-PG 3| e. Was the child premature, full term, or overdue? Premature. ( ^o( Full term..... Overdue (GO TO Q.f) 254 -1 -2 -3 lALSO RECORD IN S.R.B.-PG 31 e. Was the child premature, full term, or overdue? Premature. ( ,J(/( Full term. • • . . (GO TO Q.f) -1 -2 �CARD FOURTH CHILD 68f. Where arc (CHILD) 's birth registration records located? In what city and state is that? IRECORD IN S.R.B. PC 3 | 028 FIFTH CHILD SIXTH CHILD 69f. Where are (CHILD) 's birth registration records located? In what city and state is that? IRECORD IN S.R.B. PC 3 I g. Where are (CHILD) 's current medical records located? In what city and state is that? .812039 g. Where are (CHILD) 's current medical records located? In what city and state is that? 70f. Where are (CHILD) 's birth registration records located? In what city and state is that? 1 iREcdRD IN"SVR:B. PC'S' r g. Where are (CHILD) 's current medical records located? In what city and state is that? IRECORb ih S..R.B. PC 3 1 IRECORD IN S.R.B. PC 3 1 IRECORD IN S.R.B. PG 3 1 h. What was (CHILD) 's mother's full name? h. What was (CHILD) 's mother's full paiiie? h. What was (CHILD) 's mother's full name? (RECORD IN S.R.B. PC 3 1 IRECORD IN S.R.B. PG 3 1 1 1 i. How old was the mother when (CHILD) was born? i. How old was the mother when (CHILD) was born? i. How old was the mother when (CHILD) was born? 1 1 IRECORD IN S.R.B. PC 3 1 1 1 1 I . I 1 Age < 3 I > <32> 1 1 Age (J].i <32) 1 1 1 1 Age <31> <32> [33-48] . . |33-48| j. Did (CHILD) have any birth defects? j. Did (CHILD) have any birth defects? j. Did (CHILD) have any birth defects? 1 1 Yes.Uq( Yes.(49( -1 (ASK Q.k) Ye8.(49( -1 (ASK Q.k) No. No No -1 (ASK Q.k) . . -2 (SKIP TO Q.L) k. What kind of birth defects did (s)he have? Any others? -2 (SKIP fO Q.L) , -2 (SKIP TO Q.L) k. What kind of birth defects did (s)he have? Any others? k. What kind of birth defects did (s)he have? Any others? L. Was (CHILD) ever diagnosed as having cancer? L. Was (CHILD) ever diagnosed As having cancer? ' t. Was (CHILD) ever diagnosed as having cancer? Yes.(50( No.. lit. 1 1 -1 <ASK Q.m) . -2 (SKIP TO Q.o) In what month and year was the diagnosis made? MONTH YEAR i l l 1 1 1 l-l 1 1 (SI) ( 52) (5)) (V.) n. What kind of cancer was diagnosed? m. In what month arid year was the diagnosis made? 1 I MONTH VfiAR 1 } i \ 1 l l I l l (SI) (Sl>) (S3) (Vi) n. What kind of cancer was diagnosed? <5.5-5h) Not sure..( 57( (GO TO Q.o) -1 . (5.5-5.6) . Not sure. . ">l( ( -1 (GO TO Q.o) 255 Yes.(3Q( -1 (ASK Q.m) No Yes. (JQ(_ . -1 (ASK Q.m) -2 (SKIP TO Q.o) m. In what month and year was the diagnosis made? MOUTH 1 ' 1 1 1 YEAR 1 1 I I l-l (50 (52) ! L (S3) (54) n. What kind, of cancer Was diagnosed? . Not sure..(^7< (CO TO Q.o) (35-56). -1 �CARD FOURTH, CHILD 680. 812039 028 FIFTH CHILD (Does/Did)(CHILD) have diagnosed learning disability? Yes.(5H( -l (ASK No SIXTH CHILD 69o. (Does/l)id)(CHlLD) have a 70o. (Does/Did)(CHILD) have a diagnosed learning disdiagnosed learning disability? ability? Q.p) -2 (SKIP TO Q.q) -1 (ASK |No Q.p) Yes.(58( -2 (SKIP TO Q.q) No -1 (ASK Q.p) __-2 (SKIP TO Q.q) I p. What kind of learning d i s a b i l i t y (does/did) (s)he hnve? p. What kind of learning d i s a b i l i t y (does/did) (s)he hnve? p. What kind of learning disability (does/did) (s)he have? q. (Does/nidMCHILD) have any physical, mental, or motor impairments? q. (Uoes/nid) (CHILD) have any physical, mental, or motor impairments? q. ( Does/Did) (CHILD) have any physical, mental, or motor impairments? Yes.(59( -1 (ASK Q.r) Yes. (39 ( -1 (ASK Q.r) No -2 (SKIP TO Q.s) No -2 (SKIP TO Q.s) r. What kind of impairment (does/did) (s)he have? ,Yes.(3^( i 'No r. What kind of impairment j (does/did) (s)he have? ; -1 (ASK Q.r) -2 (SKIP TO Q.s) r. What kind of impairment (does/did) (s)he have? j IF CHILD IS DEAD: OTHERWISE: CONTI NUE. SKIP TO NEXT CHILD DAY 1 1 OTHERWISE: ', s. On what date did (CHILD) die? MONTH 1 II 1 l-l IF CHILD IS DEAD: OTHERWISE: IF CHILD IS DEAD: CONTINUE SKIP TO NEXT CHILD s. On what date did (CHILD) die? YEAR 1 1 CONTINUE SKIP TO' NEXT CHILD s. On what date did (CHILD) die? MONTH DAY YEAR MONTH DAY YEAR 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l-l 1 l-l 1 1 1 l-l 1 l-l 1 (60) (61 ) (62) (63) (64) (6r>) : (60) (61 ) (62) (6J) (64) (65) (60) (61) (62) (63) (64) (65 I I l-l t. What was the cause of death? ! 