7 15 672 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 191 Folder The folder containing the original item. 5397 Series The series number of the original item. Series VIII 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 Jemison, Terry Source A related resource from which the described resource is derived U.S. Medicine Date A point or period of time associated with an event in the lifecycle of the resource March 15 1983 Title A name given to the resource EPA, VA Dioxin Policies Vary Widely ao_seriesVIII 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 201 Folder The folder containing the original item. 5682 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Jemison, Terry Source A related resource from which the described resource is derived U. S. Medicine Date A point or period of time associated with an event in the lifecycle of the resource 1986-04-01 Title A name given to the resource CDC-Army Breakdown Stalls Study ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/861180c028f300dcb8778ceb36f8ec62.pdf 27584294477a4abdba00dd23466d3da6 PDF Text Text item n Number °5351 Alltbor Jenkins, Dale W. CorperatB Author United States Environmental Protection Agency (EPA), D Not Scanned Report/Article TltlB Project Summary: Biological Monitoring of Toxic Trace Elements Journal/Book Title Year 1981 Month/Day February Color D Number of Images ° Descripton Notes EPA-eoo/sa-so-ogo Tuesday, March 05, 2002 Page 5351 of 5363 �United States Environmental Protection Agency Environmental Monitoring Systems Laboratory Las Vegas NV 89114 Research and Development EPA-600/S3-80-090 Feb. 1981 Project Summary Biological Monitoring of Toxic Trace Elements Dale W. Jenkins The contamination of plants, animals, and humans by toxic trace elements poses a continuing and increasing threat to our environment. Of 20 trace elements that are known to be toxic, this report identifies 14 as particularly noxious because of their widespread production and use, and subsequent discharge and persistence in the environment. These 14 elements of particular concern are antimony, arsenic, beryllium, boron, cadmium, chromium, cobalt, copper, lead, mercury, nickel, selenium, tin, and vanadium. Trace elements are known to accumulate in plants and animals. An a n a l y s i s of data about their accumulation, together with a review of the organisms used for monitoring environmental gradients of trace elements, resulted in an extensive list of potential biological monitoring organisms. Volume 1 of the two volumes in this report treats the problems, needs, selection, and use of living organisms for monitoring. The author also reviews global and national integrated exposure assessment programs and environmental and food surveillance programs. Volume 2 contains comprehensive data about bioaccumulation and bioconcentration of these 14 trace elements. This Project Summary was developed by EPA's Environmental Monitoring Systems Laboratory, Las Vegas, Nevada, to announce key findings of the research project that is fully docu- mented in a separate report of the same title (see Project Report ordering information at back). Trace Element Selection Before selecting the toxic elements for study, data were compiled about the toxic effects on humans, and major exposure routes and pathways; relative abundance and concentrations, production, and use; emissions into the environment for each trace element; and, the major sources and pathways of each trace element into plants and animals. After the data compilation, the following selection criteria were used to identify toxic trace elements: 1. The elements are toxic to humans and/or their food organisms; 2. The elements are fairly common and widespread; 3. The exposure to the elements is sufficient to cause a hazard, or potential hazard; 4. The elements are persistent in the environment; and 5. The elements are becoming more important because of increased production and/or use. On the basis of the criteria, the following 14 trace elements or metalloids were selected: antimony, arsenic, beryllium, boron, cadmium, chromium, cobalt, copper, lead, mercury, nickel, selenium, tin, and vanadium. �Trace Element Toxicity Certain chemical forms of each of the 14 trace elements that were selected for evaluation are dangerous to humans and other biological organisms. A summary of the toxicity of the element, its most toxic compounds, food hazards, and air pollution hazards, and primary route of entry into humans is presented in Table 1. When ingested, antimony, arsenic, beryllium, cadmium, chromium, lead, mercury, selenium, and tin are toxic; when injected, most of the 14 elements are toxic. Some of the trace elements—arsenic, cadmium, lead, and mercury—function as cumulative poisons; others, including beryllium, chromium, nickel, and arsenic, are carcinogenic. Cadmium and lead shorten life and cause Table 1. neurological disruptions in experimental animals; at low concentrations, copper, mercury, boron, tin, cobalt, and nickel are toxic to certain species and have been used as biocides for certain plants and animals. Of the 14 elements with known degrees of toxicity to humans, only five presently constituteanimportantthreat to humans. These are, in order of their importance, lead, cadmium, mercury, arsenic, and nickel. Elements in Earth's Crust Toxic Trace Elements in the Environment Data on abundance, production, uses, environmental emissions and The abundance of the selected trace elements in the earth's crust, in dry soil, in the ocean, and in freshwater is s u m m a r i z e d in Table 2. This information is related to base lines and b a c k g r o u n d c o n c e n t r a t i o n s per element in non-polluted areas. Toxic Toxicity of selected trace elements Toxicity to humans Antimony dispersion, sources and pathways to humans are important to establish the extent of environmental pollution and the priorities for each trace element. This information is essential in order to select appropriate biological monitoring organisms, the type of monitoring required, and the number and geographic distribution of monitoring sites. Most toxic compounds toxic Air pollution hazard Primary route of entry into humans halides and stibnine L low potential hazard occupational exposure; therapeutic use; food As2O3, arsine Arsenic Food hazard hazard hazard occupational exposure; poisoning occupational inhalation Beryllium toxic all low potential hazard Boron low borane none unknown Cadmium toxic all hazard, L hazard occupational inhalation; food Chromium S hexavalent Cr low none occupational inhalation Cobalt low hazard none Copper low — none occupational exposure Lead toxic all, especially tetraethyl Pb hazard, L hazard occupational exposure Mercury toxic methyl Hg and alkyl Hg hazard hazard occupational exposure; food S nickel carbonyl low hazard occupational exposure hazard none occupational exposure L none occupational exposure; food none none occupational exposure Nickel Selenium Tin Vanadium toxic S low triethyltin tetraethyltin S—Specific compounds toxic or highly toxic, but element moderate to low toxicity. L—Leaching from container contaminates food. �Table 2. Abundance of trace elements in the environment Igneous rocks, earth's crust Ippm) Antimony Arsenic Beryllium Boron Cadmium Chromium Cobalt Copper Lead Mercury Nickel Selenium Tin Vanadium 0.2 1.8 2.8 10.0 0.2 100.0 25.0 55.0 12.5 0.08 75.0 0.05 2.0 135.0 Soil (dry) (ppm) 2.0-10.0 6.0 6.0 10.0 0.06 100.0 0.00033 0.003 0.0000006 4.6 0.00011 0.00005 0.00027 0.003 0.00003 0.0001 0.0054 0.00009 0.003 0.002 8.0 20.0 10.0 0.03 40.0 0.2 10.0 100.0 trace abundance in the earth's crust for the 14 selected elements ranges from 0.08 to 135.0 ppm and 0.03 to 100.0 ppm in dry soil. Toxic trace abundance in the ocean ranges from 0.0000006 to 4.6 ppm, and in freshwater from 0.00004 to 0.02 ppm. These estimates show the relative abundance of toxic trace elements when compared with iron (56,300 ppm) in the earth's crust, and calcium (41,500 ppm) or sodium (10,500 ppm) in seawater. Ocean (ppm) Table 3. Freshwater (ppm) — 0.0004 0.001 0.013 0.08 0.00018 0.0009 0.01 0.005 0.00008 0.01 0.02 0.00004 0.001 Toxic element production and use (1,000 metric tons) World production 1976 To provide an indication of pollution or potential pollution for each trace element, Table 3 presents annual world productions, annual U.S. consumption, and estimates of the annual air pollution emissions in the United States. More than one million metric tons of boron, chromium, copper, and lead are mined annually, and morethan one million metric tons of copper and lead are used annually in the United States. Less than one million metric tons of the other trace elements are 'mined or used in the United States annually. Antimony 75.065 Arsenic Toxic Element Production and Use Beryllium Boron Cadmium Chromium Cobalt Plants and animals acquire trace elements from naturally occurring U.S. consumption 1976 Human-caused air pollution in U.S. 1 year 34.581 0.25 39.28 4.196 9.76 10.0 0.519 0.45 2,263.68 18.594 4,203.9 28.228 1 10.23 5.886 284.44 8.287 10.5 0.4 16.46 0.75 Copper 8,083.09 2,206.60 200.0 Lead 4,625.79 1,466.47 234.5 Mercury Nickel Selenium Contamination of Biological Organisms environmental sources, such as soil or water (Table 4), through uptake in plants and ingestion in animals, as well as from exposure to environmental contamination. Normally, plants take up minute amounts of trace elements through the roots.The concentrations of a particular trace element in soil may be augmented through (1) the deposition of industrial air pollutants (fallout), (2) the addition of sewage sludge as a soil amendment, (3) the application of biocide to control unwanted weeds or pests, and (4) the direct application of industrial wastes. All these influences may cause elevated element concentrations in plants. The concentration of trace elements on plant surfaces also increases because air pollution fallout accumulates on leaves and stems (Table 4). Terrestrial animals normally ingest trace elements through food and water. Additionally, they may be acquired through breathing contaminated air, Tin Vanadium 8.429 872.342 1.364 225.24 25.537 2.241 160.35 0.486 20.0 4.885 0.50 61.934 30.0 4.65 20.0 �Table 4. Sources and pathways of toxic trace elements to plants Sb* As Be B Cd Cr Co Cu Pb Hg Ni Se Sn V X X Uptake by roots A. Soil or groundwater X X X X X X X X X X X X B. Fallout to soil from air pollution X X X X X X X X X X X X C. Sewage sludge soil amendments X X X X X X X X X X X D. Biocides applied to soil and/or seed X E. Surface water contamination X F. Fertilizers X X X X X X G. Industrial pollution X X X X X X X X X X X Uptake by leaves and stems A. Pollutant fallout from industrial sources B. Pollutant fa/lout from auto emissions C. Biocide applications to plants x x D. Pollution fallout from incineration of fossil fuels and refuse x x *Key: Sb = Antimony; As = Arsenic; Be = Beryllium; B = Boron; Cd = Cadmium; Cr = Chromium; Co - Cobalt; Cu = Copper; Pb ~ Lead; Hg = Mercury; Ni = Nickel; Se - Selenium; Sn = Tin; V = Vanadium. eating contaminated food, licking fur or feathers that are contaminated, receiving therapeutic drugs, or eating biocides or poison baits. Aquatic animals are subjected to similar sources of toxic trace elements when sewage and industrial effluents are discharged into aquatic environments (Table 5). All of the toxic elements under consideration are also found in, and distributed as, air pollutants. Biological Monitoring Biological organisms of many types have been used to measure the concentrations of trace elements at the pollutant site; these measurements are compared with different distances and with a control. Gradients of trace e l e m e n t s may be d o w n w i n d , downstream, or in a circumferential zone around a pollution source. The results of biological monitoring of e n v i r o n m e n t a l g r a d i e n t s were compiled from the published literature for each of the selected trace elements. Detailed documentation concerning the organism selected for gradient analysis, the pollution source, the tissue analyzed, the concentration gradient in the organism, and the authority are presented in this report. The p r i o r i t i e s for biological monitoring of toxic trace elements on a global, regional or national, and local scale are based on the toxicity of the element and of its most toxic compounds, its world production and use, its emissions intothe environment, its degree of hazard to humans and domestic animals, and its hazards when ingested as trace elements in food (Table 6, page 6). Value of Biological Monitoring If biological organisms are used for monitoring, they provide important data about, and actual responses to, pollutant exposure, instead of predicting biological responses and effects from physical and chemical measurements of the environment. Using living organisms as monitors is advantageous because they: 1. Provide early warning and help researchers to discover and identify toxic trace element pollutants in the early stages of environmental contamination; 2. Identify impact areas and sites of accumulation, and help researchers to measure the environmental levels of trace elements from sources; 3. Define critical pathways' of pollutants to humans from water, air, and food; �Table 5. Sources and pathways of toxic trace elements to animals Sb* As Be B Cd Cr Co Cu Pb Hg Ni Se Sn V Terrestrial A . Breathing contaminated a i r x B . Eating contaminated plant o r animal tissue C. Drinking contaminated water D. x x x x x x x x x x x x x x x x x x x x x x x x x Licking or preening fur or feathers E. x Receiving therapeutic drugs (domestic animals) x x x x x x x x x x x x x x x F. Eating biocides or poison baits Aquatic A. Metal in water B. Runoff and fallout C. Sewage and industrial waste outfalls D. Mine tailings or smelter waste leachate E. Contaminated plants, animals, or sediment x X X X X X x X X X X X X X X X X X X X X X X X X X X X X X X X X X x F. Biocides or runoff G. *Key: Lead shot Sb = Antimony; As = Arsenic; Be = Beryllium; B = Boron; Cd = Cadmium; Cr = Chromium; Co = Cobalt; Cu = Copper; Pb = Lead; Hg = Mercury; Ni = Nickel; Se = Selenium; Sn = Tin; V = Vanadium. 4. Establish base lines and integrated levels of trace elements over geographic areas; 5. Help show trends and rates of changes in levels of toxicity over time; 6. Integrate biological exposure of toxic trace elements with physical and chemical measurements in the environment; 7. Provide continuous monitoring systems over time, correlated with many other factors in the environment; and 8. Help researchers to evaluate the effectiveness of control measures. Both plants and animals can serve as biological monitors for showing changes in levels of pollution and environmental contamination, and as reference points for determining direct or indirect effects on humans. Types of Biological Monitoring Biological monitoring can be conducted using two fundamentally different but complementary methods. The first method is to measure the accumulation or concentration of toxic trace elements in selected biomonitoring organisms. The second method is to measure impacts or effects of toxic trace elements on organisms, or on populations and communities of organisms. Only the first method, that is, biological monitoring using accumulator and concentrator organisms, is considered here. Bioaccumulator organisms take up, accumulate, and store specific pollutants so that the organism's intake is g r e a t e r t h a n its o u t p u t . Bioconcentrator organisms selectively accumulate and concentrate specific pollutants by sequestering them into �Table 6. Priorities for Biological Monitoring Environmental monitoring Global Regional, national Local Food surveillance Antimony 0 0 X 0 Arsenic X XX XX XX Beryllium o o X o Boron o o o o Cadmium X XX XXX XXX Chromium 0 0 X o Cobalt 0 o X X Copper o o X X XXX XXX XXX XX Mercury o XX XX XX Nickel o X XX X Selenium o X X X Tin o o X X Vanadium 0 X X 0 Lead o—no biological monitoring (except minimum global base line surveillance) x—limited surveillance monitoring only of selected sites or foods xx—surveillance and exposure assessment monitoring of suspected sites and foods xxx—surveillance and exposure assessment monitoring of all needed sites and foods in a coordinated program certain organs or tissues, such as hair, bone, feathers, and so forth. A summary of the bioaccumu lation and bioconcentration of toxic trace elements for plants and animals is presented in Table 7. All elements, except antimony and beryllium, are concentrated by some biological organisms present in the environment. Recommendations for Biological Monitoring Recommendations for biological monitoring of the selected trace toxic elements are presented in Table 6. Of those evaluated, there is an urgent, high-priority, and well-defined need to monitor lead, primarily because of air pollution that comes from the c o m b u s t i o n of l e a d - c o n t a i n i n g Table 7. gasoline. The author also recommends global environmental monitoring for arsenic and cadmium at a lower priority than for lead. No global monitoring is recommended for the remaining 11 elements. On a regional or national, and a local basis, biological monitoring for arsenic, cadmium, lead, mercury, and nickel is recommended, especially around mining sites, near smelters, or near industries using these elements. Nationally, no monitoring is necessary for antimony, beryllium, boron, chromium, cobalt, copper, or tin. All elements, except boron, require at least surveillance monitoring at the local level. Biomagnification and Biominification Biomagnification is the increase in concentration of trace elements in organisms with increasing or higher trophic levels in food chains. This trophic level magnification indicates that concentrations of trace elements are higher in predators than they are in prey. For biomagnification to occur, the prey (or food organism) must contain a trace element in its edible or digested parts so that the element can be absorbed and retained by the predator. Biomagnification has been documented for all the toxic trace elements that were selected for this study. In terrestrial habitats, biomagnification results in increasing levels of trace elements from soil to plants for Bioaccumulation and bioconcentration of toxic trace elements Mollusks, Mosses, Mammals, Crustacea, and Higher lichens, and birds, and fish lower animals plants algae Antimony Arsenic Beryllium Boron Cadmium Chromium Cobalt Copper Lead Mercury Nickel Selenium Tin Vanadium x—Low or limited. X X X X XX xxx xxx XX X X X X X XX xxx xxx xxx xxx xxx xxx xxx xxx xxx XX X X xxx xxx xxx XX xxx xxx xxx xxx XX X XX X xx—Moderate. X xxx xxx X xxx xxx X X xxx X xxx xxx xxx xxx XX XX xxx—High to very high. X X �beryllium, boron, cadmium, lead, mercury, nickel, selenium, tin, and vanadium. In herbivores, increases take place for antimony, arsenic, cadmium, chromium, copper, lead, mercury, and selenium; only mercury and cadmium biomagnify in terrestrial carnivores. In aquatic habitats, levels of trace elements increase from seawater to algae for all except beryllium and boron. In herbivores, increases occur for antimony, arsenic, cadmium, and mercury; from prey to carnivores, an increase occurs for cadmium and mercury. Biominification is the decrease in concentration of trace elements in organisms with increasing or higher trophic levels in food chains. This phenomenon will occur if: 1. The trace element concentration is low in the edible or digested parts of prey or food organism, 2. The element has been sequestered or stored in inedible tissues, such as bones, Selecting Suitable Organisms The procedure for selecting biological monitoring organisms was accomplished in the following steps. First, data about concentrations of trace elements in various species were compiled, in comparable terms. Second, the data were evaluated, with regard to (1) the species that are suitable for monitoring, that bioaccumulate and bioconcentrate the elements, and that retain or store them in organs or tissues, (2) the usefulness of various species for monitoring the environmental gradients of trace elements, (3) the species' biomagnification and biominification in food chains, and (4) the species' biotransformation of trace elements to other forms or compounds. This information allowed a tentative selection of potential monitoring organisms. Finally, a selection was made based on several criteria; the organism must: • accumulate various elements 4. The element has been transformed from a highly absorbed to a poorly absorbed chemical form, and 5. The retention time of the trace element is short. For terrestrial organisms, frequently all 14 trace elements are biominified in trace level between soil and plants. H e r b i v o r e biomin if i c a t i o n is documented for all trace elements except beryllium; no change is documented for cobalt and mercury. Biominification from prey to carnivore occurs for antimony, arsenic, cadmium, lead, and mercury; no change occurs for chromium. In aquatic habitats, biominification from sediment or soil occurs only in freshwater rooted plants for cadmium, copper, and lead; no change occurs for mercury. In herbivores, decreases occur for arsenic, cobalt, copper, lead, and nickel; in carnivores, decreases occur for antimony, arsenic, cadmium, copper, lead, nickel, and selenium. No change occurs for mercury in herbivores or carnivores. Beryllium—In the only known research where environmental gradients of beryllium were measured, Spanish moss was used. Only three sites contained over 2 ppm ash weight out of 123 samples. Boron—Three species of higher plants have been used successfully to measure environmental gradients of boron. No animals, mosses, lichens, or algae have been tested. Cadmium—Forty-five species of plants and animals have been used to measure environmental gradients of cadmium pollution. The results show that several species of plants and animals appear to be excellent candidates for biological monitoring. • be common • be geographically widespread 3. The absorption of the trace element in the digestive system of the predator is low, measure environmental gradients of arsenic pollution. Horse and cow hair, and plants appear to be fairly effective as monitors. • be easily collected • be of adequate size to permit resampling of tissue • occur in the impact and unpolluted area • show correlation with environmental levels of trace elements. Also, background data for that species must be available. Major groups of organisms were considered with regard to the choice of potential monitoring organisms for each trace metal selected. The tentative selections of organisms for biological monitoring are presented in Table 8; these selections take into account the need for monitoring in the vicinity of humans; in urban, industrial, and rural areas, and in freshwater, estuarine, and marine environments. The use of biological organisms for monitoring environmental gradients of the select toxic elements is summarized next. Antimony—The experience of using biological monitors for measuring environmental gradients of antimony is limited and more studies are needed. Arsenic—Comparatively few species of plants and animals have been used to Chromium—Twenty-one species have been used to measure environmental gradients of chromium. Plants appear to be the most effective biomonitors, although some animals also show gradients. Cobalt— Environmental gradients of cobalt have been measured using two species of animals and eight species of higher plants. Gradients found in higher plants were most definitive. Chopper—Environmental gradients of copper have been monitored using 48 species of animals and plants. Certain mollusks, and both higher and lower plants, appear to be suitable for monitoring environmental gradients of copper. Lead—Over 95 species of animals and plants have been used successfully to monitor environmental gradients of lead. Mercury—Animals, mammals, birds, amphibians, fish, mollusks, crustaceans, insects, and annelid worms accumulate mercury and have been used in monitoring environmental gradients. Few species of higher plants accumulate mercury. �Table 8. Selected organisms for biological monitoring Human and animal hair Mammal organs Bird Bird feathers organs Mytilus edulis or other Fish Higher plants Mosses and lichens Algae — — Antimony F liver, G — — P P Arbacia, G Rose, F Arsenic F kidney, G G liver, P G E E Beryllium — P P P — G G — — Boron — bone, P — — — coral, Crustacea, G E — — Cadmium F kidney, G P kidney, liver, E G F E F F Chromium F spleen, F — — P G E F G Cobalt — — — — P E F — G Copper G liver, G G liver, G P oyster, E F F F Lead F kidney, bone, F E femur, E P E F F F Mercury F fiver, F F liver, E kidney, liver, G E E G G P Nickel G liver, P — kidney, G G G F E G Selenium G kidney, G G liver, G P — F — — Tin G kidney, P — kidney, G freshwater, G — F — G Vanadium G liver, G — — P tunicates, E F F G P G £ — £• = Poor. = Good. = Excellent. = Insufficient data. Nickel—for measuring environmental gradients of nickel, mollusks and crustaceans have shown limited potential. Higher plants, mosses, and lichens appear to be excellent biomonitors. Selenium—Plants appear to be the best organisms for measuring environmental gradients of selenium. Tin—Research reveals that no species of animal has been used to measure environmental gradients of tin. Higher plants have been effective, and certain species appear to have promise for biological monitoring. 8 Vanadium—Fish, mollusks, and some plants have been used as biomonitoring organisms for vanadium. Biological Monitoring Programs in the United States Under the Federal Environmental Pesticide Control Act of the United States, pesticide residues and heavy metals must be monitored in fish, shellfish, birds, small mammals, humans, and food destined for human consumption. The vehicles for monitoring are a series of networks, including the National Freshwater Fish Monitoring Network, National Ocean Monitoring Network, National Estuarine Monitoring Network, National Pesticide Network for Monitoring Birds, National Small M a m m a l s Monitoring Network, National Human Tissue Monitoring Network, and the National Food and Feed Monitoring Network. The author recommends that the National Environmental Pesticide Monitoring Networks. (NEPMN) program establish and implement monitoring that will assess the exposure to metals having a "high priority." The program proposed for biological monitoring within the NEPMN is presented in Table 9. �Table 9. Proposed biological monitoring for the National Environmental Pesticide Monitoring Networks (NEPMN) Organ or tissue Trace elements* Number of sites Monitoring frequency Status hair Pb, Cd, Hg, As 75 cities biennial proposed-new hair, liver Pb. Cd. Hg, As pilot study, then 75 cities biennial proposed-new edible tissue and dairy products Pb, Cd, Hg, As, Se, and Sn 1,200 slaughter houses bimonthly USDA program Pigeon feather, liver, kidney Pb, Cd, Hg, As pilot study, then 75 cities biennial proposed-new Starling feather, liver, kidney Pb, Cd, Hg, As 53 biennial modified-FWS program Eagle feather, liver, kidney Pb. Cd, Hg, As dead birds when available FWS program Poultry edible tissue and eggs Pb, Cd, As, Hg, Se, and Sn 1,200 slaughter houses bimonthly USDA program Food plants edible parts Pb, Cd, As, Hg, Se, and Sn 35 cities bimonthly FDA program Mosses and //chens, G whole plant Pb, Cd, As, Ni WO air monitoring sites quarterly proposed for EPA -new Fish (freshwater, few edible species) liver, muscle Hg,As 50 biennial modified FWS program Fish (marine, few edible species) muscle, liver Hg, Pb, Cd pilot study biennial modifed FDA and NOAA soft parts 12 trace elements pilot study annual "Mussel watch" global-new edible tissue Pb, Cd. Hg, As selected coastal sites annual modified FDA, FWS, NOAA Species Human, G Rat Cow, pig, sheep Mytilus edulis, G Oyster, crab, lobster G—Part of global surveillance programs of GEMS and "Mussell watch." *Key: Sb = Antimony; As = Arsenic; Be = Beryllium; B = Boron; Cd = Cadmium; Cr = Chromium; Co = Cobalt; Cu = Copper; Pb - Lead; Hg = Mercury; Ni = Nickel; Se = Selenium; Sn = Tin; V = Vanadium. �Dale W. Jenkins is a private consultant residing at 3028 Tanglewood Drive, Sarasota, FL 33579. John A. Santolucito is the EPA Project Officer (see below). The complete report, entitled "Biological Monitoring of Toxic Trace Metals," is in four parts: Volume 1. Biological Monitoring and Surveillance (Order No. PB 81 -1034 75; Cost: $18.50) Volume 2. Toxic Trace Metals in Plants and Animals of the World. Part I. (Order No. PB 81-103 483; Cost: $35.00) Part II. (Order No. PB 81-103 491; Cost: $33.50) Part I/I. fOrder No. PB 81-103 509; Cost: $21.50) These reports will be available only from: (prices subject to change) National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Environmental Monitoring Systems Laboratory U.S. Environmental Protection Agency P.O. Box 15027 Las Vegas, NV 89114 10 i us GOVERNMENT PRINTING OFFICE: 1961-757-064/0309 �United States Environmental Protection Agency Official Business Penalty for Private Use, $300 Center for Environmental Research Information Cincinnati OH 45268 Postage and Fees Paid Environmental Protection Agency EPA-335 � 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 184 Folder The folder containing the original item. 5351 Series The series number of the original item. Series VIII 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 Jenkins, Dale W. Description An account of the resource Corporate Author: United States Environmental Protection Agency (EPA), Environmental Monitoring Systems Laboratory, Las Vegas, Nevada Date A point or period of time associated with an event in the lifecycle of the resource 1981-02-01 Title A name given to the resource Project Summary: Biological Monitoring of Toxic Trace Elements ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/5ea2718ea5bcaefff964b2115df488b1.pdf 1233945ff12afb38858b03a9b141701c PDF Text Text Item D Number osssi D Author Kan Corporatfl Author Research Section, Agent Orange Projects Office, Depart 9>Han K- RODOrt/ArtiOlO TltlO Typescript: White Paper: Impact of Abolishing or Relocating the Research Section, Agent Orange Projects Office, Department of Medicine and Surgery Journal/Book Title Year Month/Day Color Number of Images D ° Descriptori Notes Tuesday, April 02, 2002 Page 5831 of 5840 �< WHITE PAPER SUBJ: Impact of Abolishing or Relocating the Research Section, Agent Orange Projects Office, Department of Medicine and Surgery I was told on February 26, 1985, by the Director of the Agent Orange Projects Office that DM&S is considering the options of either closing down or moving the Research Section of the Agent Orange Projects Office outside of the Washington area, and that he was asked to prepare a plan to relocate the Research Section by the close of business the same day. Consideration of abolishing the Research Section in the near future or relocating the Section outside of the Washington area should be viewed in conjunction with the inherent effect of terminating or substantially delaying the VA's two on-going major epidemiological studies and other research projects. Such actions will impact negatively on the following: 1. The goal of resolving the Agent Orange issues based on sound scientific and medical evidence; 2. The congressional and public perceptions of the VA's commitment to serve veterans; and 3. The efficient management of public funds. T* Impact on scientific and medical evidence The on-going Vietnam Veterans Mortality Study (WMS) is designed to provide, among other things, answers for 1) whether Vietnam veterans are at an increased risk of dying early from unexpected causes, 2) whether they are more likely to commit suicide, and 3) whether they are more likely to die from cancer, especially soft tissue sarcoma. The VA's Mortality Study is the largest and most statistically powerful study of its kind. Neither the on-going CDC epidemiology study nor the already published Vietnam veterans mortality studies (Air Force Health Study, New York State, Massachusetts State, Australian mortality study) can match the size and scope of the VA Mortality Study. Taking the issue of suicide among Vietnam veterans as an example, our preliminary data indicate that approximately 5,000 or 10% of the Vietnam era veterans in the VA Mortality Study, committed suicide. The CDC epidemiology study will have only approximately 100 suicide deaths available for analyses. The Research Section is now analyzing the �suicides with respect to Vietnam service, likelihood of exposure to Agent Orange, combat experience, length of service, rank, educational level, age, and many other factors. With Dr. Gronvall's (DCMD) concurrence, the Research Section and the Readjustment Counseling Service (10B/RC) have also planned to conduct an in-depth review of the suicides. This effort is designed to lead to developing a clinical profile of the suicides which may assist the VA in preventing suicides in the years to come. Without the completion of the VA Mortality Study no other studies, either on-going or planned, can provide such statistically powerful^ (results, and the questions about Vientam veterans' mortality patterns will be likely to remain inconclusive and unresolved. The possibility that exposure to Agent Orange may have increased the risk of soft tissue sarcoma (STS) has been of genuine concern to Vietnam veterans, their families and the Congress. The VA/AFIP Soft Tissue Sarcoma Study is designed to address this major health concern. This study is also the largest and the most statistically powerful in dealing with the question of Vietnam service and soft tissue sarcoma. This study complements the CDC epidemiology study and vice versa. The two epidemiologic studies now underway have been critically reviewed and approved by one or more of the following scientific groups: VA Advisory Committee on Health-Related Effects of Phenoxy Herbicides; Cabinet Council Agent Orange Working Group, Science Panel; Armed Forces Epidemiological Board. The scientific credibility of the study designs and the caliber of the principal investigators have never been questioned. The scientific community is anticipating that the results from the studies will resolve many Agent Orange related health questions as evidenced by numerous invitations to scientific meetings to present the status of the various research projects. Examples are American Chemical Society, Society for Occupational and Environmental Health, International Symposium on Dioxin, Armed Forces Epidemiological Board. �Obviously, no single study can provide all the answers. However, the results of4 the on-going studies being conducted by the Research Section will provide some of the answers in an unequivocal way. 2. Congressional and public perceptions of the VA's commitment The VA has repeatedly reported in congressional testimony its strong commitment to the resolution of the Agent Orange issues and on its in-house Agent Orange research efforts. Both the Senate and House Veterans Affairs Committees have requested and are receiving quarterly reports, as is the GAO, on the status of VA Agent Orange research activities. '' The status of in-house Agent Orange research is routinely published in the "Agent Orange Review" prepared by VA's Office of Public and Consumer Affairs. The review is published periodically to provide information on Agent Orange to concerned veterans and their families. Approximately 115,000 Vietnam veterans who came to VA hospitals for Agent Orange examinations, and other concerned veterans, are on the mailing list. The creation of the Research Section and the research which is now being conducted by the section is a reflection of the original and continuing congressional and service organization's concern and interest in the health issues. To terminate the research projects would cast serious doubts on the depth of the VA's original commitment and damage further the VA's image as the primary advocate of the veterans. 3. Mismanagement of public funds Approximately 2.2 million dollars have already been spent or obligated on the studies and research projects to bring them to their current status of partial and near completion. Funds have been disbursed through contracts and interagency agreements. If the Research Section were to be abolished, these studies and projects are likely to remain incomplete, and the scientific issues which they address will remain unresolved. Therefore,, an apparent saving by eliminating 5 FTEE frcm the Research Section would result in the waste of over two million dollars. �Recommendation 4« If DM&S is to implement the abolishment of the Research Section/ it should be carried out on a "phase-out" basis. Results of the mortality study and the soft tissue sarcoma study are expected to be available by the 3rd quarter of BY 1986 and reports will be written, reviewed, and published by the 4th quarter of EY 1986. The VA can then close down the Research Section by the end of the 4th quarter of EY 1986, without extraordinary loss of funds and with the satisfaction that the VA has contributed to the resolution of the Agent Orange controversy based on the scientific evidence. Relocation of the Section outside of the Washington area will substantially delay completion of the studies even if all key professional staff choose to move and adequate support services can be arranged. Such a move will also entail significant relocation expenses. However, if the relocation is only for the duration of completing the two studies some will choose not.to move and it will be very difficult to recruit competent scientists with national reputations. Relocation of the Section outside of the Washington area is not recommended. HAN K. RANG, D . . . rPH Chief, Research Section Agent Orange Projects Office (1QA7B) Department of Medicine and Surgery March 8, 1985 � 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 207 Folder The folder containing the original item. 5831 Series The series number of the original item. Series VIII 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 Kang, Han K. Description An account of the resource Corporate Author: Research Section, Agent Orange Projects Office, Department of Medicine and Surgery Title A name given to the resource Typescript: White Paper: Impact of Abolishing or Relocating the Research Section, Agent Orange Projects Office, Department of Medicine and Surgery ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/aedc6bae3ac6b6b8214dcde83b2e035f.pdf 5d339a4efb9ef44f9d004654b819fa96 PDF Text Text Item D Number oseos Author D NotSBann8I| Keller, Carl Corporate Author ROport/ArtiCiO TltlO Memorandum to Members of the Science Panel, AOWG, from Carl Keller, Chair Pro-Tern, Science Panel, Agent Orange Working Group, with subject: Science Panel Meeting, dated November 25,1983. Journal/Book Title Year 000 ° Month/Day Color D Number of Images ° DOSttPiptOfl NdtBS Memo announces the next meeting of the Science Panel and distributes the minutes of the November 2,1983 meeting for approval. Alvin L. Young filed these documents together with others in a folder labeled, "Agent Orange Working Group Science Panel, Current Folder." Tuesday, March 19, 2002 Page 5605 of 5611 �Public Health Service National Institutes of Health DEPARTMENT OF HEALTH & HUMAN SERVICES National Toxicology Proqram . . . Memorandum Date November 25, 1983 From Carl Keller, Chair Pro Tern, Science Panel, Aaent Oranqe Workinq Group/ f/l/ vw.t Subject Science Panel Meetinq To Members of the Science Panel, AOW6 There will be a meetinq of the Science Panel on December 5, 1983, from 9:30 a.m. to 12 noon, in Room 405A, Hubert Humphrey Buildino. We will finalize plans for review of the international studies of the association between soft tissue sarcoma and dioxin-contaminated products requested by Senator Cranston. I am also enclosinq the Minutes of the November 2 meetina for your approval. Attachments. RECEIVED NOV 2 81983 Special Hojews hfe.fr (1020 �DEPARTMENT OF HEALTH & H U M A N SERVICES !""bllc Health Service National Institutes of Health Memorandum Date November 21, 1983 From Carl Keller, Chair Pro Tern, Science Panel, Agent Orange Working Group Subject Minutes of November 2, 1983 Meeting TO Members of the Science Panel, AOWG The Science Panel met at 9:30 a.m., in Room 303-305A of the Hubert Humphrey Building on November 2, 1983. A list of attendees is appended to these Minutes. Several informational handouts were distributed including a copy of the Background document and Dioxin Strategy drafts prepared by EPA on August 15, and comments from the Veterans Administration on these documents.. Several announcements were made including the postponement of the release of results from the Ranch Hand Morbidity Study and the cancellation of the Advisory Committee meeting to review these results. Other items discussed followed the previously prepared agenda. Members unanimously accepted the report on the Australian Birth Defects Study prepared by Dr. Mulinare's subcommittee and offered several suggestions for clarification. These included a distinction between confounding, which must be designed out of a study, and effect modification which can be analyzed out; a more forceful statement that the Australian Study was intended to be a Vietnam Experience Study; and some explanation of the significance of underascertainment of Birth Defects which may have occurred. Other comments of an editorial nature should be submitted to the Chair as soon as possible so that a re-draft can be prepared. The letters from Senator Cranston and Dick Christian to Dr. Erickson concerning the CDC protocols were briefly discussed and it was noted that CDC is preparing a revised draft of the protocols and will respond to the comments of all reviewers simultaneously in the very near future. Members of the Science Panel unanimously felt that the proposals for further study of biopsy material obtained by Dr. Arnold Schecter following the Binghamton State office building incident would not provide a significant scientific contribution to issues of concern to the Science Panel. This was based on the likelihood of unmeasurable multiple and continuing exposure to a number of potentially toxic substances endured by the subjects involved which would make a pharmacokinetic study of dioxins unfeasible in this group. Members of the Science Panel recognized that the request by Senator Cranston to review international studies of the association between soft tissue sarcoma and dioxin-contaminated products is a substantial undertaking. It was agreed that a carefully selected subcommittee be assigned this task and the final effort reviewed by the Panel. Suggested names for this subcommittee should be submitted to the Chair as soon as possible so that we can begin this assignment. A copy of a paper on this subject, prepared by Dr. Larry Hobson in March 1983, is attached to these minutes for your information. The Science Panel adjourned at noon, November 2, 1983. �DEPARTMENT OF HEALTH & HUMAN SERVICES Office of the Secretary Washington, O.C. 20201 AGENT ORANGE WORKING GROUP/SCIENCE PANEL Meeting — November 2, 1983 — 9:30 - 12:30 ATTENDEES Carl Keller , .Chair 496-3511 William Triplett Acting DUSIGA...' 