1
15
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.
167
Folder
The folder containing the original item.
5008
Series
The series number of the original item.
Series VII
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
Wolfe, William H.
George D. Lathrop
Date
A point or period of time associated with an event in the lifecycle of the resource
1981
Title
A name given to the resource
Typescript: A Medical Surveillance Program for Scientist Exposed to Dioxins and Furans, Abstract and Presentation, dated October 28, 198
-
https://www.nal.usda.gov/exhibits/speccoll/files/original/674d80358b71120fe56fe8b7e6fd771d.pdf
cefd7b3f6ad234c3f86c4eb1650eeeda
PDF Text
Text
Item ID Number
Author
01588
Wolfe
' William H.
Corporate Author
RODOrt/ArtlClO TitlO Typescript: Serum Dioxin Levels in Air Force Health
Study Participants, Preliminary Report
Journal/Book Title
Year
Month/Day
A ril 2
Color
^
Numberoflman.es
9
DoSCrlptOll NOtOS
See item 1589 for final
P ®
Wednesday, May 23, 2001
' published version.
Page 1589 of 1608
�APR 2 61988
SERUM DIOXIN LEVELS
IN AIR FORCE HEALTH STUDY PARTICIPANTS
PRELIMINARY REPORT
WILLIAM H. WOLFE, MD, MPH *
JOEL E. MICHALEK, PhD *
JUDSON C. MINER, DVM, MPH *
DONALD G. PATTERSON, PhD +
LARRY L. NEEDHAM, PhD +
JAMES L. PIRKLE, MD, PhD +
EPIDEMIOLOGY DIVISION
USAF SCHOOL OF AEROSPACE MEDICINE
HUMAN SYSTEMS DIVISION
AIR FORCE SYSTEMS COMMAND
BROOKS AFB, TX 78235
DIVISION OF ENVIRONMENTAL HEALTH
LABORATORY SCIENCE
CENTER FOR ENVIRONMENTAL HEALTH
CENTERS FOR DISEASE CONTROL
ATLANTA GA, 30333
This document will be submitted -for publication in the
Morbidity and Mortality Weekly Report.
�2 61S88
Serum Dioxin Levels in Air Force Health Study
Participants - Preliminary Report
In 1978 the United States Air Force responded to a Congressional mandate to
initiate an epidemiologic study of the possible health effects of exposure to
herbicides and their 2,3,7,8 - tetrachlorodibenzo-p-dioxin (referred to as
dioxin in this article) contaminants in Air Force veterans who served in the
Ranch Hand defoliation operation during the Vietnam conflict. Accordingly, the
Air Force developed a protocol for a nonconcurrent prospective study (the Air
Force Health Study) of all 1267 members of the Ranch Hand unit and a series of
matched Comparisons (1). The Comparisons were selected from the population of
Air Force veterans who served in air cargo units stationed in Southeast Asia
during the same period as the Ranch Hand unit and who were individually
matched to the Ranch Hands on date of birth, rank (officer, enlisted) and
occupation (pilot or copilot officer, officer navigator, nonflying officer,
flying enlisted, ground enlisted). The Comparisons were assumed to have not
been exposed to herbicides or dioxin during the war. The Ranch Hands and their
matched Comparisons have been physically examined in 1982 (2), 1985 (3) and
were recently examined in 1987-88. They will be examined again in 1992, 1997
and finally in the concluding year 2002.
During 1986, CDC's Division of Environmental Health Laboratory Sciences,
Center for Environmental Health and Injury Control, developed a method for
measuring dioxin in human serum (H). The measurement, which is based on lipid
weight, is highly correlated with paired measurements of dioxin in adipose
tissue (r-0.98) ( )
5.
The application of this measurement to members of the
CDC Agent Orange Exposure Study has been reported ( )
6.
�This serum dloxin measurement is the basis for the pilot study of Air Force
Health Study participants reported here.
This pilot study was designed to
assess dioxin half-life in Ranch Hand veterans and the validity of the Air
Force study design. Air Force Health Study participants Invited to participate
in this study were enlisU'ed ground crew who served during 1968-69, who held
the rank of Staff Sergeant or below, and, if a Ranch Hand, had a herbicide
loader or herbicide specialist occupation during their tour. The study goal
was to obtain a serum dioxin measurement
on 200 such participants, 150 Ranch
Hands and 50 Comparisons. Two hundred study subjects satisfying these criteria
were identified to CDC for a blood draw during February and March 1987.
Shortly after CDC began the scheduling, 61 additional subjects were identified
for possible scheduling to assure that the goal of 200 assays would be met.
Compliance to the draw was high with 90$ of those called consenting to visit
one of four regional Red Cross Centers. Two hundred and thirty-one of these
261 subjects were called; 207 of the 231 consented to participate in this
study. For a variety of reasons, seven of the 207 were unable to visit the
centers. Of the 200 for which serum was obtained, 150 were Ranch Hands and 50
were Comparisons. Of the 150 Ranch Hand specimens, 117 results were quantifiable while 19 quantifiable results were obtained from the 50 Comparison
speciments. Four of the specimens were indeterminant due to a failure to meet
one or more of the quality control criteria (H) and an insufficient quantity
of serum for a repeat analysis. The blood was drawn at each clinic according
to a standardized protocol. The numbers of Ranch Hands drawn at the 1 Red
Cross Centers namely, Atlanta, Cleveland, Los Angeles and Tulsa, were 43, 39,
�37 and 31; the corresponding counts Tor the Comparisons were 19t 7, 10 and 14.
Table 1 shows selected characteristics of the participants by group (Ranch
Hand, Comparison).
TABLE 1. Selected characteristics of Air Force Health Study participants in
the pilot study by group.
Ranch Hand
(n-150)
Comparison
(n-50)
Demographic
characteristics
Age (mean)
Race (black)
46
5%
49
6%
Mean tour length (months) in
Southeast Asia
12
25*
93%
15$
5%
8%
4*
4$
51 %
11
B0%
29
21 %
37%
12
16%
30
22%
Self reported herbicide
exposure**
Military
Leisure
Civilian occupation
Health characteristics
Current
Smoking
Current
Alcohol
Percent
Smoking (cigarette)
history (pack yrs)
Alcohol use
history (drink yrs)
body fat (mean)
Comparisons were based outside of Vietnam and had tours of 2 to 3 years.
**From questionnaire
The dioxin distributions of the 147 Ranch Hands and 49 Comparisons having
quantifiable assay results are shown in Figure 1. Thirty eight percent of the
assayed Ranch Hands and 98% of the assayed Comparisons had dioxin levels
(based on llpid weight) below 20 parts per trillion (ppt), which Is considered
the upper limit for residents of the United States without known dioxin exposure
�( ) Additionally, 21.1* of the Ranch Hands and 98% of the Comparisons had
7.
dioxin levels at or below 10 ppt. There Is a statistically significant difference between the two groups (x2 test, p<0.0001).
Figure 1. Serum dioxin levels of Ranch Hand and Comparison ground crew participating in the Air Force dioxin pilot study, 1987.
(INSERT FIGURE 1 HERE)
The single Comparison having a level of 21.3 ppt reported exposure to industrial chemicals since 1980 in a steel foundry in Indiana.
Summary statistics
of current dioxin levels by group are shown in Table 2.
Table 2. Summary statistics of current dioxin levels by group.
Ranch Hand
(n-147)
Dioxin level
Mean (ppt)
Median (ppt)
Mode (ppt)
Minimum (ppt)
Maximum (ppt)
Standard
deviation
.
Comparison
(n-49)
Comparison excluding
21.3 ppt level
(n-48)
49.4
26.5
9.8
3-2
5.2
4.7
2.4
2.0
1.8
4.6
2.4
2.0
313
21.3
9.7
56.0
3.0
1.8
The distribution of dioxin levels for Ranch Hands suggests that some Ranch
Hands and one Comparison have had unusually heavy dioxin exposure. The serum
dioxin measurement quantifies actual dioxin body burden and therefore provides
a direct assessment of exposure. No threshhold level has been determined as
yet for the health effects of dioxin in humans. A full report will be published
after dioxin measurements have been completed for all Air Force Health Study
participants and after the report has been reviewed by the Agent Orange Working Group
of the Domestic Policy Council (Executive Branch).
�Sera drawn in 1982 on thirty-six
of the Ranch Hands having 1987 dioxin levels
above 10 ppt were assayed to allow estimation of dioxin half life in humans as
approximately 7 years ( )
8.
These developments suggest that only about 2 to 1
dioxin half lives have elapsed since potential Ranch Hand exposure in Vietnam
and that serum dioxin can serve as a biological marker for previous dioxin
exposure in Air Force Health Study participants. The results of the half life
study will be reported in a separate publication.
A procedure for measuring dioxin in serum is available and these measured
levels can be directly correlated to those in adipose tissue.
However, the
assay is extremely expensive and the entire procedure, including sample procurement, should be done only under the laboratory protocol.
References
1. Lathrop GD, Wolfe WH, Albanese R and Moynahan P. A nonconcurrent prospective
study of possible health effects in Air Force veterans exposed to
herbicides-Protocol. National Technical Information Service (AD A 122 250)
1982.
2. Lathrop GD, Wolfe WH, Albanese R and Moynahan P. A nonconcurrent prospective
study of possible health effects in Air Force veterans exposed to herbicides.
Baseline report. National Technical Information Service (AD A 138 3^0) 1981.
3. Lathrop GD, Machado SG, Karrison TG, Grubbs WD, Thomas WF, Wolfe WH,
Michalek JE, Miner JC, and Peterson MR. (1987). The Air Force Health Study. An
epidemiologic Investigation of Health Effects in Air Force personnel following
exposure to herbicides. Final report. National Technical Information Services
(AD A 188 262) 1987.
4. Patterson DC Jr, Hampton L, Lapeza CR Jr, Belser WT, Green V, Alexander L
and Needham L. High-Resolution gas chromatographic/high-resolution mass
spectrometric analysis of human serum on a whole-weight and lipid basis for
2,3,7,8-tetrachlorodibenzo-p-dioxin. Anal Chem 1987;59:2000-5.
5. Patterson DC Jr, Needham LL, Pirkle JL, et al. Correlation between serum
and adipose tissue levels of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in 50 persons
from Missouri, Arch Environ Contam Toxicol 1988;17:139-^3.
6. CDC. Serum dioxin in Vietnam-era veterans - preliminary report.
1987; 36 ( 8 : 1470-5
2)
MMWR
�7. Byard JL. lexicological Significance of 2,3t7,8-tetrachlorodibenzo-p-dioxin
and related compounds in human adipose tissue. J. Toxicol. Environ. Health
1987;22:381-1<03.
8. Pirkle JL, Wolfe WH, Patterson DC Jr, et al. Estimates of the Half-life of
2,3.7,88-Tetrachlorodibenzo-p-dioxin in Ranch Hand veterans. Presented at
Dioxin 87, Las Vegas, NV, October 1-9, 1987.
�Serum dioxin levels of Ranch Hand ground crew
participating in the Air Force dioxin pilot study (n-147^
201 -r
19- /
18- /, r7\
171615- ', /,
14' / /
13- / /
12- / /
11- / /
10- / /
9' '/ '/
8' / /
76'
54
3
2
1
'/
P
E
R
C
E
N
T
^
/
//
/
/
/
/
/
/
/
-7
7
/ ', \
y^
'/
' / /*~/
^*~" s
X
/
s
'j
/
// /s
.
i— -
t
7^
X
/ / 7^ /
-
>
7^ /S
/
/
/
j--
,^._
u
7^
/ / / / XiTipn / Xpn / [7i
PI
7^
{7i /
.-^-^
pH
I/I
5 1 2 3 4 5 6 7 8 9 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3
5 5 5 5 5 5 5 5 5 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0
5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5
dioxin levels (Darts Der trillion)
�Serum dioxin levels of Comparison ground crew
participating in the Air Force dioxin pilot study (n=49)
1001
90
80*
70'
p
E
R
C
E
N
T
60'
50
40
30
20
10
0
5 1 2 3 4 5 6 7 8 9 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3
555555555012345678901234567890
555555555555555555555
dioxin levels (oaris oer trillion)
�
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.
1588
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
Wolfe, William H.
Joel E. Michalek
Judson C. Miner
Donald G. Patterson
Larry L. Needham
James L. Pirkle
Date
A point or period of time associated with an event in the lifecycle of the resource
April 26 1988
Title
A name given to the resource
Typescript: Serum Dioxin Levels in Air Force Health Study Participants, Preliminary Report
Subject
The topic of the resource
Air Force Health Study
CDC
ao_seriesIII
-
https://www.nal.usda.gov/exhibits/speccoll/files/original/f71efe00ef0e23a97920020525184974.pdf
eec375a4d3e2e149cf799c9da6235ae0
PDF Text
Text
Item ID Number
01533
Wolfe
> William H.
Corporate Author
RepOrt/ArtfClB Title
Air Force
Health Study: Results of the 1985 Morbidity
Study
Journal/Book Title
Year
1987
Month/Day
September 25
Color
n
Number of Images
22
Descrlpton Notes
sAM-FM-34-1
Wednesday, May 23, 2001
Page 1584 of 1608
�RESULTS OF TH
Y STUDY
BRIEFER: COLONEL WILLIAM H. WOLFE
SAM-FM-34-l
25
SEP 1987
�AIR FORCE HEALTH STUDY
PHASE 111 RESULTS
MAY 1985 - MARCH 1986
EXAMINATION OF 2309 PARTICIPANTS
• 93% OF ALL PHASE I PARTICIPANTS
- 1016. RANCH HANDS
- 1293 COMPARISONS
CONTRACTORS:
• SAIC, MCLEAN, VA
• SCRIPPS CLINIC, LA JOLLA, CA
• NORC, CHICAGO, IL
SAM-FM-34-2
25
SEP 1987
�CHARACTERISTICS OF THE PARTICIPANTS
RANCH HANDS
COMPARISONS
46.9 YRS
46.8 YRS
CURRENT SMOKERS*
40.1%
35.0%
ALCOHOL USE
SINCE 1982
88.7%
88.7%
COLLEGE EDUCATED
48.6%
49.3%
RETIRED MILITARY
54.4%
MEAN AGE
•
,
52.8%
s
I
* STATISTICALLY SIGNIFICANT
SAM-FM-34-3
25 SE?
1987
�PHYSICAL EXAMINATION FORMAT
• GENERAL EXAMINATION
• NEUROLOGICAL EXAMINATION
• DERMATOLOGICAL EXAMINATION
• DOPPLER EVALUATION OF PULSES
• CHEST X-RAY
• PSYCHOLOGICAL STUDIES
• 33 CLINICAL LABORATORY TESTS
• IMMUNOLOGICAL STUDIES
SAM-FM-34--S
25 SE? 108
�QUALITY ASSURANCE PROGRAM
PREDETERMINED STATISTICAL PLAN
STRICT LABORATORY QUALITY CONTROL
MARK-SENSE FORMS USED FOR DATA COLLECTION
•
EXAMINERS UNAWARE OF PARTICIPANT STATUS
ON-SITE MONITOR PRESENT THROUGHOUT
EXAMINATION PERIOD
SAM-FM-34-5
25 SEP 1987
�GENERAL HEALTH
• OVERALL GROUP SIMILARITY
• NO DIFFERENCES IN APPEARANCE OF ILLNESS OR AGE
• MORE RANCH HANDS PERCEIVED HEALTH AS
FAIR OR POOR (9.1% VERSUS 7.3%)
• PRIMARILY SEEN IN ENLISTED PARTICIPANTS
• IMPROVED PERCEPTION OF HEALTH FROM THAT IN 1982
• RANCH HANDS HAD LOWER MEAN % BODY FAT
• SEDIMENTATION RATE RESULTS MIXED
• MEANS EQUAL (5.05 MM/HR VERSUS 4.93 MM/HR)
• MORE "ABNORMALS" IN RANCH HANDS (5.8% VERSUS 3.6%)
SAM-FM-34-6
25 SEP 1987
�MALIGNANCY
• NO GROUP DIFFERENCES OVER INTERVAL PERIOD FOR SKIN OR
SYSTEMIC MALIGNANCY
• LIFETIME HISTORY OF BASAL CELL SKIN CANCER
• CONTINUES TO BE ELEVATED IN RANCH HANDS
• RELATIVE RISK = 1.56
•
• SYSTEMIC CANCER EQUIVALENT IN BOTH GROUPS
•
1 SOFT TISSUE SARCOMA IN EACH GROUP
• 1 LYMPHOMA IN EACH GROUP
• FINDINGS NOT DISTURBING BUT ARE CAUSE FOR
CONTINUED SURVEILLANCE
SAM-FM-3-5-7
25 SEP 1987
�NEUROLOGY
FEWER NUMBER OF ABNORMALITIES THAN AT BASELINE
AGE, ALCOHOL, DIABETES HAD SIGNIFICANT IMPACT
TESTS OF CRANIAL, PERIPHERAL AND CENTRAL NERVOUS
SYSTEMS ARE EQUIVALENT
BABINSKI REFLEX DIFFERENCES AT BASELINE NO LONGER SEEN
SAM-FM-34-8
25 SEP 1987
�PSYCHOLOGY
HISTORY OF PSYCHOLOGICAL ILLNESS EQUIVALENT
MMPI RESULTS:
• INCREASED DENIAL AND MASCULINITY/FEMININITY
SCALE ABNORMALITIES IN COMPARISONS
• MARGINAL INCREASES IN HYSTERIA AND INTROVERSION
SCALE ABNORMALITIES IN RANCH HANDS
MORE CORNELL MEDICAL INDEX ABNORMALITIES IN RANCH HANDS
(COMPLAINTS OF MEDICAL PROBLEMS)
;
EQUIVALENT RESULTS ON HALSTEAD-REITAN PERFORMANCE TESTS
SIGNIFICANT EFFECTS OF EDUCATION, AGE AND ALCOHOL
SAM-FM-34-9
25 SEP 1987
�GASTROINTESTINAL/HEPATIC
• EQUIVALENT HISTORY OF LIVER DISEASE AND ULCERS
• 11 LABORATORY ASSAYS OF HEPATIC FUNCTION
• INCREASED SGPT AND UROPORPHYRINS IN COMPARISONS
• INCREASED ALKALINE PHOSPHATASE IN RANCH HANDS
•
• BORDERLINE ELEVATION OF COPROPORPHYRIN IN RANCH HANDS
• NO EVIDENCE OF PORPHYRIA CUTANEA TARDA
• ABNORMALITIES EQUALLY DIVIDED IN THE TWO GROUPS <
SAM-FM-34-10
25 SEP 1987
�DERMATOLOGY
NO CHLORACNE DIAGNOSED
EQUIVALENT HISTORY OF SKIN DISEASE
NO GROUP DIFFERENCES IN DIAGNOSED SKIN DISORDERS
HEMATOLOGY
EQUIVALENCE IN 8 LABORATORY TESTS
/
AGE, RACE, OCCUPATION SMOKING HAD SIGNIFICANT EFFECT
SAM-FM-34-11
25 SEP 1987
�CARDIOVASCULAR
NO DIFFERENCES IN EGG OR BLOOD PRESSURE
PULSE DIFFERENCES SEEN AT BASELINE NO LONGER SEEN
INCREASED HISTORY OF HEART DISEASE IN RANCH HANDS
(24% VS 20%) NOT ACCOMPANIED BY OTHER ABNORMALITY
CHOLESTEROL AND TRIGLYCERIDE LEVELS EQUIVALENT
SAM-FM-34-12
25 SEP 1987
�RENAL
• NO DIFFERENCES IN HISTORY OF RENAL DISEASE
CHANGE NOTED FROM BASELINE
• EQUIVALENCE IN 5 LABORATORY TESTS
PULMONARY
NO DIFFERENCES SEEN EXCEPT IN INCREASED RALES IN
RANCH HANDS
NO PATTERN SUGGESTING DISEASE DIFFERENTIAL
SAM-FM-34-13
25
SE? 1987
�ENDOCRINOLOGY
NO DIFFERENCE IN HISTORY OF DISEASE
TSH AND TESTOSTERONE INCREASED IN RANCH HANDS
(O.7% VS 0.5% AND 1.3% VS 1.1%)
INCREASE IN IMPAIRED GLUCOSE TOLERANCE
IN COMPARISONS (10.9% VS 14.3%)
SAM-FM-34-14
25
Sir
1987
�IMMUNOLOGY
6 CELL MARKER AND 3 FUNCTIONAL TESTS PERFORMED
ON 47% OF PARTICIPANTS
GROUP EQUIVALENCE ON ALL TESTS
SKIN TEST DATA NOT ANALYZED DUE TO EXCESSIVE
INTER-READER VARIATION
QUALITY CONTROL ENHANCED FOR PHASE III
SAM-FM-34-15
25 SEP 1987
�LONGITUDINAL
ANALYSES
19 VARIABLES STUDIED
EACH PARTICIPANT ACTED AS HIS OWN "CONTROL"
GROUP CHANGES OVER TIME ASSESSED
5 SIGNIFICANT CHANGES
•
SEDIMENTATION RATE
,•
BABINSKI REFLEX "N
•
DEPRESSION (MMPI)
I
GROUP EQUIVALENCE
•
PLATELET COUNT
f
AT FOLLOWUP
PULSES
(METHOD CHANGE)
J
NO DETRIMENTS TO RANCH HANDS
SAM-FM-34-:6
SEP 19B"?
�EXPOSURE
INDEX ANALYSES
SPORADIC DIFFERENCES SEEN BUT NO CONSISTENT
PATTERN EMERGES IN ANY CLINICAL AREA
SAM-FM-34-17
25 SEP 1987
�CURRENT EXPOSURE INDEX
ESTIMATE BASED ON:
• GALLONS OF'HERBICIDE SPRAYED PER MONTH (1962-1971)
• LEVELS OF DIOXIN CONTAMINATION
• NUMBER OF MEN ASSIGNED EACH MONTH
A DIFFERENT INDEX WAS MADE FOR EACH OCCUPATION GROUP NO
METHOD TO RELATE ONE TO THE OTHER
BETTER ESTIMATE OF EXPOSURE NEEDED
SAM-FM-34-19 25 SEP 1987
�t
'.
„
USAF/CDC COOPERATIVEEFFORTS
PILOT STUDY OF 200 GROUNDCREW PERSONNEL
DIOXIN HALF-LIFE STUDY OF 100 PAIRS OF SERA (1982/1987)
LARGE SCALE TESTING OF AFHS PARTICIPANTS
SAM-FM-34-20 25 SEP 1987
�AFHS/CDC
SERUM DIOXIN ASSAY STUDIES
FIRST 76 ASSAY RESULTS
• 23 COMPARISONS
- MEAN:
5.3 PPT
- RANGE:
3 - 9 PPT
•
• 53 RANCH HANDS
- MEAN:
55.3 PPT
- RANGE:
3 - 3 1 4 PPT
SAM-FM-34-21 25 SEP 1987
�AFHS SERUM DIOXIN TESTING
DRAWING 450cc OF BLOOD AT SCRIPPS CLINIC FROM
ALL PARTICIPANTS
PERFORM DIOXIN ASSAYS AT CDC
SAM-FM-34-22
�CONCLUSIONS
SUBTLE BUT CONSISTENT NARROWING OF DIFFERENCES SEEN
AT BASELINE
REASSURING EVIDENCE THAT CURRENT HEALTH OF RANCH HANDS
IS NOT DIFFERENT THAN COMPARISONS
NO EVIDENCE TO IMPLICATE LINK BETWEEN HERBICIDE AND
ADVERSE HEALTH
CONTINUED SURVEILLANCE STILL REQUIRED
SAM-FM-34-18
25
StP
1 98 /
�
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.
1583
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
Wolfe, William H.
Date
A point or period of time associated with an event in the lifecycle of the resource
September 25 1987
Title
A name given to the resource
Air Force Health Study: Results of the 1985 Morbidity Study
Subject
The topic of the resource
Air Force Health Study
study protocol
morbidity trends
mortality trends
ao_seriesIII
-
https://www.nal.usda.gov/exhibits/speccoll/files/original/63c98e0182bd8fd489e404ac83d38177.pdf
09482fb66b304808fde6a87a07835e1c
PDF Text
Text
Item ID Number
01582
Author
Wolfe
CorpOPatB Author
United States Air Force School of Aerospace Medicine (
RBpOrt/APtlclB TltlB
' William H.
An
Epidemiologic Investigation of Health Effects in Air
Force Personnel Following Exposure to Herbicides:
Mortality Update-1986
Journal/Book Title
Year
1986
Month/Day
December 26
Color
D
Number oflmaoes
12
Descrlpton Notes
Wednesday, May 23, 2001
Page 1583 of 1608
�AIR FORCE HEALTH STUDY
(PROJECT RANCH HAND II)
AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH
EFFECTS IN AIR FORCE PERSONNEL FOLLOWING
EXPOSURE TO HERBICIDES
MORTALITY UPDATE - 1986
26 DECEMBER 1986
WILLIAM H. WOLFE, COLONEL, USAF, MC
JOEL E. MICHALEK, PhD
JUDSON C, MINER, LIEUTENANT COLONEL, BSC
MICHAEL R. PETERSON, LIEUTENANT COLONEL, BSC
Prepared for:
THE SURGEON GENERAL
UNITED STATES AIR FORCE
WASHINGTON, D.C. 20314
EPIDEMIOLOGY DIVISION
USAF SCHOOL OF AEROSPACE MEDICINE (AFSC)
BROOKS AIR FORCE BASE, TEXAS 78235
�UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE
REPORT DOCUMENTATION PAGE
1a. REPORT SECURITY CLASSIFICATION
1b. RESTRICTIVE MARKINGS
UNCLASSIFIED
2a. SECURITY CLASSIFICATION AUTHORITY
3 . DISTRIBUTION /AVAILABILITY OF REPORT
Approved for public release;
distribut ion is unlimited
2b. DECLASSIFICATION / DOWNGRADING SCHEDULE
5. MONITORING ORGANIZATION REPORT NUMBER(S)
4. PERFORMING ORGANIZATION REPORT NUMBER(S)
USAFSAM - TR - 86 - 43
6a. NAME OF PERFORMING ORGANIZATION
6b. OFFICE SYMBOL
USAF School of Aerospace
Medicine
(if applicable)
USAFSAM/EK
6c. ADDRESS (City, State, and ZIP Code)
7a. NAME OF MONITORING ORGANIZATION
The Surgeon General
7b. ADDRESS (City, State, and ZIP Code)
Aerospace Medical Divison (AFSC)
Brooks Air Force Base, Texas 78235-5301
Sa. NAME OF FUNDING /SPONSORING
ORGANIZATION USAF School Of
8b. OFFICE SYMBOL
(If applicable)
Aerospace Medicine
United States Air Force
Washington, D.C. 20314
9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
USAFSAM/EK
8c. ADDRESS (City, State, and ZIP Code)
Aerospace Medical Division (AFSC)
Brooks Air Force Base, Texas 78235-5301
i
10. SOURCE OF FUNDING NUMBERS
PROGRAM
PROJECT
TASK
ELEMENT NO.
NO.
NO.
65306F
2767
WORK UNIT
ACCESSION NO.
00
01
11. TITLE (Include Security Classification)
An Epidemic! ogic Investigation of Health Effects in Air Force Personnel
Exposure to Herbicides - Mortality Update 1986
12
Following
WoTf??AWl¥l1ain%., Michalek, Joel E. , Miner, Judson C., Peterson, Michael R.
13b. TIME COVERED
13a. .TYPE OE REPORT
Interim
FROM
1979
„ „„
TO 1985
14. DATE OF REPORT (Year, Month, Day)
1986
December
2o
Il5. PAGE .COUNT
j
Oy
16. SUPPLEMENTARY NOTATION
17.
COSATI CODES
FIELD
06
GROUP
05
18. SUBJECT TERMS (Continue on reverse if necessary and identify by block number)
SUB-GROUP
Epidemiologic Investigatiori; Air Force Health Study;
Matched Cohort Design; None:oncurrent Prospective Design;
Mortality Study
19. ABSTRACT (Continue on reverse if necessary and identify by block number)
\
The purpose of the Air Force Health Study is to determine whether those individuals involved
in the 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 is designed to evaluate both the mortality (death) and morbidity (disease) in these Individuals
over a 20-year period beginning in 1982.
The Baseline Mortality Report was released in June 1983, the Baseline Morbidity Report in
February 1984. Follow-up mortality reports were released in 1984 and 1985. This study has
so far failed to demonstrate health effects which can be conclusively attributed to herbicide or dioxin exposure.
This report contains summary statistics only for cumulative deaths occurring up to 31 December 1985. These data were not statistically analyzed since the 4 Ranch Hand and 27 Comparison deaths since the last report would not alter the results of the last analysis.
20. DISTRIBUTION /AVAILABILITY OF ABSTRACT
jp UNCLASSIFIED/UNLIMITED D SAME AS RPT.
22a. NAME OF RESPONSIBLE INDIVIDUAL
WILLIAM H. WOLFE, Col, USAF, MC
DD FORM 1473,84 MAR
21. ABSTRACT SECURITY CLASSIFICATION
D DTIC USERS
UNCLASSIFIED
22b. TELEPHONE (Include Area Code) 22c. OFFICE SYMBOL
(512) 536-2604
83 APR edition may be used until exhausted.
Alt other editions are obsolete.
USAFSAM/EK
SECURITY CLASSIFICATION OF THIS PAGE
UNCLASSIFIED
�UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE
As of 31 December 1985 59 Ranch Hand and 312 Comparison subjects had died.
The small number of deaths since the last analysis suggests that the findings of the last
report are still valid: there are no overall statistically significant differences between
the Ranch Handers and their matched comparisons as regards mortality.
Continued mortality surveillance is recommended, since these 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..
UNCLASSIFIED
�Air Force Health Study Mortality Update - 1986
EXECUTIVE SUMMARY
BACKGROUND
The purpose of the Air Force Health 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 is designed to
evaluate both the mortality (death) and morbidity (disease) in these individuals over a 20 year period beginning in 1982.
The Baseline Mortality Report was released in June 1983, the Baseline
Morbidity Report in February 1984. Follow-up mortality reports were released
in 1981 and 1985. This study has so far failed to demonstrate health effects
which can be conclusively attributed to herbicide or dioxin exposure.
The
reader is referred to the study reports for further details (1, 2, 3, 4).
RESULTS
This report contains summary statistics only for cumulative deaths occurring up to 31 December 1985. These data were not statistically analyzed since
the 4 Ranch Hand and 27 Comparison deaths since the last report would not
alter the results of the last analysis. Current percent dead by group, rank
and occupation are shown below.