1 1 t. What was the cau?,: of death? u. Where is (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. PG 3 I u. Wh^.e is (CHILD) 's death registered? In what city and state is that? 1 RECORD IN S.R.B. PC 3 I (GO TO NEXT CHILD Q.69a) (GO TO NEXT CHILD Q.70a) t. What was the cause of death? u. Where is (CHILD) 's death registered? In what city and state is that? [RECORD IN S.R.I. .f<5.^ (GO TO NEXT CHILD Q.71a) 6Hn. (66-67) 68u. (66-67) 6MI . (!-)) («(') I,H, . (h«-M) f.Hf. (6K-69) drin. (./0-7J) (iHn. (711-71) i.Hn. (70-71) .a,,. </.',-7/) 6Hp. (7.'i-7'7) OHp. (7/,-77) 06 /M-Hil �CARD . 028 71-73. 812039 Additional Children F.ICillTII CM (I,I) SEVENTH CHILD NINTH CHILD i NAME: NAMK; 71a. How old is (CHILD) now?j 72a. I.. .. .1. - J Age (16) Child died..(l?( Male Female. c. How much did (CHILD) weigh at birth? I T POUNDS "(21) (22) (20) I I Don't know...(j>j( i -1 : b. (Is/Was) (CHILD) male or female? ( | c. How much did (CHILD) weigh at birth? OUNCES T -I b. (Is/Was) (CHILD) male or female? .(JH( 73a. How old is (CHILD) now? I I I Age (15) (16) I l Child died..(17( -1 • I I Age (13V (16) Child died,.(l7( -1 Female. POUNDS How old is (CHILD) now? I b. (Is/Was) (CHILD) male or female? Male NAME: I OUNCES r (19) ( 20) ( 21) (22) Don't know. ..(2J< -1 c. How much did (CHILD) weigh at birth? T i POUNDS i i ri i-i ( 19) ( 20) OUNCES r i r i (21) (22) Don't know...( 23( -1 j d. What is (CHILD)'s birthdate? d. What is (CHILD)'s birthdate? d. What is (CHILD)'s birthdate? YEAR MONTH DAY YEAR MONTH DAY I I 1 I I 1 I I 1 1 1 1 I I 1 1 I I 1 1 1 1 l-l 1 l-l 1 1 1:1 1 l-l I ' l -1 (24) (25) (26) (27) (28) C29) (24) (25) (26) (27) (28) (29) (24) (25) (26) (2?) (28) (2?) 1 1 MONTH 1 1 1 1 l-l. DAY 1 1 1 1 1 -1 YEAR 1 1 IALSO RECORD IN S.R.B.-PG 4| lALSO RECORD IN S.R.B.-PG 4| IALSO RECORD IN S.R.B.-PG 4 | i e. Was the child premature, full term, or overdue? ; e. Was the child premature, full term, or overdue? -1 Premature. (jp( -1 Full terra.. . . ._ -2 -7 -4 e. Was the child premature, full term, or overdue? -4 _ (GO TO Q.f) ' Premature. ([( j) (GO TO Q.f) 257 Premature. ( -jf)( Full term (GO TO Q.f) -1 -2 �CARD SEVENTH CHILD EIGHTH C1UI.1) 71 f. Where are (CHILD)'s birth registration records located? In what city and state is that? ; [RECORD IN S.R.B. PC~VT TRECQRD IN S.R.B. pgjtj i g. Where are ( C H I L D ) ' s current medical records located? In what c i t y and state is t h a t ? IRF.IIORO I N S . R . B . I'C I, [ TKECOHD IN S.R.'B. re '< I h, What was (CHILD)'s mother's lull name? I RECORD IN S.B.B. PC It \ h. What was (CHILD)'s mother's full name? I RECORD IN S.R.B. PC k \ i. How old was the mother when (CHILD) was horn? 73f. Where are (CHILD)'s birth registration records located? In ' what city and state is that? IRECORD IN S.R.B. PC V | g. Where are (CHILD)'s current medical records located? In what city and state is that? TRECORD IN S.R.B. roT | h. What was (CHILD)'s m o t h e r ' s f u l l name? IRRCORD IN S.R.B. PC a I i. How old was the mother when (CHILD) was born? . Age I I I i. How old was the mother when (CHILD) was born? Age Age ( 31) ( 32) j. Did (CHILD) have any birth defects? No (3.1) (32) j. Did (CHILD) have any birth defects? j. Did (CHILD) have any birth defects? Yes.(A9( (SKIP TO Q.L) -I (ASK Q.k) Yes.(49( -1 (ASK Q.k) No -1 (ASK Q.k) _-2 812039 NINTH CHI 1.11 T2{. Where are (CHILD)'s birth registration records located? In what city and state is j that? g. Where are (CHILD)'s . current medical records: located? In what city and state in that? Yes.(49( (K'H -2 (SKIP TO Q.L) No -2 (SKIP TO Q.L) k. What kind of birth defects did (s)he have? Any others? k. What kind of birth defects did (s)he have? Any others? k. What kind of birth defects did (s)he have? Any others? L. Was (CHILD) ever diag- j nosed as having cancer? i L. Was (CHILD) ever diag- : nosed as having cancer? | L. Was (CHILD) ever diagnosed as having cancer? Yes. (''"(-] (ASK Q.m) No iYes.(r''K -1 (ASK Q.m) -2 (SKIP TO Q.o) I No m. In what month and year ^ was the diagnosis made?) MONTH YKAK T—i—ri—i—r i (51.) i (52) i-i (53) i TV.)i _ n. What kind of cancer was diagnosed? Not sure..O7( (GO TO Q.o) m. . Yes.(r)()( -2 (SKIP TO O.o) In what month and year was the diagnosis made? MONTH YKAR -1 (ASK Q.m) No m. -2 (SKIP TO Q.o) In what month and year was the diagnosis made? MONTH YKAR T (.',]) f : >0 (53) (54) n. What kind of cancer was diagnosed? Not sure..