245-6156 Peter E.M. Beach OVA/DUSIGA 245-2210 Lt. Col. Robert Capell.....Air Force Surgeon General..767-5078 Renate Kimbrough ...CDC PTS-236-4102 Joe Mulinare. .CDC.... .PTS-236-4080 Dick Christian .. ...AAOTP 653-1828 Doug Clark Cpt. Rob Lipnick Ed Weiss - AADTF AAOTF .OGC/HHS Michael Gough... .OTA. .226-2070 OTA VA/AOPO ., .226-2070 389-5534 VACO/AOPO EPA 389-5411 382-2897 Hellen Gelband Al Young Lawrence B, Hobson.. Donald G. Barnes 653-1829 653-1829 .. 245-1920 � 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 199 Folder The folder containing the original item. 5605 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Keller, Carl Title A name given to the resource Memorandum to Members of the Science Panel, AOWG, from Carl Keller, Chair Pro-Tem, Science Panel, Agent Orange Working Group, with subject: Science Panel Meeting, dated November 25, 1983. ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/95b51612285ea826290a89dbb22fbd6c.pdf 5cb332ceb4875607b8f5fd016d10240c PDF Text Text Item n Number °5698 Author Keller, Carl G j^t Scanned Corporate Author Report/Article TltlB Report of the AOWG Science Subpanel, June 3, 1986, Appendix I: Methodology and Results of Pilot Tests for Agent Orange Exposure Among Vietnam Veterans Journal/Book Title Year 000 ° Month/Day Color D Number of Images ° Descrtoton Notes Tuesday, March 26, 2002 Page 5698 of 5743 �REPORT OF THE AOWG SCIENCE SUBPANEL June 3, 1986 APPENDIX I METHODOLOGY AND RESULTS OF PILOT TESTS FOR AGENT ORANGE EXPOSURE AMONG VIETNAM VETERANS Prepared by Carl Keller, Ph.D. and Richard Christian, C.R.M. �SUMMARY OP RESULTS OP PILOT TESTS AMONG VIETNAM VETERANS Military records were examined by the U.S. Army and Joint Services Environmental Support Group in order to estimate the amount of daily troop dispersion and the distribution of encounters with Agent Orange applications during the period from October 1, 1966 through March 31, 1969, in Corps Tactical Zone III, Vietnam. Procedures and results of these tests were as follows: Number of Encounters Seven Battalions were selected for this test, each consisting of five companies, making 35 companies in all. The daily locations for each of these companies were determined from existing military records and matched against all known Agent Orange applications during the designated 30-month period of the test. Matching was done for encounters within three different time-distance combinations, including: 1) Within 500 meters on the day of application ("Wet" Encounters) 2) Within 2 kilometers and within 3 days of application 3) Within 2 kilometers and within 6 days of application. A summary of the numbers of encounters among the 35 companies and within each of these time-distance combinations is shown in Table I. Data are included for ranch hand (Herbs) and perimeter (Services Herbs) as well as combined (either tape) and for the cumulative combined, e.g., 5 companies had 14 or more encounters within 2 kilometers and within 6 days of an Agent Orange application. Although 5 out of 35 is 14%, this may be an overestimate of what might be found among all companies in III Corps since the 7 Battalions were chosen from among those thought to be the most likely exposed to ranch hand applications. Among approximately 5000 veterans known to have been assigned to these seven Battalions, the number of individuals with various numbers of encounters were determined as shown in Table II. Since there are some veterans who were assigned to one of these Battalions who have not yet been identified, it is estimated that some of the proportions of men with more than a few encounters will be greater than indicated in the Table, although how much is unknown. A more detailed list of procedures and results of the number of encounters pilot tests follow Table II in this Appendix. �Troop Dispersion Six companies, one from each of six of the Battalions used in the other pilot tests, and six days in 1967 and 1968 were selected at random. On each of the 36 company-day combinations (six companies times six days), the recorded location for each element of the company was plotted on the appropriate topographic map of Vietnam. The distance between various elements of the same company on the same day were^ measured from the maps. The maximum distance between any two elements of the same company on the same day were according to the following distribution: NUMBER OF OCCURRENCES 17 1 5 4 5 4 36 DISTANCE IN KILOMETERS 0 (Company all in one location) <2 2-4 5-9 10-19 20-29 6 (Average) �RESULTS OF PILOT TESTS FOR NUMBER OF ENCOUNTERS WITH AGENT ORANGE APPLICATIONS TABLE I NUMBERS OF COMPANIES WITH VARIOUS NUMBERS OF ENCOUNTERS # OF ENCOUNTERS 0 1 500 M / 1 DAY 1 *HERBS 33 **SERVICES HERBS 33 2 EITHER TAPE 31 3 4 CUMULATIVE 35 CUMULATIVE % 100% 11% 2 3 4 5 6 7 8 2 6% 1 2 6% 9 10 11 13 14 15 1 1 1 3% 2 KMS / 3 DAYS 24 6 I 1 HERBS 3 7 SERVICES HERBS 17 7 2 1 EITHER TAPE 14 4 4 5 3 3 CUMULATIVE 9 35 21 16 12 6 CUMULATIVE % 100% 60% 46% 34% 26% 17% 2 6% 1 3% 1 3% 1 1 1 3% 2 KMS / 6 DAYS 1 HERBS 21 5 4 1 1 2 4 SERVICES HERBS 17 2 2 2 2 3 1 1 1 14 3 1 4 EITHER TAPE 3 1 1 2 1 1 4 CUMULATIVE 35 21 18 15 14 10 10 9 4 8 6 6 6 6 5 CUMULATIVE % 100% 60% 51% 43% 40% 29% 29% 26% 23% 17% 17% 17% 17% 14% 11 TABLE II NUMBERS OF INDI VIDUALS WITH VARIOUS NUMBERS OF ENCOUNTERS # OF ENCOUNTERS 0 9 10 11 12 13 15+ 500 M / 1 DAY EITHER TAPE <4834 104+ 62 CUMULATIVE "5000 166+ 62 CUMULATIVE % 100% 3%+ 1J 2 KMS / 3 DAYS EITHER TAPE ~4360 268 157 69 47 31 27 I 3 35 41 40 37 37 CUMULATIVE "5000 640 372 215 146 99 68 CUMULATIVE % 100% 13% 7% 4% 3% 2% 1% 1% 1% 1% 1% 2 2 0% 2 KMS / 6 DAYS 4 37 1 EITHER TAPE "4277 252 33 38 154 57 11 72 30 6 28 CUMULATIVE "5000 723 471 438 400 246 189 178 106 76 70 70 42 41 37 CUMULATIVE % 100% 14% 97, 9% 8! 5% 4% 4% 2% 2% 1% 1% 1% 1% 1 *HERBS (refers to Computer Listing of Ranch Hand Applications, Fixed-Winged and Helicopter) **SERVICES HERBS (Selected Fixed-Winged, ground applications not listed on HERBS) �DEPARTMENT OF THE ARMY US ARMY ft JOINT SERVICES ENVIRONMENTAL SUPPORT GROUP 1790 K STREET N.W. ROOM 210 WASHINGTON. DC 2OO06 MCPIY TO ATTENTION TO April 10, 1986 DAAG-ESG • ESG PILOT STUDY OF THE MAIN AGENT ORANGE EPIDEMIOLOCAL STUDY 1. The Science Panel, Agent Orange Working Group tasked the Environmental Support Group to perform a Pilot Study for the Congressionally mandated Agent Orange Epidemiological study. ESG was given the mission due to the concerns expressed by the Centers for Disease Control and other scientific agencies that the study had not produced enough qualified heavily exposed personnel. 2. There are many detailed functions in which ESG will have to perform for the Agent Orange Epidemiological Pilot Study. Those functions are listed below. a. CDC has provided ESG a list of Infantry Battalions that had the highest likelihood for exposure after being matched against the Ranch Hand Herbs Tape. ESG selected 6 of the units recommended by CDC and added one additional Artillery Battalion. ESG will identify the gaps and fill in the days in which no grid coordinates were previously reported on the 7 battalions for all companies covering the period 1 October 66 to 31 March 69. This will require a new tape (gap fillers) that must be integrated with the old CDC tape to establish a complete tape with all the tracking data. b. When all the daily location data has been accomplished ESG will computer match the complete grid coordinate locations of each company by date against the Ranch Hand and Service Herbs Tapes to produce an exposure opportunities score using the below listed criteria. - 2 kilometers, 3 days - 2 kilometers, 6 days c. ESG will produce a listing for each company documenting by date sequence each herbicide encounter and type of exposure (i.e. Ranch Hand, Perimeter, Abort, Ground Spray). d. ESG will review morning reports to insure the veteran was in the unit and on duty in the field when a spray mission occurred. A veteran could have been in the hospital or on R & R in Bangkok, Thailand, when his unit experienced a herbicide encounter and hence would not have had an exposure opportunity. �e. Our goal is to identify and qualify every soldier assigned to each of the 7 battalions that were selected for the study. The qualification criteria used for the Pilot Study are listed below. - Single tour enlistee - E5 or below - The veteran must have been in Vietnam sometime during or between 1 October 1966 to 31 March 1969 to qualify as a heavily exposed cohort. - Served anytime in a tracked unit. Non-exposed personnel must have served their entire tour in a tracked unit. 3. To date, ESG has completed filling the location gaps on all 7 battalions. Extensive quality control functions were performed on the data and quality control was initiated on the existing location data that had previously been provided CDC. This information is keypunched and is almost ready to be matched against the spray mission tapes with respect to the aforementioned time and distance requirements. 4. ESG has requested approximately 2,500 military personnel files from the National Archives Records Administration in St. Louis, Missouri. Since ESG has already qualified 15,900 study subjects, 5,000 will be targeted for use in the pilot study. Once the records are obtained the records will be researched and personnel data elements abstracted on each qualified exposed and non-exposed study subject. RICHARD S. CHRISTIAN, C.R.M. Director � 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 201 Folder The folder containing the original item. 5698 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Keller, Carl Richard Christian Title A name given to the resource Report of the AOWG Science Subpanel, June 3, 1986, Appendix I: Methodology and Results of Pilot Tests for Agent Orange Exposure Among Vietnam Veterans ao_seriesVIII 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 198 Folder The folder containing the original item. 5575 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Keyworth, George A. Title A name given to the resource Memorandum for Ann Fairbanks from George A. Keyworth, Subject: Agent Orange, July 1981 ao_seriesVIII 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 194 Folder The folder containing the original item. 5485 Series The series number of the original item. Series VIII 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 Keyworth, Jay Title A name given to the resource Memorandum for David Stockman, from Jay Keyworth, subject: Enrolled Bill H. R. 5172, dated October 19, 1984 ao_seriesVIII 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 185 Folder The folder containing the original item. 5365 Series The series number of the original item. Series VIII 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 Klein R. Source A related resource from which the described resource is derived Wounded Men, Broken Promises 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 Chapter 6: The Poison Orange ao_seriesVIII 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 199 Folder The folder containing the original item. 5634 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Kogan, Michael D. Richard W. Clapp Date A point or period of time associated with an event in the lifecycle of the resource January 25 1985 Title A name given to the resource Mortality Among Vietnam Veterans in Massachusetts, 1972-1983 ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/0ff6559eef5044078c01d024edba6a5a.pdf 6d63cfc8de133f19a050f0a242bf6f22 PDF Text Text Item D Number °5601 Author KuH Bart D not Scanned - Corporate Author RepOTt/ArtlGlB TltlO Memorandum to the Under Secretary/Chairman, Agent Orange Working Group from Bart Kull, Special Assistant to the Under Secretary for Intergovernmental Affairs, with Subject: Agent Orange Working Group (AOWG) Membership and Scope and Purpose of AOWG, dated August 18, 1983 Journal/Book Title Year 000 ° Month/Day Color D Number of Images ° DOSOripton Notes Includes attachments. Tuesday, March 19, 2002 Page 5601 of 5611 �DEPARTMENT OF HEALTH & HUMAN SERVICES Office of the Secretary Washington, D.C. 20201 August 18, 1983 MEMORANDUM TO: The Under Secretary/Chairman Agent Orange Working Group PROM : Bart Kull <£*, SUBJECT : Agent Orange Working Group (AOWG) Membership Special Assistant to the Under Secretary for Intergovernmental Affairs and Scope and Purpose of AOWG. At the meeting of the AOWG on August 10, 1983, you raised two issues: (1) the scope and purpose of AOWG; and (2) the apparently large membership of the AOWG. As to the first issue, there is attached for your consideration a decision memorandum presenting alternatives regarding the clarification of the AOWG's functions* w - " N .-'•'-' ."•'••"" ' '•" . '; •' •" '•" Regarding issue 2,, the precursor of AOWG, the Interagency Working Group (IWG) was established on December 11, 1979. It was composed of the the following: ; . - . '. Members ' -.' •' ' : -: • . " ' . . . ' • . . • " . - . " ' . . . ' • Department x>f Health and Human Services Department of Defense . Veterans Administration Observers Department of Agriculture Environmental Protection Agency Department of Labor (OSHA) Sx-Officio Participant • White House Office of Science and Technology Policy �On July 17, 1981, President Reagan raised the IWG to Cabinet Council status and renamed it the AOWG. In his implementing memorandum of August 21, 1981, secretary Schweiker established the following: Members Department of Health and Human Services Department of Defense Veterans Administration Department of Agriculture Department of Labor Environmental Protection Agency ACTION Office of Management and Budget White House Office of Science and Technology Policy White House Office of Policy Development White House Council of Economic Advisors Advisor General Accounting Office Observer - : Office of Technology Assessment Most agencies listed have had a pertinent, and at times vital, input into the Working Group*s functions. Those agencies involved.in research naturally dominate the meeting discussion. Others, such as ACTION, Labor, the White House Office of Science and Technology Policy and Office of Policy Development attend in order to keep abreast of developments in research for policy purposes. The Office of Management and Budget is important inasmuch as it is familiar with potential research funding needs. The Department of State is useful in providing members with facilitated access to foreign research and policy developments. The White House Council of Economic Advisors has sent a representative only once to my knowledge. The impression of •overcrowding" at the meetings is often due to the practice of the more concerned agencies sending one or more alternates and some observers to each meeting. You may wish to consider having only the principal from each agency at the conference table with alternates and observers seated apart. You may also wish to consider dropping ACTION, Labor and the White House Council of Economic Advisors from membership. �I would recommend writing to each present member agency head advising them of the present representatives and requesting they issue, or re-issue, a letter designating one principal representative and one alternate. Attachment �Office of the General Washington, D.C. 20201 MEMORANDUM TO : The Under Secretary Thru: ES FROM : Edwin M. Weiss ffiy Legal Counsel Agent Orange Working Group SUBJECT: Options Available to Clarify the Purpose and Sonnft of +-h*> ADWfi Scope nf the AOWG Purpose To address the question raised during the August 10, 1983, meeting on the specific mission of the Agent Orange Working Group (AOWG). Background On December 11, 1979, Stuart Eizenstat, Assistant to President Carter, requested HHS to take the lead in convening an Interagency Work Group (IWG) to study possible long-term health effects of phenoxy herbicides and contaminants. The IWG was to oversee, coordinate and set priorities among federal government research activities in this area. In designing a research agenda, the IWG was to take into consideration the duration of exposure to Agent Orange by Vietnam veterans (Tab A). President Reagan announced on July 17, 1981, that the Administration has re-established an expanded IWG as the Agent Orange Working Group (AOWG) and raised its status to Cabinet Council leve. In his implementing memorandum of August 21, 1981, Secretary Schweiker stated that the President had been motivated because he shared the widespread public and Congressional concern over possible adverse health effects among Vietnam veterans exposed to Agent Orange and other substances. (Tab B, p. 1) In the same memorandum the Secretary reaffirmed the board charter language of December 11, 1979. (Tab B, p. 2) �Page 2 — The Under Secretary Discussion The charter of IWG, and. now of AOWG, is a broad mandate which provides in explicit terms the authority to design and direct all research activities on the health effects of exposure to phenoxy herbicides and contaminants, with exposure to Agent Orange by Vietnam veterans as one part of the total research design. The contaminants mentioned include that class of chemicals known as the dioxins produced during the manufacture of these herbicides. The renaming of the IWG by President Reagan appears to focus the Work Group's primary responsibility on Agent Orange, but without a change in the'language of the charter, a lack of clarity regarding the AOWG's purpose and scope has resulted. Both the Chair (Pro-Tern) of the Cabinet Council on Human Resources and the Chair, AOWG, have made previous efforts to resolve this uncertainty of purpose. For example, in a report on the activities of AOWG released to the press on October 21, 1981, the Chair, AOWG, stated that AOWG "coordinates all Federal Agent Orange research. It does not undertake any research on its own but rather acts as the coordinator and monitor." (Tab C, p. 2) This purpose was affirmed by Secretary Schweiker in a letter dated April 8, 1982 to the Secretary of Defense in which he wrote that AOWG "has the responsibility for overseeing, coordinating and setting priorities among federal research relating to Agent Orange." (Tab D). 'Nevertheless, the language of the charter is clear that the AOWG's purpose and scope is to respond to the array of issues regarding the long-term health effects of phenoxy herbicide exposure. Selecting only the Agent Orange segment of this task may subject the AOWG, to criticism for non-performance of the other elements of its mission. Options The following options are available to clarify the purpose and scope of the AWOG. . 1. The AOWG charter could remain undisturbed with the scope and purpose to be defined through authoritative statements at the discretion of either the Chair (Pro-Tern) of the Cabinet Council on Human Resources or the Chair, AOWG. �Page 3 — The Under Secretary 2. The AOWG charter could be amended to state that AOWG is specifically concerned with research related to Agent Orange exposure in the context of the Vietnam experience. Discussion Option 1 by retaining the broad mandate of the charter, permits the Chair authority to widen or narrow the scope of AOWG at its discretion. Inherent in this option is the possible dilution of AOWG's concentration on Agent Orange as public pressure mounts on dioxin as a domestic issue. The potential for duplication and conflict is also present in Option 1 because HHS has in place a PHS-headed interagency group, The Committee to Coordinate Environmental and Related Programs, to consider the health-related effects of dioxin. Option 2 while limiting the authority and discretion of the Chair, will make clear that AOWG is, as its name indicates, concerned with Agent Orange. It will also permit the charter to reflect the President's purpose in establishing the AOWG, i.e., that the Administration is giving special consideration to the concerns of Vietnam veterans. Recommendation I recommend that you approve Option 2 which cures the current ambiguity resulting from the specific purpose inherent in the name of the Group and the expansive purposes contained in the charter language. It will send a clear and timely signal to Vietnam veteran groups and to the Congress by reaffirming HHS commitment in this area just as CDC is undertaking the Vietnam veterans' study mandated by statute, 95 Stat. 1061; 38 U.S.C. 219 note. If you select Option 2, I suggest you appoint an ad hoc committee to draft alternative charter amendments to be submitted to you for consideration. Decision / Approve Option 1: Approve Option 2: Other : Date : Attachments Tabs A-D �^4^^^h ^^^ 7&E W H I T E HOUSE 1 /• ^^ WASHINGTON '~1r!? />-. December 11, 1979 MEMORANDUM FOR SECRETARY OF DEFENSE SECRETARY OF HEALTH, EDUCATION, AND WELFARE ADMINISTRATOR OF VETERANS AFFAIRS SUBJECT: Interagency Work Group to Study the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants In recent months the public and the Congress have become concerned about adverse health effects to veterans following their possible exposure to herbicides, particularly Agent Orange, while serving in Vietnam. Although there are suggestions of adverse health effects of human exposure to such herbicides and contaminants, there is currently an inadequate scientific basis for concluding that health problems experienced by Vietnam veterans were caused by previous exposure to herbicides. Moreover, there is inadequate information on the long-term health effects of phenoxy herbicides in general. Individually, each of your agencies has a strong interest in resolving this issue. Several studies have been initiated to answer 'questions about the possible health effects of exposure to herbicides and more generally to the class of substances called the dioxins. Collectively, the Federal government needs to have reliable data and criteria on which to base decisions and policies which affect the entire country. Although I' am aware that there has already been extensive interagency cooperation on these issues, I believe there is a need for formal interagency coordination. Therefore, I request that you establish an interagency work group to coordinate agency efforts to determine if there are long-term health effects following exposure to phenoxy herbicides and contaminants, with special immediate focus on exposure of Vietnam veterans to Agent Orange. This interagoncy group should: 1. Oversee, coordinate, and set priorities among Federal government research activities designed �-o r-late exposure to phenoxy herbicides to longrorni"hcalth effects. t *" "; ,- ...;..n a research agenda to assure that the p'-^-rnl government conducts comprehensive . ::irch on the long-term health effects of ' ^ com^unds, in response to both scientific r " a'-'d liolicy needs. The type and duration of , .' ,r-urn to Agent Orange by Vietnam veterans 1u^i be "considered in the research agenda design . so that the Veterans Administration will be able •to establish sound policies for determining - • • • compensation for veterans exposed to Agent •:: Vietnam, should a relationship ....... smruicide exposure and long-term adverse hoaitn'crfects be established. The research aacnda should build on current agency activities, including the Department of Defense's Ranch Hand study. The interagency work group should identify the appropriate agencies to conduct the recommended research, either individually or through ;joint efforts. 3. Provide technical support to individual agencies and independent researchers in the formulation, development, and implementation of research on the biomedical effects of phenoxy herbicides and contaminants. 4. Assure that the protocols and methodology of ongoing and proposed Federally funded research studies will produce valid, reliable, timely, and relevant data, and periodically review the status of such research. 5. Assure that all relevant research findings, whether rsublicly or privately financed, are promptly made available to the public and the Congress, in . a comprehensible and comprehensive fashion. The work group should establish a working relationship with the Veterans Administration's Advisory Committee on Health-Related 'Effects of Herbicides and should promptly provide the Committee all relevant information as it becomes available. I am asking Secretary Harris to take the lead in convening the interagency group and would like to have an initial report on the progress of the group submitted to me by �February 15. The of current agency progress reports, other agencies on initial report should indicate the status activities, a proposed schedule for public and any recommendations for inclusion of the work group. I have asked the Office of Science and Technology Policy to be an ex-officio participant on the work group. In addition, the Department of Agriculture, the Environmental Protection Agency, and the Occupational Safety and Health Administration will initially participate on the work group in an observer status. . Stuart E. Ei^nstat Assistant to the President for Domestic Affairs and Policy cc: Secretary of Agriculture Administrator, Environmental Protection Agency Assistant Secretary of Labor for Occupational Safety and Health Director, Office of Science and Technology Policy �THE WHITE HOUSE WASHINGTON AU621 1981' MEMORANDUM TOR: SECRETARY OF DEFENSE SECRETARY OF AGRICULTURE SECRETARY OF LABOR DIRECTOR, OFFICE OF MANAGEMENT AND BUDGET ASSISTANT TO THE PRESIDENT FOR POLICY DEVELOPMENT CHAIRMAN, COUNCIL OF ECONOMIC ADVISERS DIRECTOR OF ACTION ADMINISTRATOR, ENVIRONMENTAL PROTECTION AGENCY ADMINISTRATOR OF VETERANS AFFAIRS fRfCTOiT, OF/Jq£ OF SCIENCE AND TECHNOLOGY :Y - FROM SUBJECT (a SS5RSTARY" RICHAROlCHWllKER CHAIRMAN PRO-TEM, CABINET COUNCIL ON HUMAN RESOURCES : Agent Orange Working Group The Administration has reviewed the excellent work of the Intaragency Work Group to Study tne Possible Long-Term Health .Effects of Phenoxy Herbicides and Contaminants and believes that it has Bade significant progress toward fulfilling its important mandate. By bringing together knowledgeable scientists from the various Federal departments and agencies the Work Group has identified ongoing research activities on phenoxy herbicides and contaminants and begun to develop anJ organize the means to carry out additional needed scientific research. President Reagan shares the widespread public and congressional concern over possible adverse health effects among Vietnam veterans exposed to Agent ..Orange and other substances. The President stated, during his meeting with national veterans organization leaders at the White House on July 17, 1581, that the Administration is giving special consideration to those concerns of Vietnam veterans. At the White House meeting, the President announced that the administration had re-established an expanded Working Group as the. Agent Orange Working Group and raised its status to Cabinet Council level. The President is personally determined to assure that the full .resources of the Federal government are �' * • available to support the Working Group's continuing efforts. The deiisisr. ta rs-^stablish and expand the membership of the working i-'jvry and to make it an integral part of the Cabinet Council on Human Resources reflects the President's commitment and accords the highest priority to its mission. As Chairman Pro-Ten of the Cabinet Council on Human . Resourcest 1 am, accordingly, reaffirming by this memorandum the Agent Orange Working Group's mandate of December 11, 1579 and providing specific guidance as to how that mandate is to be carried out in accordance with the Cabinet Council *s decisions. The Department of Health and Human Services shall continue to have lead responsibility for overall direction: and management of the Agent Orange Working Group. The Secretary of Defense and the Administrator of Veterans Affairs shall continue to assure that their respective agencies participate fully in *11 Working Group* activities. The Departments of Agriculture and Labor and the Environmental Protection Agency, each or wnicn have until now been observers, shall assume full membership and their respective agency heads shall assure that those agencies participate fully in all Work Group activities. Zn addition, ACTION, the Office of Management and Budget, and the Council of Economic Advisers, as well as the White House Office of Science and Technology Policy and the Office of Policy Development, shall assume membership on the Working Group and tftc htf&ds of those agencies and offices shall assure that the resources of their respective -agency or office are fully available to support it. Also, the congressional Office of Technology Assessment, which has been actively involved in all Working Group activities as an observer, will be invited to continue to participate in that capacity, and the General Accounting Office, which has been extremely helpful to the Working Group in the past, will continue to be kept abreast of developments and invited to advise and assist as appropriate. The "cjrrinQ Group has initiated research efforts designed to fino answers to many of the questions surrounding Agent Orange that have been raised. These 1-efforts include the birth' defects study being conducted by HHS Centers for Disease Control, ths Ranch Hand Study being conducted by the Air Force, the epio'emiological study being planned by the Veterans Administration pursuant to P.L. 96-151, and the compilation by HHS1 National Institute of Occupational Safety and Health of a* national rsgistry of workers exposed to dioxins. Each of these research, activiti*?. as well as the other important research �-3- activities being conducted under the overall guidance of the Working Crcup, are to be continued without. interruption or delay . The Working Group has developed an impressive record of scientific objectivity, impartiality and integrity and it. is imperative to the success of the Working Group effort that this record and the Group's credibility be maintained. In this regard, i £««!£* progress reports to the Cabinet Council, the Congress and the public will continue to be made by the Agent Orange working Group., To assure effective leadership of the working Group, I am hereby appointing James Stockdale, HHS Deputy Under Secretary for Intergovernmental Affairs, as Chair. Also, I am appointing Or. Vernon N. Houk of the Center for Environmental Health of the Centers for Disease Control as Chair of the Working Group's Science Panel. In addition, I am appointing HHS Legal Counsel Leslie A. riitt, who* has served as legal -adviser to and staff director of the Working Group since its inception, to continue in those capacities. I know and believe you will find that .these individuals share «y commitment to carrying out this important mission. •Please review your representation on the Working Group to assure that your agency or office is adequately represented by appropriate technical- experts, scientists .and policy-level officials. In order to facilitate the Group's effectiveness, it is of course -important that each agency's total membership be limited. The first meeting! of the full Working Group has been * scheduled for Friday, August 28, 1981 and a meeting of the Science Panel will be scheduled for shortly thereafter. Accordingly, please let Mr. Bart Kull, Special Assistant to the Deputy Under Secretary for Intergovernmental Affairs .Peter Beach, HHS Director of Veterans Affairs (24.5*2210), know as soon as possible the name(s)- of your designated representative(s) so that briefing materials may be foi-iiicu to them. Attached for your information is a copy of the memorandum of the Executive Secretary to the Cabinet Council on Human Resources establishing the Working Group. ^ Attachment cc: Comptroller General of the United States Director, Congressional Office of Technology Assessment Mr. Robert Carleson �activities being conducted under the overall guidance of the Working Group, are to be continued without interruption or delay. ? The Working Group has developed an impressive record of scientific objectivity, impartiality and integrity and it i* imperative to the success of the Working Group effort that t&ls record and the Group's credibility be maintained. In this regard, regular progress reports to the Cabinet Council, the Congress and the public will continue to be made by the Agent Orange Working Group. To assure effective leadership of the Working Group, I hereby appointing James Stockdale, HHS Deputy Under Secretary for Intergovernmental Affairs, as Chair. Also, I am appolntlsuj Dr. Vernon N» Houk of the Center for Environmental Health of the Centers for Disease Control as Chair of the Working Groups Science Panel. In addition, I ara appointing HHS Legal Counsel Leslie A. Platt, who' has served as legal adviser to and staff director of the Working Group since its inception, to continue in those capacities. I know and believe you will find that these individuals share my commitment to carrying out this Important mission. 'Please review your representation on the Working Group to assure that your agency or office is adequately represented by appropriate technical experts, scientists and policy-level officials. In order to facilitate the Group's effectiveness, it is of course -important that each agency's total membership be limited, The *irst meeting -of the full Working Group has been scheduled for Friday, August 28, 1981 and a meeting of the Science ;3anel will be scheduled for shortly thereafter. Accordingly, please let Mr. Bart Kull, Special Assistant to the Deputy Under Secretary for Intergovernmental Affairs (245-6156), or Dr. Peter Beach, HHS Director of Veterans Affairs (245-2210), know as soon as possible the name(s) of your designated representative(s) so that briefing materials may be forwarded to them. -Attached for your information is a copy of the memorandum of the Executive Secretary to the Cabinet Council on Human Resources establishing the Working Group. Attachment cc: Comptroller General of the United States Director, Congressional Office of Technology Assessment Mr. Robert Carleson Mr. Edwin Gray �UJS. DEPARTMENT OF HEALTH AND HUMAN SERVICES Laura Genero—(202) 245-6343 Richard McGowan— (202) 245-7204 FOR IMMEDIATE RELEASE . Wednesday, October 21, 1981 HHS Secretary Richard S. Schweiker today made public the attached Chronology of Activities on Agent Orange. * I# �PROM: TO Janes S. Stockdale Deputy Under Secretary for Intergovernmental Affairs : The Secretary ;.» -•- - ^ CHRONOLOGY OF ACTIVITIES RE: AGENT ORANGZ The first meeting of .the re-established and expanded Agent. Orange Working Group was held on August 28, 1981. The first task of the Working Group was to review the status of all ongoing and planned Federal research and related activities. Each member agency was directed to provide updated reports on the status of its current or planned research activities. A proposed protocol for the design of the Veterans Administration epidemiological study will be reviewed by the Working Group's Science Panel. A number of veterans organizations have been briefed on t%3> continuing military records search that is being conducted by the Array Agent Orange Task Force. Preparations are continuiris for the Air Force Ranch Hand Study. That study involves pilcubjs and maintenance personnel engaged in the spraying of herbicides during the Vietnam conflict. Concern was expressed that the fullest possible participation by Ranch Hand personnel be obtained for this study df possible health effects related to exposure to Agent Orange. This is critical because the Rancn • Hands are a relatively .small group of approximately 1200. i , A public affairs panel was created and will develop plane for a public meeting of the Working Group to be scheduled lafcar this.year. The Working Group also agreed to establish a resource development panel to assure adequate funding and personnel resources. Dr Vernon Houk, Chair of the Working Group's Science Panel, has plans to review all research. Dr. Kouk and several other members of the Working Group visited the Army Agent Orange Task Force Office for a 'briefing on tne status of the Department of Defense records search. During the briefing, it became apparent that a potentially �page 2 existed *« • « v mr »^BP-^^ promising new concept/for the identification of people to Agent Orange in addition to the Air Force Ranch Hand ^ _ personnel or broadly defined unics of ground troops. The -u^. research panel was briefed on this new information. The information may provide the basis for a new .approach c< €inding answers to .some of the serious scientific questions before t^a working Group. It opens the possibility of developments regarding in the next Working Group report. The Agent Orange Working Group's predecessor, the Interagency Work Group to Study the Possible Long-Tens Health Effects of Phenoxy Herbicides and Contaminants (IW<3)/ also undertook a number of activities during the transition period between April, 1981, when it transmitted its seventh report, and August, when the expanded Working Group convened. At the May meeting of the IWG, a status report was givea the birth defects study being conducted by the Centers for It was reported that the Office of Management tDisease Control. 'and Budget had approved the study and preparations for the study were under way, with completion likely in the summer or fall of 1983. • ' At the same meeting, it was reported that a representative of the Medical Follow-Up Agency of the National Academy of Sciences' National Research Council had been briefed on and £t« reviewed the Defense Department's records search effort and tia<S concluded it could be difficult to identify a population of ground troops the nature -and extent of whose exposure to Agens. Orange could be reliably reconstructed and documented. At its June meeting, the IWG was honored by a visit toy the Australian Minister of Veterans Affairs, Senator Anthony Messner. Senator Messner told the Group of his government"* s Agent Orange-related.research and urged continuing between our countries in the area of research. During meeting, the IWG was assured of the Administration's s support for Agent Orange research. As you know, Agent Orange efforts of the various F bodies include research as follows: 1. .The Agent Orange Working Group coordinates all Federal .Agent Orange research. It does not undertake research on its own but rather aces as the and monitor. �' Page 3 2. 3. The Air Force Ranch Hand Study is to be monitored by an independent advisory committee in addition to the Agent Orange Working Group. This advisory committeewill meet publicly (probably beginning in November) and will include scientists from inside and outside the Federal government. The Veterans Administration Advisory Committee on • Health-Related Effects of Herbicides meets periodically to review all VA herbicide-related research. The committee includes scientists from inside and outside the government as well as representatives of veterans organizations. 4. The Office of Technology Assessment of the Congress has established a scientific review panel to review the proposed protocol for the design of the VA epidemiological study and will provide the VA with its conclusions and recommendations regarding the protocol. 