Percent Dead
Rank
Officer
Ranch Hand
3.6
Comparison
4.7
Enlisted
5.3
5.3
Flying
4.0
5.5
Ground
5.4
4.6
4.7
5.1
Occupation
Total
The small number of deaths since the last analysis suggests that the
findings of the last report are still valid: there are no overall statistically significant differences between the Ranch Handers and their matched comparisons as regards mortality. These data will be subjected to extensive
statistical analysis in the next report, planned for 1987.
CONCLUSION AND RECOMMENDATION
Continued mortality surveillance is recommended, since these 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.
�Table of Contents
Page
Executive Summary
i
Tabl e of Contents
ii
1. Introduction
1
2. Data Summary and Interpretation
1
3. Conclusion
4
References
5
Principal Investigators
6
Appendix - Noncause-specific Statistical Summary
7
ii
�Air Force Health Study Mortality Update - 1986
1.
Introduction
This report updates the findings of the last mortality report (4) released
on 29 November 1985. This report contains summary statistics only, without
statistical analysis, since so few deaths occurred during 1985. Noncausespecific findings, taken from the last mortality report, are summarized in the
Appendix. A complete analyses of these data will be presented in the 1987
report. The reader is referred to the baseline mortality report (1), released
on 30 June 1983, for information regarding the study design and mortality
determination process.
2.
Data Summary and Interpretation
Summary counts of the population at risk and the number of deaths in each
of the two groups (Ranch Hand and Comparison) stratified by rank and occupation are shown in Table 1. Table 2 contains the counts of new deaths in the
populations since the last report. Table 3 contains summary counts and death
rates by occupation, race and group. All tables in this report correspond to
similar tables in the last annual report. These counts reflect cumulative
mortality as of 31 December 1985 (certified as of 15 April 1986).
Table 1
Summary Counts of Death by Rank and Occupation
Ranch Hand
Rank
Officer
Enlisted
Occupation
Flying
Ground
Total
At risk
Rate (%}
0.036 (3.6)
0.053 (5.3)
At risk
2278
3893
Comparison
Dead
Rate (*)
106 0.047 (4.7)
206 0.053 (5.3)
646
611
26
33
0.040 ( 4 . 0 )
0.054 (5.4)
3163
3008
173
139
0.055 ( 5 . 5 )
0.046 ( 4 . 6 )
1257
59
0.047 ( 4 . 7 )
6171
312
0.051 ( 5 . 1 )
In Table 2, the number "At risk" is the number alive on 1 January 1985.
Table 2
Deaths During 1985 by Rank and Occupation
Ranch Hand
Rank
Officer
Enlisted
Occupation
Flying
Ground
622
580
Total
1202
Comparison
Dead
Rate per 100
8
0.4
19
0.5
Rate per 100
0.2
0.4
At risk
2
2
At risk
2180
3706
0.3
0.3
3002
2884
12
15
0.4
0.5
0.3
5886
27
0.5
1.
�Table 3
Occupational and Race-Specific Cumulative Mortality
Ranch Hand
Race
Non-Black
Occupation
At
Officer-Pilot
Officer-Nav.
Officer-Other
Enlisted-Fit Eng.
Enlisted-Other
Black
Officer-Pilot
Officer-Nav.
Officer-Other
Enlisted-Fit Eng.
Enlisted-Other
risk
350
82
25
191
533
Dead
12
4
1
7
30
Comparison
Rate
per 100
3.4
4.9
4.0
3.7
5. .6
At risk
1740
390
123
935
2628
Rate
Dead per 100
~~56
4.9
4.1
16
4
3.3
61
6.5
4.7
123
0
0
0
3
2
0.0
0.0
0.0
20.0
3.8
13
10
2
75
255
0
0
0
10
12
13.3
4.7
1257
Total
6
2
1
15
52
59
4.7
6171
312
5.1
0.0
0.0
0.0
Summary counts by rank, occupation and group are shown in Table 4.
Table 4
Summary Counts by Rank, Occupation and Group
Officer
Group
Ranch Hand
Comparison
Flying Personnel
Enlisted
Total
At
risk
646
3163
Rate
Dead per 100
440
16
3.6
102
4.7
2153
At
risk
Dead
26
173
Rate
per 100
4.0
5.5
Ground Personnel
Enlisted
Total
At
Rate
At
Rate
Rate
per 100 risk Dead per 100 risk Dead per 100
3.8
32
611
33
5.4
585
5.5
3.2
3008 139
4.6
2883
135
4.7
Officer
Group
Ranch Hand
Comparison
At
risk
26
125
Dead
1
4
The death rates in Tables 1 through 4 are very similar to the corresponding rates in the 1985 report. For example, the overall Ranch Hand ground
personnel rate of 5.4/5, shown in Table 4, has increased from 5.1$, shown in
the last report, due to 2 Ranch Hand ground deaths during 1985. Although the
overall Ranch Hand and Comparison death rates of 4.7 and 5.1 percent, shown in
Table 1, are very comparable, Table 4 suggests that group contrasts change
with rank and job. That is, the Ranch Hand flying officer death rate (3.6) is
smaller than that of their matched comparisons (4.7), while the situation is
reversed for enlisted ground personnel with the Ranch Hand rate (5.5) being
greater than that of the Comparisons (4.7). The statistical significance of
this reversal will be assessed in the 1987 report.
2.
�Survival curve estimates for all Ranch Handera and all comparisons are
shown in Figure 1.
Figure 1
Survival Curve Estimates for All Ranch Handers and All Comparisons
0.9
0.9
2 0-7
>
•> 0.6L
0.5 -
a
cc
0.4 _
0.3 _
O.I _
10
30
40
so
70
flGE
Deaths by cause, group, rank and occupation are shown in Table 5. This corresponds to Table 16 in the 1985 report.
3.
�Table 5
Deaths by Cause, Group, Rank and Occupation
Officer
Accident
Suicide
Homicide
Parasitic Infection
Malignant Neoplasm
Uncertain Neoplasm
Endocrine
Mental Disorder
Nervous System
Circulatory System
Respiratory System
Digestive System
Genitourinary System
Congenital Abnormalities
111 Defined
Total
Flying
RH
c
8
33
0
5
0
0
2
0
1
18
1
0
1
0
0
0
0
1
32
5
0
2
2
5
1
0
0
0
1
0
102
rr
Enlisted
Ground
RH
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
Flying
c
RH
1
27
1
3
0
2
0
0
1
11
0
0
1
0
0
0
0
0
1
15
0
3
1
2
0
0
1
0
1
1
10
71
C
1
1
0
0
1
0
0
0
0
0
0
1
0
0
0
i
Ground
RH
8
1
2
0
5
0
1
0
0
13
0
2
0
0
0
32
C
36
10
3
2
25
1
0
1
1
17
2
5
2
0
0
135
Total
RH
C_
20
97
19
3
2
5
1
0
58
7
0
2
1
2
1
0
0
2
19
91
0
7
6
15
0
3
1
0
1
_2
312
59
Deaths during 1985 by cause are shown in Table 6.
Table 6
Deaths During 1985 by Cause
Cause
Accident
Suicide
Malignant Neoplasm
Endocrine
Circulatory System
Digestive
Congenital Anomalies
Total
Ranch Hand
1
0
1
0
1
1
0
Comparison
1
1
7
1
11
2
J_
27
3. Conclusion
No statistical inference can be drawn from these summary counts since
these data were not analyzed, although the low counts in Table 2 suggest that
the conclusions of the last report are still valid. These data will be subjected to regression analyses of the hazard ratio and of the SMR in the 1987
mortality report.
4.
�References
1.
Lathrop, G. D., Moynahan, P. M., Albanese, R. A., Wolfe, W. H. (1983). An
Epidemiologic Investigation of Health Effects in Air Force Personnel
Following Exposure to Herbicides: Baseline Mortality Study Results.
(NTIS Order Number: AD-A130 793)
2. 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-A130 340)
3. Wolfe, W. H., Michalek, J. E., Albanese, R. A., Lathrop, G. D., Moynahan,
P. M. (1984). An Epidemiologic Investigation of Health Effects in Air
Force Personnel Following Exposure to Herbicides: Mortality Update 1984. (NTIS Order Number: AD-A162 68?)
4. Wolfe, W. H., and Michalek, J. E. (1985). An Epidemiologic Investigation
of Health Effects in Air Force Personnel Following Exposure to Herbicides: Mortality Update - 1985. (NTIS Order Number: AD-A163 237)
5.
�PRINCIPAL INVESTIGATORS
William H. Wolfe, MD, MPH, FACPM
Colonel, USAF, MC
Chief, Epidemiology Division
Joel E. Michalek, PhD, GM-14
Chief, Biometrics Branch
Epidemiology Division
Judson C. Miner, DVM, MPH, ACVPM
Lt Col, USAF, BSC
Chief, Epidemiology Services Branch
Epidemiology Division
Michael R. Peterson, DVM, MPH, Dr.PH.
Lt Col, USAF, BSC
Chief, Occupational Epidemiology Section
Epidemiology Division
6.
�APPENDIX
Noncause-Specific Statistical Summary
1985 Mortality Update
Age at Death
Group
Officer
Enlisted
Flying
Ground
Total
Logrank
Value
P-Value
-0.835
0.187
-1.34
0.976
0.10
0.85
0.18
0.33
0.305
0.76
Deaths to Date
RR
P-Value
SMR
P-Value
Officer
Enlisted
Flying
Ground
0.715
0.987
0.692
1.21
0.26
0.91
0.12
0.35
0.791
1.03
0.726
1.23
0.37
0.89
0.13
0.33
Total
0.915
0.57
0.954
0.73
7.
�
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.
1582
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
Wolfe, William H.
Joel E. Michalek
Judson C. Miner
Michael R. Peterson
Description
An account of the resource
<strong>Corporate Author: </strong>United States Air Force School of Aerospace Medicine (AFSC), Epidemiology Division, Brooks AFB, Texas
Date
A point or period of time associated with an event in the lifecycle of the resource
December 26 1986
Title
A name given to the resource
An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Mortality Update - 1986
Subject
The topic of the resource
Air Force Health Study
ao_seriesIII
-
https://www.nal.usda.gov/exhibits/speccoll/files/original/83b69d30063b533807899b556076f45a.pdf
9c2fde559fa4c367f1167180d4bafbe0
PDF Text
Text
Item ID Number
°1581
Wolfe, William H.
Corporate Author
Report/Article Title
Air Force Health stud
V (Ranch Hand II) Update Briefing
Journal/Book Title
Year
1986
Month/Day
September 29
Color
Number of Images
n
27
Desoripton Notes
Wednesday, May 23,2001
Page 1582 of 1608
�AfRjl=QRCE HEALTH STUDY
HAND II)
UPDATE BRIEFING
29 SEPTEMBER 1986
BRIEFER;: COLONEL WILLIAM H. WOLFE
I
�STUDY DEVELOPMENT
19-78-1979
PROPOSAL DEVELOPED
1979
PEER REVIEW AND REFINEMENT
19*79-1982
STUDY POPULATIONS IDENTIFIED
QUESTIONNAIRES DEVELOPED
1982
PHYSICAL EXAMINATIONS ACCOMPLISHED
1983-PRESENT
MORTALITY REPORTS RELEASED
1984
BASELINE MORBIDITY REPORT RELEASED
1985
FIRST FOLLOWUP PHASE BEGUN
�STUDY DESIGN
IDENTIFY EXPOSED POPULATION
IDENTIFY AND SELECT COMPARISON POPULATION
DETERMINE BASELINE HEALTH STATUS
MORTALITY
MORBIDITY
COMPARE FINDINGS STATISTICALLY TO DELINEATE
POSSIBLE HERBICIDE
EFFECTS
ACCOMPLISH FOLLOWUP STUDIES OF
POPULATIONS
�MATCHING PROCESS
8 COMPARISONS SELECTED FOR EFlCH EXPOSED
INDIVIDUAL
MATCHED FOR SEX. AGE. DUTY CODE. RACE
RANDOM COMPARISONS FROM EACH MATCHED SET USED
5:1
IN FIRST FOUR MORTALITY REPORTS
ALL COMPARISONS AFTER
1:1
1386
IN BASELINE AND SUBSEQUENT
MORBIDITY
OTHERS AVAILABLE FOR REPLACEMENT
REPORTS
�SUMMARY OF MORTALITY RESULTS
CA/O 31 DEC 84)
EVALUATION OF 55 EXPOSED AND 285 COMPARISON DEATHS
RELATIVELY SMALL NUMBERS EMPHASIZE PRELIMINARY
NATURE OF THESE RESULTS
MORTALITY EXPERIENCE NEARLY IDENTICAL IN BOTH GROUPS
CAUSE-SPECIFIC ANALYSES NOT STATISTICALLY DIFFERENT
�ADDITIONAL MORTALITY CONTRASTS
DOD RETIRED POPULATION
ALL RANK-EXPOSURE
GROUPS DOING BETTER
EXPOSED-ENLISTED GROUP NOT STATISTICALLY BETTER
ACTIVE C I V I L SERVICE PERSONNEL
ALL SUBGROUPS
EQUIVALENT TO C I V I L SERVICE
1978 U S WHITE MALES
ALL SUBGROUPS DOING SIGNIFICANTLY BETTER
�I -"-^
SUMMARY COUNTS OF DEATHS
RANCH HAND
AT RISK
DEAD
PERCENT
OFFICERS
46
6
16
3 4
.
ENLISTED
791
3
9
FLYING
66
4
GROUND
COMPAR I SON
AT RISK
DEAD
PERCENT
2 7
2 8
9
8
4 3
.
4 9
.
3 9
S 3
187
4 8
.
24
3 7
.
3163
161
5. 1
611
31
5. 1
3OO8
124
4. 1
1 5
2 7
55
4 4
.
6171
2 5
8
4 6
.
RANK
OCCUPATION
TOTAL
�DEATHS BY CRUSE
r
EXPOSED
•
COMPARISONS
19
9
6
SUICIDE/HOMICIDE
5
23
MALIGNANCY
6
51
CIRCULATORY
18
80
RESPIRATORY
O
DIGESTIVE
5
13
OTHER
2
15
TOTAL
5
5
2 5
8
ACCIDENTAL
7
1
�BASELINE MORBIDITY RESULTS
r
NO DEFINITIVE CLINICAL ENDPOINTS CONCLUSIVELY
ATTRIBUTABLE TO HERBICIDE EXPOSURE
NO STS, PCT, OR CHLORACNE
IN THE EXPOSED GROUP
SOME CLINICAL AND SUBCLINICAL DIFFERENCES
OBSERVED
RELEVANACE OF SOME DEPENDENT ON ADDITIONAL
COVARIATE DATA
MOST VALUES WITHIN NORMAL RANGES
SCHEDULED
FOLLOWUP EXAMINATIONS WILL PROVIDE DATA
NECESSARY TO DEFINE RELEVANCE
�SIGNIFICANT GROUP DIFFERENCES
SELF-PERCEPTION OF HEALTH
SKIN CANCER
REPORTED BIRTH DEFECTS AND NEONATAL DEATHS
BABINSKI
REFLEXES
SUBJECTIVE PSYCHOLOGICAL MEASURES
HEPATIC FUNCTION TESTS (GGTP, LDH. CHOLESTEROL)
PERIPHERAL PULSES
THYROID AND TESTOSTERONE
�PHASE II EXAMINATIONS
PILOT EXAMINATIONS IN APRIL 85
1O EXAMINEES
INITIAL EXAMINATIONS BEGAN MAY 85
2 GROUPS PER WEEK
9 GROUPS LOST DUE TO HOLIDAYS
2 0 PARTICIPANTS
3 9
COMPLETED MARCH 86
�PROGRAM MODIFICATIONS FROM BASELINE
INTERVAL HISTORY (SUBJECT AND SPOUSE)
PHONE INTERVIEW WITH ENTIRE COMPARISON GROUP
BASELINE QUESTIONNAIRES TO NEW SUBJECTS AND SPOUSES
DELETIONS:
PULMONARY FUNCTION STUDIES
NERVE CONDUCTION STUDIES
SEMEN STUDIES
IQ TESTING
�PROGRAM
ENHANCEMENTS
SKIN CANCER EVALUATION
IMPROVED ALCOHOL AND SMOKING
ASSESSMENT
COMBAT STRESS ASSESSMENT
BIRTH DEFECT SEVERITY ASSESSMENT
DOPPLER EXAMINATION OF PULSES
IMPROVED
IMMUNOLOGIC ASSESSMENT
PORPHYRIN PROFILE BY HPLC
�EXAMINATION QUALITY CONTROL.