('>/( (GO TO Q.o) (51) (52) . n. What kind of cancer was diagnosed? Not sure. .(r17 (GO TO Q.o) -1 �CARD SEVENTH CHILD 812039 ()2« EIGHTH CHILD NINTH CHILD 71o. (Does/Did)(CHILD) have diagnosed learning disability? 72o. <Does/Did)(CHILD) have a 73o. (Does/Did)(CHILD) have S diagnosed learning disdiagnosed learning disability? ability? Yes.(j«( -1 (ASK Q.p) Yes.<58( -1 (ASK Q.p) Yes.(58( No...,.i -2 (SKIP TO Q.q) No -2 (SKIT TO Q.q) No -1 (ASK Q.p) _-2 (SKIP TO Q.q) i p. What kind of learning disability (does/did) (s)he have? p- What kind of learning disability (does/did; (s)he have? p. What kind of learning disability (does/did) (s)he have? ' q. (Does/Did)(CHILD) have any physical, mental, or motor impairments? Yes.(39( -1 (ASK q. (Does/Did)(CHlLD) have I q. (Does/Did)(CHILD) have any physical, mental, or j any physical, mental, or motor impairments? j motor impairments? Yes.(59( Q.r) No r. What kind of impairment (does/did) (s)he have? CONTINUE SKIP TO NBXT CHILD -1 (ASK Q.r) |Yes.(59< -2 (SKIP TO Q.s) r. What kind of impairment (does/did) (s)he have? N" -1 (ASK Q.r) —— -- - (SKIP Th n.oi ' r. What kind of impairment (does/did) (s)he have? IF CHILD IS DEAD: IF CHILD IS DEAD: IF CHILD IS DEAD: OTHERWISE: OTHERWISE: OTHERWISE: s. On what date did (CHILD) die? CONTINUE SKIP TO NEXT CHILD 1 s. On what date did (CHILD) die? CONTINUE SKIP TO NEXT CHILD s. On what date did (CHILD) die? MONTH DAY YEAR MONTH DAY YEAR MONTH DAY YEAR 1 1 ! 1 1 1 1 1 1 1 1 1 1 1 1 1 I I 1 I I 1 1 l-l 1 l-l 1 I I 1 l-l 1 1 1 l-l 1 l-l 1 1 1 (id) <6|) <62) ( 63) ( 64> (6r>)(ftO) <oO (62) (63) (64) (65) (60) (6l) <62> (63> (64> <6i t. What was the cause of death? t. What was the cause of death? t. What was the cause of death? u. Where is (CHILD) 's death registered? In what city and state is that? RECORD IN S.R.B. fG 4. 1 u. Where is (CHILD) 's death registered? In what city and state is that? IRECORD IN S.R.B. PG A i u. Where is (CHILD) 's death registered? In what city and state is that?. . (GO TO NEXT CHILD Q.72a) (GO TO NEXT CHILD Q.73a) IRECORD IN S.R.B. PG.A T (GO TO NEXT CHILD Q.74a) dftu. 6Hu. (>Hf. (66-67) fi8f . llK "'. (>»p. O'f '70-80 _08 /74-77J . , 79-80 �CARD 74-76. <>2« 812039 Additional Children ELEVENTH CHILD TENTH CHILD NAME: NAME: 74a. How old is (CHILD) now? TWELFTH CHILD 75a. How old is (CHILD) now? 1 r I I .._ I (l >) (Tf>) i Child died..(I;( I I Age r I I I I Age r h i l d died..(I/( -1 POUNDS I L ! (21) ( 22) D o n ' t know...(_,M( Male... Female. c. How much did (CHILD) weigh at hirth? OUNCES I I I I I i_ J_ _l-l _ I -1 -I b. (Is/Was) (CHILD) male or female? Male (J Female... c. How much did (CHILD) weigh at birth? f I I I A g e (I >) ( I ft) r Child died..(I/( -1 b. (Is/Was) (CHILD) male or female? -1 Male... Female. POUNDS I I 76a. How old is (CHILD) now? (I'i) (I <S) b. (Is/Was) (CHILD) male or female? (20) NAME: OUNCES I I T l \ I I I l-l I I Tn) ( • > ( { ) (717 (22r -2 c. How much did (CHILD) weigh at birth? : i , POUNDS 1 J OUNCES I I I 1_ 1-J (19) (20) I I ( 2'l) ( 22) Don't know...( :'X D o n ' I know. . .(JJ(_ -1 I d. What is (CHILD)'s birth-; date? MONTH DAY YEAR d. What is (CHILD)'s birth-' date? MONTH DAY YEAR d. What is (CHILD)'s birthdate? j ' MONTH 1 DAY I T YEAR I I T I r (24) (25) ( 2(J ( 27) ( 28) ( 29) j ( 2-'.) ( 2 >) ( .If.) ( 27) C!H ) (29 ) (24) (21) (26) ( 27) (28) ( 29) RECORD IN S.R.B.-PG A I I j"Al.SQ^R'ECORb_IN" S.R.R.-PC 4| jlALSQ RECORD IN S.R.B.-PG A l e. Was the child premature, full term, or overdue? Premature. (_ilK Full term Overdue Not sure (GO TO Q.f) -1 -2 -3 -A e. Was the child premature,' e. Was the child premature, full term, or overdue? ! full term, or overdue? Premature. (___»<_ Full term _ Overdue _ Not surr (GO TO Q.f) 260 Premature. (_UX Full term Overdue Not sure (GO TO Q.f) -1 -2 -3 -A �812039 (I2R TENTH CHILD TWELFTH. CHILD ELKVKNTH CIULD 7/.f. Win-re a r c ( C H I L D ) 's 7 5 f . Where are ( C l J I L D ) ' s hi r t l i r o g i i i t rat i o n i h i r t l i ri'Ri ^ t r i l l i o n ri'ctmls l o c a t e d ? In j records located? In w h a t c i t y and s t a t e i s ! w h n t c i t y a n d s t a t e Is t h a t ?_ ! that? |RECO;Rp.lH..S.>ft.B. PC k \ '; IRECORD IN S.R.B. P G 4 f g. Where are ( C H I L D ) ' s c u r r e n t medical records located? In what c i t y artd s t a t e is that? I RECORD,IN S.T.B. PC: 4 f g. Where are (CHILD) 's current medical records! located? In what city I ---RECORD 11 _ S . R . B . PG It T • ".................