5. There are also a number of State-level Agent Orange commissions charged with undertaking and/or monitoring Agent Orange research. The report notes in conclusion that a number of veterans organizations, members of the public and Congress have expressed support for the Administration's actions regarding Agent Orange and related research. f Attachments: (A) 'Memorandum of August 21, 1981, Re-establishing the Agent: Orange Working Group, from HHS Secretary Richard S. Schweiker in his capacity as Chairman Pro-Tern of the Cabinet Council on Human Resources. (B)' Opening Remarks of Agent Orange Working Group Chairman James S. Stockdale at the Working Group's August 28, 1981 meeting. (C) Report by Dr. Vernon Houk, Chair of the Working Group's' Science Panel. �WHITE HOUSE HI N C T O N PR Honorable Caspar W. Weinberger Secretary of Defense The Pentagon Washington, B.C. 20301 Dear Mr. Sec The Agent Orange Working Group (AOWG) of the Cabinet Council on Human Resources has the responsibility for overseeing,, coordinating and setting priorities among Federal research relating to Agent Orange. In partial fulfillment of izs , charter, the AOWG has been assisting the Veterans Rdrr.ir.istration (VA) in implementing an extensive epidemiology study to determine whether Agent Orange caused long terar. health effects in exposed veterans. This study was mandate! .ay Congress in Public Law 96-151. It appeared until r^-sntly that it would be impossible to identify, with sufficient reliability, the exposure cohorts for this study. Within the last six months, Department of Defense (DOD) representatives to the AOWG working with the staff of the Army Agent Orange Task Force have succeeded in developing a model for searching DOD records to identify-exposed a.-d nonexposed Vietnam veteran cohorts. This concept was presented to the AOWG in December, and after continued refinement by the^DOD, the concept was endorsed as a reliable method for identifying the Agent Orange exposure of certain units serving in Vietnam. We would like to compliment your staff for developing the solution to this critical problem. « The AOWG has recommended that DOD proceed with the search, retrieval, and establishment of the needed exposed and nonexposed Vietnam veteran cohort lists. The AOWG recommends that sufficient DOD personnel and computer support be committed to begin the records search necessary to provide the Veterans Administration with subjects for the initial pilot study. After the pilot is completed, additional DOD staff and computer support will be needed to identify the 36,000 potential research subjects needed for the full Agent Orange epidemiology study. The AOWG estimates the total effort will require a DOD budget of $2 - 3 million over 1 1/2 years. �.Page 2 - Caspar W. Weinberger I now forward to you the request that you review our recommendation and help to assure the priority, staffing and other support necessary to enable the DOD to develop these cohort lists in an expeditious manner. Sincerely, Richard S. Schweiker Chair, Pro-Tern, Cabinet Council on Human Resources V»£*x <%&?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/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 199 Folder The folder containing the original item. 5601 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Kull, Bart Title A name given to the resource Memorandum to the Under Secretary/Chairman, Agent Orange Working Group from Bart Kull, Special Assistant to the Under Secretary for Intergovernmental Affairs, with Subject: Agent Orange Working Group (AOWG) Membership and Scope and Purpose of AOWG, dated ao_seriesVIII 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 191 Folder The folder containing the original item. 5393 Series The series number of the original item. Series VIII 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 Lacey, Pamela Vincent A. Lacey Source A related resource from which the described resource is derived The Journal of Legal Medicine Date A point or period of time associated with an event in the lifecycle of the resource 1982 Title A name given to the resource Agent Orange: Government Responsibility for the Military Use of Phenoxy Herbicides ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/18e316bef07da71f9df02c5519c300e1.pdf 9cbfa08f7cf70f8ad0a01bd3e24864b4 PDF Text Text Item D Number °5598 D Not Scanned Lathrop, George D. Epidemiology Division, Data Sciences, Division, USAF S Report/ArtldO TitlB Pr °iect Ranch Hand II: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Baseline Mortality Study Results Journal/Book Title Year 1983 Month/Day June3 Color ° D Number of Images ° DOSCrlptOD NOtQS Also included are summaries of the baseline mortality study results (June 30,1983), the baseline morbidity study results (February 24, 1984), and the mortality update (December 10, 1984. Alvin L. Young filed these documents together with others in a folder labeled, "Agent Orange Working Group Science Panel, Current Folder." Tuesday, March 19, 2002 Page 5598 of 5611 �I PROJECT RANCH HAND II AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES BASELINE MORTALITY STUDY RESULTS 30 JUNE 1983 Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 Approved for public release; distribution unlimited EPIDEMIOLOGY DIVISION DATA SCIENCES DIVISION USAF SCHOOL OF AEROSPACE MEDICINE (AFSC) BROOKS AIR FORCE BASE, TEXAS 78235 �Unclassified SECURITY CLASSIFICATION OF THIS PAGE (When DatafEntered)i REPORT DOCUMENTATION PAGE 1. REPORT NUMBER «. Tl T L t 2. GOVT ACCESSION NO READ INSTRUCTIONS BEFORE COMPLETING FORM 3. RECIPIENT'S CATALOG NUMBER 5. TYPK OF REPORT A PKRIon C O V K R I n («"</ S AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES: BASELINE MORTALITY STUDY RESULTS Annual Report, Initial Report 6. PERFORMING OHG, REPORT NUMBER 7. AUTHORfs) 6. CONTRACT OR GRANT NUMBERS; 9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT, PROJECT, TASK AREA ft WORK UNIT NUMBERS George D. Lathrop, Colonel, USAF, MC; Patricia M, Moynahan; Colonel, USAF, NC; Richard A. Albanesei M.D.; William H. Wolfe, Lt Colonel, USAF, MC USAF School of Aerospace Medicine (EK) Aerospace Medical Division (AFSC) Brooks Air Force Base, Texas 78235 11. CONTROLLING OFFICE NAME AND ADDRESS The Surgeon General United States Air Force Washington, D.C. 20311* 1Z. REPORT DATE U. MONITORING AGENCY N'AME ft ADDRESSfff dltterent Irom Controlling Ottice) 15. SECURITY CLASS, (ol thti report) 30 June 1983 13. NUMBER OF PAGES 61 Unclassified I5«. DECLASSIFI CATION/DOWN GRADING SCHEDULE 16. DISTRIBUTION STATEMENT (oi this Report) Approved for public release; distribution unlimited. 17. DISTRIBUTION STATEMENT (of the abstract entered In Block 20. It different from Report) IB. SUPPLEMENTARY NOTES 19. KEY WORDS (Continue on reverse side H necessary and Identity by block number) Epidemiologic investigation Air Force Health Study Mortality study RANCH HAND Matched cohort design Nonconcurrent prospective design '0. ABSTRACT (Continue on reverse side if necessary and identify by block number) In 1979 the United States Air Force (USAF) made the commitment to Congress and to the White House to conduct an epidemiologic study of the possible health effects from chemical exposure in Air Force personnel who conducted aerial herbicide dissemination missions in Vietnam (Operation RANCH HAND). The purpose of this epidemiologic investigation is to determine whether long-term health effects exist and can be attributed to occupational exposure to herbicides. This study uses a matched cohort design in a nonconcurrent prospective setting, incorporating mortality, morbidity, and follow-up studle; DD , F O R M73 <|.|« EDITION OF 1 NOV 65 IS OBSOLETE JAN WO Unclassified 7 SECURITY CLASSIFICATION OF THIS PAGE (When Data Entered) �unclassified SECURITY CLASSIFICATION OF THIS PAGE(HTi«n Data Entertd) The purpose of this report is to present the baseline mortality study results. As of December 31» 1982, 50 Ranch Hand and 250 comparison subjects had died (certified on/before April 27, 1983). Data analysis showed that the mortality experience of the Ranch Hand group is nearly identical to that of this comparison group. However, this mortality report can in no way be regarded as conclusively negative since the study population may not yet have reached the latency period. Subsequent mortality reports will include additional anaylses and will be updated annually for the next 20 years. Unclassified S E C U R I T Y C L A S S I F I C A T I O N OF T U I ' PAGEfWien Dalf Entered) �EXECUTIVE SUMMARY Baseline Mortality Study The Ranch Hand II epidemiclogic study uses a matched cohort design in a nonconcurrent prospective setting, incorporating mortality, morbidity, and follow-up studies. The purpose of this report is to present the baseline mortality study results. Since 1979, a detailed population ascertainment process has enumerated a total of 1269 Ranch Hand personnel who served in Vietnam during the period of 1962-1971. As described in the protocol, this total is believed to comprise, the entire exposed study population. The eligibility of each Ranch Hander was verified by a hand review of his personnel record, A comparison group was formed by identifying all individuals assigned to selected Air Force organizational units with a mission of flying cargo to, from, and in Vietnam during the same period. All Ranch Hand and comparison subjects designated as killed in action were removed from the study population. By a computerized nearest neighbor selection process, up to 10 comparison individuals were matched to each Ranch Hander by job category, race, and age to the closest month of birth. A hand record review of the matched comparison sets revealed that on the average, 8.2 comparison individuals were fully suitable for study. From each matched comparison set, five individuals were randomly selected for the mortality study, yielding a 1:5 design. Every Ranch Hander and his set of comparisons will be the subjects of annual mortality updates throughout the entire 20 years of the follow-up study so that emerging mortality patterns or disease clusters may be detected with maximal sensitivity. Each living Ranch Hander and his first and willing comparison match were selected to participate in a comprehensive physical examination and an in-home interview; the results of this study will be presented in a subsequent report in late 1983. A mortality determination on 1,247 Ranch Handers and 6,171 comparison subjects was made, sequentially using the data sources of the Air Force, Veterans Administration, Social Security Administration, Internal Revenue Service, and personal contact efforts. As of December 31, 1982, 50 Ranch Hand and 250 comparison subjects had died (certified on/before April 27, 1983). Death certificates were obtained on all 300 deceased subjects and were coded by an Air Force nosologist (ICD, 9th ED). All codings were verified by the National Center for Health Statistics. Autopsy results are currently being sought for future analyses. Statistical analyses of noncause specific death emphasized survival curve estimates, linear rank procedures, relative risk estimates, and standardized mortality ratios (SMRs). Cause specific analyses were limited to relative risk estimates because of small cell sizes. In addition to these approaches, three other data bases were contrasted to the Ranch Hand population, where possible; the 1978 US White Male Mortality experience, the 1978 Department of Defense (DoD) Nondisability Retired Life Table, and the mortality experience of the West Point Class of 1956. These additional comparison groups have substantial comparability or sample size limitations, rendering conclusions to the weakest order. Analyses with these "external" comparison groups were accomplished to crudely define the healthy worker effect and to determine if the Ranch Hand group mortality was drastically out of line with that of other military populations. �Data analysis showed that the mortality experience of the Parch Hand group is nearly identical to that of the comparison group. Analyses showed that, officers are living longer than enlisted personnel in both Ranch Hand and comparison groups. This difference between officers and enlisted personnel was statistically significant in the comparison group whereas it was not in the Ranch Hand cohort. A contrast of the Ranch Hand and comparison group to the 1978 DoD Life Table showed significantly less mortality for Ranch Hand officers, comparison officers and comparison enlisted men, however, there was not a statistically significant favorable mortality rate for Ranch Hand enlisted personnel. This pattern of mortality was also seen in a contrast of the Ranch Hand and comparison groups to the 1978 U.B. white male mortality experience. That is, highly favorable mortality differentials for Ranch Hand officers, comparison officers and comparison enlisted personnel were observed, but not for Ranch Hand enlisted. This trend is consistent with the self perceptions of differential herbicide exposures reported by many of the Ranch Hand subjects. The reason(s) for these observations are speculative at present, but may include the related items of sample size, socioeconomic differences, access to medical care, and health education and possible herbicide effects. Cause specific analyses were statistically nonsignificant. The Ranch Handers showed a relative paucity of overall cancer but an excess of digestive disorder deaths, both statistically nonsignificant. No soft tissue sarcoma deaths were detected in either group. Analyses of both the Ranch Hand and the comparison groups to the 1978 US White male mortality experience showed highly significant favorable findings. Most of these differences are speculatively attributed to the healthy worker effect. A contrast of the Ranch Hand and comparison groups to the 1978 DoD Life Table showed significantly less mortality for Ranch Hand officers and comparison officers and enlisted men. The West Point comparison showed nonsignificant SMRs of 0.530 and 0.778 for the Ranch Hand officers and the comparison group officers, respectively. Overall, the limitations of the statistical power calculations in most of these analyses were substantial in most analyses due to 1) the low mortality rate (4%) in the Ranch Hand and comparison groups to date, 2) the inherently small group of Ranch Handers (as described in the study protocol), and 3) the observed relative risks which approached unity in most categories. This baseline mortality report can in no way be regarded as conclusively negative because this small, young, and relatively healthy cohort may not have yet reached the latency period wherein attributable fatal disease might be expected and detected within limited power boundaries of this study. Future commitments for the annual mortality updates include detailed covariate analyses for disease risk factors, herbicide exposure, and confounding industrial chemical exposures. Further, subsequent morbidity reports will include full spectrum, disease specific analyses, e.g., cancer (fatal, ongoing, cured) in an effort to enhance study sensitivity to emerging herbicide effects, if they occur. ii �PREFACE In October 1978, the United States Air Force (USAF) Surgeon General made the commitment to the Congress and to the White House to conduct an epidemiologic study of the possible adverse health effects arising from the herbicide exposure of Air Force personnel who conducted aerial dissemination missions in Vietnam (Operation Ranch Hand). The purpose of this epidemiologic investigation is to determine whether long-term adverse health effects exist and whether they can be attributed to occupational exposure to herbicides and their contaminants. The study protocol (1) for this effort incorporates a matched cohort design placed in a nonconcurrent prospective setting. The study approach includes mortality, morbidity, and follow-up elements linked tightly in time, in order to produce the most data in the shortest period of time. The study addresses the question: Has there been, or are there currently, or will there be any adverse health effects among former Ranch Hand personnel caused by repeated occupational exposure to 2,4,5-T containing herbicides and the contaminant, TCDD? At the request of the Principal Investigators (see Appendix I) the study protocol was extensively and independently peer reviewed. The review agencies included: The University of Texas School of Public Health, Houston Texas; the USAF Scientific Advisory Board; the Armed Forces Epidetniological Board; and the National Research Council of the National Academy of Sciences. In 1980, the Science Panel of the Agent Orange Working Group was created as an additional peer review agency. This group, redesignated as the Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants, has consented to the oversight responsibility of the Ranch Hand study and continues to monitor the conduct of this epidemiologic investigation (see Appendix II). The approved and official protocol for this effort is available to the public through the National Technical Information Services, 5285 Port Royal Road, Springfield, Virginia 22161. The Ranch Hand II Study protocol heralds the suboptimal statistical power of the mortality study. The mortality study was motivated by the desire to use a full spectrum epidemiologic approach to the herbicide question. Additionally, the investigators were scientifically obliged to pursue the mortality study because of previous and emerging studies (some with small sample sizes) which suggested the possibility of a soft tissue sarcoma end point (2,3,4). Within the inherent sample size limitation of the Ranch Hand population, detection of such a rare condition will be missed unless there is marked case clustering and correspondingly high relative risks. Also, because of sample size limitations as well as the myriad of proposed clinical end points, a case-control design was not entertained. The investigators have attempted to enhance statistical power and analytic sensitivity where possible by using ] a large comparison group, 2) precise ) matching procedures, 3) annual mortality updates, A) mortality-morbidity linkages, 5) a lengthy follow-up study, 6) external comparison groups, and 7) state-of-the-art statistical methodology. A final assessment of overall mortality must necessarily await substantially more data and covariate approaches to identify and isolate unusual emerging mortality patterns, if they occur. iii �This report is primarily directed to individuals with statistical and epidemiologic backgrounds. It also assumes that the reader has a familiarity with the herblcide/dioxin issue and a detailed 'knowledge of the protocol of the Air Force study. In the interest of brevity, the reader is referred to the protocol published as US Air Force School of Aerospace Medicine Technical Report 82-44. iv �Page Executive Summary Preface ............................. ............................... Table of Contents Chapter I. . . .iii ............................. THE MORTALITY STUDY DESIGN 1 . The Study Population 2. The Mortality Population Chapter II. i .......................... ........................ THE MORTALITY DETERMINATION PROCESS 1. Introduction ....................... . ...... 2. United States Air Force Data Sources .................. 3. Veterans Administration Death Beneficiary Identification and Record Location Subsystem ........................... 1. Other Governmental Data Sources .................... 5. Morbidity Population Tracking ..................... 6. Receipt and Coding of Death Certificates ................ 7. Results ................... ............. Chapter III. RANCH HAND VERSUS COMPARISON GROUP ANALYSES 1. Introduction v 1 1 2 4 4 4 6 6 6 7 8 9 9 2. Overall Comparisons 9 3. Noncause Specific Occupational Comparisons 15 4. 17 Cause Specific Ranch Hand Versus Comparison Mortality Chapter IV. NONCAUSE SPECIFIC COMPARISONS WITH EXTERNAL POPULATIONS 1. Background and Motivation 2. Adjustment Difficulties A. Adjustment for Calendar Year of Death 20 20 20 20 B. Adjustment for Military Status (Active Duty, Separated, Retired). . 20 C. Adjustment for Selection D. Adjustment for Branch of Service 21 21 �3. Comparisons with 1978 DoD Life Tables ................. 4. Comparisons with U.S. 1978 White Male Life Table ......... 21 . . . 24 Chapter V. COMPARISONS WITH THE WEST POINT STUDY GROUP 1 . Background and Motivation ..... . ................. 2. Noncause Specific Comparisons of Ranch Hand and Comparison Subgroups with the West Point Study Group .................... 3. Cause Specific Comparisons ....................... Chapter VI. STATISTICAL ASPECTS 1. 2. 3. 4. 5. 6. 28 28 31 34 Purpose ................................ 3^ Survival Curve Estimates and Confidence Bands ............. 34 Linear Rank Procedures ......................... 35 Relative Risk Estimation ....... . ............ .... 36 Indirect Standardization ........... . ............ 37 Comparing Observed Life Table Data with a Known Survival Curve. .... 37 Chapter VII. CONCLUSION 1. 2. 3. 4. 5. 28 39 Introduction .................. . ...... . . . . . 39 Internal Comparison Group . . . . . . . ..... . .......... 42 External Comparisons. . . . . ......... . . . . . . . . . . . . 42 Power Considerations. . . . ....... . . . . . . . . ....... 42 Consistency Patterns ...... . ................... 43 APPENDICES I. Ranch Hand II Principal Investigators, Coinvestigators, Contributors, and Management Personnel 45 II. Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants . . 47 III. Matching Results in the Mortality Population 48 IV. Year of Birth, Occupational and Race Specific Mortality 51 V. The Extended E j i gou-McHugh Relative Risk Estimator ..53 VI. Survival Curves and Confidence Bands for Ranch Hand and Comparison Subgroups 54 REFERENCES 59 VI �Chapter I THE MORTALITY STUDY DESIGN 1. The Study Population The exposed study population, termed "Ranch Hand", was defined as those individuals who were formally assigned to the USAF organizations responsible for the aerial dissemination of herbicides and insecticides in the Republic of Vietnam from 1962 to 1971. These individuals were identified from historical data sources at the National Personnel Records Center (NPRC), St. Louis, Missouri and the USAF Human Resources Laboratory, Brooks Air Force Base, Texas. A total of 1,269 Ranch Hand personnel were eventually identified through this process. The comparison population was defined as those individuals who were assigned to a variety of cargo mission organizations throughout Southeast Asia during the same time period. Cargo mission aircrew members and support personnel were selected because of sufficient population size, similar training and military background experiences, and psychologic similarities to the Ranch Hand group. The comparison population was not occupationally exposed to herbicides or insecticides in the Republic of Vietnam. Identification of this population was completed using the same historical data sources as were used with the Ranch Hand population; 2*1,971 individuals were so identified. In preparation for matching the study and comparison populations, all subjects killed in action (KIA) were removed from the data base. The rationale for this action is the assumption that combat death in the Ranch Hand group was not caused by the immediate effects of herbicide exposure; KIA's were removed from the comparison group for comparability purposes. A KIA analysis will be performed in a subsequent report. The Ranch Hand KIA subgroup, numbering 22 individuals, although not matched, was maintained in the data base but was deleted from the mortality analysis, leaving 12M7 Ranch Hand subjects. The Ranch Hand population was matched to the comparison population with an iterative nearest-neighbor computer program (1). Up to 10 comparison subjects were matched to each Ranch Hander by year of birth, race (Black versus non-Black), and occupational category (officer pilot, navigator and other; enlisted flight, engineer and other), thus creating matched sets of one study subject and up to 10 comparison subjects. All subjects are males. The mean age of the study subjects is *J5 years. Following the original match, the majority of Ranch Handers had 10 comparisons. The exceptions were the group of non-Black pilots who had a mean of only 9.5 comparisons per exposed subject due to the extreme ages of several individuals, and the strata of Black pilots and other Black officers who only had means of 2.7 and 5.0, respectively. In December 1981, the USAF Principal Investigators learned that several morbidity study comparison subjects had reported no experience in Southeast Asia, suggesting that overselection of the comparison population had occurred (1). Manual review of the comparison subjects' military personnel records revealed that 18 percent of the 12,193 comparison individuals in the original match were ineligible for study. The inadvertent inclusion of several non-Southeast Asia organizations resulted in the selection of these inappropriate individuals. These ineligible subjects were found to be randomly distributed throughout the matched sets and were removed from the study. Following the removal of the �ineligible subjects, the study was reduced to a 1:8 design. Also during this period, five Ranch Hand subjects were identified through personnel record sources and Veterans Administration Education Benefits and Financial Records. These five individuals had not been identified earlier because the majority of their military personnel records had been destroyed in a fire at the NPRC in St. Louis. Three of these five were newly found Ranch Handers and two were comparisons subsequently identified as Ranch Handers. No attempt was made to match comparisons to these five new Ranch Handers. During the removal of ineligible subjects, one Ranch Hander, a Black officer pilot, lost his only comparison and remains unmatched, giving a total of six unmatched Ranch Handers. All six of these unmatched Ranch Handers are included in the morbidity and mortality studies. They were used in the analyses where appropriate, in order to improve statistical power. 2. The Mortality Population Five comparisons per exposed subject were considered more than adequate for mortality analyses; this estimate has recently been verified under a multiplicative model by Breslow, et al. (5). Up to five comparisons in each matched set, were identified from the 1:8 cohort as the mortality comparisons. Since the positions of the individuals in the matched sets had already been randomized in the data file, the selection of the first five positions in each matched set array for membership in the mortality comparison resulted in a random selection of the mortality comparison cohort. If a Ranch Hander had at least one but no more than five comparisons after removal of the ineligibles, then all of his matched set were used in the mortality component of this study. The mortality population is, therefore, defined as the 1241 matched Ranch Handers and their randomly chosen mortality comparisons (6171 individuals) and the six unmatched Ranch Handers. Table 1 summarizes the mortality population by occupational category and race. Here, and elsewhere in this report, non-Black is defined as Caucasian, Mexican or Oriental. Table 1 MORTALITY POPULATION SUMMARY BY OCCUPATION AND RACE Occupation, Race Officer-Pilot, Non-Black Officer-Pilot, Black Ranch Hand Counts Comparison 349 6 1740 13 Officer-Navigator, Non-Black Officer-Navigator, Black 80 2 390 10 Officer-Other, Non-Black Officer-Other, Black 25 1 123 2 189 15 935 75 528 52_ 1247 2628 255 6171 Enlisted-Fit Eng, Non-Black Enlisted-Fit Eng, Black Enlisted-Other, Non-Black Enlisted-Other, Black �The overall match ratio, 6171/1247=4.95, reflects the lack of suitable controls in some strata, the subsequent removal of ineligible comparisons and the addition of five unmatched Ranch Handers. A detailed description of the matching results is given in Appendix III. Those Ranch Handers having fewer than five matched mortality controls are summarized in Table 2. Table 2 RANCH HAND SUBJECTS WITH LESS THAN FIVE COMPARISON SUBJECTS Counts Occupation, Race Officer-Pilot, Non-Black Ranch Hand Comparisons-^ Notes Officer-Navigator, Non-Black Officer-Other, Non-Black Officer-Other, Black Enlisted-Fit Eng, Non-Black Enlisted-Other, Non-Black Enlisted-Other, Black 2 3 1 1 0 1 1 2 1 Officer-Pilot, Black 1 1 1 2 3 1 1 1 2 2 1 2 12 1 0 4 2 0 J» 0 2 1 1 2 1 2 27 Note 1. Lack of suitable comparison subject or loss due to ineligibility. Note 2. New Ranch Hander, no attempt to match. Note 3. Comparisons per Ranch Hander 1 �Chapter II THE MORTALITY DETERMINATION PROCESS 1. Introduction The mortality status of the Ranch Hand group and their mortality comparisons are, and will continue to be, ascertained using four major data sources: USAF, Veterans Administration (VA), other Governmental and morbidity population tracking. The mortality determination process using these data sources is presented in Figure 1. Figure 1. RANCH HAND II MORTALITY DETERMINATION ALGORITHM RANCH HAND II MORTALITY STUDY ENTIRE STUDY POPULATION 1 U.S. AIR FORCE ACTIVE DUTY ^ RETIRED PERSONNEL CENTER ACCOUNTING t FIKAHM CASUALTY BRANCH WSC FACILITY IHFUR. BENEFICIARY^ DIED ACTIVE DIED AF HQSP^ ALL I i DEATH BFJiEFIT ' SOCIAI SECURITY AOVIS. CLOSED BY DEATH INTEMIAL RLTOUE ACTIVE ACCT. CLOSED BY DEATH 3E 5 JE 3 X Ut ALIVE f ,__ TRACKING CONTACT UTTER LOWS HARRIS DEAD EAO - REOtiESf DEATH CERTIFICATE VETERml'S ADMIN. (GIRLS, SGI.I) The entire study population was matched or checked against the first three sections of this algorithm while only the morbidity population was contacted and tracked. A description of the data sources within the algorithm follows. 2. United States Air Force Data Sources The USAF data sources include the USAF Military Personnel Center (MFC) records, the USAF Accounting and Finance Center records, and the USAF Medical �Service Center Facility Use Data. The USAF MFC records include the individual's military personnel record and the data accumulated by the Casualty Branch of the MFC. Individual military personnel records are created at the time of induction into the USAF, and reflect a chronological history of the individual's military career. Epidemiologically, these records are an invaluable data source as they can be used for the development of occupational histories, identification of race, sex, and date of birth as well as for location of personnel, and for determining vital status. Hard copy records of these data are maintained at the individual's base of assignment while on active duty; a computer copy of these records is maintained at the USAF military personnel center, Randolph AFB, Texas. Following retirement and/or separation from the USAF, these records are forwarded to the National Personnel Records Center (NPRC), St Louis, Missouri, the record repository for all military personnel records. They are indexed by Social Security Account Number or Air Force Serial Number at the NPRC. If an individual should die while on active duty, after retirement, or within 120 days of separation from active duty, it is the responsibility of the Casualty Branch of the USAF Military Personnel Center to update the hard copy military personnel record and the MFC computer data base and to inform the USAF Accounting and Finance Center of this fact. At the same time, USAF MFC personnel initiate a copy of the USAF Form 1312, Report of Retired Casualty, or Department of Defense DD Form 1300, Report of Casualty. The selection of the appropriate form is based on the current status of the individual concerned. The DD Form 1300 also clarifies an individual's casualty status which can be either battle or nonbattle. Copies of the appropriate death form are sent to appropriate agencies while the original is placed in the individual's military personnel record. Since the initial review of military personnel records, a system has been established with the Casualty Branch of the Military Personnel Center wherein all active duty and retired death forms are forwarded monthly to the Occupational Epidemiology Section of the USAF School of Aerospace Medicine Epidemiology Division. In this way, the mortality status of all active duty and retired study subjects is systematically determined on a continuing basis. The USAF Accounting and Finance Center data base was used as a resource to update individual Air Force serial numbers to Social Security numbers. The Social Security number is required for all other aspects of the mortality algorithm. The Air Force Medical Service Center (AFMSC) Facility Use Data is a computer data base containing information regarding all active duty and retired deaths that .occur in Department of Defense (DOD) Medical Facilities. This data base identified no additional deaths in the mortality population, but did verify the deaths known to have occurred in DOD hospitals. In addition to the USAF data bases, the Ranch Hand Association, a reunion association of approximately 850 Ranch Handers, has contributed to the success of this study. This group has assisted the Principal Investigators in the ascertainment of the exposed population, and in the determination of the current location and the mortality status of the group. The association contacts all of its members yearly through newsletters and provides updated information to the Air Force investigators. �3. Veterans Administration Death Beneficiary Identification and Record Location Subsystem The Beneficiary Identification and Record Locator Subsystem (BIRLS) is a Veterans Administration data base generated by the Veterans Administration for determination of funeral allowance. If the family of the deceased informs the funeral director that the deceased served in the US military, the funeral director submits the required data to the Veterans Administration. In January 1981, August 1982,and January 1983,the BIRLS data base was searched for Ranch Hand and comparison deaths. In addition to these searches, the Department for Veterans Benefits, Veterans Administration, coordinated the gathering of death certificates from VA regional offices. 4. Other Governmental Data Sources A. Internal Revenue Service Public Law 96-126, Section 502, 28 November 1979, authorized the use of Internal Revenue Service (IRS) addresses for individuals who had been exposed to occupational hazards in order to determine the status of their health. The National Institute for Occupational Safety and Health (NIOSH) coordinated the USAF requests for these IRS addresses. This system is based on the address shown on individual tax returns and is corrected once a year. The addresses are verified by NIOSH through use of a post card mailed to the post office responsible for the individual's mail delivery. NIOSH assumes that the individual is alive if he files a tax return and if the verification scheme confirms his address for mail delivery. The IRS assumes an individual ia dead if the individual is so reported on a joint tax return. The IRS data base search provides an incomplete mortality determination, however, since absence of an individual tax return does not necessarily imply death of that individual. B. Social Security Administration The Social Security Administration (SSA) is a source of mortality information based on data maintained by the Office of Remuneration and Earnings. The basis for this data is employer-reported earnings. The SSA assumes that an individual is living if there is no indication of death on the individuals record and earnings are recorded for the last calendar year or retirement, disability, black lung or supplemental security income payments are being made. The SSA did inform us that they do not conduct an exhaustive search, and all deaths are not necessarily reported to SSA. Therefore, this mortality information may not be complete. 5. Morbidity Population Tracking Individual tracking techniques apply only to the morbidity population, defined as those selected and compliant to questionnaire. The morbidity population for this effort is defined as all Ranch Handers and their morbidity comparisons. The morbidity comparisons are, in general, also mortality comparisons. The selection procedure for the morbidity study is presented in Figure 2. �Figure 2. SELECTION PROCEDURE FOR THE QUESTIONNAIRE, PHYSICAL EXAMINATION, AND FOLLOW-UP STUDY Living Ranch Hand Comparison Individuals Randomly Ordered Mortality Comparisons r -A—"^ r~ 1 t i 1 11 +-* #* 1:1 Morbidity Match + Dead - Unwilling * Volunteer ** Replacement candidates In this figure, the first randomly ordered comparison was found to be dead. The second was contacted but was unwilling to participate and the third volunteered to participate in the questionnaire component of the morbidity effort. This contacting process for the morbidity effort was the final step in the baseline mortality determination. The original contact was made by certified mail. Each Ranch Hander and a random living comparison were sent an introductory letter and fact sheet signed by the USAF Surgeon General. A Louis Harris and Associates (LHA) interviewer then accomplished an in-home interview. LHA identified two Ranch Handers and nine comparisons who could not be located. All eleven unlocatable subjects were assumed living and remain included in the mortality study. 