OBSERVATION
BY MONITOR WITH CHECKLIST
MARK-SENSE EXAMINATION FORMS
MANUAL AND ADP QC REVIEWS
STRICT LABORATORY
QC
TIGHT CV%
FIRCUSUM
NO PRIOR STANDARDS FOR
IMMUNOLOGY QC
PATIENT CRITIQUE FORMS
BLINDNESS OF EXAMINERS
�DAILY SCHEDULE
O53O
START 12-HOUR URINE COLLECTION
O63O
DEPART FOR CLINIC
07OOO8OO
INITIAL BLOOD DRAW
O800O9OO
SKIN
O8OO150O
EXAMINATION
TESTING
PSYCHOLOGICAL TESTING
INTERVIEW
1730
TURN IN URINE COLLECTION
�RESTRICTIONS
3 DAY CARBOHYDRATE LOADING DIET
3 DAY ABSTINENCE FROM ALCOHOL
NO ALCOHOL
FOR 2 DAYS ONSITE
FASTING AFTER MIDNIGHT
NO CAFFIENE OR NICOTINE AFTER MIDNIGHT
RESTRICTED UNTIL AFTER ECG AND DOPPLER
�PARTICIPATION
RANCH HANDERS
ORIGINALS
REPLACEMENTS
PARTIAL
39
62
32
REFUSAL
6
9
9
NEW
-
3
39
GAINS
45
74
80
LOSSES
74
64
21
971
872
267
PHASE II
TOTAL
1016
946
3-47
BASELINE
TOTAL
1045
936
288
RETURNEES
�IMMUNOLOGY STUDIES
RANDOM SELECTION OF 1082 PARTICIPANTS
B AND T CELL STUDIES PERFORMED
SKIN TESTS TO 1768 ( * %
77)
541 DELETED TO AVOID CONFOUNDING
�SPOUSE QUESTIONNAIRES
INTERVAL HISTORY
1915 COMPLETED
979 ONSITE
936 MAIL/PHONE
BASELINE GIVEN TO 7-4
FERTILITY REVIEW
CONTACTED
2108 SPOUSES
1613 RESPONDED
(76.5%)
�TELEPHONE
r
QUESTIONNAIRES
TOTAL SUBJECTS
7.963
INTERVIEWS COMPLETE
7. -411
( 3 1 )
9 . %
REFUSALS
335
< 4.2%)
UNLOCATABLES
190
( 2.4%)
26
( O.3%)
DECEASED
�THIRD PHYSICAL EXAMINATION
SCRIPPS CLINIC
LA JOLLA, CA
MAY 1 ^Q'?
�ANTICIPATED EXAMINATION ADDITIONS
COMPLIANCE
ENHANCEMENT
TONOMETRY
VISUAL ACUITY
AUDIOGRAMS
STOOL FOR OCCULT BLOOD
< PROCTOSCOPIC FOLLOWUP)
MEDICAL CONTACT CARDS
�ANTICIPATED EXAMINATION ADDITIONS
SCIENTIFIC
ISSUES
PULMONARY FUNCTION
SCLR-90
STUDIES
PSYCHOLOGICAL. BATTERY
MILLON CLINICAL SCALE
AUTOMATED BLOOD PRESSURE MONITOR
PROTEIN
PROFILE
PHENOTYPE MARKER FOR ACTIVATED T CELLS
OPTIMIZE NATURAL KILLER
WITH INTERLUKIN-2
CELL ASSAY
�ANTICIPATED EXAMINATION DELETIONS
MMPI
CORNELL
INDEX
POKEWEED ANTIGEN
HALSTEAD-REITAN
FROZEN MLC POOL
TIMED URINE COLLECTION
PERIPHERAL DOPPLER
PORPHYRIN
PROFILES
CORTISOL DETERMINATIONS
�ADDITIONAL CONSIDERATIONS
CARDIAC STRESS TESTING
PERSONALLY
FUNDED IF DESIRED
D-GLUCARIC ACID ASSAY OF BANKED URINE SPECIMENS
SERUM DIOXIN ASSAY OF SUBSET OF EXPOSED GROUP
�MILESTONES
JUNE 1386
FALL 1936
MAY
1987
FALL 1 8
9 7
FALL 1 8
9 7
FALL. 1 8
9 9
ANALYSIS BEGUN
1986
MORTALITY UPDATE
THIRd EXAMINATION TO BEGIN
RELEASE OF SECOND MORBIDITY REPORT
MORTAL I TY REPORT
RELEASE OF THIRD MORBIDITY REPORT
�
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.
1581
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
Wolfe, William H.
Date
A point or period of time associated with an event in the lifecycle of the resource
September 29 1986
Title
A name given to the resource
Air Force Health Study (Ranch Hand II) Update Briefing
Subject
The topic of the resource
Air Force Health Study
study protocol
morbidity trends
mortality trends
ao_seriesIII
-
https://www.nal.usda.gov/exhibits/speccoll/files/original/5c524e91822588d283b6b10233e0216f.pdf
55d2f6a948282ec28f629df56b9f3f7e
PDF Text
Text
Item ID Number
0158
°
Wolfe
. William H.
Corporate Author
United States Air Force School of Aerospace Medicine i
Report/Article TitlU An Epidemiologic Investigation of Health Effects in Air
Force Personnel Following Exposure to Herbicides:
Mortality Update-1985
Journal/Book Title
Year
1985
Month/Day
November 29
Color
' '
Number of linages
49
Descriptor! Notes
Wednesday, May 23, 2001
Page 1581 of 1608
�PROJECT RANCH HAND II
AN EPIDEMIOLOG1C INVESTIGATION OF HEALTH
EFFECTS IN AIR FORCE PERSONNEL FOLLOWING
EXPOSURE TO HERBICIDES
MORTALITY UPDATE - 1985
29 NOVEMBER 1985
WILLIAM H. WOLFE, COLONEL, USAF, MC
JOEL E.MICHALEK, PhD
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
�UNCLASSIFIED
SKURlTY CLASSIFICATION OF THIS PAGE
REPORT DOCUMENTATION PAGE
0
]
^BHHrW *^™
1b. RESTRICTIVE MARKINGS
2«. SECURITY CLASSIFICATION AUTHORITY
3. DISTRIBUTION /AVAILABILITY OF REPORT
I
Approved for public release;
distribution is unlimited
2b. DECLASSIFICATION/ DOWNGRADING SCHEDULE
5. MONITORING ORGANIZATION REPORT NUMBER(S)
4. PERFORMING ORGANIZATION REPORT NUMBER(S)
t"
6*. NAME OF PERFORMING ORGANIZATION
6b. OFFICE SYMBOL
USAF School of Aerospace
Medicine
(if appWcaote)
USAFSAM/EK
6c ADDRESS (Ctty. S«a«e, and ZIP Code)
Aerospace Medical Division (AFSC)
Brooks Air Force Base, Texas 78235-5301
8a. NAME OF FUNDING /SPONSORING
ORGANIZATION
8b. OFFICE SYMBOL
(If tpplktble)
Aerospace Medicine
7«, NAME OF MONITORING ORGANIZATION
The Surgeon General
7b. ADDRESS (Cfty, Sttte, *nd ZIP Code)
United States Air Force
Washington, D.C. 2031 1
9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
USAFSAM/EK
8c ADDRESS (City, State, and ZIP CoeW
10. SOURCE OF FUNDING NUMBERS
PROGRAM
PROJECT
TASK**
ELEMENT NO.
NO.
NO.
WORK UNIT
ACCESSION NO.
Aerospace Medical Division (AFSC)
Brooks Air Force Base, Texas 78235-5301
65306F
2767
00
01
11. TITLE <lte/«fc Securty Ctotttffcattonj An Epldemlologic Investigation of Health Effects In Air Force
Personnel Following Exposure to Herbicides - Mortality Update 1985
Wolfe, William H., Michalek, Joel E.
12. PERSONAL AUTHOR(S)
{
I3b, TIME COVERED
FROM 1Q7Q
TO 1Q8*
13«. TYPE OF REPORT
Interim
14. DATE OF REPORT (Vear.Montn.Oay) hs. PAGE COUNT
1985 November 29
I
47
16. SUPPLEMENTARY NOTATION
17.
COSATI CODES
GROUP
SUB-CROUP
,05
18. SUBJECT TERMS (Continue on reverie if necessary and MentnV by block number)
EpidemiolOKic InveatlRation: Air Force Health Study:
Matched Cohort Design; Nonooncurrent Prospective Design:
Mortality Study
19. ABSTRACT (Continw* on <wv«nt If necenwy and Identify by block number)
The purpose of th<9 Air force Health Study la to determine whether individuals Involved in
the aerial sprajring of herbicides in Vietnam during the Ranch Hand operation have
experienced 'inbre aaed deaths as a result of their participation in that program. The
Baseline Mortallt y Report was released in Juine 1983, the Baseline Morbidity Report in
February '1984,, an d the first follow-up mortality study in December 1984. Neither study
demonstrated, heallth effects 'that oould be conclusively attributed to herbloide, *>r dioxin
' 4nrtvMnn*A " 'ftimm npMwent report deaorlbea the third mortality analyses. Deaths in the 1257
• Ranoh Hand and 611^1 comparison «ubjaota :were determined. ^Aa of 31 December 198^1, 55 Ranoh
Hand and 285 wmparlaon aubjeota *ad ^ied. Death oertifloates were obtained on all
eubjtota. <tna <Mirrent mortality attajjraaa did not reveal any atatietioally significant
differences In mot•tmllty between th«t«Kpoaed and oompariaon groups. Continued mortality
aurvtlllanoe i» r«^commended ainoe tht^%tudy groups are still relatively ^young and healthy.
While aufflcient time may have .alapwd f or «OM olinioally algnlf ioant conditions to
FIELD
06
JM.DI$T1WUTK>N/AVAMAMUTY OF ABSTRACT
DUNOASSIFIEDVONOMITID
D SAME AS RPT.
i2*. NAMI Of MifPOHSIMf IHOIVIOUAL
..
WILLIAM H. HOUPB. Col. USAF. MC ^
(X>KMtM147M4MAft
1 t •••v^&^iffi^
pOTIC .USERS
-• v
V,^V
3 1. ABSTRACT SECURITY CLASSIFICATION
UNCLASSIFIED
22b.'HUPHONf0rK«u*>U9*C0d!0 Mt OWCI f VMtOL
' (912) 536-260* v . "
USAFSAM/EK
.-.tf.v 1 :-•: ***&•********** ***>ndMNMMl. „•. • jjjCAJplTY OAiSmCATk^l (Of THIS PAGE
otnumm,»<?*<v. •<*<>.» , . --H-. s?.vfwp.-; .:~iiX&^wl^>:'-^ • ' '
••'•'.'•. •r^%Y^;.^^^^^^'^'/i(:i^''l^~-.-^. •: . "•''• ^J'l'^&i^^jH' ••••'. •
;
�nr Force Health Study Mortality Update - 1985
EXECUTIVE SUMMARY
BACKGROUND
The purpose of the Air Force Health 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 after the studies were initiated.
The Baseline Mortality Report was released in June 1983, the Baseline
Morbidity Report in February 1984, and the first follow-up mortality study
in December 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, 2, 3).
METHOD
The present report describes the third mortality analyses. Deaths in the
1257 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 1984, 55 Ranch Handers and 285 Comparison subjects had died.
Death certificates were obtained on all subjects.
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, and the active U.S. civil
service population as discussed in the 1984 mortality report (3). The West
Point class of 1956 and the active duty USAF population are not appropriate
groups for comparing to the study population and, consequently, they have
not been used in the analyses in this report.
RESULTS
As was the case in the last 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. Within categories of rank and occupation none of the differences between the Ranch Hand and Comparison groups
are statistically significant.
�Percent Deaths
Ranch Hand
Comparison
Rank
Officers
Enlisted
3.4
4.9
4.3
4.8
Occupation
Hying
Ground
3.7
5.1
5.1
4.1
Ranch Hand
Total
Overall
4.4
Cgrnpari^son
4.6
As was reported in the 1984 mortality study, the Ranch Hand officers had a
nonstatistically significant though slightly lower death rate than their
Comparisons. There is an interaction in these data, however. 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 deaths 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. Additionally,
Ranch Hand officers experienced fewer deaths after age 35 years than
did Comparison officers, while Ranch Hand officers experienced more deaths
before age 35 years than did Comparisons. The relevance of these observations is unclear at this time.
Ranch Hand flyers had a nonstatistically 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 1984 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, all subgroups are living
longer than expected. All groups had a mortality experience similar to the
civil service population.
11
�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 as a result of herbicide exposure in those
individuals who accomplished the Ranch Hand spray operations in Vietnam.
�Table of Contents
Page
Executive Summary
i
Table of Contents
iv
1. Introduction
1
2. Ranch Hand Versus Comparison Group Analyses
2
3. Within-Group Analyses of Mortality
13
4. Cause-Specific Analyses
14
5. Noncause-Specific Comparisons with External Population
21
6. Further Covariate Adjustments
32
7. Statistical Aspects
34
8. Future Commitments
37
9. Summary and Conclusions..*
37
References
,
39
Principal Investigators
41
Appendix - Mortality by Year-of-Birth.
42
iv
�Air Force Health Study Mortality Update - 1985
1. Introduction
This report updates the findings of the last mortality report (3) released on
1C December 1984. The reader is referred to the baseline mortality report (1),
released on 30 June 1983, for information regarding the study design, statistical
procedures and the mortality determination process. One newly identified non-Black
enlisted-ground Ranch Hander has been added to the data file since the last report.
This individual was previously known, but confirmation of his eligibility was
delayed. Summary counts of the population at risk and the number of deaths in each
of the two groups (Ranch Hand and Comparison) stratified by rank and occupation are
shown in Table 1. The analyses in this report are based on this data and the data
in Table 4. Table ?. 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 occupation, race and group. In the
December 1984 report, the mortality experience of the study population was contrasted with data from West Point graduates and the active duty Air Force population. As
noted in that report, the West Point group consists only of current and former
officers, and with respect to the active duty Air Force population, individuals with
serious illness are generally not allowed to remain on active duty. Therefore,
contrasts with these groups are not appropriate in the context of this study and
have riot been used. All tables in this report correspond to similar tables in the
last annual report. These counts reflect cumulative mortality as of 31 December
1984 (certified as of 15 April 1985).
Careful interpretation of the findings in this and previous reports in this
series requires consideration of the large sample approximations and assumptions
associated with the statistical procedures. Current knowledge regarding these
statistical aspects is presented in Section 7 of this report.
Table 1
Summary Counts of Death by Rank and Occupation
r
iinA
Ranch Hand
Atjysk. Dead RateJX).