• '- 76f. Wliere are (CHILD) 's birth registration records located? In what city and state is that? _ 1 RECORD I N S_. R . B . PC It I g. Where are (CHILD) 's current medical records located? In what city and state io that? h. What was (CHILD)'s m o t h e r ' s f u l l name? h. What was (CHILD) 's mother's . f u l l _ name? IRECORD IN S.R.B. PG 4 I h . What was (CHILD) 's m o t h e r ' s f u l l name? [RECORD I N S.R.B. PC |_ &_ I IRECORD IN S-R-B/PG u \ i. How old was the mother whnn (CHILD) was born? i. How old was the mother when (CHILD) was born? i . How old was the m o t h e r when ( C H I L D ) was h o r n ? J L.._ I Age ~(t'i) TuV No j. Did (CHILD) have any birth defects? Yes.(4g( (SKIP TO Q.L) -1 (ASK Q.k) Yes.(4?( -1 (ASK Q.k) No -1 (ASK Q.k) _-2 (32> j. Did (CHILD) have any birth defects? j. Did (CHILD) have any birth defects? Yes.(/it)( G It \ IJJECQfrP IN f -2 (SKIP TO Q.L) No -2 (SKIP TO Q.L) k. What kind of birth defects did (s)he have? Any others? k. What kind of birth defects did (s)he have? Any others? k. What kind of birth defects did (s)he have? Any others? L. Was (CHILD) ever diag- | nosed as having cancer? > L. Was (CHILD) ever diagnosed as having cancer? L. Was (CHILD) ever diagnosed as having cancer.? Ves.(^( -1 (ASK Q.m) ' Yes.(^p(._-l (ASK Q.m) No No -2 (SKIP TO Q.o) m. In what month and year was the diagnosis made?: 1 1 MONTH 1 1 I I l-l (51) (!>::) ( YEAR 1 1 1 1 53) ( 54) n. What kind of cancer was diagnosed? (55-56) Not sure. . '>?( ( (GO TO Q . o ) -1. Yes.(5Q( -? (SKIP TO Q.o) m. 1 .1 In what month and year was the diagnosis made?' YEAR MONTH 1 I I 1 l-l ( r il ) ( 51') ( 53) ( 54) I 1 ! I n. What kind of cancer was diagnosed? .• (55-56) Not sure. . ( ,';( (GO TO Q . o ) 201 -1 No -1 (ASK Q.m) -2 (SKIP TO Q.o) in. 1 1 In what month and year was the diagnosis made? MONTH YEAR 1 1 1 1 1 1 l-l 1 1 (51) ( 52) (53) (54• n. What kind of cancer wa-s diagnosed? (55- 56) Not sure. . ( ^?( (GO TO Q . o ) -1 �OJH CARD EI.KVKNTH C H I L D TENTH CHILI) 812039 TWELFTH CHILD 74o. (Does/Did)(CHILD) have a 75o. (Does/Did)(CHILD) have £ 76o. (Does/DidMCHILD) have a diagnosed learning disdiagnosed le.irning dindiagnosed learning disability? I ' abiIity? ability? Yes.(58( -1 (ASK Q . p ) Yes.(r,n( -1 (ASK Q.p) No. ._-2 (SKIP TO Q.q) No. ! Yes.(38( -2 (SKIP TO Q.q) No. -1 (ASK Q.p) . - (SKIP TO Q.q) _ 2 p. Whnt kind of learning disability (does/did) (s)he have? p. What kind of learning disability (does/did) (s)hc have? p. What kind of learning disability (does/did) (s)he have? q. (Does/Did)(CHILD) have any physical, mental, or motor impairments? q. (Does/Did)(CI!ILD) have any physical, mental, or motor impairments? q. (Does/Did)(CHILD) have any physical, mental, or motor impairments? Yes.(59 ( -1 (ASK Q.r) | Yes. (Jig Yes.(^q( -1 (ASK Q.r) -1 (ASK Q.r) i No -2 (SKIP TO Q.s) ~ No I r. What kind of impairment > (does/did) (s)he have? IF CHILD IS DEAD: CONTINUE OTHERWISE: SKIP TO NEXT CHILD s. On what date did (CHILD) die? MONTH DAY i i i: i 1 l-l -2 (SKIP TO Q.s) No ...... -2 (SKIP TO Q.s) i ; IF CHILD IS DEAD: OTHERWISE: : YEAR M i j r. What kind of impairment i (does/did) (s)he have? ! CONTINUE SKIP TO NEXT CHILD s. On what date did (CHILD) die? MONTH i 1 1 - 1 1 i i 1 l-l DAY i 1 i i IF CHILD IS DEAD: CONTINUE OTHERWISE: SKIP TO NEXT CHILD s. On what d a t e did (CHILD) die? . YEAR l-l r. What kind of impairment (does/did) (s)he have? MONTH i 1 1 1! i 1 i i l-l DAY i 1 i i YEAR l-l i 1 (ilj ( 61) ( 62) ( 63) ( 64) ( f>5) ; ( 60) ( 61) ( 63 ( 6'D ( 64) ( 6 ? j ( 60) ( 61) ( 6 2 ) ( 6 3 ) ( 64) ( 6 t. What was the cause of death? t. What was the cause of death? t. What was the cause of death? u . Wh.M-r i » ( c i l l l . D ) 's d e a t h registered? I n what c i t y and state is that? u . Wlu-n> i s ( C H I L D ) '« d.-..tli r i ! n i n t e r e d ? I n what c i t y and s t a t e is that? n . Win-re I H (.CHILD) 'f d e a t h registered? In what c i t y and s t a t e is that? RECORD IN S.R.B. PG 4 I IRECORD IN S.R.B. PG 4 [ IRECORD IN S.R.B. PG A I (GO TO NEXT CHILD Q . 7 5 a ) (GO TO NEXT CHILD Q . 7 6 a ) (RETURN TO Q . 