6. Receipt and Coding of Death Certificates Death certificates were ordered from the vital statistics department of the appropriate state, trust territory, or foreign country. Death certificates or their equivalent were obtained on all appropriate subjects. All death certificates were coded by two individuals, trained by the National Center for Health Statistics (NCHS) in underlying and multiple cause of death coding procedures, using the International Classification of Diseases, Ninth Edition (1977) coding system. Classification of the underlying cause of death was in accordance with NCHS decision tables. Each coder independently classified the underlying and multiple, causes of death and gave the coding worksheet, with each corresponding death certificate, to the coding supervisor, a trained nosologist, for reconciliation. Following reconciliation, one of the coders placed the death code information, by computer terminal, in the death certificate mortality file via a blind verification program designed to mimic the NCHS underlying multiple cause of death coding sheet. At the �conclusion of this initial input of the death codes, a copy of the death certificate was forwarded to NCHS for further validation. The NCHS returned coded death certificates, which were then compared with the Air Force classification. Discordances were resolved in cooperation with NCHS and entered into the data base. 7. Results Chapter II has reviewed the comprehensive, cohesive, sequential ascertainment process of death in the study populations. This process has resulted in the identification of 50 dead Ranch Hand subjects and 250 dead comparison subjects. Although it is understood that early differential ascertainment occurred in the Ranch Hand members (because of detailed knowledge of the study group), it is judged that the overall comprehensive ascertainment process is currently balanced with respect to the two groups. Table 3 and Appendix IV contain summary counts by age, job, and race category for all Ranch Handers and their mortality comparisons; these counts reflect mortality as of 31 December 1982, as known on 27 April 1983. In the stratified analyses, the term "at risk" is defined as simply the number of subjects within a specific stratum, and in life table analyses, as the number of subjects entering a specific age bracket. The term "rate" is the proportion of those individuals "at risk" who are dead. Table 3 OCCUPATIONAL AND RACE SPECIFIC MORTALITY Race Occupation Non-Black Officer-pilot Officer-navigator Officer-other Enlisted-fit eng Enlisted-other Black Officer-pilot Officer-navigator Officer-other Enlisted-fit eng Enlisted-other TOTAL Ranch Hand At Risk Dead Rate Comparisons At Risk Dead Rate 3*»9 80 25 189 528 12 2 1 6 25 .03* .025 .040 .032 .047 1740 390 123 935 2628 72 13 3 46 97 .041 .033 .024 .049 .037 6 0 0 0 2 2 50 .000 .000 .000 .133 .038 .040 13 10 2 75 255 6171 0 0 0 9 10 250 .000 .000 .000 .120 .039 .041 2 1 15 52 1247 �Chapter III RANCH HAND VERSUS COMPARISON GROUP ANALYSES 1. Introduction Overall survival comparisons, without regard to cause of death, were made via survival curve estimation, linear rank procedures, relative risk estimation and standardized mortality ratios. Survival curves were estimated and plotted using the method of Kaplan and Meier (6); 95% confidence bands (7) for each survival curve estimate were also plotted on each graph. Linear rank testing was carried out using the logrank test and Prentice's censored data extension of the Wilcoxon test (8). All linear rank tests were carried out with matched sets merged when Ranch Hands differed by less than one year relative to date of birth, within each stratum of job and race (9). These merged matched sets were regarded as separate strata for testing purposes (9, 10, 11). Relative risk estimates and confidence intervals were computed using an extension of the method of Ejigou and McHugh (12) to variable length, one-to-many matched sets (see Appendix V). Here, due to the one-to-many limitation of the algorithm, matched sets were not merged as when testing procedures were performed. Standardized mortality ratios and associated tests and plots were carried out as in Gail (13). These analyses are fully adjusted for the matching variables, age, race and occupation, but are unadjusted for other variables of interest, such as length of time in Vietnam or Southeast Asia, herbicide dose, time since exposure, time in active duty military, and other medical or occupational risk factors. Some of these variables, such as herbicide dose and time since exposure will be adjusted for in the next analyses, after such data become available. In particular, latency analyses cannot be undertaken at this time but will be included in the next mortality report. In these analyses, we have used summary statistics for which underlying modeling assumptions can be tested. For this reason, we have used the Breslow-Day (13) approach to SMR calculation, rather than the more traditional person-years method. A detailed explanation of this choice is given in Chapter VI. 2. Overall Comparisons Survival time in these analyses was regarded as independent of censorship, if any, and was taken to be age at death. All subjects not certifiably dead, as of 31 December 1982, at the time of analysis, were considered censored at their age on that date. Contact has been lost with two Ranch Handers and nine comparisons as described in Chapter II, but these are not assumed lost to follow-up for the purpose of mortality determination. They are assumed to have been alive on 31 December 1982. With this assumption, no subjects were lost to mortality follow-up before 31 December 1982 in this study. Ranch Hand and comparison group survival curve estimates and their associated 95% confidence bands are shown in Figure 3 and Appendix VI for the five groups: pooled, officers, enlisted, flying and ground personnel, as defined in Table 4. The curves for the pooled groups are shown in Figure 3 with the 95% confidence interval bands deleted in the interest of legibility, but they are included in the group1 specific curves in Appendix VI. Review of �Ranch Hand operations has strongly suggested that Ranch Hand enlisted personnel were more heavily exposed to herbicide than Ranch Hand officers. Further, there is a perception of possible exposure differential between flying and ground Ranch Hand personnel. These notions prompted the above groupings and analyses seen in this and subsequent chapters. Analyses of latency are not possible at this time due to the as yet incomplete nature of the military service data base. These analyses will be performed after the hand review of military tour records has been completed. Figure 3 SURVIVAL CURVE ESTIMATES FOR POOLED RANCH HANDERS AND COMPARISON SUBJECTS 1 0.9 0.8 Z 0.7 t—i > > 0,6 a: co 0.5 0 « * 0.4 •-• 0.3 0.2 0-1 0 R - RANCH HAND C - COMPARISON 10 20 30 YEflRS 70 �Table 4 GROUP DEFINITIONS Group Definition Officer Enlisted Officer-pilot, navigator, other Enlisted-flight engineer, other Flying Officer-pilot, navigator Enlisted-flight engineer Ground Officer-other Enlisted-other Pooled All occupational categories Summary counts by group are shown in Table 5. Ignoring the matching, interaction between officer-enlisted categories and Ranch Hand membership, and interaction between flying-ground categories and Ranch Hand membership was evaluated using log-linear models. No statistically significant interactions were detected. Table 5 SUMMARY COUNTS BY GROUP Group Ranch Hand At Risk Dead Rate Comparisons At Risk Dead Rate Officer Enlisted 463 784 15 35 .032 .045 2278 3893 162 .039 .042 Flying Ground 641 606 22 28 .03** .046 3163 3008 140 110 .044 .037 Pooled 1247 50 .040 6171 250 .041 88 Linear rank procedures were carried out on the same five groups. The results, summarized by test statistics and two-sided P-values, are shown in Table 6. Small P-values, less than .05, indicate significant differences, at the 5% level, between the two groups. These procedures are designed so that the statistic will be positive when the Ranch Handers are dying before the comparison subjects and negative when the comparisons are dying prior to the Ranch Handers. The null hypothesis is that the actual survival distributions of Ranch Handers and their matched comparisons are identical. Each statistic is approximately null distributed as a standard normal random deviate. 11 �Table 6 TEST RESULTS AND P-VALUES FOR OVERALL COMPARISONS Group Logrank (Value) P-Value Wilcoxon (Value) P-Value Officer Enlisted (-0.631) (0.383)' .526 .702 (-0.722) (0.33D .470 .741 Flying Ground (-1.021) (1.023) .307 .306 (-1.116) (0.950) .264 .342 Pooled (-0.047) .962 (-0.123) .902 There is no significant difference, based on these data, between the Ranch Handers and their mortality comparison group. This means that, in particular, the mean ages-at-death of the Ranch Handers and their matched comparisons are not significantly different. In some groups, pooled, officer and flying, the statistics are negative, indicating that the Ranch Handers are living longer than the comparisons, but the differences are, again, insignificant, as evidenced by the large P-values. The situation is reversed for enlisted and ground personnel. These findings are consistent with the observation that, within each group, the comparison confidence bands are contained within the Ranch Hand confidence bands. When matched sets are stratified by five year intervals on year of birth, the same procedures give larger P-values than those in Table 6. Relative risk estimates, the associated 95$ confidence intervals, twosided P-values for testing the null hypothesis of relative risk equal to unity and the associated power are given in Table 7. Here, the power of the test is defined as the conditional probability of rejecting the null hypothesis at the 5% level of significance, given that the relative risk is equal to its estimated value. Table 7 RELATIVE RISKS AND 95$ CONFIDENCE INTERVALS, P-VALUES AND POWER Group Relative Risk 95% Confidence Interval P-Value Power Officer Enlisted 0.763 1.065 (.320- 1.207) (.660- 1.471) .373 .742 .105 .072 Flying Ground 0.734 1.232 (.387 - 1.081) (.694- 1.769) .211 .337 .197 .195 Pooled 0.964 (.658- .819 .051 12 1.269) �The confidence intervals and P-values in Table 7 indicate no significant difference, at the 5% level, between the mortality of the Ranch Handers and comparisons in each of the five groups. Year-of-birth specific mortality rates for each of the five groups are given in Tables 8 through 12, with the corresponding standardized mortality ratios (SMR). In each group, the comparisons are the internal standard. The SMR estimates relative risk in these comparisons if the year-of-birth specific relative risks are all equal (13). A likelihood ratio test for the hypothesis of equal year-of-birth specific relative risks was carried out for each comparison; its P-value is denoted by P1 . In addition, the hypothesis that relative risk is unity, given that relative risk is constant across strata, was tested via a likelihood ratio procedure (13); its P-value is denoted by P2. The SMR and both P-values are given with each comparison. Here, and elsewhere in this report, the denominator of the SMR where n^j is the number of individuals for the ith stratum population and r^ is the death rate, per person, in the standard for the ith stratum. In these calculations the data is stratified birth. Table 8 POOLED SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR - .996; P1-.389, P2-.955) Birth Year Ranch Handers At Risk Dead Rate Comparison At Risk Dead Rate 1905-14 5 2 .1100 1*1 2 .143 1915-19 17 4 .235 96 11 .115 1920-24 1925-29 48 84 3 2 .063 .024 241 501 24 40 .100 .080 1930-31 1935-39 304 207 15 7 .049 .034 1389 1020 67 33 .048 .032 1940-44 1945-54 208 374 5 j_2 50 .024 .032 1096 1814 23 50_ 250 .021 .028 13 is of the jth population on year of �Table 9 OFFICER SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR - .827; P1-.233, P2-.490) Ranch Hand Officers Rate At Risk Dead Birth Year 1910-24 1925-34 1935-39 1940-44 191)5-49 1*1 194 93 90 45 3 4 4 2 2 15 Comparison Officers At Risk Dead Rate 205 930 458 495 190 .073 .021 .043 .022 .044 17 49 11 6 5 88 .083 .053 .024 .012 .026 Table 10 ENLISTED SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR = 1.074; P1=.733, P2=.722) Birth Year Enlisted Ranch Handers At Risk Dead Rate 1905-14 1915-19 1920-24 1925-29 1930-34 1935-39 1940-44 1945-54 4 9 16 41 153 114 118 329 2 1 3 2 11 3 3 10 35 Enlisted Comparisons At Risk Dead Rate .500 .111 .188 .049 .072 .026 .025 .030 12 54 80 211 749 562 601 1624 2 7 11 22 36 22 17 45 .167 .130 .138 .104 .048 .039 .028 .028 162 Table 11 FLYING SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR = .769; P1-.678, P2-.238) Birth Year 1915-24 1925-34 1935-39 1940-44 1945-49 Flying Ranch Handers At Risk Dead Rate 44 272 142 120 63 4 8 6 2 2 22 .091 .029 .042 .017 .032 14 Flying Comparisons At Risk Dead Rate 220 1316 698 653 276 23 71 22 14 10 140 .105 .054 .032 .021 .036 �Table 12 GROUND SPECIFIC MORTALITY RATES BY YEAR OF BIRTH (SMR = 1.257; P1-.535, P2=.302) Birth Year Ground Ranch Handera At Risk Dead Rate Ground Comparisons At Risk Dead Rate 1905-14 1915-24 1925-29 1930-34 5 21 31 85 2 3 2 7 .1400 .143 .065 .082 14 117 151 423 2 12 19 17 .143 .103 .126 .040 1935-39 1940-44 1945-54 65 88 311 1 3 1£ 28 .015 .034 .032 322 443 1538 11 9 40 110 .034 .020 .026 These SMR comparisons are in agreement with the preceding relative risk and linear rank analyses; there is no significant difference in mortality, based on these data, between the Ranch Hand group and the comparison group. 3. Noncause Specific Occupational Comparisons Within-group comparisons by occupation via SMR's, with P-values for testing constant relative risk across year of birth strata (P1) and for testing relative risk equal to unity (P2) are given in Tables 13 through 16. The enlisted and ground personnel are the internal standards in these comparisons. Comparisons via the logrank procedure are given in Table 17. Table 13 RANCH HAND OFFICERS VERSUS RANCH HAND ENLISTED MORTALITY BY YEAR OF BIRTH (SMR «= .544; P1-.280, P2= .087) Birth Year 1905-24 1925-34 1935-39 1940-44 1945-54 Ranch Hand Officers At risk Dead Rate 41 194 93 90 45 3 4 4 2 _2 15 .073 .021 .043 .022 .044 15 Ranch Hand Enlisted At Risk Dead Rate 29 194 114 118 329 6 13 3 3 lp_ 35 .207 .067 .026 .025 .030 �Table 14 RANCH HAND FLYING PERSONNEL VERSUS RANCH HAND GROUND PERSONNEL MORTALITY BY YEAR OF BIRTH (SMR - .581; P1-.382, P2-.100) Birth Year Ranch Hand Fliers At Risk Dead Rate 1905-24 1925-34 1935-39 1940-44 1945-54 44 272 142 120 63 4 8 6 2 _2 22 Ranch Hand Ground At Risk Dead Rate .091 .029 .042 .017 .032 26 116 65 88 311 5 9 1 3 1P_ 28 .192 .078 .015 .034 .032 Table 15 COMPARISON GROUP OFFICERS VERSUS COMPARISON GROUP ENLISTED MORTALITY BY YEAR OF BIRTH (SMR - .697; P1-.640, P2-.015) Birth Year 1905-19 1920-24 1925-29 1930-34 1935-39 1940-44 1945-54 Comparison Officers At Risk Dead Rate 44 161 290 640 458 495 190 4 13 18 31 11 .091 .081 .062 .048 .024 .012 .026 6 5 Comparison Enlisted At Risk Dead Rate 66 80 211 749 562 601 9 11 22 36 22 17 1624 45 T52 M .136 .138 .104 .048 .039 .028 .028 Table 16 COMPARISON FLYING PERSONNEL VERSUS COMPARISON GROUND MORTALITY BY YEAR OF BIRTH (SMR - .930; P1-.305, P2-.867) Birth Year 1905-19 1920-24 1925-29 1930-34 1935-39 1940-44 1945-54 Comparison Fliers At Risk Dead Rate 45 175 350 966 698 653 276 6 17 21 50 22 14 10 140 .133 .097 .060 .052 .032 .021 .036 16 Comparison Ground At Risk Dead Rate 65 66 151 423 322 443 1538 7 7 19 17 11 9 40 110 .108 .106 .126 .040 .034 .020 .026 �Table 17 LOGRANK WITHIN GROUP COMPARISONS Comparison Logrank P-Value RH Officer vs RH Enlisted RH Flyers vs RH Ground Comparison Officer vs Comp Enlisted Comparison Flyers vs Comp Ground -1.468 -1.455 -2.597 -0.363 0.1*12 0.146 0.009 0.717 The SMR and logrank analyses are somewhat in agreement, with both procedures finding significant differences between comparison officers and comparison enlisted, with the officers living longer. The two methods approximately agree on the Ranch Hand fliers versus ground personnel and on Ranch Hand officer versus enlisted personnel with the logrank result near significance at the .10 level; the fliers appear to be living longer than the ground personnel within the Ranch Hand group. 4. Cause Specific Ranch Hand Versus Comparison Mortality Cause specific mortality, relative risks, two-sided P-values for testing relative risk equal to unity, power and 95? confidence intervals for relative risks are summarized in Table 18 for the 1241 matched Ranch Handers and their mortality comparisons. Mortality data for the six unmatched Ranch Handers were not used in this analysis. Of the six, one has died of an accident and the rest are still alive. In some categories, the data were too sparse for relative risk estimation. Table 18 CAUSE SPECIFIC MORTALITY AND RELATIVE RISKS Cause Dead Relative RJH Comparison Risk 95% Conf Int. Accidental 18 Suicide 3 Homicide 2 Infectious, Parasitic 0 Malignant Neoplasm 4 Uncertain Neoplasm 0 Endocrine 1 Mental Disorder 0 Nervous System 0 Circulatory 16 Respiratory 0 Digestive 5 Genitourinary 0 111 Defined 0 Unknown _0_ 92 14 3 49 250 P-value Power .959 1.071 3.333 (.466 - 1.453) (0 - 2.407) (0 - 9.297) .875 .913 .099 .047 .061 .489 .503 (0 - 1.024) .205 .153 5.000 (0 -18.859) .102 .562 1.002 (.411 - 1.594) .994 .050 2.273 (0 - 4.675) .085 .457 3 39 2 1 1 2 70 4 11 3 2 3_< 17 �The low powers in Table 18 reflect the sparseness of data or the fact that some of the observed relative risks approach unity. However, two categories do stand out as deserving further attention: malignant neoplasms and digestive system deaths. It should be noted that if matched sets are ignored and relative risk is estimated using the method of Mantel and Haenszel (14), these results remain essentially unchanged; using this'approach, the relative risk for malignant neoplasms, for example, is .506 with a P-value of .195 and power equal to .254. The 95/& confidence interval for relative risk using this approach is .180 to 1.419. The Mantel-Haenszel relative risk for the digestive system comparison is 2.254, with a P-value of .132 and a power equal to .325; the 95$ confidence interval for relative risk is .782 to 6.501. The digestive system deaths are further defined in Table 19. There has been an increase in deaths due to liver disease among the Ranch Handers; however, this observed difference is not statistically significant. These data are also based on death certificate diagnoses and will be subjected to verification and validation from medical record and autopsy reports. When all deaths from liver disease are considered as a whole, a relative risk of 2.50 is found, with a 95% confidence interval of 0 to 5.501. The P value is 0.083. Similarly, the relative risk for pancreatitis is 2.50 with a 95% confidence interval of 0 to 8.501.; the P value is 0.386. These observations are of interest and will be pursued in depth in subsequent reports. Table 19 DIGESTIVE SYSTEM MORTALITY Deaths Ranch Hand Comparison ICD Code (9th Ed) Pancreatitis (5770) Alcoholic cirrhosis (5712) Nonalcoholic cirrhosis (5715) Nonalcoholic fatty liver (5718) Chronic liver disease (5728) Alcoholic liver disease (5711) Duodenal ulcer (5325) 1 0 3 0 0 1 0_ 5 2 3 3 1 1 0 _J_ 11 Table 20 SITE SPECIFIC MALIGNANT NEOPLASM MORTALITY Deaths Ranch Hand Comparison Site ICD Code (9th Ed) Lip, oral cavity, Pharynx (140-149) Digestive organs, peritoneum (150-159) Respiratory, intrathoracic (160-165) Bone, connective tissue, skin, breast (170-175) Genitourinary organs (179-189) Brain (191-192) Lymphatic and hematopoietic tissue (200-208) No site specification (199) 18 0 0 2 4 8 15 0 1 0 0 J_ 4 1 3 3 4 J_ 39 �The malignant neoplasms are detailed in Table 20, the cell types or the neoplasms, as recorded on the death certificates, are summarized in Table 21. Table 21 MORPHOLOGY OF NEOPLASMS ICD Code 9th Ed. M800 M801-80*» M805-808 M8HJ-838 M872-879 M905 M938-948 M959-963 M965-966 M986 Deaths Ranch Hand Comparison Nomenclature Neoplasms not otherwise specified (NOS) Brain Bronchus and Lung Colon Intestinal Tract Epithelial neoplasms (NOS) Bronchus and Lung Esophagus Kidney Nasopharynx Pancreas Unspecified site Papillary and Squamous Cell Nasal Sinus Lip Tongue Tonsil Andenomas and Adenocarcinomas Appendix Bronchus and Lung Colon Kidney Stomach Nevi and Melanomas Skin (NOS) Mediastinal Mesothelioma Bronchus and Lung Gliomas Frontal Lobe Brain (NOS) Lymphomas NOS and Diffuse Lymphomas (NOS) Hodgkins disease Hodgkin's (NOS) Myeloid Leukemias Acute Myelocytic Leukemia 0 0 0 0 1 3 1 1 1 0 1 0 0 1 8 1 1 1 2 1 0 0 0 0 1 1 1 1 0 0 0 0 0 1 2 1 2 1 0 1 1 0 0 1 0 0 1 1 0 1 0 2 0 4 Epithelial, papillary and adenomas account for 6H% of the comparison neoplasms. Three Ranch Hand neoplasms arose from epithelial cells. There were no tumors in either group which were classified as soft tissue sarcoma. 19 �Chapter IV NONCAUSE SPECIFIC COMPARISONS WITH EXTERNAL POPULATIONS 1. Background and Motivation It is important to know, not only how the Ranch Handers and their matched comparisons relate to each other, but also how they compare with general military and male United States populations. Pitfalls inherent in these comparisons are well known and are briefly reviewed below for specific comparisons with 1978 DoD period life tables for nondisability retired military officer and enlisted personnel (15) and the 1978 U.S. White Male Life Table (16). Although there are difficulties in the use of these comparisons, their use does provide an additional indicator of trends in mortality when viewed in the context of the total analytic process. 2. Adjustment Difficulties Mortality rates in any military population are strongly dependent upon 1) calendar year of death, 2) military status (active duty, separated, retired), 3) selection and retention, and 4) branch of service. Adjustment for these effects was not made in these comparisons because published select Air Force life tables, by calendar year and by status, are not available. In addition, there is also a problem with the statistical method used, since the Gail and Ware (17) procedure assumes constant relative risk with respect to age; the selection effect has been shown to diminish sharply with time making this assumption untenable in these comparisons. The adjustment difficulties (1-4), and their likely consequences, are detailed below. These difficulties apply to all of the comparison groups, but these concerns have less effect on the comparisons of the Ranch Hand group to their matched cohort since these two groups are generally equivalent, relative to these key factors. A. Adjustment for Calendar Year of Death Due to the continuing decrease in overall mortality in the military (18) and in the United States (19), the referenced external age-specific rates are appropriate only for the calendar period of the referenced external life table, that is, 1977-79 for the 1978 period military table used in this analysis. The 1977-79 period rates would, for example, be too low for comparison with subjects dying in 1970 at the age of 40. These subjects would more properly be compared with the death rate for 40 year olds in a 1970 period life table or with a death rate for 40 year olds in a cohort military life table for subjects born in 1930. Calendar time is not taken into account in this analysis because period life tables covering the three decades from 1950 to 1980, for the the active duty, separated and retired Air Force subpopulations, are not currently available. This discrepancy is serious because the decline in death rates in the active duty Air Force during the period 1966 to 1980 has been very substantial (18). B. Adjustment for Military Status (Active Duty, Separated, Retired). The only published military life tables available at this writing are 1978 period tables for DoD nondisability retired officer and enlisted personnel (15) and a series of yearly abridged tables for the active duty Air Force, the first covering the period 1966-1968 and the last, 1978-1980 (18). With these data limitations, adjustment for military status is not possible. It is 20 �olear, however, that there are substantial differences between active duty and retired death rates with the active duty rates being lower than retired rates (15). C. Adjustment for Selection Entry into the military carries with it an effect known as selection, a lengthening of life expectancy due to health prerequisites upon entry into select status and periodic health checks thereafter. This effect is well known to insurance actuaries who have observed that, in insured populations, the effect diminishes as time passes unless there are continued checks on the state of health of the insured persons (20). If selection is to be adjusted for in this analysis, it would be necessary to know Air Force death rates as a function of both age and of time elapsed since entry into the Air Force. It would also be necessary, therefore, to know enlistment and discharge or retirement dates for all study subjects. It is the lack of these data that makes this adjustment impossible at this time. The consequences of this lack of adjustment are not known at this writing. D. Adjustment for Branch of Service Age specific active duty Air Force death rates are substantially lower than the corresponding rates for other services (18). Nonservice specific death rates are therefore too high for appropriate comparison with these two study groups. 3- Comparisons with 1978 DoD Life Tables In Tables 22 and 23, Ranch Hand officers and comparison group officers are contrasted to a 1978 DoD nondisability retired officer life table (15) and in Tables 24 and 25, Ranch Hand and comparison group enlisted personnel are compared with a 1978 DoD nondisability retired enlisted life table (15). In each table, the column labeled "At Risk" lists the number of subjects entering each five year age interval, the column labeled "Deaths" tabulates the numbers of deaths in the age intervals and the column labeled "Expected Deaths" gives the expected numbers of deaths in the age intervals if the study subjects had experienced the same death rates as those specified by the DoD table. The value of the test statistic (17) for testing the null hypothesis of equality is denoted by T; its two-sided P-value is denoted by P. While each table summarizes the findings with five year age intervals for ease of presentation, one year age intervals were used for the computation of the statistic T. All comparisons are conditioned on survival to age 35, since the DoD tables begin at that age. All comparisons are unadjusted for race since the DoD tables are not race specific. 21 �Table 22 RANCH HAND OFFICER VERSUS DOD NONDISABILITY RETIRED OFFICER LIFE TABLE (T - -3.962, P < .001) Age At Risk Deaths Expected Deaths 35-39 40-44 45-49 50-54 456 366 288 173 2 1 1 1 4.183 4.244 4.578 3.099 55-59 57 1 2.043 60-64 30 2 .823 65-68 1 £ 8 .076 19.046 Table 23 COMPARISON OFFICERS VERSUS DOD NONDISABILITY RETIRED OFFICER LIFE TABLE (T - -2.402, P = .016) Age At Risk Deaths Expected Deaths 35-39 40-44 45-49 50-54 2264 1822 1365 842 12 13 24 12 20.837 20.703 21.920 15.901 55-59 60-64 308 145 9 4 10.265 4.377 65-68 .19 JD .601 74 94.604 Table 24 ENLISTED RANCH HANDERS VERSUS DOD NONDISABILITY RETIRED ENLISTED LIFE TABLE (T = -.239, P = .811) Age At Risk Deaths 35-39 668 6 40-44 45-49 50-54 55-59 60-64 392 287 140 41 20 5 5 5 2 2 65-69 6 0 70-71 1 J_ 26 22 Expected Deaths �10UJ.C C-J ENLISTED COMPARISON SUBJECTS VERSUS DOD NONDISABILITY RETIRED ENLISTED LIFE TABLE (T = -3.214, P = .001) Age At Risk Deaths Expected Deaths 35-39 40-44 45-49 50-54 55-59 60-64 65-59 70-74 3299 1945 1437 695 203 103 35 5 21 20 31 14 12 3 1 0 102 33.370 27.681 31.450 20.076 10.980 7.515 2.593 .646 134.311 These findings suggest that, if the effects discussed in section 2 are assumed to be negligible, Ranch Hand officers and comparison officers and comparison enlisted personnel are living longer than expected relative to their respective external populations. Enlisted Ranch Hand personnel are not different from DoD enlisted personnel. In the above DoD comparison there is a suggestion of interaction between officer-enlisted categories and Ranch Hand versus comparison group membership. If matching and time of death are ignored, the following table can be constructed. The term "rate" is as defined on page 8 of this report. Table 26 DEATH AFTER 35 YEARS Ranch Hand Alive Dead Rate Officer Enlisted 448 642 8 26 Alive .018 .039 Comparison Dead Rate 2190 3197 74 102 .033 .031 Analysis using log-linear models shows a statistically significant interaction with pS 0.05. It appears that Ranch Hand officers have a lower mortality after age 35 than Ranch Hand enlisted or comparison officers or enlisted. However, the converse situation is noted considering mortality prior to age 35 and is significant with pSO.05. The data for this analysis of mortality prior to age 35 is set out below. Table 27 DEATHS BEFORE AGE 35 YEAR Alive Officer Enlisted 456 775 Ranch Hand Dead Rate 7 9 > Alive .015 .011 2264 3833 23 Comparison Dead Rate 14 60 .006 .015 �These interactions will require further detailed analysis and evaluation, with specific consideration of medical covariables including risk taking, other life patterns and herbicide. 4. Comparisons with U.S. 1978 White Male Life Table Non-Black Ranch Handera and non-Black comparisons are compared in this section with the population of White males, as represented by the 1978 U.S. White Male Life Table (16). Two serious and well known problems with the use of this table are the lack' of adjustments for the calendar year and selection effects just described; when comparing occupational cohorts with national populations, the selection effect is known as the "healthy worker" effect. The pitfalls of these kinds of comparisons are well documented (21, 22, 23). In Tables 28 and 29, non-Black Ranch Handers and non-Black comparisons are compared, via the method of Gail and Ware (17), with the 1978 U.S. White Male Life Table (16). In Tables 30 through 33, non-Black officers and enlisted personnel in both study groups are compared with the same 1978 U.S. White Male Table. Table 28 NON-BLACK RANCH HANDERS VERSUS 1978 U.S. WHITE MALE LIFE TABLE (T--4.588, P <.001) Age At Risk Deaths Expected Deaths 21-24 25-29 1171 1169 2 6 9.003 9;783 30-34 35-39 1163 1054 7 7 9.396 9.256 40-44 722 5 10.381 45-49 50-54 549 304 6 5 12.085 8.114 55-59 60-64 65-69 70-71 98 50 7 1 3 4 0 1 W 5.039 2.790 0.669 0.089 76.605 24 �Table 29 NON-BLACK COMPARISONS VERSUS THE 1978 U.S. WHITE MALE LIFE TABLE (T = -11.230, P <.001) Age At Risk Deaths Expected Death 19-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 5816 5815 5799 5772 5245 3593 2675 1487 509 248 54 5 1 16 27 23 31 29 50 26 20 7 1 0 231 10.325 55.444 48.592 46.719 46.124 51.041 58.810 40.529 25.210 11:161 3.403 0.601 354.540 Table 30 NON-BLACK RANCH HAND OFFICERS VERSUS 1978 U.S. WHITE MALE LIFE TABLE (T «= -4.575, P < .001 ) Age 25-29 30-31 35-39 40-44 45-49 50-54 55-59 60-64 65-68 At Risk Deaths 151 151 147 362 285 172 57 30 1 3 1 2 1 1 1 1 2 _0 15 25 Expected Deaths 3.794 3.710 4.42Q 5.304 6.370 4.541 3.019 1.302 0.110 32.570 �Table 31 NON-BLACK COMPARISON OFFICERS VERSUS 197$ U.S. WHITE MALE LIFE TABLE (T - -7.923, P < .001) Age At Risk Deaths Expected Deaths 25-29 2253 9 18.880 30-34 2244 5 18;530 35-39 2239 12 22.137 40-44 45-49 50-54 55-59 60-64 65-68 1801 1352 834 308 145 19 13 24 12 9 4 _0 25.841 30.468 23.328 15.157 6.923 0.887 88 162.151 Table 32 NON-BLACK RANCH HAND ENLISTED PERSONNEL VERSUS 1978 U.S. WHITE MALE LIFE TABLE (T - -1.753, P - .080) Age At Risk Deaths 21-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-71 717 715 712 607 360 264 132 41 20 6 1 2 3 3 5 4 5 4 2 2 0 J_ 5.510 5.988 5.686 4.836 5.077 5.716 3.573 2.020 1.488 0.588 0.089 33 40.571 26 Expected Deaths �Table 33 NON-BLACK COMPARISON ENLISTED PERSONNEL VERSUS THE 1978 U.S. WHITE MALE LIFE TABLE (T =-5.923, ? < .001) Age At Risk Deaths Expected Death 19-19 20-21 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 3563 3562 3546 3528 3006 1792 1323 653 201 103 35 5 1 16 18 18 19 16 26 14 11 3 1 0 6.325 33-938 29.713 28.189 23.987 25.200 28.341 17.201 10.053 7.538 2.515 0.601 T4~3 213.601 Given the cautions just described, these findings suggest that the non-Black Ranch Handers and comparisons are living much longer than expected relative to the 1978 U.S. White Male Life Table. The ratios of the observed to the expected deaths described in Tables 28 and 29 reveal that the Ranch Hand and comparison subjects are experiencing death at only 60 to 6556 of the rate of the U.S. White male population. The ratio is 0.461 for the subset of Ranch Hand officers, 0.543 for comparison officers, 0.813 for enlisted Ranch Handers, and 0.669 for enlisted comparison subjects. The healthy worker effect is very likely a major contributor to the undoubtedly real differences between these study groups and the general population. 27 �Chapter V '. COMPARISONS WITH THE WEST POINT STUDY GROUP 1. Background and Motivation The statistical and epidemiological literature is replete with warnings against the uncritical use of the SMR and related summary measures for comparing study groups with published vital statistics for national populations or subpopulations (5), (24), (25). Those cautions are based on the adjustment difficulties described in Chapter 4, Section 2, and departures from the assumption of constant relative risk across age intervals between the study group and the external population. These drawbacks can be avoided by not referencing an external standard at all, by using one of the study groups as the standard (13). or by using as an external standard a group of military personnel, born during approximately the same years, with the same mortality follow-up, as the Ranch Hand and comparison groups. An external group of sufficient size for meaningful statistical comparisons is not available at this time. Mortality and year of birth data are available, however, on a small group of West Point graduates, the subjects of the West Point Follow-up. Study. Although this group is too small for all but very crude statistical comparisons (1), it is the only known external data available at this time. The following comparisons are, therefore, primarily descriptive. The West Point Study Group consists of W members of the West Point graduation class of 1956. These men have been followed up since then for morbidity and mortality. All members of that class were, or still are, officers in the U.S. Armed Forces. The purpose of the West Point study is to investigate the relationship between blood lipid levels and cardiovascular disease. Each study subject is physically examined biennially and blood samples are obtained for lipid and lipoprotein analyses at the USAF School of Aerospace Medicine (26). 2. Noncause Specific Comparisons of Ranch Hand and Comparison with the West Point Study Group Subgroups For the purpose of these mortality comparisons, 15 of the 36 known West Point deaths occurring on or before 31 December 1982 were deleted, 9 of the 15 were killed in action, one was killed in 1959 in the line of duty and 5 were killed in automobile crashes prior to 1962. These deletions imitate the deletion of personnel killed in action from the Ranch Hand and comparison groups. Noncombat or accidental deaths prior to 1962 were deleted because death prior to 1962 would have precluded membership in the Ranch Hand or comparison groups. In addition, one West Pointer who is also a Ranch Hander, was deleted; that individual was alive on 31 December 1982. 28 �A summary of the remaining 21 deaths among the 458 West Point subjects used in these analyses is given by year of birth in Table 34 and by age in Table 35. In Table 35 the column headed "censored" lists by age, the number of West Pointers alive on 31 December 1982. Table 34 WEST POINT DEATHS BY YEAR OF BIRTH Year of Birth At Risk Dead 1930 1931 1932 1933 1934 1935 20 59 90 136 141 12 458 0 2 6 8 4 J_ 21 Table 35 WEST POINT DEATHS BY AGE Age 25-29 30-34 35-39 HO-44 H5-H9 50-52 At Risk Censored 458 456 451 4 4 8 lji)6 162 Dead 0 0 0 0 276 161 437 2 5 3 2 8 J_ 21 In this analysis, non-Black Ranch Hand and comparison officers are compared, without regard to cause of death, with the West Point study group; all of the West Point subjects are non-Black. Non-Black Ranch Hand Officers were matched, one-to-one, by year of birth, to West Point subjects. Due to the relatively small number of Ranch Hand officers and the limited year of birth range imposed by the age of the Class of 1956, only 283 of the 458 West Point subjects received a matched Ranch Hander. Matched sets with West Pointers having the same year of birth were then merged to create six matched sets, corresponding to the six years of birth, 1930 through 35, of the West Pointers. To compare West Pointers with comparison officers, two non-Black comparison officers were matched to each West Pointer by year of birth. All West Pointers received two matched comparison individuals. Matched sets with West Pointers having the same year of birth were merged, giving six. matched sets containing a total of 916 comparisons. 29 �Logrank teats were carried out on these two matched data sets, and the results are summarized in Table 36. In these analyses, survival time is age at death. Censorship is due to survival to 31 December 1982. For those still alive on 31 December 1982, censoring time is age on that day. Table 36 STUDY GROUP VERSUS WEST POINT GROUP LOGRANK COMPARISONS WITH TWO-SIDED P-VALUES Comparison P-Value Ranch Hand officer versus West Point Comparison officer versus West Point .218 .528 An SMR analysis, with the West Pointers being the standard, is summarized in Table 37. Table 37 SMR COMPARISON OF NON-BLACK RANCH HAND AND COMPARISON OFFICERS WITH THE WEST POINT STUDY GROUP (SMR = .530) Birth Year 25-31 32 33-34 35-40 (SMR = .778) Ranch Hand (SMR = 1) Comparison West Point At Risk Dead Rate At Risk Dead Rate At Risk Dead Rate 95 35 60 93 2 1 1 1 8 .021 .029 .017 .043 272 164 257 223 19 7 6 _5 37 .070 .043 .023 .022 79 90 277 12 2 6 12 J_ 21 .025 .067 .043 .083 The test for constant relative risk across year of birth strata gives a P-value of .229. Further, a likelihood ratio test suggests that these SMR's are not different (P = .392). �j. Cause Specific Comparisons The <VHi.se .sponi f \.r. death count,0, for the West Point Study Group are givon in Table Table 38 WEST POINT MORTALITY BY CAUSE Cause Count Accidents Infectious disease Malignant neoplasms Circulatory Digestive Genitourinary 111 defined 6 1 6 5 1 1 1 21 Cause specific comparisons are carried out with three causes, cancer (malignant neoplasms), other diseases, and nondisease (accidents, suicides, homicides and ill-defined), with an adjustment for year of birth by stratification on year of birth. Relative risks are calculated using the method of Mantel and Haenszel (14). These results, based on the counts in Tables 39 and 40, are shown in Table 41. Table 39 CAUSE SPECIFIC COMPARISONS RANCH HAND OFFICERS VERSUS WEST POINT Ranch Hand At Risk Dead West Point At Risk Dead Cause Birth Year Nondisease 1925-1933 1934-1940 166 117 1 M 305 153 5 1 Cancer 1925-1930 1931 1932 1933 1934 1935-1940 72 23 35 36 24 93 0 0 0 0 0 0 20 59 90 136 141 12 0 1 3 1 1 0 Other diseases 1925-1934 1935-1940 190 93 2 1 446 12 8 1 31 �Table 40 CAUSE SPECIFIC COMPARISON COMPARISON OFFICERS VERSUS WEST POINT Birth Year Cause Nondisease Comparisons Dead Number Other diseases 1 11 2 1 2 79 90 136 153 2 2 1 272 164 148 332 2 2 1 79 90 136 153 3 1 1 436 148 109 223 9 1 1 169 136 141 12 1 5 2 1 1929-1931 1932 1933 193^-1937 272 164 148 1929-1931 1932 1933 1934-1937 Cancer West Point Dead Number 1929-1932 332 1933 1934 1935-1937 2 3 1 Table 41 CAUSE SPECIFIC RELATIVE RISKS, P-VALUES 95$ CONFIDENCE INTERVALS FOR LOG RELATIVE RISK 95% Conf Interval Cause Comparison Nondisease RH vs WP Comp vs WP 1.072 0.841 (-1.504 - 1.643) (-1.354 - 1.009) .931 .775 .051 .059 Cancer RH vs WP Comp vs WP 0.690 (-1.634 - .891) .564 .089 Comp vs WP 0.474 0.779 (-3.540 - 2.047) (-2.367 - 1.867) .600 .817 .082 .056 RH vs WP Comp vs WP 0.539 0.728 (-2.191 - .954) (-1.940 - .306) .441 .702 .120 .067 Other diseases RH vs WP All causes RR for Log Rel Risk 32 P-value Power �While the Ranch Hand versus West Point cancer comparison cannot be assessed using the Mantel-Haenszel procedure, the absence of Ranch Hand cancer deaths in this analysis is of interest. This finding is consistent with the apparent but nonsignificantly decreased Ranch Hand cancer mortality noted in the Ranch Hand versus matched comparison group analysis (Chapter III). 33 �Chapter VI STATISTICAL ASPECTS 1. Purpose The purposes of this chapter are 1) to briefly describe each statistical procedure used in the preceding chapters 2) to state the underlying assumptions of each procedure and 3) discuss the validity of those assumptions in this study. The procedures used in this analysis were survival curve estimates and confidence bands, linear rank tests, relative risk estimation and standardized mortality ratios. Points 1-3 are addressed for each procedure in Sections 2 through 5. 2. Survival Curve Estimation and Confidence Bands The survival function of a homogeneous population, S(t), is defined as the probability of surviving t years. The problem is to estimate S(t) and make a confidence statement about that estimate based on randomly censored data. Randomly censored data occur in survival studies since analyses are usually carried out before all subjects have failed. In the present application, failure is defined as death and censorship occurs because most subjects are still living at the time of analysis. Other causes for censorship in this kind of epidemiological study are loss to follow-up or death from causes other than those of interest. Thus far in this study, there have been no subjects lost to follow-up, and all causes of death are of interest. The survival function is estimated here by the product limit estimate K(t), also called the Kaplan-Meier estimate (6). This estimate is derived under the assumption that, in a life testing experiment with n subjects on test, exactly k subjects, with k less than n, are observed to fail; the other n-k remaining are observed only until they are censored. The subjects are assumed drawn randomly from a homogeneous population. Censorship is assumed to be independent of failure. The Kaplan-Meier estimator is asymptotically unbiased and reduces to one minus the empirical distribution function in the absence of censoring. In the present application, the homogeneous populations are the Ranch Handers, the comparisons and various subgroups of these two groups. Death time is taken as age at death measured to the nearest month; censoring time is age" on 31 December 1982, measured to the nearest month. Survival time is age at death or age on 31 December 1982 for those subjects still living. The process n[K(t)-S(t)] converges weakly to a zero mean Gaussian process, as n tends to infinity, under random censorship when the underlying survival function S(t) and the censoring distribution are continuous on a bounded interval (27). This convergence is the theoretical basis for the confidence band algorithm (7) used in Figures 2 and 3, Chapter III and Appendices VI. �The independence of death and censorship can be assumed to hold horc since censorship (survival to December 31, 1982) is not being invoked on individuals because they appear to be at unusually high, or low, risk of death (28). Direct contact has been lost with two Ranch Handers and nine comparisons as described in Chapter II, but these are assumed to be alive, and hence censored at their age on 31 December 1982. The reason for this assumption is that the extensive death ascertainment system is believed to be thorough enough so that, had any of these subjects died, the death would have been detected. Hence, while contact has been lost, loss to follow-up for the purpose of mortality determination has not occurred (29). All other subjects still alive on 31 December 1982 are censored at their age on that date. The validity of inferences based on the estimate K(t) and its associated confidence band depends on the sample size and the observed number of deaths. The sample sizes and numbers of deaths in every stratum used in these analyses exceed the minimum requirements for these procedures (7). The survival curve estimates and confidence bands displayed in Figures 2 and 3 and Appendix VI are not adjusted for year of birth. To do so would have required stratification on year of birth, creating many small strata with associated sample size difficulties. Some year-of-birth adjusted plots in the larger occupational strata will be presented in the next report. 