Comparison
At Risk Dead Rate (%}
Officers
Enlisted
466
791
16
39
0.034 (3.4)
0.049 (4.9)
2278
3893
Flying
Ground
646
611
24
31
0.037 (3.7) 3163
0.051 (5.1) 3008
98
187
0.043 (4.3)
0.048 (4.8)
161 0.051 (5.1)
124 0.041 (4.1)
1257
55
Mil
°-044 (4-4) 6^71 285 °-046 (4
In Table 2, the number "at risk" is the number alive on 1 January 1984.
�Table 2
Deaths During 1984 by Rank and Occupation
Rank
Officer
Enlisted
Ranch Hand
1984
Rate
At Risk Deaths Per 100
Comparison
Rate
1984
At Risk Deaths Per 100
451
752
1
0
0.2
2187
3719
7
13
0.3
0.3
623
580
1
0
0.2
3014
2892
12
8
0.4
0.3
0.1
5906
20
0.3
Occupation
Flying
Ground
Total
1203
Since so few deaths have occurred during 1984, the statistical findings and
interpretations presented in this report are very similar to those in the 1984
mortality update (3).
Table 3
Occupational and Race-Specific Mortality
Race
Occupation
Non-Black
Officer-Pilot
Officer-Nav
Officer-Other
Enlisted-Fit Eng
Enlisted-Other
Officer-Pilot
Officer-Nav
Officer-Other
Enlisted-Fit Eng
Enlisted-Other
Black
Total
At Risk
Ranch Hand
Dead"Rate Per 100
350
82
25
191
533
6
2
1
15
52
55
Comparisons
D e a d R a t e Per 100
79
15
4
57
108
0
0
0
10
12
12
3
1
7
28
0
0
0
2
_2
1257
At Risk
6171
4.5
3.8
3.3
6.1
4.1
0.0
0.0
0.0
13.3
4.7
285
4.6
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 by the product-limit estimate of Kaplan and Meier (4). Linear rank
testing was carried out using the logrank test and Prentice's censored data extension of the Wilcoxon test (5). All linear rank tests were carried out with matched
sets merged when Ranch Handers 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 Egigou and McHugh (6), is applied only to the 1241 Ranch Handers with
matched Comparisons, and the stratified relative risk or SMR estimate is
applied to all 1257 Ranch Handers.
Group contrasts were made on officers, enlisted personnel, flying personnel,
ground personnel and the total group. Summary counts are shown in Table 4.
Table 4
Summary Counts by Rank, Occupation and Group
FlyingPersonnel
Officer
Enlisted
Total
Rate
At Risk Dead Per 100 At Risk Dead
Rate
At Risk Dead
Ranch Hand
Comparisons
440
2153
15
94
3.4
4.4
206
1010
9
67
4.4
6.6
646
3163
24
161
Rate
PerJLOO
3.7
5.1
Ground Personnel
Groups
Ranch Hand
Comparisons
Officer
~ Rate
_A_t Rjjk Dead Per 100
lead
26
1
4
3.8
3.2
Enlisted
Total
Rate
At Risk Dead Per 100 At Risk Dead
585
2883
30
120
5.1
4.2
611
3008
31
124
Rate
Per 100
5.1
4.1
Survival curves were estimated only for officers, enlisted, flying, ground
personnel and all personnel in Ranch Hand and 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, flyers and ground personnel are shown in Figures 2 through 5.
�Figure 1
Survival Curve Estimates for All Ranch Handers and All Comparisons
RH
0.9
o.a
COMP
> 0.8
C£
OT
0.6
z
o
~
0-4
0-2
0.1 JL
30
20
10
60
40
70
flCE
Figure 2
Survival Curve Estimates for Ranch Hand and Comparison Officers
i
RH
0.9
COMP
0.8
o
! 0-7
?
0-6
QC
to
o
—
0-4
o
oc
cc 0-3
0.2 .
0.1 .
01
10
20
30
50
flGE
4
60
70
�Figure 3
Survival Curve Estimates for Enlisted Ranch Handers and Comparisons
s
0.9
0.8
COMP
O
5 0.7
a:
0.6
O-b
6
~ o. < _
CJ
0-3
0-2
0.1
0
10
20
40
30
60
60
70
flCE
Figure 4
Survival Curve Estimates for Ranch Hand and Comparison Flyers
i _.
0.9
COMP
RH
0-6
o
or
0.6
0.1
o
-•>
O.-l
LJ
0.3
0.2
0.1
0
10
20
30
40
flGE.
5
60
60
70
�Figure 5
Survival Curve Estimates for Ranch Hand and Comparison
I
0.9
0-8
O
COMP
> 0.6
o:
Z)
CO
0-6
o
LJ
d
Ctl
a. 3
0-2
O.I
20
SO
30
60
_J
70
RCE
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 linear rank statistic used is a valid measure of group difference only when
this difference occurs consistently across strata. Since the strata in these
analyses were formed by date-of-birth, occupation and race, the linear rank statistic is valid only when the direction of the group difference in death times does not
change with date-of-birth, race and occupation. As discussed in Section 7, there is
currently no statistical procedure available for testing the assumption that differences in group survival distributions remain constant across strata. As will be
shown later, there is an indication, however, that there is an effect of date-ofbirth on relative risks in the officer subgroup. Thus, the logrank and Wilcoxon
tests on officers must be interpreted carefully. However, these data suggest that
the sumrnan/ statistics for the remaining subgroups are valid. Further, since there
is an inc cation that mortality contrasts change with rank and occupation, the
overall (t,tal) logrank and Wilcoxon values and p-values, shown in Table 5 are not
valid summary statistics.
�Table 5
Test Results and P-Values for Noncause-Specific Survival
Logrank_
J Value);
T^-VaTue'
Vlilcoxon
HUyH P-Value
Officer
Enlisted
Flying
Ground
(-0.835)
( 0.187)
(-1.34 )
( 0.976)
0.40
0.85
0.18
0.33
(-0.903)
( 0.161)
(-1.42 )
( 0.093)
0.37
0.87
0.16
0.34
Total
(-0.305)
0.76
(-0.344)
0.73
Table 5 suggests that ground personnel in the Ranch Hand group are dying sooner
than their matched Comparisons (logrank = 0.976), but again the difference is not
statistically significant (p=0.33). The negative values of the logrank and Wilcoxon
statistics for officers (logrank * -0.835) and flying personnel (logrank = -1.34)
suggest that Ranch Handers in this group may be living longer than their matched
Comparisons, but not to a statistically significant degree.
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
Wjlcgxon
(Value") P^Value
Officer
Enlisted
Flying
Ground
(-0.819
( 0.211
(-1.43
( 1.10
0.41
0.83
0.15
0.27
(-0.885) 0.38
( 0.192) 0.85
( 0.
(-1.50 ) 0.13
(-1.
( 1.08 ) 0.28
Total
(-0.286)
0.78
(-0.320) 0.75
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 1n 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. The estimated relative
risks are valid summary statistics only when relative risk can be assumed to be
constant across date of birth strata. Again, there is indication that this
assumption is not met in the officer cohort so their estimated relative risks
must be viewed with caution. On the other hand, the assumption appears to be
�met in the flying, ground and enlisted subgroups so these relative risk estimates do
appear to be valid. Similarly, since there is an indication that relative risk
changes with rank and occupation, the overall relative risk, 0.915, is not a valid
summary statistic.
Table 7
Relative Risks, 95% Confidence Intervals, P-Values and
Power for Noncause-Specific Deaths to Date
(1241 Ranch Handers Versus 6171 Matched Comparisons)
Rel Risk
Conf Int
P-Value
Power
Officer
Enlisted
Flying
Ground
0.715
0.987
0.692
1.21
0.311,
0.622,
0.377,
(0.708,
.12)
,35)
.01
,72
0.26
0.94
0.12
0.35
0.90
0.99
0.98
0.94
Total
0.915
(0.636, 1.20)
0.57
1.0
Table 7 shows that Ranch Hand flyers are experiencing fewer deaths than their
matched Comparisons (relative risk = 0.692), but this group difference is not
statistically significant (p=0.12). The Ranch Hand ground personnel experienced
more deaths (relative risk = 1.21) than their matched ground Comparisons, but again,
this excess is also not statistically significant (p=0.35). The statistical power
to detect a relative risk of two is quite strong (equal to or greater than 90%).
Year-of-birth specific mortality rates are given in Tables 8 through 12, with
the corresponding standardized mortality ratios (SMR) and associated p-values (7).
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 analyses were conducted and are presented at the end of this
section. They indicate that the hypothesis of equal year-of-birth specific relative
risks may not be met in the officer cohort.
�Table 8
Year-Of-Blrth Specific Mortality Rates
(1257 Ranch Hariders Versus 6171 Comparisons)
(SMR = 0.954, PI = 0.22, P2 = 0.73)
Ranch Hand
Comparison
Birth
Dead
5
17
48
84
305
211
210
377
2
5
3
2
18
7
5
13
40.0
29.4
6.3
2.4
5.9
3.3
2.4
3.4
1257
e
Rate
Per 100
55
A
1905-1914
1915-1919
1920-1924
1925-1929
1930-1934
1935-1939
1940-1944
1945-1954
Total
s
A t Risk
Dead
14
96
241
501
1389
1020
1096
1814
3
14
30
46
79
39
24
_50
6171
Rate
Per 100
285
21.4
14.6
12.4
9.2
5.7
3.8
2.2
2.8
Table 9
Officer-Specific Mortality Rates by Year-Of-Birth
(SMR =0.791, PI = 0.41, P2 = 0.37)
Ranch Hand
Birth
Year
At Risk Dead
1910-1924
1925-1934
1935-1939
1940-1944
1945-1949
41
194
95
91
45
3
5
4
2
_2
Total
466
16
Comparison
Rate
Per 100
7.3
2.6
4.2
2.2
4.4
At Risk Dead
205
930
458
495
190
21
52
13
7
_5
2278
98
Rate
Per 100
10.2
5.6
2.8
1.4
2.6
�Table 10
Enlisted-Specific Mortality Rates by Year-Of -Birth
(SMR = 1.03, PI = 0.67, P2 = 0.89)
Comparison
Ranch Hand
Birth
Year
At Risk Dead
1905-1914
1915-1919
1920-1924
1925-1929
1930-1934
1935-1939
1940-1944
1945-1954
4
9
16
41
154
116
119
332
791
At Risk Dead
39
50.0
22.2
18.8
4.9
8.4
2.6
2.5
3.3
12
54
80
211
749
562
601
1624
Rate
Per 100
3
10
13
26
47
26
17
45
3893
2
2
3
2
13
3
3
11
Total
Rate
Per 100
187
25.0
18.5
16.?
12.3
6.3
4.6
?.8
2.8
Table 11
Flying-Specific Mortality Rates by Year-Of-Birth
(SMR =0.726, PI = 0.85, P2 = 0.13)
Comparison
Ranch Hand
Birth
_Yea_r
Rate?
Per 100
At Risk Dead
1915-1924
1925-1934
1935-1939
1940-1944
1945-1949
44
272
145
121
64
Total
646
Rate
At Risk Dead Per 100
9.1
3.7
4.1
1.7
3.1
10
220
1316
698
653
276
26
84
26
15
10
3163
4
10
6
2
JL
24
161
11.8
6.4
3.7
3." 6
�Table I?.
Ground-Specific Mortality Rates by Year-of-Birth
(SMR = 1.23, PI = 0.59, P2 = 0.33)
Ranch Hand
Rate
Birth
1905-1914
1915-1919
1920-1924
1925-1929
1930-1934
1935-1939
1940-1944
1945-1954
5
8
.13.
31
86
66
89
313
2
1
3
2
8
1
3
11
Total
611
40.0
12.5
23.1
6.5
9.3
1.5
3.4
3.5
31
Comparison
Rate
Dead er100
14
51
66
151
423
322
443
1538
3
8
10
20
21
13
9
40
3008
21.4
15.7
15.2
13.2
5.0
4.0
2.0
2.6
124
Additional log- linear analyses of the data in Tables 9 through 12 were carried
out. These analyses are directed at the hypothesis already tested and reported, via
the p-value (PI), but have an advantage in that they are more powerful. They have a
disadvantage in that, since they were carried out after the data had already been
tested, the overall level of significance is higher than the nominal 5%. The extent
of the increases in power and significance level is not known. 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 borderline
significant four-way interaction is evident (p=0.054). The officer and enlisted
relative risks are 0.53 and 1.10 in the 1905-1934 year-of-birth stratum and 1.58 and
0.95 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=0.085), and no significant
group by flying status by birth-year interaction (p=0. 92) 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=0.67), flying (p=0.30) or ground personnel (p=0.28) but
there is a significant three-way interaction for the officers (p=0.044). That is,
the survival status by group relationship changes with year-of-birth in the officer
cohort. Two-factor p-values are 0.87 for enlisted, 0.12 for flying, and 0.077 for
ground personnel. These findings are consistent with previous analyses.
Taken together, these log-linear analyses suggest that relative risk changes
with year-of-birth in the officer cohort. 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).
11
�These findings cast doubt upon the validity of the SMR and, possibly, the linear
rank procedures, as summary statistics for the officer cohort. The SMR appears to
be a valid summary statistic for Ranch Hand and Comparison contrasts within the
enlisted, flying and ground cohorts.
A summary of logrank, relative risk and SMR results obtained is shown in
Table 13.
Table 13
Noncause-Specific Statistical Summary
Age at Death
Logrank
Value
P-Value
-0.835
0.187
-1.34
0.976
Total
0.40
0.85
0.18
0.33
0.305
Officer
Enlisted
Flying
Ground
0.76
Deaths to Date
RR
P-Value
SMR
P-Value
Officer
Enlisted
Flying
Ground
0.715
0.987
0.692
1.21
0.26
0.94
0.12
0.35
0.791
1.03
0.726
1.23
0.37
0.89
0.13
0.33
Total
0.915
0.57
0.954
0.73
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, and the
Ranch Hand flying cohort exhibits fewer deaths, but again these group differences
are 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, these data appear to suggest that favorable mortality experience occurs
in those officers born before 1935, while Ranch Hand officers born after 1935 appear
to have experienced the same or greater death rate than their Comparisons.
12
�Table 14
Death Rates by Group, Rank, Occupation and Year-of-Birth
Death Rate per 100
Ranch Hand Comparison
Year~of~Birth
Rank
Relative
Risk
Officer
1905-1934
1935-1954
3.4
3.5
6.4
2.2
0.53
1.58
Enlisted
1905-1934
1935-1954
9.8
3.0
9.0
3.2
1.10
0.95
Year^jBirth
Death Rate per 100
Ranch Hand Comparison
Relative
Risk
Flying
1905-1934
1935-1954
4.4
3.0
7.2
3.1
0.62
0.97
Ground
1905-1934
1935-1954
11.2
3.2
8.8
2.7
1.27
1.19
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 Ana^lyses^of Mortaljty_
Within-group year-of-birth adjusted contrasts by occupation and rank via SMR's
are summarized in Table 15. The data supporting these SMR analyses are shown in
Appendix Tables 1 through 4.
Table 15
Summary of Within-Group SMR Analyses
Subgroups
SMR
PI
P2
Officers versus Enlisted
Ranch Hand
Comparison
0.515
0.648
0.27
0.88
0.047
0.001
Flying versus Ground
Ranch Hand
Comparison
0.572
0.909
0.41
0.46
0.067
0.65
13
�Table 15 shows that Ranch Hand officers are having significantly fewer deaths
(SMR=0.515, p=0.047) 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=0.648, p=0.001). The table also suggests a favorable mortality experience of Ranch Hand flyers relative to that of the
Ranch Hand ground personnel (SMR=0.572, p=0.067), although this difference is not
statistically significant. A flyer versus ground differential is not apparent in
the Comparison group (SMR=0.909, p=0.65).