2 7 a ) 68u. (66-67) 68u. (66-67) 68u. (66- 6 '71 "Hi- (68-M) hHf. (hH-6<J) (,«f (f)H-<,<)) 681,. (7u- / 0 (1811. _ ( 7 0 - 7 t) 6811. ( 7 0 - ; 1) 68p. (74-77) _ (//-//J ')-»(! J02 �CAJD._ .-~*-~"''l.?>' •..'.". «, ,_ 812039 ! « T ~ — 77. Additional pneumonia (Q. l») Fourth Time 77a. Fifth Tiiw During what months and yi'«r t. d id hi* have pneumonia (the fourth r i me ) ? 77f. .RECORD IN. S.R.B. PG 5 1 During what months, and . 77k. yearn <1 i<l In hafiv pneumonia dhe fifth time)? IRECORI) IN S.R.H. PG 5 1 IF BEFORE 1961 , SKIP TO Q.f. h. lylxch Time IRECORD IN S.R.B. PG 5 1 IF BKFORE 1961, SKII* TO IF BEFORE 1961, RETURN TO Q.29a. Q.k. What i s th<; f u l l name, of t h e d n c t o r who made ' t h f d i a g n o s i s or I IIP ', meilii'al f a c i l i t y whi-rc I h e d i a >MIO s i ,s Wcis nuidt- ? '• g. Wliat is th« full name of rhc doctor who made the diagnosis or thf ' medical . f a c i l i t y when! tin' diagnosis was made? ^ ^^ / ^ t IHE(X)KI) IN S.K.H. PU 3 f Wliat prescribed medicim> did he take for the pniMimonia he had that lime? I.. h. iRKCORI) IN S.K.B. PO 5.-1 m. ( ( ,2. ( .< 3. ( ( J3. ( ( Was ht> hospitalized . for clip pneumonia he _ had thai time? Yes.( _ (__-! (ASK Q.u) No ...... _^-'2 (SKIP TO Q.f) e. What w.is the f u l l nan of that hospital? Tm-xcfRD 'IN sTk.¥. pcTTl ( ( 13. ( ( Was he hospitalized for the pneumonia he had that time? ' Yes . ( No \>/hal was the f u l l name of t h a t h o s p i t a l ? i I N S . R . B . ,PG 5J J65 What prescribed medicine did he take for the pneumonia he had t h a t time? 2. Was lie hospitalized for the pneumonia he had that t ime? JYes.( ( -1 (ASKQ.J) JNo ...... __-2 (SKIPTOQ.k) What is the full name of the doctor who made thp diagnosis or the medical facility where thf diagnosis was made? S.R.B. PC TT What prescribed medicine did he take for the pneumonia he had that t in.e? 2. d. During what months and yearn ct id ho hnye pneumonia (the sixth t imp)? o. .- 1 (ASKfl.o) -2 I RETURN TO Q.29a) What was th'e f u l l name of t h a t h o s p i t a l ? 1RKCOBD IN S . R . B . PG �CARD 78. 817030 Additional Cancer (Q.29) Part t, Part Part 6 7tic. . I n what, month and year S i . In what m o n t h a n d year 78o. In what month and year was cancer of the ( B O D Y J was c a n c e r of the (BODY was cancer of the (BODY f.jxst. diajjjnos e d ? PART) f i r s t diagnosed? PART) first diagnosed? RECORD S . R . B . PC 6 I TRECORrijN _S.R.E.^ PC: 6J i I RECORD IN S.R.R. PC 6 I What i s tin 1 f u l l min»> of the d o c t o r or tinm e d i c a l f a c i l i t y where _ the d i a g n o s i s was ma d e ? I RECORD IN S .R_. B . PC I (1 . e. What is the f u l l name | of the doctor or f l i p mo d i c a 1 f a c i l i t y he '. l a s t c o n s u l t e d about i cancer of the (BODY | _ PART)? __ ! I RECORD IN S . R . H . PG ft I | f. D u r i n g what month and ! year di<3 he last con- ; suit (NAME FKOM Q. e ) ? ! [RECORD IN S . R . R . PC.; h \ g. j. Wlint is tlu> f u l l rinmp l of tin- doctor or the ! m e d i c . i l f a c i l i t y where I t h e d i a g n o s i s waa m a d e ? ; iRECORirTN s-R-B r ., Pf:.. 'Q' k. _.. PART)?_ IRECORD I N S . R . B . PC ft r I Other (SPECIFY) _ I RECORD L. During what month and year did he last consult (NAHK FRO_M Q.kj? j RECORD lj) S.R.B. PC b'T Radiation. (l5.( Chemotherapy.. .(|(,( Surgery (y (_ Other ( S P E C I F Y ) What is t lit1 t u l I nnmt' of the doctor or HIP medical L a c i l i t y where the diagnosis was made? RECORD IN S.K.b. "I'l; t, T Wliat i s t h e f u l l name of the d o c t o r or the m e d i c a l f n r i l i t y he l a s t c o n s u l t f d ,iho>it cancer of the (BODY Wliat is the full name of thr doctor or tinmpdic.'il facility hr last consulted about cancer of the (BODY PART_)_? 1 r. S .K._B_._. PC fr I DurinR what month and year did he last connult'._(HAME FROM Q.9>? I RECORD/IN' s.j^.BT^ PC & F What t r e a t m e n t s or , in. What treatments or medicines did he take medicines did he take ! for cancer of the (BODY for cancer of the (BODYPART)? ; PART)?_ . I [MULTIPLE RECORD BELOWf MULTIPLE RECORD BELOW I | Radiation......(15 ( Chemotherapy . . . (jj ]j Surgery ........ (17 ' P i. What treatments or medicines did he take for cancer ot the (BODY PART)? _ I MULTIPLE RRCORI) BKLOWl Radiation (L;.I.( Chemotherapy. .. (|ii ( -1 -1 -1 : Surgery i -1 ~1 Other (SPECIFY) (L/ ( -1 MONTH Radia1 tion. .. . 1 1 1 YEAR 1 1 1 l-l I J I ! I-J (21) C!H) ( MONTH Other I 1 I I 1 ! l-l l r i i 2<O (.»