3. Linear Rank Procedures The hypothesis of interest in this analysis is that the actual survival distributions of the Ranch Handers and their matched comparisons are identical. The procedures of choice for testing equality of the two unknown survival distributions based on the matched and censored data in this study are the censored data extensions of the exponential scores and Wilcoxon tests, due to Prentice (8). The first of these is widely known as the logrank test. The test statistics, T, are of the form given by equation 6-23 of (28), where the summands are calculated on matched sets consisting of survival information on one Ranch Bander and his matched mortality comparisons. The statistic T, for either logrank or generalized Wilcoxon summands, is approximately standard normal under the null hypothesis (9). tions other study favor The large sample normal approximation for T will hold when all distribuare continuous and all censoring times are mutually independent of each and independent of death. These assumptions are well satisfied in this since the censorship mechanism, survival to time of analysis, does not one group over the other. In these procedures, the sampling unit is a matched set, so that these tests are adjusted for all matching variables. Prior to calculation, matched sets with Ranch Handers in the same race and job classification having the same year of birth are merged. 35 �The logrank and extended Wilcoxon tests are locally most powerful when the logarithm of the survival times are distributed as extreme value or logistic random variates, respectively. While the efficiency of these procedures peaks at these two underlying distributionsf they have been shown to be robust against departures ( ) These distributional assumptions, however, are 8. not viewed as strictly valid in this study since there is good evidence in the literature that survival time due to certain cancers and other diseases is log normally distributed (30, 31, 32, 33). A linear rank procedure of the Prentice form, whose efficiency peaks under the lognormal distributional assumption, can be constructed" (3*0, but this algorithm is not available at the present time; it will be included in the next analysis. The effect of this departure from the assumptions is considered mild. It should also be noted that these distributional assumptions cannot be checked since these match sets are small and the observations in the combined samples of all matched sets cannot be assumed to have a common distribution. Therefore, reliance must be placed on historical data to determine which linear rank procedure to use. The logrank and Wilcoxon procedures are used here because they are powerful and widely accepted in epidemiology and statistics. 4. Relative Risk Estimation Two relative risk estimators are used in this analysis, a generalization of the Ejigou-McHugh estimator for one to many matched data (12) and the Mantel-Haenszel estimator for stratified data (14). The Ejigou-McHugh estimate was chosen because it allows full adjustment for the one-to-many year-of-birth matching in this study, it is asymptotically as efficient as the maximum likelihood estimator and it is noniterative. The Mantel-Haenszel estimate was chosen because of its ease of calculation, efficiency (35), and general acceptance. It's variance is estimated according to the advice of Anderson et al. (36). Recent work suggests that the variance of the Mantel-Haenszel statistic might be better estimated by a jack-knife procedure (37); this newer method will be carried out in the next mortality report. The Ejigou-McHugh estimator in its published form is suitable only for 1 to R matched designs in which the number, R, of controls matched to each case is the same for all cases. Since the number of controls matched to each Ranch Hander is not the same for all Ranch Handers, the Ejigou-McHugh estimate and its variance was extended to a one-to-many matched design in which the number of comparisons is allowed to vary from case to case. Since this extension is unpublished it is stated in Appendix V for reference. The extended estimate and its variance reduces to the Ejigou-McHugh estimate and variance when all matched sets contain an equal number of comparisons. It is asymptotically efficient and consistent and is noniterative. The Ejigou-McHugh estimate and the Mantel-Haenszel estimate are based on the assumption that relative risk is constant across levels of the matching variable. Some indication that this assumption holds in this study when the data is grouped, by stratifying on year of birth, is furnished by likelihood ratio testing; there is no evidence in this study to suggest that relative 36 �risk is not constant across levels of the matching variables when the event of interest is death from any cause. Therefore, the Ejigou-McHugh and Mantel-Haenszel estimates are appropriate for these data. 5. Indirect Standardization With either an external or internal standard, the SMR is a good summary mortality index for comparing two or more populations, provided the product model, P.jj=ripj, holds, where PJJ is the probability of death in stratum i of population j, r^ is a set of standard stratum specific rates and PJ characterizes the mortality of population j, 1=1,2, ..., I, j=1,2, ..., J, (38, 13). If standard rates are known from some external source and if the product model holds, the best estimate of pj is proportional to the SMR. If J=2, the product model holds, and if one of the two groups is used as the standard, the SMR estimates relative risk. In any case, any SMR summary of mortality data should be preceded by analytical and graphical tests of fit of the product model. Because one of the study groups was always used as the standard in these analyses, the test of fit of the product model was, equivalently, a test of constancy of relative risk across year of birth strata. The fit of the model was verified in each analysis. Further, a likelihood ratio test for equality of population was carried out as described by Gail (13). The results of both tests are summarized by their P-values in each application. The sample sizes in every application are large enough so that chi-square approximations hold; these analyses are, therefore, valid and appropriate. The expected number of deaths in the SMR used in these analyses was calculated as In^rj, where n^j is the number of subjects in the ith stratum of the jth population. The person-years SMR was not used here for two reasons. First, its validity as an estimator of relative risk is dependent upon the fit of the proportional hazards model for which an omnibus test is not currently available. Secondly, the person-years calculation is typically carried out from entry into follow-up (5); in this study, follow-up begins at first entry to Vietnam or Southeast Asia and these entry dates are being verified at this writing. 6. Comparing Observed Life Table Data with a Known Survival Curve The procedure of Gail and Ware (17) is used in these analyses to compare Ranch Hand and comparison group survival data with published period life tables. The basic assumptions of this procedure are that death and censorship are independent competing risks and that the reference curve is a survival distribution for some external population. The test is of the form e v wnere an e are ^°j~ j }/(I j ^ ^» °j $ j observed and expected numbers of deaths in age interval j, and Vj is the variance of OJ-BJ. The statistic is not an omnibus goodness-of-fit test consistent against all alternatives to the null hypothesis that the observed sample comes from a known survival distribution. Rather, it has good power against proportional hazards alternatives or, more loosely, against alternatives for which the observed survival is better (or worse) in every interval than predicted by the known survival curve. 37 �The independence of death and censorship assumption is well satisfied in these data, as discussed in Section 2 of this chapter. The life tables used in these analyses do not, however, represent the survival distribution of any population since they are period, not cohort, life tables. The appropriateness of this procedure is, therefore, dependent upon the extent to which these period life tables approximate the survival distribution of some relevant reference population. These period tables were used because the more appropriate cohort life tables were not available at the time of analysis. �Chapter VII CONCLUSION 1. Introduction The mortality analyses described in this report have not revealed any adverse death experience in the herbicide/dioxin exposed cohort. The results of the analyses, regardless of the source of the comparison data, were consistent: at this time, there is no indication that operation Ranch Hand personnel have experienced any increased mortality or any unusual patterns of death in time or by cause. They are not dying in increased numbers, at earlier ages, or by unexpected causes. The fact that only a relatively small number of Ranch Hand deaths were available for analysis is reassuring in itself. However, the fact that adverse effects have not yet been detected does not imply that an effect will not become manifest at a future time or after covariate adjusted analyses. For this reason, further analyses are intended and mortality in the study population will be ascertained annually for the next 20 years. A summary of the statistical techniques applied to each source of comparison data is presented in Table 42. It should be noted here that these analyses have been carried out without knowledge of covariate information, such as herbicide exposure, industrial chemical exposure, or other risk factors and that these analyses were carried out at a time when approximately 96? of Ranch Handers and their matched comparison subjects were still living. The data, therefore, must be viewed as preliminary to more definitive analyses, which will be performed over the next 20 years. Table i»3 summarizes the results of the noncause specific analyses by source of the comparison data, and Table M presents the results of the cause specific analyses. 39 �Table 42 SUMMARY OF STATISTICAL PROCEDURES USED IN ANALYSIS Comparison Database Internal Comparison Group 1978 1978 DoD U.S. Life West Point White Males Tables Class of 1956 Noncause Speoifio Analyses Logrank & Wilcoxon Procedures Ejigou-McHugh Relative Risk Mantel-Haenszel Relative Risk SMR/Breslow-Day Product Model Gail-Ware Procedure + + •«• + + + + + + Cause Specific Analyses Ejigou-McHugh Relative Risk Mantel-Haenszel Relative Risk + + + Procedure usage is indicated by a "+" symbol. 40 �SUMMARY OF NONCAUSE SPECIFIC MORTALITY ANALYSES BY SOURCE OF COMPARISON DATA Internal Comparison 1978 US Group White Males^ Ranch Hand Group RH = C RH0 = C0 RHE = CE RHp = Cp RHG = CG Comparison Group RH <«US RH0 «<US RHE SUS C <«US C0 <«US CE <«us Internal Occupational Group Specific « £ < « <« P P P P P 1978 DoD West Point Life Tables1 Class of 19562 RH0 «<DoD0 RHE = DoDE CQ < RH0 - WP0 C0 = WP0 RH0 $ RHE RHp $ RHG C0 < CE valued greater value equal to value equal to value equal to value equal to than .10 or less than .10 or less than .05 or less than .01 or less than .001 RH Ranch Hand Group C Comparison Group 0 Officers E Enlisted F Flying G Ground 1 2 Validity of these comparisons is questionable (see Chapter H) Statistical inference is limited by small sample size 3 All P value symbols are based upon SMR and Gail-Ware analysis Table W SUMMARY OF CAUSE SPECIFIC ANALYSES BY SOURCE OF COMPARISON DATA RH Versus Internal Comparison No significant difference in cause specific relative risks RH Versus West Point* No significant difference in cause specific relative risks * Statistical inference is limited by small sample size �2. Internal Comparison Group Based on these early results, there appears to be no significant difference between Ranch Handers and comparisons as regards mortality. This null finding holds for both cause specific and noncause specific comparisons. One within group comparison did yield a significant difference, however. The non-Black comparison officers are living significantly longer than the non-Black comparison enlisted personnel. This may reflect the underlying health care and socioeconomic differences between these two groups. Non-Black Ranch Hand officers also appear to be living longer than non-Black Ranch Hand enlisted personnel, but this finding cannot be viewed as significant, with a P-value of .142 (Table 17). This lack of significance in the Ranch Hand analysis might be attributed to the smaller group sizes within the Ranch Hand cohort in contrast to the comparison cohort. 3. External Comparisons As outlined in the study protocol, considerable effort was expended in the selection of the study comparison group. While the chosen comparison group appeared closest to the Ranch Hand cohort except for herbicide exposure, it seemed appropriate to also contrast the Ranch Hand mortality experience to that of additional comparison groups. Three additional comparison data sets were then selected: mortality data from the West Point Class of 1956, the DoD Nondisability Retired Officer and Enlisted Life Tables for 1978, and the U.S. White Male Life Table, also for 1978. These data sets were chosen in a hierarchical fashion with the expectation that, in the absence of a herbicide effect, the Ranch Handers would have: 1) a mortality pattern comparable to the West Pointers, 2) a lower mortality than the DoD group due to the healthy worker effect, and 3) a still lower mortality than the U.S. male cohort due to healthy worker and military selection effects. These expectations were reassuringly fully realized with respect to overall mortality. Additionally, interesting officer-enlisted differentials emerged. As discussed below, these officer-enlisted differentials may have resulted from sample size effects or from covariable effects, potentially including herbicide exposure. 4. Power Considerations The power limitations of this study, specifically regarding mortality from rare conditions, such as soft tissue sarcoma, were fully acknowledged and described in the protocol (Ref 1, page 67). For example, a fatal disease with an incidence of .001 would require an approximate risk of 4 for a power of 0.8. Power calculations, while desirable for planning and study design, are also revealing at analysis. They are, however, sometimes difficult to carry out without further assumptions. The powers of the logrank and Wilcoxon tests �and the likelihood ratio tests in the SMR analyses are not calculable at this time due to the lack of appropriate methodology. The powers of the tests for cause specific mortality were calculated at the estimated relative risk. The values were low because the estimates of relative risk were close to unity and/or the data were sparse. The null findings in this report are unlikely to have been observed by chance had the true group differences been substantial. For example, if the true overall relative risk were in fact equal to 2, a crude calculation gives a probability of .0007 of observing a relative risk smaller than the observed .96*1 (Table 7). This probability is less than .001 if the true relative risk is 1.5. These findings are, therefore, very likely reflective of a near overall equivalence between Ranch Handers and their matched comparisons. Finally, these unadjusted findings do not preclude the possibility of the emergence of significant differences after adjustment for risk factors. 5. Consistency Patterns When the analysis of each external comparison data base is considered separately, the restrictions inherent in each source limit the strength of the inferences which can be made. However, when the results of all internal and external comparison data bases are considered in context, some patterns of consistency emerge. While some of these patterns may not have firm statistical underpinnings, they still may provide epidemiologic clues to the dynamics of the mortality process. The Ranch Hand officers exhibit a very consistent and predictable pattern across all analyses. As shown in Table ^43, their mortality is nearly the same as that of their most equivalent comparison groups (the matched comparison group officers and the West Point group). As the comparison groups become progressively less equivalent to the Ranch Hand group, the relative mortality of the Ranch Hand officers improves, presumably due to selection comparability (healthy worker effect, etc.). Their mortality is lower than that of their enlisted counterparts; however, this difference is not as striking as is the statistically significant comparable analysis between the matched comparison officers and the matched enlisted personnel. Unfortunately, the cross-comparison trends for the enlisted Ranch Handers are not as clearcut. Their mortality is greater,though not significantly different from their matched comparisons. The enlisted comparison group had a highly significant underrepresentation of mortality against both the DoD and US life tables, whereas the Ranch Handers are equivalent to the DoD group and only marginally better than the 1978 US White males. The consistent observation that the enlisted Ranch Handers appear to demonstrate less of a difference in relative mortality than do their matched comparisons is intriguing. This may reflect an actual increase in mortality due to herbicide exposure or some other factor, or it could be an artifact of small sample size created by the 1:5 matching or basic comparability problems as previously described. The inclusion of substantially more subjects in one group than another can have a profound effect on the significance level of a �statistical technique. Nevertheless, these observations are of interest, and will continue to be subjected to detailed analysis throughout the course of the follow-up study. This trend is consistent with self-perception of herbicide exposure held by many of the Ranch Hand group. Covariate analyses will be conducted, the herbicide exposure index will be applied to these data, and the effects of interaction will be assessed to determine whether the Ranch Hand enlisted findings are real or artifactual. The next mortality assessment will include analyses by person-year of follow-up, adjusted for age in an effort to better address the issue of latency. As the number of deaths in the study population increases with the passage of time, all of the statistical approaches outlined in the protocol (1) will be applied to the data. �Appendix I RANCH HAND II PRINCIPAL INVESTIGATORS COINVESTIGATORS, CONTRIBUTORS, AND MANAGEMENT PERSONNEL A. Principal Investigators George D. Lathrop, MD, MPH, PhD, FACPM Colonel, USAF.MC Chief, Epidemiology Division Patricia M. Moynahan, RN, MS Colonel, USAF, NC Chief, Occupational Epidemiology Section Richard A. Albanese, MD, GM-15 Chief, Biomathematical Modeling Branch Data Sciences Division William H. Wolfe, MD, MPH, FACPM Lt Colonel, USAF,MC Chief, Epidemiology Services Branch B. Coinvestigators Joel E. Michalek, PhD, GS-13 Mathematical Statistician Data Sciences Division Richard C. McNee, MS, GM-13 Chief, Advanced Analysis Branch Data Sciences Division Alton J. Rahe, MS, GS-13 Mathematical Statistician Data Sciences Division �Appendix I (Continued) C. Contributors William J. Besich, BS, GS-12 Computer Systems Analyst Data Sciences Division Vincent V. Elequin, BS, RRA, GS-11 Medical Record Librarian Occupational Epidemiology Section William E. Nixon, BS, GM-13 Computer Systems Analyst Data Sciences Division Thomas J. White, MA Senior Subject Matter Specialist Data Sciences Division i Alvin L. Young, BS, MS, PhD Major, USAF Special Assistant in Military Herbicides D. Management Personnel Project Director: Roy L. DeHart-, MD, MPH, MS, FACPM Colonel, USAF, MC Commander, USAF School of Aerospace Medicine Project Resource Manager: Melvin B. Dobbs Colonel, USAF, BSC Director, Systems Acquisition Research, Development and Acquisition Aerospace Medical Division Air Force HQ Systems Command Coordinator: Ronald D. Burnett Colonel, USAF, BSC Command Bioenvironmental Engineer Office of the Command Surgeon Air Force Surgeon General Coordinator: Robert A. Capell Major, USAF, BSC Assistant for Bioenvironmental Engineering Office of the Surgeon General 46 �Appendix II SCIENCE PANEL Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants. Dr. G. W. Comstock Johns Hopkins Research Center Box 2067 Hagerstown MD 217^0 Dr. John Doull Professor Department of Pharmacology and Toxicology University of Kansas Medical Center Kansas City KA 66103 Dr. John A. Moore (Chair) Deputy Director National Toxicology Program P.O. Box 12233 Research Triangle Park NC 27709 Dr. Richard Monson Professor of Epidemiology Harvard School of Public Health 677 Huntington Avenue Boston MA 02115 Dr. Norton Nelson Professor and Chairman Department of Environmental Medicine New York University School of Medicine New York NY 10016 Dr. Alan Poland Associate Professor of Oncology McCardle Laboratory University of Wisconsin Madison WI 53706 Dr. Irving Selikoff Director, Environmental Sciences Laboratory Mt Sinai School of Medicine 5th Avenue and 100th Street New York NY 10029 47 �Appendix III MATCHING RESULTS IN THE MORTALITY POPULATION The matching results are described here for the mortality population consisting of 1241 Ranch Hands, their 6171 matched mortality comparisons, and the six unmatched Ranch Hands. The matching procedure is described in the Protocol (Ref. 1, pages 23~26). All study subjects were matched perfectly on job category. Three mismatches occurred on race because the recorded race designations for three study subjects were found to be incorrect at the LHA interview. These three subjects were comparisons, two were in the enlisted-other stratum (one was originally recorded as Black and was discovered to be non-Black, the other was originally recorded as non-Black and was discovered to be Black), and one was in the enlisted-flight engineer stratum (he was originally recorded as Black and was discovered to be non-Black). Matching on date of birth was carried out by first expressing date of birth in months from 1 January 1900, to the nearest month; the result is termed month-of-birth. Six discrepancies occurred in matching on month-of-birth due to erroneous months-of-birth for one Ranch Hand and one comparison. These were discovered at the LHA interview. The Ranch Hand, in the non-Black enlisted-other stratum, was discovered to be 72 months older than was recorded prior to the matching. The comparison, in the non-Black officer-pilot stratum, was found to be 15 years younger than was originally recorded. The erroneous Ranch Hand month-of-birth put all five of his matched comparisons 12 months out of range since he was originally perfectly matched to all five mortality comparisons. The erroneous comparison month-of-birth put that comparison 119 months out of range. Given the very small number of mismatches on age and race relative to the number of subjects, their effect was assumed negligible. The matching by month-of-birth, overall, and within each of the ten job and race categories within the mortality population is summarized in this Appendix. The column headed "Age Difference" lists absolute differences of months-of-birth of Ranch Hands and comparisons. The column headed "Number of Comparisons with RH younger (older)" gives, at each level of age difference, the number of comparisons within the level of age difference and older (younger) than the Ranch Hand to whom they are matched. The column headed "Total Count" gives the total numbers of comparisons having the absolute age differences with their matched Ranch Hand given in the first column; in "Total Percent", these counts are expressed as percentages of 6171. These are cumulated in the last two columns. �Appendix III MATCH SUMMARY FOR THE MORTALITY POPULATION Strata Age Difference Comparisons with RH Total Younger Older Count Cumulative Total % * Overall Officer-pilot Non-Black Officer-Pilot Black OfficerNavigator Non-Black 0 1-6 7-12 13-18 19-21 25-30 31-36 37-42 13-18 19-51 55-60 72 179 0 1-6 7-12 13-18 19-21 25-30 31-36 37-12 13-18 19-51 55-60 179 0 1-6 7-12 31-36 37-12 13-18 19-51 55 0 1-6 7-12 713 77 10 22 12 16 10 9 13 17 0 0 706 102 36 22 19 11 19 13 7 7 5 1 272 33 20 8 9 13 7 7 11 11 0 259 32 17 12 11 10 18 11 7 7 1 3 3 0 2 1 0 0 1 0 0 0 0 2 1 70 6 71 0 19 1261 1119 179 76 1 1 31 30 29 22 20 21 5 1 1261 5710 69.0 23.5 2.9 1.2 0.7 0.5 0.5 0.5 0.1 0.3 0,1 0.1 0.0 5889 5965 6009 6010 6070 6099 6121 6111 6165 6170 6171 961 531 65 37 20 20 23 25 18 18 21 1 55.2 30.5 3.7 2.1 1.1 1.1 1.3 1.1 1.0 1.0 1.2 0.1 961 1192 1557 1591 1611 1631 1657 1682 1700 1718 1739 1710 89.5 91.6 92.8 93.9 95.2 96.7 97.7 98.7 99.9 100.0 0 3 3 1 0.0 23.1 0.0 23.1 16.2 53.9 69.2 76.9 92.3 100.0 61.5 98.5 100.0 2 1 2 1 7.7 15.1 7.7 15.1 7.7 0 3 6 7 9 10 12 13 210 111 6 61.5 36.9 1.5 210 381 390 23.1 69.0 92.5 95.1 96.7 97.1 97.9 98.1 98.8 99.2 99.5 99.9 100.0 100.0 55.2 85.8 �Appendix III (Continued) MATCH SUMMARY FOR THE MORTALITY POPULATION Strata OfficerNavigator Black Number of Total Comparisons with RH Age Younger Older Count % Difference 0 1-6 7-12 13-18 19-21 25-30 31-36 Officer-Other Non-Black 0 1-6 7-12 13-18 19-21* 25 Officer-Other 13-18 1 0 2 0 0 0 0 2 1 1 1 1 38 2 1 1 0 57 8 1 0 1 2 0 Cumulative Total % 1 1 1 1 1 1 1 10.0 10.0 10.0 10.0 10.0 10.0 10.0 1 2 6 7 8 9 10 10.0 20.0 60.0 70.0 80.0 90.0 100.0 11 95 11.1 77.2 10 2 1 1 8.1 1.6 0.8 0.8 11 109 119 121 122 11.1 88.6 96.8 2 98.1 123 99.2 100.0 100.0 2 100.0 516 306 55.2 32.7 55.2 87.9 6.7 516 822 885 Black EnlistedFlight 0 1-6 Engineer Non-Black 7-12 13-18 19-21 25-30 31-36 37-12 46 EnlistedFlight Engineer Black 111 34 11 0 2 2 1 1 7 6 1 0 0 63 30 7 8 3 1 1 26 7 22 5 0 0 10 18 12 3 2 13.3 61.0 16.0 16 0 1-6 Non-Black 7-12 13-18 19-21 25-30 31-36 37-12 13-18 72 0 1-6 7-12 3.2 91,7 97.9 98.6 99.5 99.8 99.9 3 2 116 1 1 91 11 0.7 915 922 0.9 930 0.3 933 931 935 100.0 1.0 2.7 10 58 70 73 75 13.3 77.3 93.3 97.3 100.0 2382 0 1-6 7-12 19-21 55-58 EnlistedOther EnlistedOther Black 165 29 90.6 2382 7.9 0.5 0.2 0.5 0.0 0.1 0.0 0.1 0.2 2589 90.6 98.5 99.0 99.1 99.6 99.6 53.7 11.7 10 1 1 0 0 0 2 0 1 1 2 5 207 12 1 12 1 2 1 2 5 19 0 65 1 137 111 1 3 50 0.1 0.1 1.6 2601 2605 2617 2618 2620 2621 2623 2628 99.7 99.7 99.8 100.0 137 251 255 53.7 98.1 100.0 �Appendix IV YEAR OF BIRTH, OCCUPATIONAL AND RACE SPECIFIC MORTALITY Birth Ranch Hand Job Category, Race Year At Risk Dead Death Rate Officer-Pilot, Non-Black .375 1 915-19 1 920-21 1 925-29 1 930-31 1 935-39 1 910-11 1 915-19 8 31 31 113 66 60 10 3 0 0 3 3 1 Comparison At Risk Dead 1 13 11 Death Rate 23 8 5 5 .103 .081 .060 .050 .025 .011 .028 TOTAL 319 Officer-Pilot, Black 2 .027 .015 .017 .050 39 155 232 156 326 351 178 12 .031 1710 72 .011 0 1 3 2 0 0 0 0 3 1 6 0 0 0 0 0 6 0 13 0 9 35 21 13 17 163 105 67 8 3 7 3 0 0 .061 .013 .029 2 0 1 1 0 0 80 2 390 13 .033 1 1 0 0 6 1 0 0 2 0 10 0 1 0 1 3 2 1 0 0 0 0 0 0 2 3 6 11 0 0 0 1 1 0 1 0 1 930-31 1 935-39 1 910-11 1 915-19 TOTAL Officer-Navigator 1 925-29 Non-Black 1 930-31 1 935-39 1 910-11 1 915-19 TOTAL Officer-Navigator 1 930-31 Black 1 935-39 TOTAL Officer-Other, Non-Black 1 910-11 1 915-19 1 920-21 1 925-29 1 930-31 1 935-39 1 910-11 1 915-19 TOTAL 13 1 .029 .018 .025 12 .077 19 66 1 .010 1 123 0 25 51 .091 .083 .015 .021 �Appendix IV (Continued) Job Category, Race Officer-Other, Black Birth Ranch Hand Year At Risk Dead Death Rate Comparison At Risk Dead Death Rate 1940-44 TOTAL Enlisted-Fit Eng Non-Black 1 0 0 3 2 0 0 1925-29 1930-34 1935-39 1940-44 TOTAL Enlisted-Other Non-Black 1905- 9 1910-14 1915-19 1920-24 1925-29 1930-34 1935-39 1940-44 1945-49 1950-54 1 6 5 3 0 1 0 1 15 2 0 4 8 12 28 76 52 67 270 11 0 2 0 3 2 6 1 2 9 0 25 Enlisted-Other Black .333 .200 .049 .049 .041 .033 .056 935 46 .049 1 5 1 .333 10 34 16 15 2 .100 .150 .063 .133 .133 75 9 .120 2 10 48 60 140 0 .167 .500 .250 .071 .079 .019 .030 .033 376 263 340 1333 56 2 5 7 18 14 8 7 36 .200 .104 .117 .129 .037 .030 .021 .027 0 52 52 2628 97 .037 .143 35 40 35 .034 145 2 3 1 4 .057 .075 .029 .028 .038 1 0 0 1 TOTAL 2 4 3 15 10 7 5 .032 Enlisted-Fit Eng Black .0*17 .042 6 20 61 304 243 211 90 1.000 255 10 .039 .047 �Appendix V THE EXTENDED EJIGOU-McHUGH RELATIVE RISK ESTIMATOR Let R k , k=1,2, ..., K, denote the distinct numbers of comparisons matched to the cases and let nk denote the number of matched sets with exactly Rk comparisons. A matched set is defined as the case and his matched comparisons. Let n=ni+n2+ ... +nj( denote the total number of matched sets. Define Zk«i,T» k-1 .2 ..... K, i-0,1, by Zk 0 T * tne number of matched sets, among those having exactly Rk comparisons, in which the case is alive and exactly T of the R k comparisons have died, T=1,2, ..., Rk Zk,1 T * the number of matched sets, among those having exactly Rk comparisons in which the case has died and exactly T of the Rk comparisons have died, T=0,1,2, ..., Rk-1 . The extended estimate, \p, is given by K 1 Rk I zk,0,Tzk,1 ,T-1/(zk,0,T+zk,1 ,T-l) K Rk Z * TZ2k,0,T/(Rk-T+1)(Zk,0,T+Zk,1,T-l) k-1 T-1 and its variance is estimated by K Rk Z Z Zkf0>T/[i|;+(Rk-T+1)/T] k-1 T-1 Since K is finite, this estimate has the same distributional properties as the Ejigou-MoHugh estimator; it is asymptotically efficient and unbiased. The underlying assumptions used in its derivation are that the disease under study is of low incidence and that relative risk is constant over the levels of the matching variables. The Ejigou-McHugh estimate and the above extension are equivalent in asymptotic efficiency to maximum likelihood estimation (12). 53 �Appendix VI Figure 4 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR POOLED RANCH HANDERS C.9 0.3 Ci) z 0.7 - 14 - O.G - CO C'.V> - 2: C3 C.4 CT ce: 0 . 3 C.I . 0I 0 <Q 10 bO 70 RGF. Figure 5 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR POOLED COMPARISONS I 0.9 0.0 § 0«7 »—i > £ 0.6 W Q.& 2 0.4 I— o CE or. ° •3 u. 0.2 0.1 0 0 J_ _. 10 :ji"" fiC-f 54 I '"0 70 �Figure 6 SURVIVAL CURVE ESTIMATE AND'95% CONFIDENCE BANDS FOR RANCH HAND OFFICERS 0.9 O.Q 0.7 0.6 9s 0.5 •z. CJ 0 0.3 0.2 0.1 0 0 10 20 30 60 70 ROE Figure 7 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR COMPARISON OFFICERS I 0-9 o.e 0.7 0.6 z H- 0.4 0.3 0.1 0 . IK 30 60 ROE 55 70 �Figure 8 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR RANCH HAND ENLISTED PERSONNEL 1 0.9 0.8 g 0.7 oc to 0.5, z 2 Q.4 Iu S 0.3 u. 0.2 O.I 0 10 30 60 70 RGE Figure 9 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR COMPARISON ENLISTED PERSONNEL I 0.9 O.Q g 0.7 o.e cn z S 0.4 0.3 0.2 0.1 0 1— 10 no f. 56 60 70 �Figure 10 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR RANCH HAND FLYING PERSONNEL 1 0.9 0.0 0.7 > 0.6 or O.b z o £ 0.3 U_ 0.2 0.1 0 10 20 30 E>0 60 70 flGF. Figure 11 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR COMPARISON FLYING PERSONNEL 1 0.9 0.0 Z 0.7 U 0.6 W 0-b CJ 0.3 0.2 0.1 0 10 20 30 bO 57 60 70 �Figure 12 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR RANCH HAND GROUND PERSONNEL 0.9 0.9 g 0.7 £ 0.5 to 0.5 z •-• 0.4 ho £ 0.3 u_ 0.2 0.1 10 20 30 60 70 ROE Figure 13 SURVIVAL CURVE ESTIMATE AND 95% CONFIDENCE BANDS FOR COMPARISON GROUND PERSONNEL 1 0.9 0.8 Si 0.7 > £ 0.6 W 0.5 Z •-< 0.4 o £ 0.3 u_ 0.2 0.1), 0 0 10 30 40 flOE 58 £.0 70 �REFERENCES 1. Lathrop, G. D., Wolfe, W. H., Albanese, R.A. and Moynahan, P. M. Epidemiclogical investigation of health effects in Air Force personnel following exposure to herbicides: Study Protocol. SAM-TR- 82-4*1, Dec 1982. 2. Honchar, P. A., and Halperin, W. E. 2,4,5-T, trichlorophenol, and soft tissue sarcoma. Lancet Jan 31:268-269 (1981). 3. Hardell, L., and Sandstrom, A. Case-control study: soft tissue sarcomas and exposure to phenoxyacetic acids or chlorophenols . Br. J. Cancer 39:711-717 (1979). 4. Eriksson, M. , Hardell, L., Berg, N. 0., Moller, T., and Axelson, 0. Case-control study on malignant mesenchymol tumors of the soft tissue and exposure to chemical substances. Lakartidningen 76:3872-75 (1979). 5. Breslow, N. E., Lubin, J. H., Marek, P. and Langholz, B. Multiplicative models for cohort analysis. Journal of the American Statistical Association 78:1-12 (1983). 6. Kaplan, E. L. and Meier, P. Nonparametric estimation from incomplete observation. J. Am. Stat. Assoc. 53:1»57-II81 (1958). 7. Hall, W. J, and Wellner, J. A. Confidence bands for a survival curve from censored data. Biometrika 67:133-1^3 (1980). 8.. Prentice, R. L. Linear rank tests with right censored data. Biometrika 65:167-179, (1978). 9. Michalek, J. E. and Mihalko, D. Linear rank procedures for matched observations. SAM-TR-83-16 (In Press), May 1983. 10. Michalek, J. E. and Mihalko, D. Matched survival analysis (MSURV). SAM-TR-83-15 (In Press), May 1983. 11. Michalek, J. E. and Mihalko, D. Linear rank procedures for litter matched data. Biometrics (to appear). 12. Ejigou, A. and McHugh, R. Relative risk estimation under multiple matching. Biometrika 68:85-91 (1981). 13. Gail, M. The analysis of heterogeneity for indirect?- standardized mortality ratios. Journal of the Royal Statistical Society, Series A (1978). . Mantel, N. and Haenszel, W. Statistical aspects of the analysis of data from retrospective studies of disease. Journal of the National Cancer Institute 22:719-748 (1959). 59 �15. Valuation of the Military Retirement System FY 1980. Office of the Actuary, Defense Manpower Data Center, 300 North Washington Street, Alexandria, Virginia 22314. 16. Vital Statistics of the United States, 1978 Vol II - Section 5, Life Tables; U.S. Dept of Health and Human Services, DHHS Publication No (PHS) 81-1104; Hyattsville, Maryland:1980. 17. Gail, M. H. and Ware, J. H. Comparing observed life table data with a known survival curve in the presence of random censorship. Biometrics 35:285-391 (1979). 18. Annual Report, 1966-1981, Servicemen's Group Life Insurance Program, Insurance Service, Department of Veterans Benefits, Veterans Administration. 19. Cohort Mortality and Survivorship: United States Death-Registration States, 1900-1968. Public Health Service. DHEW Publication No (HSM) 73-1400. 20. Elandt-Johnson, R.C. and Johnson, M.L. New York: John Wiley, 1980. Survival Models and Analysis 21. McMichael, A. J., Standardized Mortality Ratios and the "Healthy Worker Effect": Scratching Beneath the Surface. J. of Occupational Medicine, Vol 18, No. 3, March 1976. 22. Tabershaw, I. R., Special Communications, "What Do We Expect from an Occupational Cohort?", J. of Occupational Medicine, Vol 17, No. 2, February 1975. 23. Enterline, P. E. ; Pitfalls in Epidemiological Research; J. of Occupational Medicine, Vol 18, No. 3, March 1976 24. Gaffey, W. R. A critique of the standardized mortality ratio. Journal of Occupational Medicine 18:157-160 (1976). 25. Wong, 0. Further criticisms of epidemiological methods in occupations studies. Journal of Occupational Medicine 19:220-222. 26. Clark, D., Allen, M. and Wilson, F. Longitudinal study of serum lipidstwelve year report. American Journal of Nutrition 20:743-752 (1967). 27. Breslow, N. and Crowley, J. A large sample study of the life table and product limit estimates under random censorship. The Annals of Statistics 2:437-453 (1974). 28. Kalbfleisch, J. D. and Prentice, R. L. The Statistical Analysis of Failure Time Data. New York: John Wiley, 1980. 29. Monson, R. R. Occupational Epidemiology, CRC Press, Inc., Boca Raton FL (1980). 60 �30. Feinleib, M. A method of analyzing log-normally distributed survival data with incomplete follow up. Journal of the American Statistical cal Association 55:53*1-535 (1960). 31. Feinleib, M. and McMahon, B. Variation in the duration of survival of patients with chronic leukemias. Blood 15:332-3*19 (1960). 32. Nice, C. M. and Stenstrom, K. W. Radiology 62:6*11-652 (195*1). Irradiation therapy in Hodgkins disease. 33- Osgood, E. E. Methods for analyzing survival data, illustrated by Hodgkin's disease. American Journal of Medicine 2*4:*IO-*I7 (1958). 3*1. Mehrotra, K. G., Michalek, J. E., Mihalko, D. and White, T. Score computation for linear rank procedures. Journal of Statistical Computation and Simulation 16:201-211 (1982). 35. Breslow, N. Odds ratio estimators when the data are sparse. Biometrika 68:73-84 (1981). 36. Anderson, S., Auquier, A., Hauck, W., Oakes, D., Vandaele, W., and Weisberg, H. Statistical methods for comparative studies, techniques for bias reduction. New York: John Wiley, 1980. 37. Breslow, N. E. and Liang, K. Y. The variance of the Mantel-Haenszel estimator. Biometrics 38:9-43-952 (1982). 38. Breslow, N. E. and Day, N. W. Indirect standardization and multiplicative models for rates, with reference to the age adjustment of cancer incidence and relative frequency data. Journal of Chronic Diseases 28: 289-303 (1975). 61 �PROJECT RANCH HAND II AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES BASELINE MORTALITY STUDY RESULTS 30 JUNE 1983 Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 Approved for public release; distribution unlimited EPIDEMIOLOGY DIVISION DATA SCIENCES DIVISION USAF SCHOOL OF AEROSPACE MEDICINE (AFSC) BROOKS AIR FORCE BASE, TEXAS 78235 �EXECUTIVE SUMMARY Baseline Mortality Study The Ranch Hand II epidemiologic study'uses a matched cohort design in a nonconcurrent prospective setting, incorporating mortality, morbidity, and follow-up studies. The purpose of this report is to present the baseline mortality study results. , ,>••• Since 1979, a detailed population ascertainment process has enumerated a total of 1269 Ranch Hand personnel who served in Vietnam during the period of 1962-1971. As described in the protocol, this total is believed to comprise the entire exposed study population. The eligibility of each Ranch Bander was verified by a hand review of his personnel record. A comparison group was formed by identifying all individuals assigned to selected Air Force organizational units with a mission of flying cargo to, from, and In Vietnam during the same period. All Ranch Hand and comparison subjects designated as killed in action were removed from the study population. By a computerized nearest neighbor selection process, up to 10 comparison individuals were matched to each Ranch Hander by job category, race, and age to the closest month of birth. A hand record review of the matched comparison sets revealed that on the average, 8.2 comparison individuals were fully suitable for study. From each matched comparison set, five individuals were randomly selected for the mortality study, yielding a 1:5 design. Every Ranch Hander and his set of comparisons will be the subjects of annual mortality updates throughout the entire 20 years of the follow-up study so that emerging mortality patterns or disease clusters may be detected with maximal sensitivity. Each living Ranch Hander and his first and willing comparison match were selected to participate in a comprehensive physical examination and an in-home interview; the results of this study will be presented in a subsequent report in late 1983. A mortality determination on 1,247 Ranch Banders and 6,171 comparison subjects was made, sequentially using the data sources of the Air Force, Veterans Administration, Social Security Administration, Internal Revenue Service, and personal contact efforts. As of December 31, 1982, 50 Ranch Hand and 250 comparison subjects had died (certified on/before April 27, 1983). Death certificates were obtained on all 300 deceased subjects and were coded by an Air Force nosologist (ICD, 9th ED). All codings were verified by the National Center for Health Statistics. Autopsy results are currently being sought for future analyses. Statistical analyses of noncause specific death emphasized survival curve estimates, linear rank procedures, relative risk estimates, and standardized mortality ratios (SMRs). Cause specific analyses were limited to relative risk estimates because of small cell sizes. In addition to these approaches, three other data bases were contrasted to the Ranch Hand population, where possible; the 1978 US White Male Mortality experience, the 1978 Department of Defense (DoD) Nondisability Retired Life Table, and the mortality experience of the West Point Class of 1956. These additional comparison groups have substantial comparability or sample size limitations, rendering conclusions to the weakest order. Analyses with these "external" comparison groups were accomplished to crudely define the healthy worker effect and to determine if the Ranch Hand group mortality was drastically out of line with that of other military populations. �Data analysis showed that the mortality experience of the Ranch Hand group is nearly Identical to that of the comparison group. Analyses showed that officers are living" longer than enlisted personnel in both Ranch Hand and comparison groups. This difference between officers and enlisted personnel was statistically significant in the comparison group whereas it was not in the Ranch Hand cohort. A contrast of the Ranch Hand and comparison group to the 1978 DoD Life Table showed significantly less mortality for Ranch Hand officers, comparison officers and comparison enlisted men, however, there was not a statistically significant favorable mortality rate for Ranch Hand enlisted personnel. This pattern of mortality was also seen in a contrast of the Ranch Hand and comparison groups to the 1978 U.S. white male mortality experience. That is, highly favorable mortality differentials for Ranch Hand officers, comparison officers and comparison enlisted personnel were observed, but not for Ranch Hand enlisted. This trend is consistent with the self perceptions of differential herbicide exposures reported by many of the Ranch Hand subjects. The reason(s) for these observations are speculative at present, but may include the related items of sample size, socioeconomic differences, access to medical care, and health education and possible herbicide effects. Cause specific analyses were statistically nonsignificant. The Ranch Handers showed a relative paucity of overall cancer but an excess of digestive disorder deaths, both statistically nonsignificant. No soft tissue sarcoma deaths were detected in either group. Analyses of both the Ranch Hand and the comparison groups to the 1978 US White male mortality experience showed highly significant favorable findings. Most of these differences are speculatively attributed to the healthy worker effect. A contrast of the Ranch Hand and comparison groups to the 1978 DoD Life Table showed significantly less mortality for Ranch Hand officers and comparison officers and enlisted men. The West Point comparison showed nonsignificant SMRs of 0.530 and 0.778 for the Ranch Hand officers and the comparison group officers, respectively. Overall, the limitations of the statistical power calculations in most of these analyses were substantial in most analyses due to 1) the low mortality rate (4%) in the Ranch Hand and comparison groups to date, 2) the inherently small group of Ranch Handers (as described in the study protocol), and 3) the observed relative risks which approached unity in most categories. This baseline mortality report can in no way be regarded as conclusively negative because this small, young, and relatively healthy cohort may not have yet reached the latency period wherein attributable fatal disease might be expected and detected within limited power boundaries of this study. Future commitments for the annual mortality updates include detailed covariate analyses for disease risk factors, herbicide exposure, and confounding industrial chemical exposures. Further, subsequent morbidity reports will include full spectrum, disease specific analyses, e.g., cancer (fatal, ongoing, cured) in an effort to enhance study sensitivity to emerging herbicide effects, if they occur. ii �AIR FORCEHEALTH STUDY (PROJECT RANCH HAND II) AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES BASELINE MORBIDITY STUDY RESULTS 24 FEBRUARY 1984 Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 George D. Lathrop, Colonel, USAF, MC William H. Wolfe, Lieutenant Colonel, USAF, MC Richard A. Albanese, M.D., GM-15 Patricia M. Moynahan, Colonel, USAF, NC Approved for public release; distribution unlimited. .