4. Cause-Specific Analyses
Table 16 shows death counts by cause and subgroup (flying officer, ground officer, flying enlisted and ground enlisted). Counts are shown for all 1?57 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 16 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 the total population.
One cell in Table 18, the analysis of malignancy deaths in flying personnel,
contains 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 non-null variance estimate. A null
variance is defined as a variance derived upon the assumption that the true relative
risk is unity. A non-null variance is derived without any assumption about the true
value of the relative risk. The choice of which variance estimate to use in the
standardization of the test statistic is currently a point of research in
theoretical statistics. We have chosen to use the null variance when computing
p-value 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.
14
�Table 16
Deaths by Cause and Subgroup
Officer
Flyjng_
ITrouncT
11 C
8 33
Cause
Enlisted
Ground
Flying
Total
C
RH
C
RH
_C
RH
C
0
4
27
7
35
19 96
1
3
1
9
3 18
RH
Suicide
0
5
1
1
1
Homicide
0
0
0
0
0
2
2
3
2
5
Parasitic
infection
0
2
0
0
0
0
0
2
0
4
Malignant
neoplasm
0 15
0
1
1
14
5
21
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
5 28
0
0
1
14
12
38
Respiratory
System
0
2
0
0
0
3
0
2
0
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
anomal ies
0
0
0
0
0
0
0
0
0
0
_0 _1
_0
_g
_!
J.
JO
0
J. _?.
15 94
1
4
9
67
30 120
55 285
Accident
111 defined
Total
15
6 51
18 80
7
�Table 17
New Deaths by Cause
Cause
Ranch Hand
Comparisoni
Accident
0
2
Suicide
0
2
Homicide
0
1
Malignant
Neoplasm
0
8
Circulatory
System
1
5
Respiratory
System
0
2
Totals
20
16
�Table 18
Cause-Specific Age Adjusted Relative Risks by Group
(1241 Ranch banders versus fi!7! Comparisons)
Statistic
Accident
Suicide
Cause
Homicide
Mai ignancy
Circulatory
Digestive
0.577
(0.000, 1.27^
0.38
0.54
2.00
(0.000, 5.280)
0.36
0.28
1.08
(0.361, 1.78)
0.82
0.72
1.88
(0.000, 4.36)
0.31
0.3?
Officer
Rel Risk
Conf Int
P-Value
Power
0.968
0.833
(0.160, 1.78) (0.000, 2.60)
0.94
0.87
0.61
0.29
Enlisted
Rel Risk
Conf Int
P-Value
Power
0.830
0.833
(0.275, 1.38) (0.000, 2.08)
0.59
0.81
0.37
0.79
Flying
Rel Risk
Conf Int
P-Value
Power
0.905
0.625
(0.316, 1.49) (0.000, 1.9?)
0.67
0.77
0.78
0.32
0.172
0.500
(0.000, 0 .5161 (0.000, 1.01)
0.069(0. 000) 0.20
0.57
0.66
2.14
( . 0 , 5.04)
000.
0.2?
0.3!
Ground
Rel Risk
Conf Int
P-Value
Power
0.803
l.CO
3.333
(0.090, 1.5?) (0.000, ?.52) (0.000, 9.30)
1.0
0.63
0.099
0.62
0.25
0.35
1.08
1.34
(0.009, 2 .15) (0.367, 2.31)
0.88
0.42
0.62
0.49
1.67
(0.000, 4.33)
0.51
0.29
Total
Rel Risk
Conf Int
P-Value
Power
0.917
0.833
(0.447, 1.39) (0.000, 1.8R)
0.74
0.77
0.90
0.45
0.579
(0.087, 1 .07)
0.22
0.73
1.92
(0.000, 3,91)
0.17
0.39
P-value compatible with the confidence interval.
2.00
(0.000, 5.28)
0.36
0.28
2.00
(0.000, 5.28)
0.36
0.28
O.R34
(0.104, 1 .57)
0.69
0.62
1.02
(0.459, 1.58)
0.95
0.85
�Tables 16 and 18 must be interpreted with care since the data are very sparse
in some categories. The behavior of the Ejigou-McHugh estimate, like that of other
relative risk estimates, has not been investigated when the death probabilities are
very small, as is the case for the causes analyzed in Table 18. The analyses of
malignant neoplasm and circulatory system deaths are more reliable than the other
cause-specific analyses because these two categories contain more deaths than the
others.
Digestive system mortality by ICD cpde is shown in Table 19, site-specific
malignant neoplasm mortality is shown in Tijible 20 and the morphology of neoplasms is
shown in Table 21. There was one case jof soft tissue sarcoma in a Comparison
individual, but none in the Ranch Hand group. There have been no cancer deaths in
the Ranch Hand group and eight in the Comparison group during 1984.
Table 19
Digestive System Mortality
Deaths
Ranch HancfTjomparison
ICD Code
Pancreatitis (5770)
Alcoholic cirrhosis (5712)
Nonalcoholic cirrhosis (5715)
Nonalcoholic fatty liver (5718)
Chronic liver disease (5728)
Alcoholic liver disease (5711)
1
0
3
0
0
1
2
4
3
1
2
0
Duodenal ulcer (5325)
Peptic ulcer (5334)
0
0
1
0
J3
JO
5
13
Hepatocellular disease (573a)
Total
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. It is of interest that during 1984, there were no new
deaths attributable to the digestive system in either group.
18
�Table 20
Site-Specific Malignant Neoplasm Mortality
Sj_fce I CD Code
Ranch 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) '
Other and ill-defined sites (195)
Lymphatic arid hematopoietic
tissue (200-208)
No site specification (199)
Comparison
19
4
12
21
0
1
0
0
1
3
3
1
0
__1
5
_1
6
Total
0
1
3
51
�Table 21
Morphology of Neoplasms
ICD Code
9th Ed.
M800
Deaths
Ranch Hand
Nomenclature
Neoplasms not otherwise specified (NOS)
Brain
Bronchus and Lung
M801-804
M805-808
Colon
Pancreas
Intestinal Tract
Head and Neck
Epithelial neoplasms (NOS)
Bronchus and Lung
Esophagus
Kidney
Nasopharynx
Pancreas
Stomach
Unspecified site
Papillary and Squamous Cell
Nasal Sinus
1
1
0
0
1
1
0
0
0
0
0
M905
M938-948
M959-963
M964
M965-966
M986
1
1
1
1
1
0
0
0
0
0
0
1
2
1
2
1
1
0
1
1
0
0
0
1
1
0
1
0
1
0
2
J)
_!
6
Tongue
Lung
Tonsil
Adenomas and Adenocarcinomas
Appendix
Bronchus and Lung
Colon
Kidney
M872-879
1
6
2
2
1
1
10
1
1
1
2
0
1
51
0
Lip
M814-838
0
1
0
0
0
0
Stomach
Rectum
Nevi and Melanomas
Skin (NOS)
Mediastinal
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
20
�5•
N o n c 9 uJ±T_S£§cjrfj c_ Cpmpjnrjs o n s wiJth_ J[xt ernal Popul at ions
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. 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 (8), the 1978 Department
of Defense period life tables for nondisability retired military officer and enlisted personnel (9), and a 1974 U.S. active male civil service life table (10). All
analyses in this section depend on the assumption, that relative risk is constant
across age strata (Section 7). This assumption is not currently testable.
5 l
-
Cgmparisons with 1978 POD Li fe Tables
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. The magnitude of the statistic T is sample-size dependent. All contrasts are unadjusted for race since the DOD tables are not race-specific. All
analyses are conditioned on survival to age 35, since active duty personnel are not
eligible for retirement prior to that age and, therefore, the DOD tables do not
begin until that age. The totals in Tables 22 through 25 do not, therefore, agree
with Table 1.
�Table 22
Ranch Hand Officer Versus DOD Nondisability Retired Officer Life Table
j>-4.43, P ^0.001)
Age
At__Rjj>Jl
Deaths
37-39
40-44
45-49
50-54
55-59
60-64
65-69
70-70
459
414
324
232
84
40
6
1
ExpectedDeaths
2
1
1
2
1
2
0
0
2.8
4.7
5.4
4.7
2.7
1.7
0.2
_JLJL
22.2
Total
Table 23
Comparison Officers Versus DOD Nondisability Retired Officer Life Table
(T=-3.71, P<0.001)
Age
At Risk
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-70
2264
2067
1565
1095
472
192
40
2
Deaths
Expected Deaths
22.2
23.1
25.5
23.0
13.9
8.2
1.9
0.0
12
14
25
15
10
8
0
117.9
84
Total
Table 24
Ranch Hand Enlisted Personnel Versus DOD Nondisability
Retired Enlisted Life Table
(T=-1.01, P=0.31)
Age
At Risk
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-71
771
454
333
214
67
26
10
3
Total
Deaths
Expected Deaths
7
5
6
6
2
3
0
30
22
35.5
�Table 25
Comparison Enlisted Personnel Versus DOD Nondisability Retired
Enlisted Life Table
(T=-4.29, P< 0.001)
Deaths
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-76
Total
3777
2233
1628
1054
331
130
57
9
2
Expec ted Pea t n s
21
21
38
21
16
6
4
0
__0
127
175.1
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 for both groups).
Tables 24 and 25 show that Ranch Hand enlisted personnel are experiencing mortality
patterns similar to the DOD retired enlisted population (p=0.31), and the Comparison
enlisted personnel are living significantly longer (p 0.001) than the DOD nondisability retired enlisted population. This, together with the nonsignificant logrank
value for Ranch Hand versus Comparison enlisted personnel shown in Table 5 (p=0.85),
suggests that the Ranch Hand versus Comparison contrasts may 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). Comparing the conditional analyses in Table 26 with the unconditional
analyses in Table 5, it appears that group contrasts change with age at death within
the officer cohort.
Table 26
Ranch Hand Versus Comparison
Test Results and P-Values for Noncause-Specific Survival
Conditioned on Survival to Age 35
Logrank
Wile ox on
( Va 1 uJP F^Val ue (Value) P^Value
Officer
Enlisted
Flying
Ground
(-1.87 )
( 0.802)
(-1.55 )
( 1.12 )
0.061
0.42
0.12
0.27
Total
(-0.481) 0.63
23
(-1.99 ) 0.047
( 0.810) 0.42
(-1.66 ) 0.097
( 1.12 ) 0.26
(-0.529)
0.60
�Additional categorical analyses described below reveal the interaction suggested
by the previous conditioned analyses. These are shown in Tables 27 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 Age 35, Ranch Hand Versus Comparisons
(Group By Rank By Status Interaction: P=0.044)
Status
Rank
Group
Dead
Alive
Total
Ranch Hand
459
7
466
Comparison
2264
14
2278
Totals
2723
21
2744
Enlisted Ranch Hand
782
9
791
Comparison
3833
60
3893
Totals
4615
69
Relative Risk
4684
Officer
2.44
0.738
Table 28
Death After Age 35, Ranch Hand Versus Comparisons
(Group By Rank By Status Interaction: P=0.039)
Status
Rank
Officer
Group
Alive
Dead
Total
Ranch Hand
450
9
459
Comparison
2180
84
2264
Totals
2630
93
2723
Enlisted Ranch Hand
752
30
782
Comparison
3706
127
3833
Totals
4458
157
4615
Relative Risk
0.528
1.16
�In Table 27 and 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 = 0.528) than their matched Comparisons
while the Ranch Hand enlisted are experiencing more deaths than their matched
Comparisons (relative risk = 1.16). This situation is reversed in those men dying
prior to age 35. The relevance of these observations is unclear at this time.
These death rates are summarized 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, as will be shown later in this report.
Further analyses in future reports will attempt to clarify these patterns.
Table 29
Death Rates by Age at Death, Group, and Rank
Death Rates per 100
Age at
Death
_ J?iQ£!'l. Nanc!
Wfi cers "
InTfste?
Before Age 35
After Age 35
1.5 (N=466) 1.1 (N=791)
2.0 (N=459) 3.8 (N=782)
5 2
•
0.6 (N=2278)
3.7 (N=2264)
1.5 (N=3893)
3.3 (N=3833)
Comparisons with the U.S. Active Male Civil SeryiceU^fe Table
To further place the Ranch Handers and their matched Comparisons in perspective, Ranch Handers, Comparisons, and officer and enlisted personnel are contrasted
with the 1974 male active U.S. civil service life table (10). These contrasts are
shown in Tables 30 through 35. There was 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 30
All Ranch Handers Versus U.S. Male Civil Service
(T=-0.313, P=0.75)
Deaths
Expected Pea ths
6.8
21-24
1257
2
25-29
1255
7
6.0
30-34
35-39
40-44
45-49
1248
1230
868
657
7
9
6
7
5.7
7.0
8.3
9.6
50-54
446
8
7.3
55-59
60-64
151
66
3
5
3.7
2.2
65-69
16
0
0.6
70-71
4
_1
0.1
55
57.3
Total
25
�Table 31
Comparison Versus U.S. Male Civil Service
(T--1.04, P=0.30)
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
70-74
6171
6169
6151
6122
6041
4300
3193
2149
803
322
87
11
2
18
29
25
33
35
63
36
26
14
4
0
10.5
43.1
29.5
28.1
34.3
40.6
46.4
35.7
18.8
11.1
3.7
0.6
285
302.5
Total
Table 32
Ranch Hand Officers Versus U.S. Male Civil Service
(T«-l.92, P=0.054)
Age
At Risk
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-70
466
463
459
414
324
232
84
40
6
1
Total
Deaths
Expected Deaths
3
4
2
1
1
2
1
2
0
0
2.2
2.1
3.0
4.0
4.8
4.0
2.2
1.2
0.2
0.0
16
23.7
26
�Table 33
Comparison Officers Versus U.S. Male Civil Service
(Comparisons: T=-1.88, P=0.060)
Acje
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-70
Deaths
9
5
12
14
25
15
10
8
0
_0
Total
10.9
10.4
14.6
19.8
22.
19.
11.
6,
1.4
0.0
98
2278
2269
2264
2067
1565
1095
472
192
40
2
Expected Deaths
116.6
Table 34
Ranch Hand Enlisted Personnel Versus U.S. Male Civil Service
(T=1.28, P=0.20)
Age
At Risk
Deaths
Expected Deaths
21-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-71
791
789
785
771
454
333
214
67
26
10
3
4
3
7
5
6
6
2
3
0
4.3
3.8
3.6
4.0
4.3
4.8
3.3
1.5
1.0
0.5
0.1
Total
39
27
31.1
�Table 35
Comparison Enlisted Personnel Versus U.S. Male Civil Service
(T=1.54, P=0.12)
Age
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
3893
3891
3873
3853
3777
2233
1628
1054
331
130
57
9
Total
Deaths
2
18
20
20
21
21
38
21
16
6
4
0
187
Expected Deaths
6.6
27.2
18.5
17.6
19.7
20.9
23.7
16.4
7.5
5.C
2.3
0.6
166.1
The Ranch Handers 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 officer personnel in the Ranch Hand and Comparison cohorts with the
male civil service reveal that the Ranch Hand and Comparison officers are experiencing a slightly, but not significantly better mortality than the civil service.
Ranch Hand and Comparison enlisted personnel are experiencing more mortality than
the civil service, but these differences are not 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.