<() YEAR r 1 1 YEAR MONTH 1 Radia1 t i on .... I 1 I I 1 MONTH MONTH YEAR Chemo1 1 1 1 1 Chemo1 therapy . 1 1 l-l 1 1 t horapv . 1 (:• i) (;'/.) ( .",) (.'») YEAR MONTH MONTH I Surgery. . 1 During what month and year did he first receive (EACH TREATMENT CODED IN Q.s) for cancer of the (BODY PART)? Durinfi what month and year did he first receive (KACII TREATMENT CODED IN Q.m) for cancer of the (BODY PART)? During what month and year did he first receive (EACH TRF.ATMKNT CODED IN Q.g) for cancer of the (BODY PART)? l-l 1 1 1 1 YEAR" 1 l-l C ! 7 ) (I'M) ( ."*) ( 1 0 ) MONTH YEAR I I l-l 1 1 YEAR MONTH 1 M l 1 1 l-l Ml 1 1 1 Surgery.. 1 i 1 ( I I ) ( ( . ' ) < M) ( l / , ) (GO TO NEXT BODY PART) MONTH 1 I I 1 l-l YEAR MONTH Chnmo1 thi-rapv. 1 YEAR ! 1 i 1 1 Radia1 tion —r~T l-l i i SurgTy . . 1 1 1 1 Other. . . . 1 ( t i ) (•!;') ( n) (t.',) (GO TO NEXT BODY HART) ! YEAR 1 1 (<'l] (.' MONTH ! 1 l-l. 1 1 1 I O (:"i) (in) YEAR 1 (RETURN TO Q.lOa) 04 7'1-HO JO-J 1 1 M i l 1 1 I-! 1 1 ( II ) ( 1 •) (u) (i/,) 1 Other 1 1 �, "L " ——— II .....:• I I CARD "I I»H 'I »• " . ' . ..III 812039 ' Q.50b-51 Medical Providers — DEATH 50b. Name of Hospital Address City ^^ _____^___,________,,_____rt_____.________ ' . State 51. ^ • .^ • ^ , ., ., Address State _ -... ^ Zip Primary/Physician Name City ..-...,,- ._ t _ _ „ _• _ _.__., .• . L1.... . .L ._ r i .--. • - - -••- -• .... 205 ' .. „„,.....-. .; . ••_,.._.. • • .. ^ ' �DEPARTMENT OF THE AIR FORCE USAF SCHOOL OF AEROSPACE MEDICINE (AFSC! BROCKS AIR FORCE BASE. TEXAS 78235 The Air Force is conducting a very comprehensive health study of certain Air Force members who served our nation in the Vietnam conflict. The purpose of the study is to determine the potential adverse health effects resulting from the complex environment of Southeast Asia. Federal record systems identified your late as having been assigned in Southeast Asia. The collection of information concerning his health prior to his death is essential to the Air Force study. You are the best individual to give us the information we need. We ask that you help us and all Vietnam veterans by voluntarily participating in this major health study. Your participation will consist of an in-depth interview-in your home. The administration of the interview will begin in a few weeks under the direction of a nationally recognized health survey organization, Louis Harris and Associates, Inc. You will be contacted by phone or letter by them to arrange a convenient time for your interview which will take about two hours to complete. Our intent is to maintain all individual health data in the strictest confidence. In case outside parties attempt to gain access to the data, the Air Force and the Department of Justice are committed to protect this individual .confidentiality. This is one of the most important health studies undertaken by the Air Force, Your voluntary participation is critical to its success. The only way we can get clarification of the difficult questions being asked by the Vietnam veterans is through your cooperation and participation. Any questions that you may have concerning this effort can be answered by letter from the United States Air Force School of Aerospace Medicine, Epidemiology Division, Brooks AFB, Texas 78235, or a collect call to Area Code 512-536-3309. .Thank you. Sincerely GEORGE D. LATHROP, M.D., Ph.n, Colonel, USAF, MC Chief, Epidemiology Division 266 �Louis HARRIS AN.O ASSOCIATES, INC. >' • ,• Vf-.". ti ' "•• - 63O FIFTH .AVEN'l/e' NEW Y O R K , NEW Y O R K I Q I M TEL (2121 975-I6OO TELEX 148383 UOUtS HARRIS INTERNATIONAL. INC. OPINION fcESEAPCH CENTRE LOUIS HARRIS FRANCE 3O weUfSCCK ST. 21 OUt VIVIENNt LONDdN WIM 8* 76OO2 PARIS, rflANCE TEU1 OI-*8«-SI TEL. O I - 2 O O - » « S 1 TttEX! PRIVACY ACT STATEMENT - EPJDEMIOIOSIC STUDY AUTHORITY: Section 133, 1071-87, 3012, 5031 and 8012, Title 10, United States Code and Executive Order 9397. PRINCIPAL AND PURPQSEJS): The purpose of requesting personal information is to assist medical/technical personnel in developing records relative to your participation in an approved epidemiologic investigation. The Soda] Security Number (SSN) and Armed Forces Service Number (AFSN) §re necessary to Identify the person and records. ROUTINE USES: This information will be used to initiate, coordinate, and conduct the