£>. USAF SCHOOL OF A E R O S P A C E MEDICINE Aerospace Medical Division (AFSC) Brooks Air Force Base, T e x a s 7 8 2 3 5 �.* EXECUTIVE SUMMARY BASELINE MORBIDITY STUDY The Ranch Hand II epidemiologic study uses a matched cohort design in a nonconcurrent prospective setting, and .incorporates mortality, morbidity, and follow-up studies. -The purpose of this report is to present the baseline morbidity study. The morbidity study design matched each living Ranch Hander (by age, Job, and race) to the first living and compliant member of a randomly selected comparison mortality set of 5 individuals, producing a 1:1 contrast. The comparison group was formed from numerous flying organizations which transported cargo to, from, and within Vietnam, but were not involved in aerial spray operations of Herbicide Orange. Of the potential study participants, 99.5$ were located. Early in the physical examination phase of the study, it was discovered that 18$ of the entire comparison group was ineligible to participate because of inappropriate selection. Thereafter, study eligibility was certified only after a hand-review of personnel records. Next-in-line compliant comparisons entered the study as replacements after fully completing the questionnaire and physical examination. Statistical analyses of these replacement individuals later showed that they differed from the original comparisons in a variety of subtle and often opposite ways. As a conservative measure to avoid possible bias by the inclusion of the replacements in the analyses, a management decision was made to base the statistical tests in this report primarily upon contrasts of the Ranch Hand group to the original comparison group. The preponderance of data was obtained from the in-home interviews and the physical examination, each conducted under contract to the Air Force by Louis Harris and Associates, Inc., New York NY, and the Kelsey-Seybold Clinic, P.A., Houston TX, respectively. All contacts with the participants were carried out with utmost professionalism and sensitivity. Other morbidity data sources included reviews of medical records, military personnel documents, and birth certificates; in-home questionnaires and telephone questionnaires of the study participant's wives, former wives and, occasionally, their next-of-kin. All aspects of the study were voluntary. As a contract requirement, data collection personnel were blind as to the exposure status of the participants. Ninety-seven percent of the Ranch Handers and 93$ of the comparisons participated in the in-home interview. For the physical examination, 87$ of the Ranch Handers and 76$ of the comparison group participated, a total of 2,272 individuals. This differential attendance at the examination may have introduced a potential participation bias that, in a military population predominantly engaged in flying duties, is multifactorial and complex. All study phases were monitored by stringent quality control standards. Statistical analyses of the data consisted primarily of log-linear models, logistic regression techniques, generalized linear models, matched covariate analyses, and Kolmogorov-Smirnov, chi-square, and t tests. The physical examination and the in-home questionnaire data were analyzed by major organ system. In terms of general health, more Ranch Handers perceived themselves to be in fair or poor health than did their comparisons. No �group differences were detected for hematocrit or percent body fat determinations. Unadjusted group differences in sedimentation rate were not observed; however, significantly more young comparisons had abnormalities in sedimentation rate than did their Ranch Hand counterparts. There were no statistically significant differences in the occurrence of malignant or benign systemic tumors between the groups. One case of soft tissue sarcoma was found in a comparison member. Significantly more nonmelanotic skin cancer was noted in the Ranch Hand group, but these - analyses have not yet considered (adjusted for) sunlight exposure, the prime etiology of these cancers. Such nonmelanotic skin cancer (predominantly basal cell carcinoma) is the most common neoplasm in the White population of the United States. Up to the statistical limits of the study there were no consistent data that showed that the Ranch Handers were developing uncommon cancers, or cancer in unusual sites, or at an unusual age. Measures of fertility and reproductive outcome showed mixed results. It is emphasized that the fertility and reproductive results are preliminary at this time as they are based largely upon subjective self reports that await full medical record and birth certificate verification. Four measures of fertility: number of childless marriages, couples with the desired number of children, the infertility index and the fertility index, showed no difference between the Ranch Hand and comparison groups. A semen specimen obtained from those willing and able to provide one showed no group differences with respect to total sperm count or percent abnormal sperm. There were no significant findings in conception outcomes for miscarriages, stillbirths, induced abortions, or live births. For live birth outcomes no differences were observed for prematurity, learning disability, or infant deaths. There was no significant disparity between groups for the classifications of severe or moderate birth defects. By parental history, however, Ranch Hand offspring showed significantly more minor birth defects (birth marks, etc). Reporte'd neonatal deaths and physical handicaps were also significantly excessive in the Ranch Hand group when contrasted to the total comparison group. All fertility and reproductive findings in the Ranch Hand group showed inconsistent relationships to the herbicide exposure index. Medical records and birth certificates are currently being chronicled for complete verification of all historical findings. A comprehensive neurological examination showed no consistent abnormalities in the cranial nerves, peripheral nerves or central nervous system function of the Ranch Handers. As expected, there was a profound influence of diabetes and alcohol in both groups upon numerous neurological tests. Detailed psychologic data were obtained on all participants at both the in-home interview and the physical examination. It is emphasized that the majority of psychological data was derived from self reported responses during interview and has not been fully assessed for the effect of differential reporting. A variety of subjective deficits (fatigue, anger, fear, anxiety, etc) were significantly more common in the high school educated Ranch Handers. Educational level significantly and consistently influenced most subjective test results. In sharp contrast, more objective performance testing by the Halstead-Reitan battery and IQ testing did not reveal any significant intergroup differences. The roles of overreporting and the Post Vietnam Stress Syndrome In these analyses have not as yet been assessed. Liver function tests and clinical history data showed mixed results. Ranch Handers had some elevated liver enzyme tests and lower cholesterol levels. More Ranch Handers were found to have hepatomegaly and verified histories of prior hepatic disease than their counterpart comparisons. Exposure to alcohol, degreasing chemicals, and industrial chemicals in general, influenced ii �the liver test results. Ranch Handers reported significantly more symptoms resembling porphyria cutanea tarda than the comparisons-, but these data have not been verified by medical record reviews nor were they substantiated by laboratory testing or by physical examination. Exposure index analyses were essentially negative. In the dermatologic evaluation, no cases of c'hloracne were diagnosed clinically or by biopsy. A thorough questionnaire analysis of acne showed that the incidence, severity, duration, and anatomic location did not differ between groups, and suggested that the historical occurrence of chloracne was highly unlikely in the Ranch Handers. Evaluation of the cardiovascular system showed equal proportions of abnormalities in blood pressures, electrocardiograms, past electrocardiograms, and heart sounds in both groups. Ranch Handers are not having premature heart attacks or generalized heart disease. However, the Ranch Handers showed significant deficits in 2 specific peripheral leg pulses and all leg pulses as a group. These puzzling findings were highly correlated with age and smoking patterns, and verified past heart disease. The assessment of the immune system by laboratory testing was compromised by excessive test variability. An independent review committee determined which test data were suitable for statistical analysis. As an unexpected finding, the test data were significantly influenced by the age and smoking history of the participant; no group differences were detected after adjustment for these factors. A hematologic test battery revealed three red cell abnormalities in the Ranch Hand group, but these were difficult to place into a clinical or epidemiologic context. Evaluation of renal, pulmonary, and endocrine functions generally disclosed small and inconsistent proportions of abnormalities between groups, and were deemed clinically unimportant. An unrefined assessment of all summed and weighted organ system abnormalities by group did not show an aggregation of multisystem disease or malfunction. Any interpretation of these study data, in whole or in part, must carefully consider the methodical steps required for a proper inference of causality. It is specifically pointed out that many group differences were largely based upon subjective data, and that a subtle effect of differential reporting is suggested but has not been fully evaluated. For objective data, group differences were generally within normal ranges and were not correlated to the herbicide exposure index, nor fell within the expected latency periods following Vietnam service. The proposed clinical end points of dioxin exposure, chloracne, soft tissue sarcoma, and porphyria cutanea tarda, were not found in the Ranch Hand group (study power limitations recognized). Overall, substantial credence is given to the objective study findings, particularly after observing the consistent duplication of the classical effects of risk factors such as age, smoking, alcohol, etc., in almost all clinical areas. Additional work with these baseline data is still required in the areas of data base refinement, statistical testing and bias analysis, exposure index refinement, establishment of the follow-up examination requirements, and collaboration with other dioxin research studies. This baseline report concludes that there is insufficient evidence to support a cause and effect relationship between herbicide exposure and adverse health in the Ranch Hand group at this time. The study has disclosed numerous medical findings, mostly of a minor or undetermined nature, that require detailed follow-up. In full context, the baseline study results should be viewed as reassuring to the Ranch Handers and their families at this time. iii �PROJECT RANCH HAND II AN EPIDEMIOLOGIC INVESTIGATION OF HEALTH EFFECTS IN AIR FORCE PERSONNEL FOLLOWING EXPOSURE TO HERBICIDES MORTALITY UPDATE -1984 10 DECEMBER 84 William H. Wolfe, Colonel, USAF, MC Joel E. Michalek, PhD Richard A. Albanese, M.D. George D. Lathrop, Colonel, USAF, MC Patricia M. Moynahan, Colonel, USAF, NC Prepared for: The Surgeon General United States Air Force Washington, D.C. 20314 EPIDEMIOLOGY DIVISION DATA SCIENCES DIVISION USAF SCHOOL OF AEROSPACE MEDICINE (AFSC) BROOKS AIR FORCE BASE, TEXAS 78235 �Project Ranch Hand IT Mortality Update - 1984 EXECUTIVE SUMMARY HAOKCR<WNI> The purpose of the Ranch Hand II Study is to determine whether those Individuals involved in the aerial spraying of herbicides in Vietnam during the'Ranch Hand operation have experienced any adverse health effects as a result of their participation in that program. The study evaluates both mortality (death) and morbidity (disease) In these Individuals 6ver a 20-year period of time after the studies were initiated. The baseline mortality study was released in June 1983 and the baseline morbidity study in February 1984. Neither study demonstrated health effects which could be conclusively attributed to herbicide or dioxln exposure. The reader is referred to reports of the studies for further details (1, 12). METHOD The present study report describes the second mortality analyses. Deaths in the 1256 Ranch Hand and 6171 comparison subjects were determined, using the data sources of the Air Force, Veterans Administration, Social Security Administration, Internal Revenue Service, and personal contacts. As of 31 December 1983, 54 Ranch Handers and 265 comparison subjects had died. Death certificates were obtained on all subjects. Autopsies were conducted on 157 of the individuals who had died. Results have been obtained for 104 of these autopsies to confirm the death certificate findings. Autopsy reports for the 53 others have been requested, but have not yet been obtained. Extensive statistical analyses were accomplished, as detailed in the report, to compare the death experience in the Ranch Hand population with the comparison group. In addition, death experience in these groups was compared to the 1978 U.S. White Male Mortality experience, the 1978 Department of Defense Nondisability Retired Life Table, the mortality experience of the West Point Class of 1956, the USAF active duty personnel, and the active U.S. Civil Service population. RESULTS As was the case in the first mortality report, the current mortality analyses did not reveal any statistically significant differences in mortality between the exposed and comparison groups. The percentages dead in each major category are summarized below. Percent Deaths Ranch Hand Comparison Rank Officers Enlisted 3.2 4.9 4.0 4.5 Occupation Flying Ground 3.6 5.1 4.7 3.9 Note: None of the above differences between the Ranch Hand and Comparison groups are statistically significant. �Ranch Hand Total Overall 4.3 Comparison 4.3 As was reported In the baseline mortality study, the Ranch Hand officers had a nonstatlstically significant though slightly lover death rate than their comparisons, Ranch Hand flyers had a nonstatlstically significant though slightly lower death rate than comparisons, and Ranch Hand ground personnel had a slightly higher but nonstatlstlcally significant death rate than the comparisons. The herbicide/dioxln exposure index described In the morbidity report was applied to the data, and no relationship between exposure and mortality experience was identified. As was also noted in the baseline mortality study, analyaes consistently demonstrated significantly better survival in the Ranch Hand officers than Ranch Hand enlisted members, as was the case with comparison officers and comparison enlisted personnel. Cause-specific analyses did not demonstrate any Increased Ranch Hand mortality for accidents, suicide, homicide, malignancy or circulatory system disease. No unusual patterns of malignancy were observed in either the Ranch Hand or comparison groups, a finding which would be expected from the small number of deaths to date. When compared to the 1978 U.S. White male population, the Ranch Rand officers, comparison officers, and comparison enlisted are living significantly longer than expected. Although Ranch Hand enlisted are also living longer, the difference is not Significant. A similar pattern was seen in analyses using the DOD retired population. All groups bad a mortality experience similar to the civil service population. As would be expected from the fact that individuals In the active duty population who develop severe chronic disease are medically retired, all groups in this study had an increased mortality when compared to the Air Fore* population currently on active duty. Both Ranch Hand and comparison officers had mortality similar to the West Point group. I i CONCLUSION AND RECOMMENDATION Continued mortality surveillance is recommended, since the study groups are still relatively young and healthy. While sufficient time may have elapsed for some clinically significant conditions to occur, additional time Is necessary for other conditions, which; nay possibly be attributable to herbicide exposure, to develop. At this time, however* there Is no evidence of increased mortality as a result of herbicide exposure in those individuals who accomplished the Ranch Hand spray operations in Vietnam. �ABSTRACT OF REPRODUCTIVE OUTCOME DRAFT REPORT, 17 December 1984 Abstract: This report was written in response to requests from the Office of Science and Technology Policy, the Chair - Agent Orange Working Group and by direction of the Air Force Deputy Surgeon General. It was intended as an "interim" analysis of the data and not as a definitive answer to the birth defects issue. Since the release of the baseline morbidity report in February 1984, birth defects and neonatal deaths reported by study participants during the baseline questionnaire have been verified by record review. This verification was accomplished by the review of birth and other medical records, birth certificates and death certificates. Since not all data can be verified for at least an additional 24 months, these interim analyses were based on the v e r i f i c a t i o n of the p o s i t i v e reports only. Verification of negative responses to the birth defect and neonatal death questions have not as yet been completed. The draft report was submitted to the Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants. Their comments have been received and will be discussed with them in a forthcoming visit to the Ranch Hand Office. This interim assessment will then be forwarded to the OSTP through the Chair, AOWG. b-s � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 199 Folder The folder containing the original item. 5598 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Lathrop, George D. Patricia M. Moynahan Richard A. Albanese William H. Wolfe Description An account of the resource Corporate Author: Epidemiology Division, Data Sciences, Division, USAF School of Aerospace Medicine (AFSC), Brooks Air Force Base, Texas Date A point or period of time associated with an event in the lifecycle of the resource June 30 1983 Title A name given to the resource Project Ranch Hand II: An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides, Baseline Mortality Study Results ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/1846f7c6353bf01b345a16c4f214cb5b.pdf a105a9490d67eb685a07b9dfcf5f2ad6 PDF Text Text Item D Number °5827 Author Lav D Not Scanned y- T - L Corporate Author Report/Article TltlB Typescript: Human Exposure to Phenoxy Herbicides Journal/Book Title Year Month/Day Color Number of Images DeecrlptOU Notes n ° Final draft ' Accompanied by letters to Barclay Shepard and Alvin L. Young from Lavy. Monograph was published in 1987 by the Veterans Administration Central Office, Department of Medicine and Surgery, Agent Orange Projects Office. Tuesday, April 02, 2002 Page 5827 of 5840 �T.L. Lavy 2208 Sweetbriar Fayetteville, AR 72701 December 3, 1984 Dr. Barclay Shephard Office of Env. Medicine Veterans Administration 810 Vermont Ave. Washington, D.C. 20420 Dear Barclay: I am pleased to send you this final copy of my monograph "Human Exposure to Phenoxy Herbicides." I hope you are able to get this published as soon as possible since some of the information is already becoming outdated. At times it seemed like this was a never ending task. Recapping for you - in July 1983 I completed the first draft of the entire monograph. It was returned to me for revisions about January 1, 1984 after Dr. Shelden Wagner edited it. I made many of the changes suggested and then returned it to the V.A. around April 1, 1984. After that time I'm not exactly sure what happened at the V.A.; why it happened and why it took so long. After finally receiving Alan Fitzgibbens comments and suggestions I was shocked and disappointed at the revisions he was demanding. In all of that confusion somehow the incorrect, unrevised version of Chapter 5 kept getting into your hands. As I indicated to you following my trip to Florida in February, major revisions were made, but somehow they didn't get to you. A significant step forward occurred when I learned I no longer had to interact with Mr. Alan Fitzgibbens. I labored extensively over his reactions, comments, suggestions, and threats. Since my April revision I have put in 96 dedicated hours to completing this monograph. Conservatively, I can truthfully say that Fitzgibbens diversions have cost me at least 60 hours of work. Originally I tried hard to attempt to revise it in the manner he was requesting. After his dismissal I was able to concentrate on completing my task. If further revisions are required I will be able to work on it further at an additional expense to the Veterans Administration. �Dr. Barclay Shephard December 3, 1984 Page 2 Please call me if you have any questions or if I can be of assistance in getting this published. Truly yours, .L.avy Professor TLL/jlu cc: Dr. Al Young Dr. Layne Dresh �T.L. Lavy 2208 Sweetbriar Fayetteville, AR 72701 December 3, 1984 Dr. Al Young 1498B Edwards Place Boiling Air Force Base Washington, D.C. 20336 Dear |\1: Pflease give me your opinion on the following: 1) At one time you indicated that we needed to indicate why I was selected to write the monograph. I am not sure it is appropriate or necessary. 2) Is the Table of Contents too detailed? I do not object to drastically cutting the Table of Contents for Chapters 1-5, but for the Appendix I would like to preserve the entire list of questions as currently listed in the Table of Contents. Note that I have also listed them at the beginning of the Appendix. Perhaps this is too redundant. Should I omit them from the beginning of the Appendix? 3) Especially look at: My interpretations of the controversy Chapter 1, Chapter 5, Appendix A-54. « •••- ,•> , -' Thanks for your assistance. T.L. Lavy Professor TLL/jlu Enclosure P.S. I wish to thank you for your part in getting me geared up to write the monograph, your suggestions, and encouragement. �HUMAN EXPOSURE TO PHENOXY HERBICIDES T.L. Lavy Professor of Agronomy University of Arkansas �PREFACE The primary intent of this monograph is not to present new information in a manner which will stimulate the scientific community but rather to communicate with those in other disciplines who have an interest in or concern for those who have been or believed to have been exposed to phenoxy compounds or the associated dioxin contaminants. The goal of this monograph is to provide information to the reader which will allow him to better evaluate the facts and fiction which surround the use of phenoxy herbicides employed in crop production. The reader will have a better understanding of the toxicity associated with these compounds after considering some of the physical and chemical properties of the phenoxy herbicides, taking an in-depth look at their fate in the environment, and evaluating animal feeding trials in which the phenoxys make up a portion of the food supply. Since the phenoxy herbicides have been used by agriculture for nearly 40 years, volumes of research projects have been completed. An evaluation of animal feeding studies and discussion of industrial exposure accidents accompanied by over 15 years of historical followup of those humans who have inadvertently been exposed to high levels of dioxin provide an extraordinary opportunity to view the effects of phenoxy compounds and their contaminants on man and other animals. Emphasis is placed on evaluating the kinds and amounts of exposure received by pesticide applicators. By thoroughly measuring the levels of exposure being received by this group of workers, who routinely apply phenoxy herbicide for their means of livlihood, we are examining �those most likely to receive health threatening levels. Others not associated with phenoxy spraying would be expected to receive much less exposure. In this monograph special effort is devoted to differentiating between exposure and dose. Clarity regarding these concepts will provide us a better opportunity to assess the potential health effects of phenoxy herbicides on the human body. Personnel involved with the manufacture of the compounds, herbicide applicators, and others in and around the agriculture community, the Vietnamese people and veterans serving the military during the Vietnam conflict deserve to have the facts on exposure, dose and risk presented to them in -a manner they can understand. �TABLE OF CONTENTS Page Preface Table of Contents Chapter 1: Phenoxy Herbicides, Their Use, Their Controversies . 1-1 Phenoxy herbicides as weed control chemicals 1-1 Chemical properties of phenoxies 1-4 Commercial uses of phenoxy herbicides in the United States . 1-8 Pastures and rangelands 1-12 Rice production 1-13 Phenoxy herbicide use in rights-of-way management in the United States 1-13 Other uses 1-14 Selectivity 1-14 How do phenoxy herbicides kill plants? 1-15 Advantages of using phenoxy herbicides 1-15 Controversies associated with phenoxy herbicides 1-16 TCDD and phenoxy herbicides news items 1-17 Chick edema disease 1-18 Findings of Bionetics Research Laboratories 1-18 Claims of Vietnamese health problems 1-19 The incident at Glove, Arizona 1-21 Reindeer in Swedish Lapland 1-22 Alsea II report 1-22 The on-going Missouri horse arena (Times Beach) Episode 1-23 Seveso, Italy, and other explosions 1-26 References 1-29 �Page Chapter 2: Factors Influencing Human Exposure to Phenoxy Compounds 2-1 How long do phenoxy compouds persist in the environment? . . 2-2 Adsorption 2-3 Leaching 2-4 Runoff 2-4 Plant uptake 2-4 Chemical decomposition 2-5 Microbial decomposition 2-5 Volatilization . . 2-5 Photodecomposition 2-6 Other important factors 2-6 Persistence of phenoxy hericides in soil 2-7 Effects of high rates of aplication or persistence 2-9 Factors affecting phenoxy apors in air 2-11 Vapor pressure 2-11 Particle size 2-13 Protective equipment 2-19 Possibilities of human exposure via phenoxy compounds in air 2-20 Man's potential for exposure to phenoxy herbicides in drinking water and other water sources 2-22 Man's exposure to phenoxy herbicides through the food he eats 2-26 Assessment of human exposure to the phenoxy herbicides . . . 2-32 References Chapter 3: 2-34 Studies of Human Exposure to Phenoxy Herbicides. . . 3-1 Phenoxys and organophosphates 3-2 �Page Prohibition of 2,4,5-T sales 3-4 Forestry applicator exposure studies 3-5 Methodology of the 2,4,5-T study 3-6 Application processes 3-9 Climate conditions 3-9 Spray material 3-9 Personnel 3-10 Criteria for obtaining air samples 3-10 Criteria for obtaining dermal exposure samples 3-10 Urine sampling criteria 3-11 Dioxin analyses 3-12 Results of the 2,4,5-T study (Lavy, 1978) 3-12 Methodology of 2,4-D study 3-18 Results of the 2,4-D forestry aerial application study (Lavy, 1980) 3-23 Air samples 3-23 Patch samples 3-23 Urine samples 3-24 Other phenoxy exposure studies 3-31 Observations, findings, and significance of applicator exposure studies 3-33 References 3-39 Chapter 4: Pharmacokinetic Studies of the Phenoxy Herbicides in Humans and Animals 4-1 Fate of phenoxy herbicides in humans 4-12 Probability of dermal absorption 4-13 Factors affecting man's absorption and excretion rates of phenoxy compounds 4-16 �Page Effects of repeated exposures 4-19 Relative toxicities of the phenoxy herbicides and TCDD . . . 4-26 Ranch Hand personnel 4-28 Domestic aerial applicators . 4-29 References 4-32 Chapter 5: Assessment of Harm Due to Phenoxy Herbicide Exposure 5-1 What research information is needed? 5-1 What does the majority of scientific information tell us about man's potential for harm from being exposed to phenoxy herbicides? 5-4 USDA R.C. Nash Study 5-4 Oregon researchers (Newton et a!., 1975) 5-4 J.M. Witt American Chemical Society Symposium, Seattle 1983. 5-5 United Kingdoms: 2,4,5-T Review of the safety of the herbicide 5-6 Information in medical literature 5-7 Relative risks of spraying 2,4,5-T 5-8 Concluding remarks 5-9 References 5-12 Appendix: Significance of TCDD in Man's Exposure to Phenoxy Herbicides A-l What is dioxin? A-4 Is there more than one dioxin? A-4 Why does TCDD exist? A-5 How do we learn about TCDD? A-6 What is being done to learn the significance of man's exposure to TCDD? A-6 Is TCDO harmful to humans? A-7 �Page Were phenoxy herbicides involved at Times Beach? Are people being harmed by the TCDD present at Times Beach? A-ll A-ll Since the discovery that soils removed from the horse arena and deposited in other Missouri communities still contain detectable TCDD levels, what is being done? A-12 What can be done to clean-up or remedy the current TCDD containing areas? A-12 How do humans come into contact with TCDD and other dioxins? A-13 Do phenoxy herbicides contain TCDD? A-17 Is the TCDD present in 2,4,5-T harmful to animals eating vegetation sprayed with the herbicide? A-17 What is Agent Orange? A-18 Why were defoliants used in Vietnam? A-20 Were defoliants other than Agent Orange used in Vietnam? . . A-20 Were other pesticides used in Vietnam? A-21 Is Agent Orange as toxic as TCDD? A-22 How long does TCDD persist in the environment? A-22 Do green plants take up TCDD from the soil? A-24 Does TCDD magnify in the food chain? A-24 Is TCDD in the environment harmful to animals? A-26 What are the short- and long-term effects of TCDD? A-27 If TCDD entered a human would it accumulate? A-30 Why is the Seveso, Italy accident important to the TCDD exposure questions? A-31 Since it has been shown that TCDD is widely distributed in the environment, does that mean that it is present in the food we eat? A-33 Can man be harmed from ingesting 2,4-D or 2,4,5-T? A-34 How toxic is 2,4-D to workers applying this herbicide in the forest? A-35 �Should a concerned pesticide applicator individual quit worrying about exposure to pesticides since the scientific facts indicate no imminent danger is present? . A-36 If ones chances of being harmed by phenoxy herbicides and TCDD are so low, why do we keep hearing about the danger on TV and reading about it in the newspapers? A-37 Other than in production plants what groups of humans are exposed to the highest amounts of pesticides? A-40 Do phenoxy herbicides cause soft tissue sarcomas in man? . . A-40 Were there dangerous components of Agent Orange besides TCDD? A-41 What happens if 2,4,5-T, which contains a trace level of TCDD, comes into contact with a human? A-41 Were U.S. ground troops in Vietnam "at risk"from Agent Orange since they could not shower daily? A-43 Could Agent Orange harm soldiers exposed to it? A-43 How could one know if he had been exposed to significant levels of Agent Orange? A-46 Chloracne has been described as the primary symptom for men exposed to chlorinated organic compounds, are there other symptoms that would accompany these symptoms?. . . . A-47 What has happened to the health of those in Vietnam who mixed and applied Agent Orange? A-47 Could Agent Orange affect the fertility of males? A-47 Can birth defects be caused by pregnant women being exposed to the phenoxy herbicides and TCDD? A-48 Does human mothers milk contain TCDD? A-53 How do the teratogenic effects of TCDD and 2,4,5-T compare?. A-53 Were U.S. military personnel who used Agent Orange well informed about its properties at the time of its use in Vietnam? A-54 What evidence do we have which indicates that use of defoliants in South Vietnam has not caused man health problems? A-55 �Why do the Veteran's Administration and other government agencies continue to study Vietnam veteran health problems if the government is not admitting that Agent Orange exposure has caused the problems? A-55 References A-57 �LIST OF FIGURES Page Figure 3-1. Daily excretion of 2,4-D in urine of batchmanloader. Exposure apparently occurred on days in addition to the spray days (Days 3 and 9). . . 3-30(F) Figure 4-1. Excretion rate of p-aminophenol in urine in a subject exposed to 1% aniline solution for 1 h (Baranowska-Dutkiewicz, 1982) 4-16(F) Average daily excretion of 2,4,5-T in urine of forest workers applying 2,4,5-T (backpack, mist blower, and.helicopter methods)(Lavy, 1978) 4-24a(F) Simulated body burden of 2,4,5-T following 13 repeated daily exposures in units of the daily dose D 0 (Lavy, 1978) 4-24b(F) Average daily excretion of 2,4-D in urine of forestry workers applying 2,4-D by helicopter (Lavy, 1980) 4-24c(F) Figure 4-2. Figure 4-3. Figure 4-4. 4-26c(F) Figure 4-5. Simulated body burden of 2,4-D in humans after a single dermal exposure and after repeated (daily) dermal exposures. The body burden is expressed in units of the dose level applied to the skin (D0) (Lavy, 1980) . . . 4-25(F) �CHAPTER 1 Phenoxy Herbicides, Their Use, Their Controversies Phenoxy herbicides as weed control chemicals Weeds have presented serious problems to human food production for centuries. They infest home lawns and gardens and cause, among others, the human allergies of poison ivy and hay fever. Commmonly, weeds invade man's recreation areas and interfere with his fishing, boating, golf courses, or parks. Rights-of-way for telephone and electrical power lines, railways, highway systems in addition to navigational waterways and irrigation systems could be drastically affected if weed growth was allowed total freedom. In recent time, we have learned to successfully control many serious weed problems with chemical control measures. Among these chemicals are phenoxy herbicides which are very effective and economical control agents. The first of the phenoxy herbicides to be discovered was 2,4-dichlorophenoxyacetic acid (2,4-D) whose synthesis was described by Pokorny in 1941. Ahlgren et al. (1951) described a series of alkaline reactions which lead to the production of 2,4-D. Use of 2,4-D as a plant growth regulator was shown in 1942. Removal of dandelion, plantain and other broadleaved plants from a bluegrass lawn following an application of 2,4-D was first reported in 1944 (Marth and Mitchell, 1944). Discovery of this concept of selec- tivity and its potential for enhancing farm productivity through weed control was a key motivating factor which prompted scientists in academia and several chemical manufacturers to initiate new research 1-1 �efforts. This series of events has played a vital role in helping the U.S. build a sound economic foundation. The American farmer's output has steadily risen as shown by the fact that one U.S. farm worker fed 16 people in 1950, 26 people in 1960, 47 people in 1970, and in 1978 fed 65 people (Farmline, 1980). The use of organic chemicals for weed control and other pests has played a major role in this marked increase in agricultural production. As shown in Table 1.1, the tonnage of phenoxy herbicides used has continued to rise (Bovey and Young, 1980). The ability of the formulation chemist to adapt the phenoxy molecule to different field applications has added to the popularity of the herbicides. For a pesticide to have an optimum biological effect on its target it must be permitted an opportunity to make contact with the organism. Some of the formulations employed to put pure chemical pesticides into a form facilitating field use include: wettable powders, emulsifiable concentrates, dusts, granules, water-soluble concentrates, flowables and aerosols (Van Valkenburg, 1967). The formulation chemist is interested in achieving several goals. Assuming he has found a compatible mix that will readily pass through the applicating equipment, he is interested in creating a spray solution with viscosity properties which will allow it to be retained on the foliage of the target plant, to penetrate the tissues and to translocate within the plant. Pesticide formulations and surfactants have primarily been designed to provide a high level of biological activity on plants or other pests. Most of the research which has been completed on the phenoxy 1-2 �Table 1.1. Production of 2,4-D and 2,4,5-T (Acid B a s i s ) , United States 1951-1971. Modified from Bovey and Young (1980). Produced 2,4-D Year 2,4,5-T x 1Q6 kg x 106 kg 1951 9.4 1.1 1958 14.0 1.7 1959 13.3 2.5 1960 16.5 2.9 1961 19.7 3.1 1962 19.5 3.8 1963 21.0 4.1 1964 24.3 5.2 1965 28.7 5.3 1966 30.9 7.0 1967 35.0 6.6 1968 36.0 7.9 1969 21.4 2.3 1970 19.8 N.A. 1971* N.A. N.A. a Production data for 2,4-D and 2,4,5-T not a v a i l a b l e after 1970 to avoid disclosure of i n d i v i d u a l company data. Source. The Pesticide Review; 1969, 1970, 1971, 1972, and 1975, U.S. Department of A g r i c u l t u r e , Agricultural S t a b i l i z a t i o n and Conservation Service. 1-3 �herbicides has centered around determining how much to use, what plants are most easily controlled, which are not affected, and how long they remain active in the soil-plant-water environment. Extensive information on the effects of phenoxy herbicides on mammals in animal feeding studies has been gathered. Extrapolation of these data from sensitive experimental animals provides much information which can add to our understanding of their effects on humans. Prior to 1978 only a few studies have been devoted to measuring the exposure received by people who apply phenoxy compounds in the field. As a result, little information is readily'available to the field applicator about which formulations are most likely to penetrate skin or those most likely to be present in the air. Chemical properties of phenoxys Although all compounds are volatile to some extent, in the case of pesticides this property may represent a significant mechanism for permitting the active component to penetrate at a biologically active concentration to the most susceptible layer of the target organism. 