5.3 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 36
Non-Black Ranch Handers Versus the 1978 U.S. White Male Life Table
(T=-5.63, P<0.001)
Expected Death s_
Rsk
21-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-71
1181
1179
1173
1155
824
627
432
150
66
16
4
Total
2
6
7
8
5
7
7
3
5
0
_1
51
90.8
Table 37
Non-Black Comparisons Versus the 1978 U.S. White Male Life Table
(T=-12.8, P<0.001)
Age
At Jisk
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
5693
4095
3047
2069
793
322
97
11
2
Total
Deaths
1
16
27
23
31
31
56
36
24
14
4
0
0
263
29
Expected Deaths
48.
47.
469.7
�Table 38
Non-Black Ranch Hand Officers Versus the 1978 U.S. White Male Life Table
(T=-5.89, P-cO.OOl)
Age
At Risk
Dea_d
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-70
457
454
450
407
321
231
84
40
6
1
3
4
2
1
1
2
1
2
0
_0
3.8
3.7
4.7
5.8
7.5
6.9
3.9
2.6
0.4
0.0
16
39.5
Total
Expected Deaths
Table 39
Non-Black Comparison Officers Versus the 1978 U.S. White Male Life Table
(T=-9.85, P<0.001)
Age
At Risk
Dead
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-70
2253
2244
2239
2042
1548
1086
472
192
40
2
9
5
12
14
25
15
10
8
0
0
Total
98
30
Expected Deaths
195.3
�Table 40
Non-Black Ranch Hand Enlisted Personnel Versus the 1978 U.S. White Male Life Table
(T—2.20, P=0.028)
At Risk
21-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-71
Dead
724
722
719
705
417
306
201
66
26
10
3
Expected Deaths
2
3
3
6
4
6
5
2
3
0
Total
47.9
35
Table 41
Non-Black Comparison Enlisted Personnel Versus the 1978 U.S. White Male Life Table
(T=-6.56, 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
Total
Dead
3563
3562
3546
3528
3454
2053
1499
983
321
130
57
9
2
16
18
18
19
17
31
21
14
6
4
0
0
Expected Deaths
1
165
13.
10.
248.9
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 1n the contrasts shown in Tables 36
through 41. Both Ranch Handers and Comparisons are seen to be living far longer
than expected relative to the general U.S. White male population. The same effect
31
�is seen in both Ranch Hand and Comparison officers (Tables 38 and 39) and in Ranch
Hand and Comparison enlisted personnel. In contrast with previous mortality analyses (1,3), the analysis of the Ranch Hand enlisted cohort has reached statistical
significance with the passage of time.
6.
Further Co van' ate Adjustments
Some of the contrasts shown in previous sections in this report were further
analyzed 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
(2) 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.
6.1 Ranch Hand and Com
Contrasts on Tour Length
In this report, some descriptive statistics on tour length are presented.
Table 42 shows the 5th, 50th, and 95th percentiles of tour length in months for
flying and ground personnel, and officers and enlisted personnel in the Ranch
Handers and Comparison groups. The effect of tour length on mortality will be more
thoroughly investigated in future reports.
Table 42
Tour Length Percentiles (in Months) for Ranch Handers and Comparisons
Percentiles
W
T5l
*
Group
Rank
Flying
Status
~5I
Ranch Hand
Officer
Flying
Ground
4
5
12
13
19
15
439
26
440
26
Enlisted
Flying
Ground
4
5
12
13
21
20
206
585
mF"
206
585
1257
Officer
Flying
Ground
11
11
19
18
46
43
2123
123
2153
125
Enlisted
Flying
Ground
10
10
19
18
49
45
995
2859
1010
2883
6100
6171
Total s
Comparison
Sample Population
Size
Size
The totals show that one Ranch Hander and 71 Comparisons have no tour data at this time.
32
�IP general, the Comparisons had longer tour lengths than did the Ranch Handers.
This is the result of longer tours of duty at noncombat zone bases (Comparisons)
relative to combat area bases (Ranch Hand).
6.2
The effect of exposure on mortality was assessed on the 1140 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 43.
Table 43
Ranch Hand Mortality Adjusted for Year-Of-Birth, Rank and Exposure *
Birth
Light
Medium
Heavy
Officer
1905-1934
1935-1954
Enlisted 1905-1934
1935-1954
Officer 1905-1934
1935-1954
Enlisted 1905-1934
1935-1954
Officer 1905-1934
1935-1954
Enlisted 1905-1934
1935-1954
Totals
Survival Sjtatus
AliveTotal""
1
2
7
3
2
2
4
6
5
3
6
7
54
61
50
121
79
66
51
214
84
73
84
155
55
63
57
124
81
68
55
220
89
76
90
162
48
1092
1140
Death Rate
IJeFTW'
1.8
3.2
12.2
2.4
2.5
2.9
7.3
2.7
5.6
3.9
6.7
4.3
4.2
*
117 Ranch Hand personnel either had a tour AFSC which removed any chance of exposure or were assigned to a Ranch Hand unit at a time when no spraying occurred or
both. Tour information is not available for one Ranch Hander.
There is no four-way interaction (exposure/rank/birth year/survival status) in
the data shown in Table 43 (p=0.40); there are no statistically significant
three-way interactions involving survival, and the two-way survival by exposure
interaction is not significant (p=0.54). These patterns do not indicate a herbicide
exposure effect.
7
• 1 ^ s_tijcaJ^Ajip_ects_
*
The purpose of this section is to update the information contained in Chapter
VI, Statistical Aspects, of the Baseline Mortality Report (1), regarding the properties of the statistical procedures used in this and all preceding mortality reports
in this series. The procedures discussed here are: linear rank tests (5),
33
�log-linear analysis (11), the SMR analysis (7), and the Gail and Ware study group
versus reference life table analysis (12) and the Ejigou-McHugh relative risk
estimator (6).
7•1 Large Sample P-Va1ue Approximatjons
P-value calculations for all five of these procedures rely on large sample
approximations of the distribution of the statistic under the associated null
hypotheses, termed the null distribution. This is because the finite sample null
distributions of these procedures have not been formulated. The relevant issue,
therefore, is the adequacy of these approximations in the context of this study.
Linear rank tests: The adequacy of the large sample p-value approximation in
certain linear rank procedures has been investigated via Monte Carlo simulation by
Latta (13) in the two-sample situation and by Michalek, Mihalko and White (14) on
one-to-many matched data. The primary goal of both of these studies was to investigate the power of certain linear rank procedures under various failure time
distributions, censoring percentages and sample size configurations. In the
two-sample case, the Prentice efficient score censored data extension of the
Wilcoxon procedure was judged to be best overall, and in the matched data case, the
logrank test with the hypergeometric variance was deemed the best overall procedure.
These are the two procedures used in this and all previous mortality reports in this
series. These simulation studies did not, however, attempt to assess the adequacy
of the large sample distributions of these procedures as a function of sample size
and percent censoring. In particular, neither study assessed the properties of
these procedures with heavy censoring (as seen in these mortality data).
Unpublished Monte Carlo studies conducted at the USAF School of Aerospace Medicine
have shown, however, that the logrank and Wilcoxon tests achieve nominal 1% and 5%
significance levels in two-sided testing on simulated 1:5 matched data with 1200
matched sets and 96% censoring when the survival distributions follow the
accelerated failure time model (5) and the censoring variable is uniformly
distributed. These results, while encouraging, are not directly applicable to this
study since all linear rank testing in these reports were carried out with the data
stratified by one-year birth intervals, race and occupation. Other simulations did
confirm the validity of the large sample null distributions in this highly
stratified case, but not with censoring percentages as high as 96%. Based on these
published and unpublished investigations and the smallest sample sizes in this study
(466 Ranch Hand officers contrasted with 2278 Comparison officers), the authors of
this report believe that the linear rank p-value approximations are adequate when
consideration is restricted to sample size and percent censoring.
Log-linear ana1yses: All p-values derived from log-linear analyses are based
on large sample chi-square approximations. The adequacy of these approximations has
generally been studied in terms of the magnitudes of the expected cell counts 'in
multiway contingency tables. There is extensive literature on this subject with
resultant guidance published in recent statistical texts. Conover (15) states that
the chi-square approximation is good if the expected cell counts are fairly large
but if some of the expected counts are small, the approximation may be poor. He
quotes Cochran (16), who concluded that, if any of the expected counts are less than
1 or if more than 20% are less than 5, the approximation may be poor. Conover views
34
�Cochran's advice as, perhaps, too conservative and renders the opinion that the
expected counts may be as small as I without endangering the validity of the test.
Since most expected counts in this report are greater than 5, the chi-square
approximations are considered adequate by the authors of this report.
SMR anajyses: Large sample chi-square approximations were used to obtain the
p-values~Tn~th"e™$MR analyses. The first of these was for a likelihood ratio test
for the hypothesis that the data satisfies the product model (7), the second was for
a likelihood ratio test that the SMR was equal to unity. The test of fit for the
product model is analogous to a test for no three-factor interaction in a log-linear
model, the factors, being survival states (dead, alive), group (Ranch Hand,
Comparison) and year-of-birth. Sample size requirements for this procedure are,
therefore, the same as those described above for log-linear analysis; that, is, that
the expected numbers of dead at each level of year-of-birth be at least 5 or at
least 1, depending on the advice of Conover and Cochran. The test for an SMR equal
to unity is not analogous to a test on the main effect in the same log-linear model.
No guidance has been published regarding the sample size requirements for the
adequacy of the chi-square approximation. In our opinion, this approximation is
adequate in these data.
Gail_and Ware analysis: The test statistic for comparing an observed survival
distrTButfon wTtF a™ "reference life table is a standardized sum of deviations between
observed and expected numbers of deaths and has, for large samples, an approximate
standard normal distribution under the null hypothesis. The minimum sample size and
maximum censoring percentage needed for this approximation to be adequate is not
knov/n. In our opinion, this approximation is adequate in these data.
Ej i gou -McHugh fel atj[ve_n_slk_ analyses^; The statistic used in testing relative
risk equaT~to "unity'Tias an approximate standard normal distribution under the null
hypothesis when the number of matched sets is large. In-house simulations have
shown that this approximation is adequate with 1200 match sets. The threshold of
adequacy has not been investigated to date. In our judgement, the approximation is
good in these analyses.
7 2
•
Alll^EtiPJl^^^^
of thei> Validity
In all studies, statistical procedures are based upon assumptions regarding the
data. Good statistical practice requires that the assumptions be checked before
proceeding to the final analysis. In most cases this is done subjectively by
examining plots of the data. For some statistical procedures, the assumptions can
be tested directly; such tests are termed pretests. When resampling is not
possible, pretesting should be accounted for in the overall inference. Unfortunately, pretests and procedures which account for pretests in the overall inference are
almost nonexistent in the field of statistics. Of the five procedures used in this
report, a pretest of assumptions exists only for the SMR analysis, and it is not
currently known, how to take that pretesting into consideration in the overall
analysis. Generally, pretesting should be carried out so that the overall
significance level of the pretests and the final inferential test should be a
prescribed value, such as 5%.
. r?.OJS_tJ?lM: The log^ank and Wilcoxon procedures are based upon the
assumptions That"tTTe underlying survival distributions are continuous, that survival
and censoring are statistically independent and that the difference in group sur35
�viva! distributions does not change with levels of the stratification variable. The
third assumption would hold if, for example, there were no interaction terms involving group membership in the accelerated failure time model (5). In our opinion,
the first two of these assumptions can be safely assumed in these analyses.
The
third must be checked. There does not exist a statistical procedure for testing the
assumption that the difference in group survival distributions does not change with
levels of the stratification variable, without making further assumptions. If
further assumptions were made and such a test were developed, there would, at this
time, be no way to adjust its critical value so that the overall significance level
was 5%. In this report, the stratification variables were year-of-birth, race, and
occupation. The no-interaction assumption was subjectively checked by comparing the
logrank and Wilcoxon values with other analyses, looking for consistency. There is
some indication that the assumption is not met in the officer subgroup and, therefore, the logrank and Wilcoxon values are misleading for contrasting Ranch Hand and
Comparison officers.
Log-linear analyses; The log-linear analyses are based upon the assumptions
that the data are" distributed as multinomials or product-multinomials, that all
interactions of order higher than the one of interest are nonexistent and that there
is no confounding. The multinomial assumption is correct in these analyses because
the data were categorized so that the multinomial or product-multinomial model would
hold. Tests for the existence of interactions of all orders are available and are
carried out in all analyses but, at this time, there is no way to adjust their
critical values so that the significance level of the overall procedure is 5%.
Statisticians typically use a 5% significance level for each pretest, but this may
vary.
SMR analyses^: The basic assumption in these analyses is that relative risk is
constant across levels of the stratification variable. In these analyses the
stratification variable is year-of-birth. A likelihood ratio test was used to check
this assumption. It is not known how to prescribe its critical value so that the
overall level of significance is 5%. This assumption was also checked using additional log-linear analyses.
Gail and Ware analyses: The basic assumption in these analyses is that the
study hazard function is proportional to the reference hazard function. There does
not exist a single sample test for the proportional hazards assumption. This
assumption was checked subjectively by computing relative risks at different ages
within the data sets.
Ejigou-McHugh relative risk analyses: This analysis assumes that relative risk
is constantwrtfirespect to the~"matching variables. A procedure for testing this
assumption has been recently developed (17) but has not yet been programmed for
inclusion in these reports. The new method does not provide for the adjustment of
the pretest critical value so that the overall significance level is 5%. This
assumption was subjectively checked in this report by comparing the Ejigou-McHugh
relative risk with the SMR, looking for consistency.
7.3 Summary
The issues regarding large sample approximations and pretesting assumptions are
intrinsic to the field of mathematical statistics and, therefore, are relevant to
36
�applications of statistical theory in any research. In this respect, the
statistical content of this report reflects the extent of current theory.
8-
Fu tu re Commi toejrts
Future work will attempt to evaluate mortality patterns as a function of
occupational subgroup in the ground cohort. This effort will require the collection
of data to delineate differential exposure between occupational subgroups. Flight
line duties arid 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. Joint morbidity-mortality
analyses, adjusting for relevant covariates will be carried out. Finally, the small
sample properties of the linear rank, relative risk, and SMR tests will be investigated by simulation and analytical methods.
9
•
^MaJlY. £ 1 Cj?£L£ IjUjy £n
1.
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 herbicide/dioxin exposed group, their matched Comparisons and other external Comparison groups.
Overall mortality of the Ranch Hand group (4.4%) is nearly identical to that of
the Comparison group (4.6%). Ranch Hand officers have experienced fewer deaths
than the Comparison group officers, but this difference is not statistically significant. There is an interaction in these data, however. 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 deaths 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. Additionally, Ranch Hand officers experienced fewer deaths after age 35 years than
did Comparison officers, while Ranch 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 and Ranch Hand flying personnel have experienced lower 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.
37
�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 Handers.
The Ranch Hand and Comparison groups were contrasted with five external populations. All study groups are experiencing significantly less mortality than U.S.
White males. All study groups except Ranch Hand enlisted personnel are experiencing statistically significantly less mortality than the corresponding
nondisability retired DOD population. The Ranch Hand enlisted mortality is not
significantly different from the nondisability retired DOD enlisted population. The
Ranch Hand and Comparison groups taken together have experienced a mortality pattern
not statistically different from civil service employees.