Investigation. It will be used to compile statistical data, but information allowing identification of the individual volunteer will not be included. Data and results from this investigation may be used to supplement other approved research studies coHduct^d at the USAF School of Aerospace Medicine or at other Federal agencies engaged in the conduct of similar studies. WHETHER DISCLOSURE IS MANDATORY OR VOLUNTARY, AND EFFECT ON INDIVIDUAL FOR NOT PROVIDING INFORMATION: Disclosure or requested information is voluntary. If the information is not furnished, acceptance as a subject ^s not possible. This is an all-inclusive Privacy Act Statement which will apply to all requests for personal information made by medical/technical personnel during the time you are a volunteer subject. A copy of this form will be placed in your investigation subject folder as evidence of this ratification. Your signature merely acknowledges that you have been advised of the foregoing. If requested, a copy of this form will be furnished to you. Signature of Volunteer SSN 267 Date �LIFE EVENTS CHART U.S. Air Force Survey YOUR AGE THEN SCHOOLS MILITARY EXPERIENCE OTHER JOBS MARRIAGE CHILDREN DEATH IN FAMILY MAJOR ILLNESS OTHER SPECIAL EVENTS 1930 31 . , 32 33 34 - „ 1935 36 37 38 _ 39 1940 41 . 42 '„ 43 44 1945 46 47 48 49 1950 51 52 ^j 54 195556 _ 57 _ 58 59 1960 61 62 63 „, . 64 1965 66 67 . 68 69 1970 t 7i 72 _ _ 73 74 1975 - 76 77 78 79 1980 81 » ~\ ^ rt r * v -----. ' ' ? " �SHOW CARD "B1 01 Aerospace 02 Aircraft 03 Agriculture 04 Automotive 05 Chemical 06 Electronic 07 Mining 08 Pest Control 09 Petroleum 10 Textile 11 None Apply 269 �SHOW CARD "E" STUDY NO. S120'59 \ 270 �LOUIS HARRIS AND ASSOCIATES. INC N £vV Y C P K . N £w Vffip* iO I II • NTCPNATiONAL , > ^ C -30IS T' -ABDIS FRANCE 3O w(CL8eCl\ ST »U£ V I V I E N N E -aooz P»(?IS. FBAMCE -EI. 01- zeo •«««•» TELEX: iooso1 r .ONOON W I M 8 A 8 ChfOLANC TCL. o : - < « o - 6 i S i TELEX zi UNITED STATES AIR FORCE H|ALTH STUDY Name of Medical Provider/Medical Facility ? '" Name of Place ' " •' Street Address State City Zip ) ( hone Number Dear Doctor or Administrator: As an authorized representative for .,,..... .1...: ....;-:,. I ani participating in a survey conducted for the lihited States Air Force to gather information on the health of current and fofffie> A'ir Force personnel. As part of this survey, medical providers who have delivered health care services to .^ . ^ are being asked to supplement information that I have already provided about him. ay this statement or a photocopy of it» I, .. hereby authorize and request you furnish to the United States Air Force Health Study any medical information in your records concerning health services received by: These services were provided during the'1 p'ertocT' tO .; . Thank you very much. Sincerely, Resp. Signature ;of Autnorized Representative FOR OFFICE USE ONLY: Full Name of Authorized Representative 271 Date M£OICAL PROVIDER POpM -- pRQYV �FOR OFFICE USE ONLY: LOUIS HARRIS AND ASSOCIATES, INC 630 Fi fth Avenue New York, New York 10111 Case # Hf 812039 Air Force Health Survey Respondent INTERVIEW,EVALUATION (NlERVIEWER: I WMP'fFJE THE FOLLOWfNGTN~PRTvATE IMMEDIA~TELY~AFTER THE"fNTERVIEW, UsTNG~| I YOURfiEST_JUpGMENT_TOANSWER F.ACH ITEM. _ J i .:acfi of respondent: Black.... Nonblar.k. 2a. Did the respondent want to terminate the interview before it was finished? No (SKIP TO Q3a) Yes _.HZ (ANSWER 2b AND 2c) 2b. -At what question number or during what question series? 'c l-'hat was the reason? la -*i?re there any (other) significant problems during '".he interview? No Yes (SKIP TO Q4a) ." (ANSWKR 3b) 3b. Describe the problems. 4a. Did respondent refer to records during the interview? No Yes _ (SKIP TO Q5a) (ANSWER 4b) 4b. What records did the respondent use? 5a. Was anyone else present at any time during the interview? No Yes (SKIP TO Q6) (ANSWER 5b and 5c) 5b. Who was present? | RECORD RELATIONSHIP | 5c. During which section(s)? 6. Length of interview: minutes 272 �lOiils H/vHKI'. AN!) A:r.ni:iAll : ->; iw:, t i . f i i I i i I li Avenue . Hew 'M,rl , ()i. w \'ork [ H i 1 1 ' ' ;.o . ; ; , • • ; • . S t u d y 1812039. . . . . . . . A I R K)k(.l III AI.III SIMVIY HAD.INO II<AM',,MI I fAl. IORM 111, Ui-,v fnrt 01 I i. >• ';- .'