1-4 �Table 1.2. Vapor pressures of esters of some phenoxyacetic acid herbi cides. C 3.5 x lO'4 1.1 x 10-3 25 9.6 x 10-6 ' 35 2,4-D n-butyl ester 25 35 2,4-D methyl ester mm Hg 2.6 x lO-5 2,4-D Methyl Amine 38 8.9 x 10-10 2,4,5-T n-propyl ester 25 7.5 x 10-5 35 1.2 x 10-4 Vapor pressure of pesticides determines the maximum vapor concentration at the target site and also governs the rate the compound may diffuse (Hamaker and Kerlinger, 1967). Volatile compounds with relatively high vapor pressures, may be more efficacious in their plant killing properties; thus, coverage within a canopy could be supplemented as the compound vaporizes. As shown in Table 1.2, the ester formulations of some of the phenoxy herbicides are markedly more volatile than amine salt formulations. Movement of phytotoxic vapors away from the site of application can damage crops and irritate neighbors, especially in areas where several different kinds of crops are grown in close proximity. Formulation of long chain ester herbicides has decreased the volatility of the compound but still facilitates the entry of the herbicide 1-5 �into the plant tissues. Formulations of the phenoxys include emulsifiable acid, amine salts, mineral salts and esters. The concentration of these materials are expressed on the basis of acid equivalents, i.e. that portion that could be converted to the acid. Since the acid form is only slightly soluble in water and the emulsifiable concentrate is relatively expensive, only limited amounts of this formulation are used (Klingman and Ashton, 1982). The most commonly used of the phenoxys are the amine salts. Characteristically these materials are white crystalline solids that are readily soluble in water and present essentially no volatility problem. As a contrast, phenoxy esters are essentially insoluble in water but dissolve in petroleum oils. Esters made by using alcohols with one to four carbons are no longer used due to the high volatility of the phenoxy moiety. Today more expensive, long chain alcohols with an ether linkages (-0-) are used in the formulations to lower the volatility hazard (Klingman and Ashton, 1982). The ester forms are often preferred over the amines since the esters are much more effective on some species - especially woody plants. It should be emphasized that the high molecular weight alcohols are now used to make phenoxy esters with reduced volatility; although these esters are low volatile they may, under hot, humid conditions, still volatilize enough to injury highly susceptible plants (Klingman and Ashton, 1982). Since esters are more phytotoxic, more caution is required to ascertain that excessive levels are not used, killing susceptible plants where specificity is desired. Three possible explanations for the enhanced toxicity to plants from ester for- 1-6 �mulations include: (1) volatility permits absorption of the gases through leaf stomates, (2) wetting action of the oil-like ester and the oil carrier may aid penetration and (3) ester forms, with their low polarity, are compatible with the cuticle and aid penetration through it (Klingman and Ashton, 1982). Thus, it seems likely that man's skin surfaces would also be more susceptible to penetration of the phenoxy compounds when formulated as the ester form. When the ester and amine formulations are compared with respect to potential for exposure to man, the more volatile ester formulations are more likely to be present in the air of man's breathing zone. If skin contact occurs, the more soluble compounds will be more likely to be absorbed. Polar perspiration would aid in solubilizing the amine salt phenoxy formulations. However, oily skin surfaces of humans would tend to solubilize ester formulations of the phenoxy compounds. Surfactants are classified as cationic, anionic or ionic. In general, they consist of a hydrophobic and hydrophilic section. This unique combination of polar and non-polar properties provide surfactants with the ability to emulsify, disperse, spread, wet, or solubilize, thus enhancing herbicidal action through increased foliar penetration (Foy and Smith, 1967). Freed and Witt (1967), in Table 3, present data showing that the energy required for plants to absorb herbicide is reduced in the presence of surfactants. 1-7 �Table 1.3. Effect of surfactant on absorption energy. (Freed and Witt, 1967). E a (kcal/mole) Without surfactant Herbicide With surfactant Picloram 18.9 - Dicamba 12.0 10.4 6.5 3.0 2,4-D Considerable time and study have been devoted to evaluating enhanced herbicidal action due to surfactants, however, little is known with regard to any inadvertent exposure man may receive during application processes or other times he comes into contact with surfactant-phenoxy herbicide mixtures. Since Smith and Bayer (1966) have shown that diuron leaching in a soil column can be increased or decreased by the kind and rate of surfactant added, several interactions are possible. Thus, theoretically someday it may be conceivable that phenoxy herbicides or other pesticides could be optimized for penetrating plant tissues and minimized for passing man's skin. Commercial uses of phenoxy herbicides in the United States Annual and perennial broadleaf weeds in non-cropped areas, as well as in tolerant crops, are routinely controlled using 2,4-D formulations. This compound is used extensively in the cornbelt and for small grain production. Much of the cropland acreage is treated with 2,4-D as a postemergence application after the weeds are rapidly 1-8 �growing. Postemergence application rates are routinely 0.5 Ib/acre or less; at rates above 1 Ib/acre this compound will serve as a preemergence herbicide, thus killing weeds as they germinate. A drawback to this means of application is the relatively short residual life of 2,4-D in the soil. The forest industry has relied heavily on the use of 2,4,5-T for vegetation control. Bans on the use of this material for forestry applications has had a negative impact on timber production. Of the nearly 500 million acres of commercial forest land in the United States less than half of this is available for timber production. In 1972 the U.S. imported 11% of the U.S. consumption of wood (USDA, 1982). As a consequence of the ban of 2,4,5-T use in forestry and accompanying problems associated with aerial spraying of U.S. forest lands, vast acreages of potentially excellent production sites became overgrown by undesirable, much less productive woody vegetation. On these sites eighteen inch tall, 4-year-old pine seedlings that could have attained heights of 5 to 7 feet or more are pressed to compete for sunlight, water, and fertility. Newer, more expensive compounds and other application means are being used to remedy the problems in some of these areas. In many cases mechanical means such as chain saws or machetes have been used in an attempt to eliminate undesirable species. In addition to this being an extremely costly and dangerous task for man, the competitive advantages seldom lasts for more than one growing season. Table 1.4 (USDA, 1982) shows aerial application if possible, would 1-9 �be the primary tool for utilizing 2,4,5-T in the South, Rocky Mountains, and Pacific Coast while in the North individual stem treatment (hand applied) would be the preferred application method. 1-10 �Table 1 4 Estimated annual potential use of 2,4,5-T on commercial .. forest land by section, application purpose, and application method. (USOA, 1982). Purpose and application method North Present Potential • ( A f* Y*rt<*^ _ \ r\L» i Co J . South AppliAppl i cation cation Present Potential rate rate (Ib ae/ (Ac (Ib ae/ acre) acre) Site preparation and rehabilitation Aerial application 1,500 Broadcast ground 1,200 Individual stem 30,000 Total 32,700 40,600 5,400 89,400 135,400 2-3 3-4 1-3 Release and TSI Aerial application 0 Broadcast ground 1,650 Individual stem 62,200 Total 63,850 32,200 9,300 256,500 298,000 1-2 2 1/4-2 Fuel breaks Aerial application Broadcast ground Individual stem 0 200 0 0 700 10 All purposes Aeri al appl i cation 1,500 Broadcast ground 3,050 Individual stem 92,200 Total 96,750 72,800 15,400 345,910 434,110 2 2 2-4 2-4 4-6 414,000 1,083,100 42,600 17,000 8,250 31,900 439,250 1,083,100 2 2-4 2-6 0 0 0 0 0 0 614,000 1,594,300 379,100 148,050 97,000 27,550 789,600 2,070,400 4,400 800 0 5,200 2-3 2 2 1-11 — Pacific Coast Rocky Mountai ns Site preparation and rehabilitation Aerial application 20 Broadcast ground 0 Individual stem 0 Total 20 585,700 336,500 65,100 987,300 200,000 131,050 19,300 350,350 29,142 1,369 489 31,000 57,184 3,670 3,746 64,600 2-4 3-4 2-3 �North Purpose and application method Present South Potential Release and TSI Aerial application Broadcast ground Individual stem Total 220 0 0 220 13,400 2,400 0 15,800 Fuel breaks Aerial application Broadcast ground Individual stem 0 0 0 All purposes Aerial application Broadcast ground Individual stem Total 240 0 0 130 Appl i cation Present rate Application Potential rate 231,872 9,519 4,609 246,000 500,330 21,126 15,844 537,300 2 2 1-3 0 0 0 0 800 70 6,000 3,700 400 2-4 2-4 4 17,800 3,200 0 1,000 261,014 11,688 5,168 277,870 563,514 28,496 19,990 612,000 1-2 2 1-2 lae = acid equivalent. Pastures and range! ands. Grazi ng lands are a vital natural resource of the U.S. They are used not only for forage production but also for watersheds, soil and water conservation, lumber, medicinal compounds, mining and recreational purposes. This land is often used only for grazing since it is too steep, too wet, too shallow, too cold, too dry, or too acid to be important for growing other crops. The relatively low cost of the phenoxy herbicides and the wide spectrum of broadleaved weed control have contributed significantly to the popularity of this group of chemicals for forage and pasture weed control. Pastures are subject to being invaded by many different spe- cies of woody plants. If an extensive woody plant problem was present on a minimum acreage ground application, equipment could be used to 1-12 �apply a broadcast treatment. More commonly, land owners spot treat woody plants in cultivated pastures as they appear, thus minimizing the amount of 2,4,5-T required (USDA, 1982). Rice production. Rice production has relied heavily on the use of 2,4,5-T for weed control. The 2.5 million acres of rice the U.S. located primarily in Arkansas, Louisiana, Texas, Mississippi, and California. Broadleaf aquatic weeds have been effectively controlled by this compound; 28 percent of the rice acres in this four-state area has relied on the use of 2,4,5-T. A rate of 1 Ib/acre is commonly used for weed control, resulting in an annual application of 300,000 pounds 2,4,5-T. Silvex, another chlorinated phenoxy herbicide, is also effective in controlling this broad!eaf aquatic weed complex (USDA, 1982). In 1983 rice was the only U.S. produced commodity for human consumption that could legally be sprayed with 2,4,5-T during its production. However, all routine domestic uses of this compound was ceased in September, 1984. Phenoxy herbicide use in rights-of-way management in the United States. Good vegetation management programs are critical to the nation as they ensure the safety, security, and reliability of our rights-of-way systems. Major rights-of-way types include railroads, highways, pipelines, and electrical transmission lines. They supply our needs for food, fuel, communication, energy, and many other items. Estimated rights-of-way acreage in each of the major categories are listed: Railroads - 2.4 million acres; highways - 21.7 million 1-13 �acres; pipelines - 2.2 million acres; and electrical utilities - 5 million acres; this makes up 31.3 million acres or nearly 1 percent of the total U.S. acreage (USDA, 1979). Annual treatment with 2,4,5-T occurred on more than 680,000 acres annually. Ground application equipment was usually used when 2,4,5-T was applied to highways and railway rights-of-way areas while electrical and pipeline utilities relied more heavily on aerial means of application. Use of 2,4,5-T resulted in decreased soil erosion since little soil erosion or compaction occurs due to a lack of site disturbance. If bulldozers or other large land clearing devices were required to clear the vegetation soil erosion would be increased markedly. Use of 2,4,5-T and other phenoxy compounds for vegetation management has generally enhanced wildlife activity in treated areas (USDA, 1982). Other uses. Other extensive areas of phenoxy herbicide use include drainage ditches, canals, channels, recreational areas, and other waterways where brush control is important. Selectivity The selectivity of phenoxy herbicides to control broadleaf plants in the presence of grasses or grain crops is one of the important reasons they phenoxy herbicides have gained in popularity. Ashton and Harvey (1971) suggested that a herbicide is selective to a particular crop only within certain limits. These limits are the result of an interaction between plants, herbicides, and the environment. Significant factors affecting plant responses to a chemical include: rate of plant growth, shape and characteristics of leaf sur- 1-14 �faces, amount of chemical taken up by the plant, how it moves in the plant, and deactivation in the plant. The molecular configuration of the herbicide and its concentration are important factors governing the ability of the compound to kill the desired plant. How do phenoxy herbicides kill plants? When these compounds kill plants they penetrate into leaves, stems, or roots; are absorbed into the plant's living protoplasm; migrate across surrounding tissues to the vascular channels; and trasnlocate from the point of entry to the actively growing portion where plant death beings. Ashton and Crafts noted that Final death of plant tissues may result from contact action; most characteristic of the 2,4-D light ester or amine formulations on foliage. It may result from extreme hormonelike response giving rise to tumorous tissues, excessive production of buds or root initials, softening of root cortex and general degeneration where concentrations are not sufficient to cause direct death, crushing and plugging of vascular tissues may result in a slower death from lack of nutrients normally supplied via these tissues (Ashton and Crafts, 1973). Although phenoxy compounds are also effective as weed killers after they have been applied to the soil, the mode of action asso- ciated with their removal from the soil-solution complex, translocation through root systems, and subsequent plant killing steps are less well understood. Advantages of using phenoxy herbicides The family of chlorophenoxy herbicides has gained worldwide acceptance for several reasons. Among these reasons are 1) low rates are effective in killing a wide spectrum of broadleaf plants, 2) they 1-15 �are readily adaptable for many different formulations, 3) they degrade rapidly in the environment and 4) the vast quantities used and the advances in technology have allowed the phenoxy compounds to compete favorably from an economical standpoint. Controversies associated with the phenoxy herbicides In view of the positive aspects of the phenoxy compounds, why is there any controversy? Confusion abounds and misinformation is common with regard to this subject. One of the goals of t h i s book is to pre- sent facts and allow the reader to place them in their proper perspective. D u r i n g production of 2,4,5-trichlorophenol which is used to make some of the chlorophenoxy herbicides an undesirable by product is formed. dioxins. As a group these compounds are commonly referred to as Often the term d i o x i n , TCOD and 2,3,7,8-tetrachlorodibenzo- p-dioxin are used interchangeably as synonyms. Since there are eight possible locations for chlorine atoms on the d i o x i n molecule there is the p o s s i b i l i t y of 75 different d i o x i n isomers being formed and 22 of these could contain 4 chlorine atoms and be labeled as tetrachlorinated d i o x i n s ( H a l p e r i n et al., 1982). The tetrachlorinated d i o x i n molecule with c h l o r i n e atoms in the 2,3,7, and 8 positions is the most toxic of the known d i o x i n members. Any of the d i o x i n s c o n t a i n i n g four c h l o r i n e atoms can correctly be called TCDD. The 1,2,3,4 TCDD is only 1/50,000th as toxic as its 2,3,7,8 relative which appears from animal studies to be the most toxic of the d i o x i n isomers studied to date (International Agency for Research on Cancer, 1978). Costly special mass spectrophotometers are 1-16 �required to quantify and differentiate between the different TCDDs. For further discussion in this monograph the term TCDD will be reservesed specifically for the 2,3,7,8 isomer. By proper regulation of pressure, temperature and solvent conditions the level of TCDD contamination in 2,4,5-T can be minimized. In 1964 the Dow Chemical Company became aware of TCDD contamination in 2,4,5-T. At that time they closed their production facilities and made major modifications in the reaction conditions for synthesis of trichlorophenol and by late 1965 Dow Chemical had developed technology that permitted production of 2,4,5-T containing 1 ppm or less TCDD (Young et al, 1978). Dioxins also result from sources other than phenoxy herbicides. Recent research conducted at the University of Amsterdam and in the U.S. has shown that chlorinated dibenzodioxins, which includes TCDD, can be formed on combustion in municipal incinerators and by heating chlorophenols (Rappe et al. 1979). Pentachlorophenol, a commonly used preservative for wood products, is contaminated during manufacturing processes with hexa, hepta, octa, and possible tetra-chlorinated dioxins (Plimmer, 1973, Buser and Bosshardt, 1976). TCDD and phenoxy herbicide news items Following is a list of items concerning the phenoxy and associated dioxins which have caught public attention during the past 30 years: Chick Edema Disease, 1955 Bionetics Institute's Report on 2,4,5-T teratogenicity Use of 2,4,5-T and TCDD in Vietnam 1-17 �Claims of Vietnamese Birth Defects Swedish Lapland Globe, Arizona EPA Alsea II Missouri Horse Arena Seveso, Italy Explosion and Other Accidents On Going Concerns for Veterans Chick edema disease. Sanger et al. (1958) reported a new (undiagnosed) disease responsible for killing millions of chickens in 1957. Only a few similar problems were described until 1969. At that time soapstock, a by product from refining crude vegetable oil, was the source of the toxic substance. Firestone et al (1971) found that chlorophenols used in antimicrobial water treatment had contaminated the feed fat. Research stimulated by these findings now allows commercial vegetable oils to be filtered through activated carbon filters to remove any dioxins which may be formed as a by product from refining crude vegetable oil (Mounts, 1976). Findings of bionetics research laboratories. Studies conducted for the National Cancer Institute by the Bionetics Research Laboratories in 1969 revealed that rats or mice given relatively large oral doses of 2,4,5-T during the early stages of pregnancy showed a higher than average number of deformed offspring (Allen et al., 1977). This research triggered Dr. Lee A. DuBridge, Science Advisor to the President and Executive Secretary of the President's Environmental Quality Council, to state that the "Department of Agriculture will cancel registration of 2,4,5-T for use on food crops effective January 1-18 �1, 1970, unless by that time the Food and Drug Administration has found a basis for establishing a safe, legal tolerance in and on foods" (Bovey and Young, 1980). Immediately after this announcement of 2,4,5-T registration cancellation, the Department of Defense suspended the use of Herbicide Orange in Vietnam (Anonymous, 1977; Bovey and Young, 1980). Following this announcement and restrictions on 2,4,5-T a close evaluation was made of the Bionetic's findings, since 2,4,5-T had had an excellent health record. It had been used in agriculture for over 20 years with no claims of human or other animal problems. Close scrutiny of the Bionetics Research Laboratory findings showed that the 2,4,5-T used was an old sample from a company no longer producing 2,4,5-T. The compound used contained 27 +8 ppm TCDD. This level was well in excess of the TCDD levels occurring in 2,4,5-T being produced by other companies at that time. In an attempt to reproduce results to those reported, the Dow Chemical Company conducted a study inconsultation with scientists at the National Institute of Environmental Health Science Laboratires at Research Triangle, North Carolina. The only planned deviation in the study was to substitute regular production grade 2,4,5-T. Results of these studies showed that 2,4,5-T of regular production grade did not cause birth defects in rats at 24 mg/kg/day administered orally on days 6 through 15 of gestation (Bovey and Young, 1980). Claims of Vietnamese health problems. The reports from the Bionetics Research Laboratories, Dr. Lee DuBridges 1 consequent actions, the fact that 2,4,5-T used in the U.S. war effort contained 1-19 �low levels of the TCDD contaminant caused concern in Vietnam for the possibility of increased birth defects due to exposure to 2,4,5-T. Independent surveys were conducted by Cutting et al. (1970) and Meselson et al. (1971). In 1971 an Advisory Committee on 2,4,5-T evaluated the two studies mentioned above for U.S. Environmental Protection Agency (Advisory Committee, 1971). A summary of the evaluation follows: Summarizing the Vietnam data on human embryotoxicity, it said that (1) the sample of births surveyed was from year to year a variable but usually very small fraction of the total number, (2) it was quite unrepresentative of the geographic and ethnic distributions, (3) the heavily sprayed and otherwise exposed areas were greatly underrepresented, and (4) the birth records were not trustworthy and, therefore, the rates of stillbirth, and especially of congenital malformation, derived from them were equally unreliable. For example, the overall congenital malformation rate found in South Vietnam, 4.91 per 1000 live births, is about half of what was reported in other studies in various parts of Asia, and possible a quarter of what might actually exist at term. A further indication that the newborn children were not carefully examined is the absence of Down's syndrome in the list of specific malformations compiled by the Army survey (Cutting et al., 1970) despite the fact that some Oriental populations have been reported to have an incidence of this condition not unlike that in Western populations. There is, and can be, no precise knowledge of the exposure to 2,4,5-T (and hence, TCDO) experienced by pregnant South Vietnamese 1-20 �women, including what amounts they ingested or absorbed and when this may have occurred during pregnancy. Consequently, attempts to accura- tely relate birth defects or stillbirths to herbicide exposure can not be made. The detailed searches reported by Cutting et al. (1970) almost certainly would have shown any marked increase in the incidence of birth defects if they had occurred. The incident at Globe, Arizona. Soon after the U.S. Forest Service applied 1670 kg Si 1 vex and 54 kg 2,4,5-T in the Kellner-Canyon Gulch Spray Project near Glove, Arizona reports of harmful effects on human and other animals were made (Tschirley et al., 1970). These claims were investigated by the U.S. Forest Service, but many of the local residents did not agree with the report of the investigation and in February, 1970 the incident attracted national attention. Television coverage showed deformities in animals which allegedly had. been caused by the herbicides. A public hearing was held on February 13, 1970. An account by Time Magazine (Anonymous, 1970) indicated that the local veterinarian insisted that he had noticed nothing out of the ordinary in local animals. Local medical doctors could find no relationship between the spraying and the illnesses. Time also reported that the investigators holding the public hearing ended up perplexed. The investigators were accused by some local townspeople of being imposters, really representatives of chemical companies. The Office of Science and Education, USDA further investigated the charges. Results of this investigation were published by Tschirley et al. (1970). Roan and Morgan (1972) in a subsequent report of the investigation concluded: 1-21 �We cannot find any evidence that there was long term exposure of resident of the Globe, Arizona area to chlorophenoxy herbicides, or significant contamination of water supplies in this area with 2,4-D, 2,4,5-T, silvex, or metabolites of these herbicides. Nor have we found contaminants such as TCOD that may be associated with one or more of the above technical grade products. Statistics on reproductive mortality and morbidity for the period 1960 through the first six months of 1970, from one hospital serving this area do not indicate any trends that are suggestive of adverse influences on human reproductive function that might be associated with herbicide use during the years 1965, 1966, 1968, and 1969. Reindeer in Swedish Lapland. Sudden deaths of reindeer grazing on 2,4,5-T treated vegetation was reported in the Swedish Lapland during the spring of 1970. Within a week of a heavy snowfall approximately 30 reindeer died without any previous signs of illness; also it was reported that about 10 reindeer cows aborted their fetuses. Veterinarian examination of the dead animals revealed empty stomachs. When food was given to the surviving reindeer the deaths stopped. To more fully evaluate the possibility that 2,4,5-T may have been the culprit, Erne (1973) conducted a controlled experiment by feeding 2,4,5-T to female reindeer. From results of his study he concluded that any toxic problems associated with the reindeer deaths were probably not caused by ingestion of 2,4,5-T. Alsea II Report. The EPA issued an order for the emergency suspension of 2,4,5-T on the last day of February, 1979. They had found it to be an imminent threat to human health and had it immediately removed from the market. The public had expressed concern over miscarriages and other adverse health effects. In the fall of 1978, the EPA began a search for answers in three areas - an area where 2,4,5-T was used in the forest in Oregon, and 1-22 �two Oregon areas where it was not used. This study followed an earlier study of the circumstances in which nine women in the area suffered miscarriages in which the women could reach no conclusion as to the cause. Upon the completion of the second study (in February of 1979), known colloquially as "Alsea II", the EPA suspended 2,4,5-T for forestry, pasture, and right-of-way uses, but not for rice or range, or for a variety of trivial uses such as industrial or vacant lots. Witt (1980) in reviewing world-wide assessment of the EPA Alsea II Report found 18 reports, assessments or reviews on the subject. Every one of these evaluations found that the EPA data in the reports did not support the EPA conclusions. The most detailed accounting of Alsea II was prepared by Wagner et al. (1979). Following are statements prepared by the Advisory Committee on Pesticides for the United Kingdom (Kilpatrick, 1980). In essence we concluded that the data reported were not valid in either scientific or statistical terms and did not provide an adequate basis on which to compare abortion rates in women exposed and not exposed to 2,4,5-T; that evidence of relevant exposure to 2,4,5-T was lacking; and that a correlation had not been established between the abortion rates and the usage of material. The on-going Missouri horse arena (Times Beach) Episode. Several of the alledged TCDD "incidents" reported in this chapter have been of the "suspected" exposure nature. The report of TCDD poisoning in this episode have been confirmed both by effects on animal health and by sound analytical determinations. In 1972, following the spraying of horse arenas with salvage oil for dust control, 54 of 57 horses exposed to the arena died. In a report by the Missouri Division of Health and the Center for Disease 1-23 �Control (CDC), Atlanta, Georgia (Anonymous, 1974) extensive investigations were completed. Symptoms of the illness of the horses were skin lesions, severe weight loss and heptotoxicity. Birds, cats, dogs, insects and rodents were also found dead in the arena. An exposed 6-year-old girl developed hemorrhagic cystitis (characterized by blood in the urine). Other horse arenas sprayed with the similar salvage oil reported horse deaths occurring. within a month of each other. The three arenas were sprayed According to CDC testing, the original soil in the arena contained 33,000 ppb of the dioxin. The problem was supposedly solved by removing the treated soil from the arenas. The report of this incident was misquoted by several newspapers across the country which implied that the herbicide 2,4,5-T was responsible for the problem (Anonymous, 1974b). Published facts (Carter et al, 1975, Kimbrough, 1977) reveal that a factory in southwestern Missouri, while producing hexachlorophene, had accumulated by-product distillate containing up to 356 ppm TCDD. Episodes such as this, vividly point out our past inadequacies for properly disposing of toxic wastes. Whether the blame should lie with the manufacturer of the toxin or with the poor judgement of the licensed vendor for proper disposal is debateable. Unfortunately the episode is not complete. The top several inches of soil from the arenas were excavated and disposed of in several places. Soil from one arena was dumped along the highway front of the arena where human exposure would be minimal. was dumped in a sanitary landfill. Soil from another stable Soil from the third arena was used as landfill for two homes located in Imperial, a small town near St. 1-24 �Louis. As of January, 1983 families in these two homes along with four other families had been allocated Superfund money for temporary relocation (Sun, 1983). A considerable expansion in use of Superfund dollars has more recently been appropriated for relocation, etc. for families and businesses in Times Beach, Missouri where similar TCDD materials have been found. A re-examination of the young girl who had been exposed in one of the horse arenas conducted 5 years later revealed that the patient had grown normally. Detailed physical, chemical and neurological examinations were also conducted and found to be normal. The researchers concluded: Our experience demonstrates that people exposed to dioxin can recover completely with no apparent sequels from the toxin. It remains to be determined whether the exposure to dioxin in these children will result in abnormal pregnancies or affect their offspring" (Anonymous, 1974). An additional 4600 gallons of salvage oil from the hexachlorophene producing plant was located in the mid 1970's. Six years later, in 1980, the material was degraded using photodegradation techniques. Up to 100 sites in the state may have been contaminated with the TCDD twelve years ago. For some of the sites the kind and amount of TCDD have been well established. The important item, its availability to man, has not been demonstrated but its water solubility and other properties indicate it is not readily available for man to receive an absorbed dose. Assume 1) that the 2,4,5-T contained 0.1 ppm TCDD (the highest allowable level) and 2) that 2,4,5-T is applied at a rate of 2 Ib per acre, 3) that an acre 6 inch slice of soil weighs 2,000,000 Ibs. If 1-25 �the 2,4,5-T was incorporated to a depth of 3 inches then 1,000,000 Ibs of soil would contain 2 Ibs 2,4,5-T. If TCDD was present in the 2,4,5-T at a 0.1 ppm level then in 2 Ib of 2,4,5-T 2 x 10~7 Ib TCDD would be present. Dispersing this 2 x 10~7 Ib TCDD in the top 3 inches of soil gives a concentration of 0.2 ppt or 2 x 10~4 ppb. Comparing this amount to the 33,000 ppb TCDD present in the horse arena 33,000 ppb in horse arena = 165 million 0.002 ppb in normal use The important findings are: 1) that the dioxin concentration in soil from the horse arena was 165 million times greater than the concentration of dioxin in a soil normally treated with 2,4,5-T, and 2) no major health problems to residents of the affected areas have been reported other than those of the young girl described above. Seveso, Italy and other explosions. Since 2,4,5-T has come into commercial production in 1946, several industrial accidents have occurred resulting in human exposure to 2,3,7,8 tetrachlorodibenzo-pdioxin and/or other chlorinated dibenzo-p-dioxins. The most common method of human exposure occurred during handling of contaminated intermediate products (e.g. trichlorophenol, TCP). Young (1978) reported 23 industrial exposure episodes; eight of these occurred during an explosion or clean-up after the accident occurred. At least 740 people have been exposed during these episodes. Although accidents involving toxic dioxins and subsequent exposure to humans are highly undesirable they do afford us with an opportunity to evaluate the health of humans who have received massive TCDD doses. By detailed investigations of the accidents and extensive .record 1-26 �keeping of the victims scientists have been able to follow the health patterns of humans who have inadvertently received documented exposure to the tetrachlorinated dioxins. A combination of any or all trichlorophenol (TCP), pentachlorophenol (PCP), or 2,4,5-T may have been involved with these exposures. The most consistent disease exhibited by the victims was a skin disease called chloracne. Crow (1979) indi- cates that men are more likely than women to be afflicted by chloracne. The most publicized chemical accident in modern times occurred in Seveso, Italy. Hundreds of physicians, scientists and veterinarians have been involved in either on-site inspections, conferences or consultations involving the accident. This accident did not involve 2,4,5-T, but was associated with the production of trichlorophenol. Details of this TCDD episode follow: The episode of TCDD poisoning occurred on 10 July 1976 in Seveso, Italy, a small town 40 kilometers (km) north of Milan (Forth, 1977; Hay, 1977). The source of the TCDO was a chemical factory that produced trichlorophenol through the alkaline hydrolysis of tetrachlorobenzene. When the temperature in a steam-heated reaction vessel rapidly increased, a safety disk ruptured sending a plume of trichlorophenol, TCDD and other products 30 to 50 meters (m) high above the factory. The cloud apparently rose into the air, cooled and came down over a cone-shaped area about 2 km long and 700 m wide. It was estimated that 650 to 1700 grams of "TCDD was discharged into the environment (Reggiani, 1978). The escaped product had a TCDD concentration of 35,000 ppm (Anonymmous, 1977). Measurements of TCDD 1-27 �on vegetation in the immediate v i c i n i t y ranged from 1 to 50 ppm. Reggiani (1978) also reported that birds, rabbits and chickens began to die 2-3 days after the accident. A few children and some adults who had been exposed to cloud of TCDD dust complained of nausea and different kinds of s k i n lesions. Wagner (1981) reported that f a m i l i e s were not evacuated from the area u n t i l 17 to 38 days after the explosion. W i t h i n the next 6 weeks (by late August, 1976) an extensive system for monitoring the health of people in the v i c i n i t y was established to cover the acute as well as the long-term effects. The most complete analysis of health data was published by Reggiani in 1978. He conclu- ded that the Seveso accident did not reveal toxic effects in humans, which had not been observed in other episodes. Chloracne, the typical skin lesion, has occurred in children with tendency to spontaneous and rapid h e a l i n g . None of the exposed c h i l d r e n showed clinical signs of pathology other than chloracne d u r i n g these years. 1-28 �Chapter 1 References Advisory Committee on 2,4,5-T. 1971. Report of the Advisory Committee on 2,4,5-T to the Administrator of the Environmental Protection Agency. Ahlgren, G. H., 6. C. Klingman, and D. E. Wolf. 1951. Principles of Weed Control. New York, John Wiley and Sons. 368 pp. Allen, J. R., et al. 1977. Morphological changes in monkeys consuming a diet containing low levels of 2,3,7,8-tetrachlorodibenzo-pdioxin. Food Cosmet.Toxicol. 15:401-10. Anonymous. 1970. Globe's mystery. Time 95;42 (February 23, 1970). Anonymous, 1974a. Illness associated with TCDD-contaminated soil-Missouri. Morb. Mort. 23:299. Anonymous. 1974b. Death of animals laid to ch emicals. Times, p. 36 (August 28, 1974). New York Anonymous. 1977. Activity of the Laboratorio di Igiene e Profilassi (LPIP) in testing consequent to the ICMESA incident. Report of 5 November 1977 to the Seveso Authority. Reporto Chimico, Lab or Atorio di Igiene e Profilassi, Milano, Italy (Italian). Ashton, F. M. and W. A. Harvey. 1971. Selective chemical weed control. Circular 558, California Agricultural Experimental Station and Extension Service. Ashton, F. M. and A. S. Crafts. 1973. Mode of Action of Herbicides. New York, John Wiley and Sons. 504 pp. Bovey, R. W. and A. L. Young. 1980. The Science of 2,4,5-T and Associated Phenoxy Herbicides. New York, John Wiley and Sons. 462 pp. Buser, Hans-Rudolph, and Hans Paul Bosshardt. 1976. Determination of polychlorinated dibenzo-p-dioxins and dibenzofurans in commercial pentachlorophenols by combined gas chromatography-mass spectrometry, J.. Assoc. Off. Anal. Chem. 59:562-69. Carter, C. D., et al. 1975. Tetrachlorodibenzodioxin: poisoning episode in horse arenas. Science 188:738-40. An accidental Cooper, J. A. 1980. Environmental impact of residential weed combustion emissions and its implications. JAPCA 30:855-61. Cutting, R. T., Tran Huv Phuoc, Joseph M. Ballo, Michael W. Beneson, and Charles H. Evans. 1970. Congenital malformations, hydatidiform moles, and stillbirths in the Republic of Vietnam, 1960-1969. Superintendent of Documents, Washington, D.C. 1.29 �Erne, K. 1973. Toxicity studies with phenoxy herbicides on reindeer. Svensk. Veterinaertidning 24:273-75. Executive Office of the President. 1971. Report on 2,4,5-T. A report of the Panel on Herbicides of the President's Science Advisory Committee. Washington, Office of Science and Technology, Executive Office of the President. Farmline. 1980. Agricultural Productivity: No new miracles. August 1980. USDA vol. 1, no. 5, Washington, D.C. Firestone, D., et al. 1971. Distribution of chick edema factors in chick tissues. J_. Assoc. Off. Anal. Chem. 54:1293-98. Forth, W. 1977. Seveso incident. 2,3,7,8-tetrachlorodibenzo-l,4-dioxin TCDD; The Deutsches Arzetblatt 44:2617-28 (German). Foy, C. L. and L. W. Smith. 1967. The role of surfactants in modifying the activity of herbicidal sprays. In: J. U. Van Valkenburg (ed.), Pesticide Formulations Research. Washington, American Chemical Society. Pp. 55-69. Freed, V. H. and 0. M. Witt. 1967. Physicochemical principles in formulating pesticides'relating to biological activity. In: J. U. Van Valkenburg (ed.), Pesticide Formulations Research, Washington, American Chemical Society. Pp. 70-80. Halperin, U. E., P. A. Honchar, and M. A. Fingerhut. 1982. Dioxin: An Overview. The American Statistician, August 1982, 36(3) -.285-289. Hamaker, J. W. and H. 0. Kerlinger. 1967. Vapor pressure of pesticides. In: 0. U. Van Valkenburg (ed.). Pesticide Formulations Research, Washington, American Chemical Society. Pp. 39-54. Hay, A. W. M. 1977. Disasters 1:289-308. Tetrachlorodibenzo-p-dioxin release at Seveso. International Agency for Research on Cancer. 1978. Long-term Hazards of Polychlorinated Debenzodioxins and Polychlorinated Dibenzofurans. Joint NIEHS/IARC Working Group Report, IARC Internal Technical Report No. 78-001, Lyons, IARC. Kilpatrick, R. 1980. Further review of the safety for use in the United Kingdom of the herbicide 2,4,5-T. Advisory Committee on Pesticides, United Kingdom. Kimbrough, R. D., et al. 1977. Epidemiology and pathology of a tetrachlorodibenzodioxin poisoning episode. Arch. Environ. Health 28:77-85. Klingman, Glenn C. and Floyd M. Ashton. New York, John Wiley and Sons. 1-30 1982. Weed Science 1982. �Marth, P. C. and J. VI. Mitchell. 1944. Comparative volatility of various forms of 2,4-D. Botan. Gaz. 110:632-36. Meselson, M. S., A. H. Westing, and J. D. Constable. 1971. Background Material Relevant to Presentations at the 1970 Annual Meeting of the AAAS. Concerning the Herbicide Assessment Commission for the American Association for the Advancement of Science. Mimeo., Washington, D.C. Mounts, T. L., H. J. Dutton, C. D. Evans, and J. C. Cowan. 1976. Chick edema factor. Removal from soybean oil. J_. Am. Oil Chem. Soc. 53:105-107. Plimmer, Jack R. 1973. "Technical Pentachlorophenol" Origin and Analysis of Base-insoluble Contaminants, Environmental Health Perspectives. Exp. Issue No. 5. Pp. 41-48. Pokorny, R. 1941. New Compounds. J_. Am. Chem. Soc. 63:1768. Rappe, C., H. R. Buser, and H. P. Bosshardt. 1979. Polychlorinated dibenzo-p-dioxins (PCDD's) and dibenzofurans (PCDF's): Occurrence formation and analysis of environmentally hazardous compounds. Presented at CIDAC - Symposium in Baltimore Md., June 5-6, 1979. Reggiani, 6. 1978. The estimation of the TCDD toxic retential in the light of the Seveso accident. Paper presented at the 20th Congress of the European Society of Toxicology. Berlin (West), June 25-28, 1978. Roan, C. C. and D. P. Morgan. 1972. Alleged effects on human health of the use of herbicides in the area around Globe, Arizona. Mimeo., Arizona Community Pesticide Studies Project. Tucson, University of Arizona, March 6, 1972. Sanger, V. L., et al. 1958. Alimentary toxemia in chickens. ^ Am. . Vet. Med. Assoc. 133:172-76. Smith, L. W. and 0. E. Bayer. Soil Sci. 103-328. Sun, Margorie. 219:367-69. 1983. 1966. Missouri's costly dioxin lesson. Science Tschirley, F. H., W. Binns, C. Cueto, B. C. Eliason, H. E. Heggestad, G. H. Hepting, P. F. Sand, and R. F. Stephens. 1970. Investigations of spray project near Glove, Arizona. Investigation conducted February 1970. Mimeo, U.S. Dep. Agric., Office of Science and Education. 29 pp. U.S.D.A. Technical Bulletin Number 1671, 1982. Economic Assessment £f 2,4,5-T. The Biological and Van Valkenburg, J. U. 1967. The physical and colloidal chemical aspects of pesticidal formulations research: A challenge. In: J. U. Van Valkenburg (ed.), Pesticide Formulations Research, Washington, American Chemical Society. Pp. 1-7. 1-31 �Wagner, S. L. and J. M. Witt, et al. Oregon State University. Environmental Health Services Center, Oct. 25, 1979. Wagner, S. L. 1981. Clinical Toxicology of Agricultural Chemicals. Corvallis, Environmental Health Sciences Center, Oregon State University, Oregon. 