In conclusion, summary counts of death by rank and occupation did not reveal any
statistically significant differences, within the power limitations of this study,
between the Ranch Hand and Comparison groups. This study has excellent power of
detecting a doubling of risk of death, and therefore it is unlikely that an effect
of this magnitude could have been missed. 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
experienced less favorable mortality relative to Comparisons irrespective of date of
birth or age at death, this difference is not statistically significant. Exposure
index analyses indicate these mortality rate differences cannot be attributed to
herbicide exposure. These analyses have identified several findings of interest,
which will be further evaluated in future mortality updates. The findings of this
report are similar to those of prior mortality analyses with the exception that the
non-Black Ranch Hand enlisted personnel now demonstrate statistically significantly
better survival than the 1978 U.S. White male population.
38
�References
1. Lathrop, G. D. , Moynahan, P. M. , Albanese, R. A., Wolfe, W. H. (1983).
An Epidemiologic Investigation of Health Effects in Air Force
Personnel Following Exposure to Herbicides: Baseline Mortality
Study Results, (NTIS Order Number: AD-A130 793)
2. Lathrop, 6. IX, 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)
3. Wolfe, W. H., Michalek, J. E., Albanese, R. A., Lathrop, G. D.,
Moynahan, P. M. (1984). An Epidemiologic Investigation of Health
Effects in Air Force Personnel Following Exposure to Herbicides:
Mortality Update - 1984. (NTIS Order Number: not yet available)
4. Kaplan, E. L. and Meier, P. (1958). Nonparametric estimation from
incomplete observation. Journa 1 of the Ameri can Stati sti cal
Association 53:457-481.
5. Prentice, R. L, (1978). Linear rank tests with right censored data.
jyometrlka 65:167-179.
6. Ejigou, A. and McHugh, R. (1981). Relative risk estimation under
multiple matching. Biometrika 68:85-91.
7. Gail, M. (1978). The analysis of heterogeneity for indirect
standardized mortality ratios. Journal of _the_ Royal Statistical
A, 141:224-234.
8. 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.
9. Evaluation of the Military Retirement System FY 1980. Office of the
Actuary, Defense Manpower Data Center, 300 North Washington Street,
Alexandria, Virginia 22314.
10. Board of Actuaries of the Civil Service Retirement System, FiftySeventh Annual Report. US Government Printing Offices; 1980.
11. Bishop, Y. M., Fienberg, S. E, , and Holland, P. W. (1975). Discrete
Multivari_ate_ Analysis: Theory a n d Prac t i ce . The MIT Press,
12. Gail, M. and Ware, J, H. (1979). Comparing observed life table data
with a known survival curve in the presence of random censorship.
Biometrics 35, 385-391.
39
�13. Latta, R. B. (1981). A Monte Carlo study of some two sample rank
tests with censored data. Journal of the Amen'can^Statistical
Association, 76, 713-718.
14. Michalek, J. E., Mihalko, D. and White, T. (1985). A Monte Carlo
study of logrank, Wilcoxon and normal scores procedures on matched
and censored data. To appear in Communications in Statistics.
15. Conover, N. J. (1980) Practical Nonparametric Statistics, John Wiley,
New York.
16. Cochran, W. G. (1952). The chi-square test of goodness of fit.
Annals of Mathematical Statistics, 23, 315-345.
17. Ejigou, A. and McHugh, R. (1984). Testing the homogeneity of relative
risk under multiple matching. Biometrika, 71, 408-411.
40
�PRINCIPAL INVESTIGATORS
William H. Wolfe, MD, MPH, FACPM
Colonel, USAF, MC
Chief, Epidemiology Division
Joel E. Michalek, PhD, GS-14
Research Mathematical Statistician
Data Sciences Division
CONTRIBUTORS
Richard A. Albanese, MD, GM-15
Chief, Biomathematical Modeling Branch
Data Sciences Division
Vincent V. Elequin, BS, RRA, GS-11
Medical Record Librarian
Epidemiology Division
Melody Prihoda, MS, GS-9
Mathematical Statistician
Data Sciences Division
Alton J. Rahe, MS, GS-13
Mathematical Statistician
Data Sciences Division
Thomas J. White, MA
Senior Subject Matter Specialist
Data Sciences Division
41
�Appendix Table 1
Ranch Hand Officers Versus Ranch Hand Enlisted
Mortality by Year-Of-Birth
(SMR = 0.515, PI = 0.27, P2 = 0.047)
Birth
Year
Ranch Hand Officers
Rate
At Risk Dead per 100
1905-1924
1925-1934
1935-1939
1940-1944
1945-1954
41
194
95
91
45
3
5
4
2
.2
Total
466
16
Ranch
Ranc Hand Enlisted
Rate
At Risk Dead per 10'Q
_At__R_
29
195
116
119
332
7
15
3
3
11
791
7.3
2.6
4.2
2.2
4.4
24.1
7.7
2.6
2.5
3.3
39
Appendix Table 2
Comparison Officers Versus Comparison Enlisted Mortality by Year-Of-Birth
(SMR = 0.648, PI = 0.88, P2 = 0.001)
Enl is ted
Officers
Birth
Year
1905-1919
1920-1924
1925-1929
1930-1934
1935-1939
1940-1944
1945-1954
Total
At Risk Dead
44
161
290
640
458
495
190
Rate
Dead per 10"0
98
9.1
10.6
6.9
5.0
2.8
1.4
2.6
66
80
211
749
562
601
1624
13
13
26
47
26
17
45
3893
4
17
20
32
13
7
5
2278
Rate
peFTOO At Risk
19. 7
16. 2
12.3
6.3
4.6
2.8
2.8
187
Appendix Table 3
Ranch Hand Flying Personnel Versus Ranch Hand Ground Personnel
Mortality by Year-Of-Birth
(SMR = 0.572, PI = 0.41, P2 = 0.067)
Ground
Flyers
Birth
Year
Rate
At Risk Dead per 100 At Risk
1905-1924
1925-1934
1935-1939
1940-1944
1945-1954
44
272
145
121
64
4
10
6
2
2
Total
646
24
Dead
42
26
117
66
89
313
6
10
1
3
U
611
9.1
3.7
4.1
1.7
3.1
31
Rate
per 100
23.1
8.5
1.5
3.4
3.5
�Appendix Table 4
Comparison Flying Versus Comparison Ground Personnel Mortality by Year-Of-Birth
Within Comparison Group
(SMR = 0.909 PI = 0.46, P.2 « 0.65)
Flyers
Birth
Year
1905-1919
1920-1924
1925-1929
1930-1934
1935-1939
1940-1944
1945-1954
Total
Ground
Rate
At Risk Dead per 100 At Risk
45
175
350
966
698
653
276
6
20
26
58
26
15
10
3163
Dead
43
65
66
151
423
322
443
1538
11
10 .
20
21
13
9
_40
3008
161
13.3
11.4
7.4
6.0
3.7
2.3
3.6
124
Rate
per 100
16.9
15.2
13.2
5.0
4.0
2.0
2.6
�
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.
1580
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
Wolfe, William H.
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
November 29 1985
Title
A name given to the resource
An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides: Mortality Update - 1985
Subject
The topic of the resource
Air Force Health Study
ao_seriesIII
-
https://www.nal.usda.gov/exhibits/speccoll/files/original/9d782020ef4c8d15142c065f02c390ae.pdf
2e520e45d86b949ad5d269455b226d8b
PDF Text
Text
Item ID Number
01534
Wolfe, William H.
Corporate Author
RBDOrt/ArtiCia TltlB
Tr
'P Report - National Academy of Sciences (TA-2091,
29 Nov 79)
Journal/Book Title
Year
198
Month/Day
January 7
Color
°
n
Number of Images
3
DOSOrlDtOn NOtOS
Tr
'P ReP°rt outlines the events of and concerns
resulting from the National Academy of Sciences (NAS)
peer review of the Ranch Hand II study protocol.
Wednesday, May 23, 2001
Page 1534 of 1608
�DEPARTMENT OF THE AIR FORCE
USAF SCHOOL OF AEROSPACE MEDICINE (AFSC)
BROOKS AIR FORCE BASE, TEXAS 78235
7 January 1980
EKS
Trip Report - National Academy of Sciences (TA-2091, 29 Nov 79)
EKO
EK
CE
CD
CC
TS
INJURN
1. Place visited: National Academy of Sciences (NAS), Washington DC,
18 December 1979, for the purpose of obtaining peer review of the RANCH
HAND II study protocol.
2. Personnel contacted/participants
a. USAF School of Aerospace Medicine
George D. Lathrop, Col, USAF, MC
William H. Wolfe, Lt Col, USAF, MC
Patricia M. Moynahan, Lt Col, USAF, NC
Alvin L. Young, Maj, USAF
Richard A. Albanese, M.D.
b. Subcommittee of the Toxicology Section, NAS
Dr. Carl M. Shy, Institute for Environmental Studies, University of
North Carolina, Chapel Hill, NC 27514
Dr0 Leon Gordis, Department of Epidemiology, Johns Hopkins School
of Public Health, Baltimore, MD 21205
Dr. William Halperin, Division of Surveillance, Hazard Evaluations
and Field Studies, NIOSH, 4676 Columbia Parkway, Cincinnati, OH
45226
Dr. Leonard T. Kurland, Department of Medical Statistics, Epidemiology and Genetics, Mayo Clinic, Rochester, MN 55901
Dr. Philip Landrigran, Director, Division of Surveillance, Hazard
Evaluations and Field Studies, NIOSH
Dr. Gordon W. Newell, NS-356, National Academy of Sciences, 2101
Constitution Avenue, NW, Washington, DC 20418
�Dr. Raymond Seltser, Department of Epidemiology, School of Hygiene
and Public Health, Baltimore, MD 20014
c. Other attendees
Ronald Burnett, It Col, USAF, HQ AFSC/SGP
Philip Brown, Maj, USAF, HQ USAF/SGES
Graduate Student, University of North Carolina
3. Events and Topics:
Elements of the RANCH HAND II study design were presented by the
USAFSAM team to the NAS toxicology subcommittee, but several factors
affected the effectiveness of the USAF presentation and the quality of the
peer review process.
a. Thirty minutes prior to the start of the review, the USAFSAM team
was informed that it could not present their prepared briefing. Instead,
they were to respond to specific questions in the general areas of the
Mortality Study, Morbidity Study, and Follow-up Study. This format did
not allow the review committee members to view the proposed study design
as an integrated effort over time.
b. Initially, the USAFSAM team had been informed that they would have
four hours to conduct the briefing, and that additional time in the evening
could be made available if needed. Upon arrival, they were informed that
only 2-1/2 hours were allowed, so that a final report by the committee
could be prepared at 1600 hours. This time limitation did not permit
completion of discussion on the Morbidity Study, and the follow-up aspects
of the proposed design were not addressed at all. The NAS chairman
privately expressed the opinion that two days were needed to properly discuss the protocol. Had the USAFSAM team been permitted to follow the
original format, most of the review committee's questions would have been
answered.
Difficulties caused by both a and b above could have been avoided
by better staffing and coordination between the USAF, NAS staff and the
subcommittee chairman.
c. Despite having the protocol in their possession 18-20 days prior
to the review meeting, many of the committee members were unfamiliar with
even basic and fundamental aspects of the design. Their questions
reflected a lack of understanding and knowledge of the design. They
repeatedly asked questions which were clearly addressed and answered in
the protocol. Two of the seven committee members present failed to ask
any questions and only four members participated significantly in the discussion.
d. Two key members of the subcommittee, Dr. Alan Poland, a recognized
TCDD expert, and Dr. Ian Higgins, a statistician/epidemiologist, were not
in attendance.
�e. An epidemiology graduate student/physician from Milan, Italy
attended the committee meeting at the request of Dr. Shy, his major professor. While he did not participate in the discussion, his presence at the
meeting was inappropriate.
f. The committee chairman did not state the intent and the purpose of
the meeting. Rather than a "Dog and Pony Show," as he put it, he stated
that the members would ask pertinent questions thereby eliminating the
need for a long session. The meeting was conducted in a manner which
discouraged the input of the peer review process, and simulated a student/
professor relationship. One of the committee members commented later that
this meeting was like a doctoral dissertation defense.
g. The subcommittee members from Johns Hopkins asked essentially all
of the questions and concentrated almost exclusively on the limited size
(and power considerations) of the exposed group for the USAF Mortality
Study as contrasted with the US Marine Corps population noted in the
Government Accounting Office (GAO) report. They did not perceive the
Mortality Study as more than a determination of deaths as of March 1980,
and did not integrate the planned 5-year follow-up analysis and analysis
of disease patterns.
h. The committee dwelled at some length on how and why the USAF
Surgeon General made the decision to conduct this study. They made no
substantive comments to improve the study design, and did not discuss/find
any element of the study that required extensive improvement.
In short, they presented no scientific debate which indicated that
the study should be conducted in any fashion other than stated in the
protocol.
4. Recommendations:
a. In view of the political orientation of the NAS subcommittee and
the cursory nature of the scientific review, the Air Force should be prepared to respond to any adverse criticism of the protocol leveled by the
NAS.
b. A consortium of university experts should be formed by the Air
Force or DOD to monitor and review all aspects of the study from design,
through data collection, to final analysis.
5. Items of interest to procurement and USAFSAM programming: None
6. This information may be of interest to AMD/SG, HQ AFSC/SGP, and
HQ USAF/SG.
WILLIAM H. WOLFE<Lt Col, USAF, MC
Chief, Disease /Surveillance Branch
�
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.
057
Folder
The folder containing the original item.
1534
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
Wolfe, William H.
Date
A point or period of time associated with an event in the lifecycle of the resource
January 7 1980
Title
A name given to the resource
Trip Report - National Academy of Sciences (TA-2091, 29 Nov 79)
Subject
The topic of the resource
Air Force Health Study
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.
050
Folder
The folder containing the original item.
1303
Series
The series number of the original item.
Series III 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
Wolfe, William H.
Date
A point or period of time associated with an event in the lifecycle of the resource
1980-05-01
Title
A name given to the resource
Typescript: Presentation/Proceedings: Human Health Effects Following Exposure to the Phenoxy Herbicides and TCDD
Subject
The topic of the resource
dioxin
VA
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.
046
Folder
The folder containing the original item.
1159
Series
The series number of the original item.
Series III Subseries I
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
Wolfe, William H.
Date
A point or period of time associated with an event in the lifecycle of the resource
December 10 1980
Title
A name given to the resource
Typescript: Presentation to the Inland Empire Agricultural Chemical Association, Spokane, Washington. 10 December 1980. "Agent Orange in Perspective"
Subject
The topic of the resource
Vietnam
dioxin
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.
046
Folder
The folder containing the original item.
1158
Series
The series number of the original item.
Series III Subseries I
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
Wolfe, William H.
Date
A point or period of time associated with an event in the lifecycle of the resource
1980-05-01
Title
A name given to the resource
Human Health Effects Following Exposure to the Phenoxy Herbicides and TCDD
Subject
The topic of the resource
dioxin
VA
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.
046
Folder
The folder containing the original item.
1157
Series
The series number of the original item.
Series III Subseries I
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
Wolfe, William H.
Date
A point or period of time associated with an event in the lifecycle of the resource
1980-05-01
Title
A name given to the resource
Diagnostic Indicators of Phenoxy Herbicide/Dioxin Toxicity
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.
046
Folder
The folder containing the original item.
1156
Series
The series number of the original item.
Series III Subseries I
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
Wolfe, William H.
George D. Lathrop
Source
A related resource from which the described resource is derived
Human and Environmental Risk of Chlorinated Dioxins and Related Compounds
Date
A point or period of time associated with an event in the lifecycle of the resource
1983
Title
A name given to the resource
A Medical Surveillance Program for Scientists
Subject
The topic of the resource
industrial exposure
ao_seriesIII