••- - I Dili" liitnis ((nil A s s o c i a t e s i KliM. . ... ' Interviewer U<i.ine; -. !'l,r'<lsc Print. . ' " 1 111! , (iji tii«|i.' (;nnlaii;. (lie I n] lowinij nidtorinl for Study "Sulj.joc't Respbndent NumtKjr W r i l i ' iii'N'DMIir.K of iMi:h ituni hoiiig sent o'n thi- line at the ricjht :.riiL'Y M!ii,ilCi l N J L k ¥ i ! W litiitJy I'uli.jt'ct Namo .A:'.:, ignnifn't- Sheet . Study ji(li.i'.'(M. Privitcy Act Statement (Signed) , . : ,.. •; ^ _ ..;... Study Sub.locf Questionnaire. Study Suliject Supplemental Recording Book. Study Subject Self Administered form w Study Subject. Medic<il Consent Form Study Subject Former W i te Consent Letter , Study Suliject: Interviewer Evaluation I'onn PRISfNT W I T K Privacy Act Statement. (Signed). Spouse line', t ionnairr Spouse 'iu|j|)K:meiitfil Uncording lioot-. Spouse MediiMl r.ori'.n'nt form Spousi; Interviewer [valuation Form. Former W i f e Name Assignment Sheet.. Privacy Art Statement (Signed) Spouse (Jues tionnai re Spouse Supplemental Recording Book. Spouse MediC'il Consent form Spouse Interviewer (.valuation Form. PROXY 1.NTI K V l l W Proxy Name Assignment. Sheet Privacy Act Statement ( S i g n e d ) . . . . Proxy Questionnaire'. Proxy Supplemental Recording Book. Proxy Medical Consent! Proxy Interviewer fvflluation Received: Jlate i het ki^cl id hv: 273 _^ �CHAPTER IV NON-COMPLIANT (MINI) QUESTIONNAIRE The following Non-compliant Questionnaire was used to collect baseline data for the Epidemiologtc Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicide Orange. This data was collected during 1981-1982. The Mini-questionnaire was used for individuals who refused the Study Subject Questionnaire (in person and telephonically). This instrument was administered in person, via telephone, and independently (mailed to study subject). The Non-compliant Questionnaire, as used in the field, fol1 ows. 274 �CONFIDENTAL O.M.B. NQ: 0701-003; APPROVAL EXPIRES: 11/30/82 UNITED STATES AIR FORCE STUDY NON-INTERVIEW HEALTH QUESTIONS CASE NUMBER 01Q2/45992A INTERVIEWER NAME: DATE OF NON-INTERVIEW HEALTH QUESTIONNAIRE: ' . «*' MONTH DAY YEAR 1. Compared to other people youh age would you say that your health is... Excellent,...........01 good, 02 fair, or . . . ; . , 0 ......3 poor? . . . . . . . . 4 .......0 2. Are you currently taking prescribed medicines for any Illness? Yes...... No .,.01 .....2 ....0 3. For what condition are you taking prescribed medicines? Any other conditions? 275 �4. Within the past three months, did illness or injury keep you from work, not counting work around the house? Yes.. ...01(A&B) No doesn't work..02 A. How many days did you miss from work within the past three months? Days B. What illness or conditions caused you to miss work? 5. Did you earn any income from any job during 1980? 01(A) Yes No.. . . 02 A. Was your income less than $20,000, $20,000 to $40,000 or more than $40,000? ' less than $20,000 $20,000 to $40,000 02 More than $40,000 6. .01 03 In order to obtain the most complete and useful information that we can, we are asking some participants to have a physical examination. The USAF will pay for all travel and per diem expenses sd that participants may go to a nationally recognized medical facility. (IF SEPARATED OR RETIRED FROM USAF, SAY: In addition, you will receive a $100.00 per day stipend.) The examination will take place over a five day period that you find convenient. If you are asked would you be willing to have a physical exam at a time most convenient for you? ..01 Yes 02(A) No A. What is your reason for not wanting to have the examination? 5 days too long from family...01 5 days too long from work 02 Don't want to travel 03 Other reason (SPECIFY) Thank you very much. ' 276 . � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 058 Folder The folder containing the original item. 1567 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. Patricia M. Moynahan Richard A. Albanese William H. Wolfe Description An account of the resource <strong>Corporate Author: </strong>United States Air Force School of Aerospace Medicine, Aerospace Medical Division (AFSC), Brooks AFB, Texas Date A point or period of time associated with an event in the lifecycle of the resource 1982-11-01 Title A name given to the resource Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Baseline Questionnaires Subject The topic of the resource Air Force Health Study ao_seriesIII