309 pp. Witt, J. M. 1980. A discussion of the suspension of 2,4,5-T and the EPA Alsea II study. Pp. 47-59. Witt, J.M. 1980. A discussion of the suspension of 2j4,5-T and the EPA Alsea II Study. In: Supplement to the 34th Annual Meeting of the Northeastern Weed Science Society. 34:9-25. Young, A. L., et al. 1978. The toxicology, environmental fate and human risk of herbicide orange and its associated dioxin. United States Air Force OEHL Technical Report 78-92. Brooks Air Force Base, Texas. 262 pp. 1-32 �CHAPTER 2 Factors I n f l u e n c i n g M a n ' s Exposure to Phenoxy Compounds Chapter 1 pointed out that the phenoxy compounds have been widely used for several purposes. Their use has not been confined to spar- sely populated forests or prairies: they have also been used on or in croplands, rights-of-way, railroads, p u b l i c u t i l i t i e s , and national parks. This chapter w i l l attempt to c l a r i f y the fate and d i s t r i b u t i o n of the phenoxy herbicides in use. If people are exposed, what factors determine what their exposure level w i l l be? The continued development of more sensitive analytical instruments allow us a new perspective in l o o k i n g for and locating trace levels of contaminants in our environment. From one standpoint these new analy- tic tools should enable us to be "safer" because we can locate and i d e n t i f y much smaller amounts of toxic compounds than we could 10 years ago. Because of these advances, we must better understand the language used to describe the new trace levels. The term " m i l l i g r a m s per kilogram" (mg/kg) is commonly used to denote the weight in m i l l i g r a m s of the compound of interest, w h i l e the kilogram f i g u r e represents the weight of the test organism in kilograms. This ratio is useful in extrapolating data from an animal of one size to another. Sometimes large numbers w i t h small u n i t s tend to be confusing. We could call the distance across the crater of a volcano 804,672 m i l l i meters or equally as true we could call the distance one-half m i l e . Measurement terms commonly encountered when reading about con- 2-1 �taminants in the environment are parts-per-million (ppm) (10~ 6 ), partsper-billion (ppb) (10~9), and parts-per-trill ion (ppt)(10~^). To gain a better concept of what 1 ppm represents imagine a full railroad box car that contains 240,000,000 soybeans; if 240 of these soybeans were somehow different (let's assume they have a different odor), then the concentration of these different soybeans would be 240 in 240,000,000 or one in every million or 1 ppm. If we were searching for 1 ppb level, this would be equivalent to finding one of these soybeans with a different odor in 4.18 full box cars; if we are looking for a 1 ppt level we must look through all of the soybeans in 4180 box cars (or a train 39.5 miles long). This example may sound somewhat absurd but we have sensitive mass spectrometers capable of identifying and quantifying dioxins down to the ppt range. The ter- minology ppm is correctly interchanged with micrograms per gram (ug/g) or milligrams per kilogram (mg/kg). The ultimate question remains: cause harm to a human? will trace levels of a chemical This monograph will address several of the factors necessary before one can make a valid answer. How long do phenoxy compounds persist in the environment? Phenoxys are usually applied in postemergence treatment of growing vegetation and directly to soil as a preemergence treatment to control weeds during the germination or small seedling stage. The special structure of many herbicide molecules gives them excellent specificity properties, i.e., they will kill a target weed species without harming the surrounding cash crop. Under some conditions these very specific molecules are also easily destroyed. 2-2 Their biologic activity is often �greatly affected by a very simple change in molecular structure. The herbicide is ordinarily at its optimal plant-killing capability (bioactivity) when it is applied in the field. Almost all environmen- tal alterations that occur in the phenoxy compounds result in their loss of bioactivity. The following discussion is designed to illustrate the persistence of phenoxy compounds in the environment. As they degrade or are removed from the environment, the changes of human exposure to them as dislodgeable residues from leaf surfaces or through food uptake also decrease. Since the soil acts as an intermediary between application of the phenoxy herbicides and their potential uptake in human food, knowledge of the factors that affect their fate in soil is important. Those factors include adsorption, leaching, runoff, plant uptake, nonbiologic chemical reactions, microbial decomposition, volatilization, and photodecomposition. Adsorption. This factor keys the importance of most of the other dissipating factors. If the compound is tightly bound, only a small fraction of the total applied material is available for the other modes of dissipation. For the phenoxy compounds, adsorption to soil colloids does not tie up a large percentage of the soil-applied material. Although both clay and organic components of soil are effective in binding phenoxys, the bond is not strong when compared to many other families of pesticides. reservoir in an equilibrium process. Adsorption may be thought of as a One of the primary reasons for the phenoxys' rapid degradation is the relatively weak bonds between 2-3 �them and soil colloids. Leaching. Since the phenoxys are not tightly bound in soil, their ability to leach throughout soil is relatively high. Comparing their mobility with other herbicide families, Helling (1971) found them moderately mobile to mobile. The amount of downward movement in a soil profile depends on several factors, among which are the compounds' water solubility, a property which varies greatly depending on the formulation; the amount and intensity of rainfall; and the physical and chemical composition of the soil. Relatively mobile compounds such as 2,4-D and 2,4,5-T are also able to move upward in soil. This can happen as a wick action which allows the wetting front to move downward or upward in a soil profile. Runoff. Because of the phenoxys1 low adsorption and relatively high water solubility, runoff may occur, especially when applied to very sloping land. Several factors, such as rapid herbicide breakdown, can prevent excessive runoff. For runoff to be of major impor- tance, rainfall or irrigation must occur within the first week after application. Plant uptake. Although this is the primary reason soil-applied herbicides are used, it is surprising that seldom is more than 5% of the herbicide taken up by plants. In silty clay loams, 2% is more common. The amount of plant uptake is often related to the amount of herbicide in the soil solution; thus, any factor limiting this value will also decrease the amount the plant takes up (Lavy et al., 1968). Once inside the plant the herbicide may be degraded or it may remain as an intact molecule. Plants are able to degrade phenoxys by several 2-4 �pathways, as reported by Kearney and Kaufman (1975). Chemical decomposition. Several chemical reactions routinely occur in higher plants as the phenoxy acetic acid compounds degrade. Among these reactions are side-chain degradation, ring hydroxylation, and beta oxidation (Kearney and Kaufman, 1975). These reactions may become important if other factors do not readily dissipate the compounds. Microbial decomposition. This breakdown mode appears to be the major one for the phenoxy compounds. Experiments with sterile and nonsterile soils produce marked differences in the rate of loss of phenoxy herbicide phytotoxicity. An interesting phenomenon occurs when one retreats soil which had earlier been treated with phenoxy compounds. The half-life of a phenoxy in previously untreated soil may be in the vicinity of one to four weeks (Klingman, 1961), but when the same soil is retreated with a similar compound the half-life may be 3 to 15 days. Walker and Smith (1979) conducted a laboratory study to evaluate rates of 2,4,5-T degradation in soil. Degradation approximated first-order kinetics and the half-life varied from four days at 35°C and 34% moisture to 60 days at 10°C and 2Q% moisture. Soil microorganisms are quite adaptable. They live in a variety of climates and can devise ways to detoxify compounds they have not encountered before. We are indeed indebted to this mass of active biologic material. Volatilization. volatile. Compounds with high vapor pressures are more Their phytotoxic properties remain with them as they escape from the soil into the air. Once the farmer applies herbicides to the 2-5 �soil, he often incorporates his herbicide in it; i.e., turns the soil over and thus covers the treated layer with an untreated portion, thereby preserving the phenoxys1 plant-killing properties for their intended purpose. Before a properly incorporated herbicide is free to leave the soil through volatilization, it must first escape the binding capacity of the adsorptive soil complex. Loss of a herbicide through volatilization is often underestimated. EPIC, not a phenoxy but one of the more volatile herbicides, was lost from a free-liquid surface at a rate of about 5 Ib/A/h (Ashton and Sheets, 1959). Even herbicides with relatively low vapor pressures may be lost from a surface in significant quantities over an extended period, especially if exposed to high temperature. Soil tem- peratures as high as 180°F have been measured (Klingman and Ashton, 1982). Photodecomposition. Photodecomposition commonly degrades many herbicides. Like volatilization, it is often thought of as a surface reaction which can be prevented by proper incorporation techniques. On soil and leaf surfaces it represents a major avenue of loss. Pesticide transformation in sunlight plays an increasingly significant role in our environment (Crosby, 1976). The photo-oxidation of phenoxy acid herbicides results in rapid deactivation and low persistence. This process also limits the phenoxys1 usefulness since they degrade so easily. Other important factors Adequate moisture and a temperature range of 20°C to 35°C greatly enhance the dissipation of most soil-borne phenoxy compounds. 2-6 The �chemistry of 2,4-D and 2,4,5-T is very similar in many respects. Studies have shown that 2,4-D is somewhat more rapidly biodegradable, but the two herbicides' mobility factors are quite similar. Another difference is that no TCDD affiliation has been shown for 2,4-D. Because of their striking number of similarities, 2,4-D and 2,4,5-T will be exemplified interchangeably herein. One study showed that shortly after 2,4-D was sprayed on growing vegetation, the leaves contained residues of 200 to 500 ppm, but 10 weeks later this decreased to 20 to 50 ppm (Ramel, 1977). In a British review of U.S. research, 2,4-D was shown to be nonpersistent in plant tissues (Turner, 1977). When 2,4-D was applied to cucumbers in a sublethal dose, more than 75% of the herbicide disappeared within 13 days. Thus, if people are to be exposed to dislodgeable residues on sprayed leaf surfaces to any appreciable extent, they must make contact with the treated plants shortly after application. Photodegradation, rainfall washoff, volatilization, and plant uptake all play roles as these 2,4-D losses from leaf surfaces occur. Persistence of phenoxy herbicides in soil Initial levels of 2,4,5-T in soils are usually low and disappear relatively rapidly. In field studies, DeRose and Newman (1947) found that 2,4,5-T at 10 pounds per acre persisted 93 days after application. The investigators concluded that persistence was determined by soil microbial activity since 2,4,5-T persisted longer in autoclaved soil than in nonautoclaved soil. Other factors affecting disappearance of 2,4,5-T in soil include soil temperature, leaching, and soil organic matter. Generally, those conditions that favor microbial 2-7 �activity will favor more rapid decomposition of 2,4,5-T. In 1954, Warren (1954) studied the leaching and rate of breakdown of several phenoxy herbicides in a fine sand and silt loam, using crabgrass as a bioassay species. He found that 2,4-D ester, 2,4,5-T amine, and si 1 vex amine moved readily in sandy soil but moved little in mineral soils or mucks. Esters of si 1 vex and 2,4,5-T were resistant to leaching in all soils, with some movement in sand only. The amine formulation decomposed in the soil faster than the ester formulation. Silvex tended to be more persistent than 2,4-D and 2,4,5-T. More recent research, using gas chromatographic analytical techniques, has generally confirmed the results of earlier investigators. Altom and Stritzke (1973) reported that the average half-life of the diethanolamine salts of 2,4-D, dichlorprop, silvex, and 2,4,5-T were 4, 10, 17, and 20 days, respectively, in three soils. Except for 2,4-D, the rate of disappearance of the other phenoxys was faster in soil from Oklahoma grasslands than in soil from forests. Lutz and others (1973) studied the movement and persistence of picloram and 2,4,5-T (2 and 4 Ib/acre) on a North Carolina watershed that averaged a 27-percent slope. Approximately 60 percent of the picloram and 90 percent of the 2,4,5-T disappeared in 15 days. Most of the 2,4,5-T was found in the top 3 inches of soil, with no movement of 2,4,5-T beyond 12 inches downslope. In Texas, Bovey and Baur (1972) applied an ester of 2,4,5-T at 0.5 and 1 pound per acre to soils at five locations. Soils were sampled to a depth of 3 feet. Similar results were obtained at other geographical locations (Norris et al. 1977). 2-8 �Plumb et al. (1977) and others reported on the persistence characteristics of 2,4,5-T applied at 3 pounds per acre to a chamise site in southern California. Residue levels immediately after application were not determined, but based on residues present 14 days after application (0.9 ppm), 2,4,5-T showed a half-life of about 19 days for the period 14 to 29 days after application. The rate of degradation changes with time, however. Approximately 1 year after application, the residue level was about 0.05 ppm. Morris et al. (1977) determined 2,4,5-T residues in forest floor and soil after two successive annual applications of herbicide at 2 pounds per acre (ae) applied as the isooctyl ester in diesel oil by helicopter in March. The study area was a cool, moist site in western Oregon. The rate of decline in 2,4,5-T levels in forest floor after the first application at this site was slower than at the hot, dry site in southern California (Plumb et al., 1977), which may reflect the importance of volatilization and photodecomposition on the loss of phenoxy herbicides from exposed soil surfaces. The rate of loss of 2,4,5-T was quite rapid the first 30 days after the second application, which indicates good adaptation of the microorganisms after the first application. One year after application, residue levels in forest floor were about 0.75 .percent of the amount of herb-icide originally applied. These data show the strong tendency of forest sites to dissipate 2,4,5-T. Residues were largely confined to the top 6 inches of soil. Effects of high rates of application or persistence Some people are concerned that residues of 2,4-D and 2,4,5-T left 2-9 �in soils in Vietnam might destroy subsequent crops. Early work by Craft (1949) DeRose and Newman (1944), and many others indicated that 2,4-D and 2,4,5-T, when applied even at high rates, usually do not persist from one growing season to another, due largely to microbial decomposition of the herbicides. Work by Bovey and others (1968) in Puerto Rico indicated that corn, sorghum, wheat, rice, soybeans, and cotton could be grown in soils without reduction in fresh weight of the crops 3 months after the application of a 1:1 mixture of the nbutyl esters of 2,4-D + 2,4,5-T at 24 pounds per acre. Similar results were obtained for a 2:2:1 mixture of 2,4-D + 2,4,5-T + picloram at 15 pounds per acre (except for soybeans, which required 6 months for the phytotoxic effect to disappear). The longer residual effect on soybeans is prob-ably due to picloram, because of its greater persistence in soils. Blackman et al. (1974) reported on recent studies in Vietnam, which indicate that sensitive crops can be safely grown 4 to 6 months after single applications of the jv-butyl esters of 2,4-D and 2,4,5-T at rates up to 12 pounds per acre. The authors indicate that the dosage of herbicides in their experiments was considerably higher than would occur in spraying forests or mangroves since their materials were applied directly to bare soil and were not intercepted by herbaceous and woody vegetation. Young and others (1974) incorporated a 50:50 mixture of the ji-butyl ester of 2,4-D and 2,4,5-T into a soil trench in Utah at the rate of 1,000, 2,000, and 4,000 pounds per acre. After 440 days, 89 percent of the herbicide was degraded at the 1,000-lb rate, 85 percent was degraded at the 2,000-lb rate, and 83 2-10 �percent was degraded at the 4,000-lb rate. The rate of loss of the herbicide was rapid considering the low temperatures that prevailed for 7 months d u r i n g the experiments. In another study, Young and others (1974) reported on the effect of massive doses of 2,4-D and 2,4,5-T sprayed on an area at Eg!in Air Force Base in Florida. About 92 acres received 1,900 pounds per acre 2,4-D and 2,4,5-T in 1962 to 1964; a second area received 1,200 pounds per acre in 1964 to 1966; and a third area received 340 pounds per acre of 2,4-D and 2,4,5-T from 1966 to 1970. Chemical analyses of soil cores collected in 1970 from the treated areas showed a maximum concentration, 8.7 ppb of either herbicide, i n d i c a t i n g that the herbicide had essentially disappeared. Factors affecting phenoxy vapors in air A human's absorbed dose of any compound w i l l be due to i n h a l a t i o n of airborne materials, ingestion of food or water, and absorption through the skin. Tobacco use many contribute to the ingested portion. Human applicator exposure studies w i l l be addressed in detail in Chapter 3. Some of the factors that affect the amount of i n h a l a t i o n are the compound's vapor pressure, the particle size produced d u r i n g application, and the amount of protective equipment used. Vapor pressure The vapor pressure of the phenoxy compounds is markedly affected by the chemical make-up of their formulation. Since the esters are more toxic to plants than acid or salt formulations, this is the preferred form for many of the use areas. 2-11 These more phytotoxic forms �have a higher vapor pressure and are thus more v o l a t i l e . This property of v o l a t i l i t y nearly eliminated the use of the phenoxy compounds in agricultural areas where crop rotations are common since broadleaf plants, be they weed or cash crop, are indiscriminately k i l l e d by this family of compounds. Formulation chemists have greatly widened the scope of their use by l i n k i n g longer chain esters w i t h the active plant k i l l i n g component. This lower vapor pressure, less volat i l e material, allows the compound to be used in the near proximity to susceptible crops. Thus, when correctly applied placed in the f i e l d , it is expected that they w i l l not move in the vapor state but w i l l remain on the target area. The necessity of decreasing the vapor pressure of the phenoxys to allow their acceptance in the agricultural community has at the same time, markedly reduced the levels of the phenoxy vapors in the air that man breathes. A factor that has led man to be wary of pesticides in general, and the phenoxys specifically is the fact that the human nose is sensitive to low levels of these chemicals and/or their carriers. Another factor contributing to man's concern over the use of phenoxys is the fact that v i s i b l e symptoms of extremely low levels of airborne phenoxys are sometimes shown by off-site broadleaf green plants. Some chemicals that are not acutely toxic to plants or a n i - mals may contaminate an area without anyone suspecting their presence, however, even the presence of trace levels of phenoxy herbicides becomes quite obvious. Phytotoxic symptoms such as leaf c u r l i n g or t w i s t i n g for sensitive species such as cotton, tomato, potato, grapes, peas, beans, and many broadleaf weeds readily reveal the presence of 2-12 �these compounds. In an analytical laboratory this property of vapor pressure among the esters of the phenoxys can be seen quite easily. In a gas chroma- tographic determination of a mixture of the compounds, the more volat i l e short chained esters are eluted before the high molecular weight long-chained materials. Use of formulations containing these low volatile esters and even lower vapor pressure amines greatly lowers the probabilities of phenoxy vapors becoming airborne. Particle size Once the herbicide has been formulated in a manner that its vapor pressure does not s i g n i f i c a n t l y contribute to it being present in the air there is still another factor that can cause it to be airborne and that is particle size of the spray material. Larger particles would be expected to f a l l faster and smaller particles to remain suspended for a longer time period. A modification of Stokes Law (Hansen, 1965) which can be used to calculate how far a compound may d r i f t is: D = 4 1.49 _2 V - H 10 104 D = 1.49 2 V 'H where: D = drift in feet, H = height above ground in feet, V = crosswind velocity in mi/h, r = droplet diameter in m The drift of spray droplets in a 5 mi/h crosswind from a height of 100 feet would be as shown in Table 2-1. The drift distances resulting from other crosswind velocities or other release heights can be determined by applying an appropriate factor to the distances given or by calculation using the modified Stokes' equation. Thus, droplet �size is the critical factor determining spray drift since halving the droplet diameter results in a fourfold increase of drift distance. A spray droplet is also subject to evaporation while falling. A very small droplet can evaporate completely before reaching the ground or a leaf surface (USDA, 1982). Assuming an air temperature of 86°F, Table 2-1. Theoretical drift of spray droplets released 100 feet above ground in a 5-mi/h crosswind (USDA, 1982). Droplet size, Theoretical diameter (urn) d r i f t (feet) 50 298.0 100 74.0 200 19.0 400 4.6 600 2.0 800 1.2 1,000 .7 1,500 . .3 relative humidity of 50 percent, and still air, the approximate lifetime and distance of fall for water droplets would be as shown in Table 2-2. The tabulation shows that water droplets less than 100 urn in diameter would probably never reach the ground or a leaf surface when applied from a height of 10 feet, which is nearly a minimum for aerial application. 2-14 �The lifetimes and fall distances for herbicide spray droplets would vary from the figures given above. The kind of carrier (oil or water), vapor pressure of the carrier and the herbicide, and the kind of emulsion (oil in water or water in oil) would all influence droplet lifetime. Air turbulence causes major deviations from calculated fallout rates. Table 2-2. Lifetime and fall distance of water droplets in air at 86 F and 50% relative humidity (USDA, 1982). Droplet size, Fall diameter Lifetime distance (urn) (seconds) (inches) 200 56.0 1,678 100 14.0 151 80 9.5 36 50 3.5 11 40 2.4 2 20 .6 <1 10 .2 <1 2 .1 <1 a6°F, 50 pet relative humidity, still air. Most of the application equipment in use today produces a range of droplet sizes. The greater the volume of spray solution found in �small droplets (less than 100 urn), the greater the drift; however, there is an upper range of droplet sizes beyond which biological effect of a herbicide is reduced. Thus, herbicide applications should have the goal of achieving a range of droplet sizes that minimizes drift without unduly sacrificing biological effectiveness. Under field conditions the effective fall distance is the vertical distance moved relative to the surrounding air. In turbulent air this may be several times the height at which the spray is released. Knowledge of the in-flight behavior of selected droplet sizes and of the micrometereological factors should allow us to make significant improvements in predicting drift potential from agricultural spray operations (Seymour, 1967). Since it has been established that use of larger droplet sizes will reduce the amount of particle potentially available to drift let us examine means of selecting the desirable particle size. Stewart and Gratkowski (1976) present five factors of importance. (1) Increasing airspeed results in smaller droplets because of the greater shear forces imposed on the spray solution as it leaves the nozzle. (2) Pressure in the spray system also affects droplet size. Higher pressure increases turbulence in the nozzle, which in turn increases shear forces at the nozzle orifice, resulting in smaller droplets being formed. (3) Orifice diameter of nozzles is directly related to droplet size. A larger orifice will reduce shear forces caused by turbulence in the nozzle, and produce larger droplets. (4) The kind of nozzle also affects droplet size. Nozzles producing narrow, cylindrical patterns form fewer small drops; thus, they are better for 2-16 �reducing drift. (5) Nozzle orientation is a major factor affecting droplet size. The smallest range of droplet sizes and the lowest volume of spray solution in small droplets is obtained when nozzles are oriented parallel to the airstream and discharge downwind to the direction of airflow. Equipment is available that will provide droplet sizes of 300 to 400 vmd (volume median diameter in m) with 70 to 90 percent recovery in a 500-foot width; 400 to 600 vmd with 85 to 95 percent recovery; 800 to 1000 vmd with 95 to 98 percent recovery; and 800 to 1000 vmd with 99 or more percent recovery (Akeson et al., 1978). Temperature and relative humidity influence drift through evaporation, which reduces droplet size and results in more drift. In prac- tice, many States impose limitations to herbicide application based upon these two factors. Limitations are also imposed in terms of maximum permissible windspeed at the time of application. A maximum windspeed of 5 mi/h is common, although up to 10 mi/h is permitted in areas where there is less hazard to sensitive vegetation. A critical atmospheric factor is the temperature gradient with height, specifically the occurrence of warm air overhead, usually referred to as an inversion condition (Akeson et al., 1978). An inversion limits vertical air circulation and acts to concentrate fumes and small particles in a cloud under the inversion ceiling, relatively close to the ground. The material thus entrapped may be transported long distances in amounts sufficient to cause damage to sensitive crops. Spray solutions can be modified to reduce the number of small 2-17 �droplets and thereby reduce drift. The principles involved are the increase of viscosity or of surface tension, each of which tends to reduce the number of small droplets. The types of preparations available to reduce drift may be classified as invert emulsions, thickeners, particulating agents, and foaming agents (Gratkowski, 1973). Although many variables affect spray drift, it is clear that elimination of small droplets, especially those less than 100 m in diameter, is the fundamental solution to the drift problem; however, the biological effectiveness of the phenoxy herbicides decreases as droplet size increases and droplet density decreases. For example, McKinlay et al (1972) found that increasing droplet size from 100 to 200 urn, with volume kept constant, required three times as much active ingredient, and when size was increased to 400 urn, six times as much herbicide was needed to give equivalent biological effects. There are two factors that tend to make smaller droplets more effective. First, the leaf area contacted by a given volume of spray solution is greater when droplets are smaller. That may enhance absorption. Second, high herbicide concentrations localized in larger droplets may so damage the underlying cells that translocation to other tissues is reduced. In practice, the lower effectiveness of larger droplets can be offset by increasing herbicide concentration of the spray solution or by increasing the total volume. Both increase costs. Drift can be reduced when conventional application equipment is used, by taking advantage of the best combination of nozzle type and 2-18 �orientation, orifice size, pressure, and spray mixture. In addition, modern engineering developments permit reduction of droplets below 100 m diameter to near zero. The microfoil boom, for example, has nozzles placed in a boom shaped like an airfoil, which minimizes turbulence at the point where droplets are formed. Primary droplets from microfoil nozzles are about twice the size of the orifice. Smaller satellite droplets are formed from thin filaments of spray between the primary droplets, but proper nozzle orientation will result in the capture of small droplets by large droplets in the smooth air behind each nozzle. The microfoil boom provides many of the desirable qualities in droplet size. However, the microfoil boom is expensive to buy and is subject to clogging and other problems if it is not properly maintained. Nevertheless, it provides the best drift control available at this time. Other application systems are in the process of development (Stewart, Gratkowski, 1976). Protective equipment Although much can and has been done to limit the number of respirable particles it is not always possible or practical to eliminate all of them. Many kinds of masks and filters have been devised to remove respirable size pesticide particles. Although it may be recommended that a mask be worn the human element plays an important role in determining whether it actually will be. If the compound of concern is known to be highly hazardous the worker is more likely to use the protective mask. Since most woody vegetation growth occurs during the warm and hot months of the year that is also when much of the phenoxy herbicide use occurs. Both the 2-19 �heat and the humidity add to the discomfort of man using the protective devices. When men engage in strenuous physical activity in the hot summer months such as during phenoxy herbicide application masks are seldom used. Many of these workers have been using phenoxy herbicides for over 25 years with no known ill effects. In spite of the fact the label on the phenoxy herbicides container indicates that masks be used, they seldom are. Possibilities of human exposure via phenoxy components in air Most dependable inhalation data are achieved by trapping exposure levels via air. Bamesberger and Adams (1966) indicated that 2,4,5-T was found infrequently and in low concentrations in air-sampling studies in Washington state. In high use areas, however, one might expect concentrations similar to 2,4-D as reported by Adams and others (1974). Average concentrations of ester of 2,4-D in air in Washington during the spraying season was 0.1 ug/m3. Assuming that a person would inhale 30 cubic meters of air per day, the exposure would be 0.003 mg per day. The threshold limit values in air adopted by the American Conference of Government and Industrial Hygienists in 1977 were 10 mg/m3 for 2,4-D or 2,4,5-T. The highest concentration of phenoxy herbicides in the air probably occurs during application. Russian workers (Fetison, 1966) reported concentrations of the sodium salt of 2,4-D up to 22.4 mg/m3 after spraying. Akesson, 1978 (communication to USDA, 1982) however, has shown a maximum of 20 ug/m3 of herbicides downwind from typical aerial .application sprays. TCDD has not been measured in the air in spray areas, but possible levels can be calculated based on an assumed 2-20 �2,4,5-T:TCDD ratio of l-.lxlO'7 and the levels of herbicide above (this assumes TCDD to be present at 0.1 ppm level). In a national air monitoring pesticide program there were 2,479 samples taken in 14 to 16 different states (Kutz :et al.). The sampling sites for 1970 and 1972 were selected for being potentially areas with a high concentration of pesticides. The analytical method had a sensitivity of 1-10 nanograms per cubic meter.- It is evident that most samples were analyzed at the lower limit of the method's capability. There is no established concentration in the air for 2,4-D that is considered to be safe or unsafe for the general population. The TLV of 2,4-D for a normal working eight-hour day without adverse effect is 10 nig/m^ Table 2-3. Three-Year Summary of 2,4-D in Ambient Air in U.S.A. in Nanograms3 per Cubic Meter from 1970-1972 (Kutz et al. 1976). Total 3-yr Average Percent positive samples 5.64 Average values 1.54 Average values of positive samples 18.33 Maximum values 68.17 a A nanogram is 1 x 10~9 grams = 1 x 10~6 mg, 68.17 nanograms = 0.000,068,17 mg. Table 2-3 shows that the maximum value found was 68.17 nanograms which is 0.00006817 mg. This amounts to a safety factor of 146,700 �before reaching the TLV value considered to be a safe exposure for a 40-hour working week. Again it should be remembered that over the three-year period covered by this report 2,4-D occurred in the air on the average only 5.64% of the time and the sites were potentially the worst possible situations. Thus, practically speaking, the general public is not exposed to 2,4-D in the air. Man's potential for exposure to phenoxy herbicides in drinking water and other water sources As discussed under photodecomposition of phenoxy herbicides in soil, UV light is effective in degrading many herbicides including the phenoxy compounds. Microbial action and other chemical reactions may also occur to rid water of trace levels of herbicide. The Water Resources Division (WRD) of the U.S. Geological Survey reported a study on 12 pesticides occurring in 11 western streams (Brown and Nishloka, 1967). The three herbicides 2,4-D, 2,4,5-T and silvex were included but none was found. The sensitivity of the method was given as 100 parts trillion (ppt) for 2,4-D and 5 ppt for 2,4,5-T and silvex. The survey was made over approximately one year. The WRD reported on the incidence of pesticides in selected western streams (Mangold and Schulze, 1969). three-year period, 1966-1968. over 15 western states. This report covered a Samples were taken from 20 locations The highest amount of 2,4-D found was 0.35 g/liter (0.35 ppb) in James River at Huron, S.D. (According to this report, the Committee on Water Quality Criteria has established a standard of 100 ug/liter (100 ppb, 0.1 ppm or 0.1 tug/liter) for all the phenoxy herbicides.) There were 321 samples analyzed for 2,4-D. 2-22 �Eighty-seven percent of the samples tested had no 2,4-D. Of the 41 samples containing 2,4-D, 29, or 70%, had less than 0.15 parts per billion (ppb). Table 3 gives a detailed breakdown of the quantities of 2,4-D found in these samples that occurred only 13% of the time. Table 2-4. Occurrence of 2,4-D in ppb in selected Western streams over a three-year period (Mangold and Schulze, 1969). 0.01- ppb 0.06- 0.11- 0.16- 0.21- 0.26- 0.31- 0.05 0.10 0.15 0.20 0.25 0.30 0.35 13 13 6 2 3 2 2 No. of occurrences3 a Remember that this is only in the samples containing 2,4-D and 87% of the samples had none. A monitoring program on water has also been conducted by the Federal Water Quality Administration of the U.S. Department of Interior (Lictenberg et al., 1970). In their five-year survey of water in the major river systems in the U.S.A., no mention is made of the phenoxy herbicides. Analyses were not made for them presumably because based on the earlier studies (Brown and Nishloka, 1967; Mangold et al., 1969) they were thought not to be of sufficient concern to warrant analysis. The U.S. government has established an acceptable level of 0.1 ppm 2,4-D as a contaminant in what they call community water systems. This information was published in the Federal Reg-ister (Anonymous, �1975) in 1975. A tolerance of 0.1 ppm of 2,4-D in potable water has also been established by EPA. EPA published in 1980 a list of 64 toxic pollutants classified under the Clean Water Act. As would be expected from the foregoing, 2,4-D is not included on this list. Since residues of phenoxy herbicides tend to remain in upper soil layers, they are rapidly degraded. It is unlikely that ground water would be polluted from current registered uses of phenoxy herbicides, and thus exposure is considered to be zero. Surveys by the U.S. Geological Survey program of the surface waters of major rivers in the Western United States over a period of years indicated that the highest concentration of a phenoxy herbicide was 0.97 ppb of 2,4-D. Researchers have found that even in streams adjacent to aerial spraying operations in the forest, concentrations of 2,4-D, 2,4,5-T, or other herbicides seldom exceed 0.01 ppm. After application, concentrations of the herbicide rapidly diminish by dilution. The preponderance of streamwater samples from operational monitoring programs in forest land has not contained detectasble residues of 2,4,5-T. Even when ditchbanks were sprayed directly, so that spray fell into the stream, the maximum 2,4,5-T found after applications at 2 pounds per acre was 0.04 ppm. Herbicide could be found only in the treated area, but none could be found one mile downstream. Therefore, considering that only small and intermittent portions of the total land area are treated, the risk of exposure of the general population in the United States to significant levels of the phenoxy herbicides in water is remote. The greatest potential for 2-24 �exposure occurs if domestic water is taken from very small streams in, or immediately downstream from, treated areas. An extensive research base shows: (1) Such exposure would be infrequent because most small watersheds are never treated and those that are treated seldom yield water contaminated with herbicides; and (2) when contamination does occur, it is low (less than 0.1 ppm, usually less than 0.01 ppm 2,4,5-T) and remains for only one hour to a few days. Sugarbeets, soybeans and dwarf corn crops were irrigated with phenoxy treated water at concentrations of 0.02, 1.1 and 5.5 ppm in 2 acre in. of water by both furrow irrigation and sprinkler irrigation (Bruns et al., 1973). These levels added to the water were much higher than those found in monitoring studies. At the two highest concentrations 2,4-D residues were detected in roots only. Sugarbeet roots contained 0.01 ppm or less. The concentrations of the residues in the crops even from this height concentration treated irrigation water, were many times below the tolerance in effect in 1972 for certain food and forages for human and livestock consumption (Bovey and Young, 1980). Averitt (1967) studied the persistence of 2,4-D amine in a bayou in Louisiana after application for control of water hyacinth. Application of 2,4-D was made as a coarse wet stream at 4.48 kg/ha by wetting the leaves with little or no runoff into the water. An evaluation of the persistence of 2,4-D amine in a bayou for water hyacinth control in Louisiana was made (Averett, 1967). He applied 2,4-D as a coarse wet stream with little run-off occurring from the leaves. 2-25 �Water samples taken below depths of 13.5 in one location indicated residues of 153, 37, 727, 38, 12, and 0.6 ppb for 1 hr, 24 hrs, 1, 2, 3, and 20 weeks, respectively, after application. Residues at a deeper depth were similar. The high concentration of 2,4-D after 1 hr is attributed to direct spraying of the water and/or "roll-off" of spray droplets from hyacinth leaves. The increased concentration at 1 week compared to the level at 24 hr was attributed to loss of chemical from the plants into the water. Three weeks after treatment the 2,4-D levels diminished considerably to 12 ppb. The body of water was relatively stagnant with somme flow from upstream, however, the banks of the bayou upstream had been treated so an influx of water should have also contained 2,4-D. Small plastic lined boxes, 60 x 120 x 30 cm deep, filled with lake or bayou water and supporting stands of water hyacinths were treated with rates of 4.5 or 5.6 kg/ha of 2,4-D amine. Loss of 2,4-D from the boxes was similar to loss in tha natural bayou. The 2,4-D had essentially disappeared in 27 days (4 and 15 ppb for 4.5 and 5.6 kg/ha application). Two lagoons chosen for study gave similar results to earlier studies. Man's exposure to phenoxy herbicides through the food he eats Evaluation of market basket surveys provide us with the best criteria for quantifying the levels man is ingesting. For definition purposes a "market basket" represents the amount of food consumed by a 16-19 year old male in a basic two-week diet. This amount represents almost twice the food intake for the average individual. Pesticide residue studies of our food has been conducted by the U.S. Food and Drug Administration since 1964. 2-26 Seventeen classes of �food are sampled: dairy products; meat; shellfish, fish; eggs and poultry; grain and cereals; potatoes; leaf and stem vegetables; beans; vine and ear vegetables; root vegetables; garden fruits large and small; oils; fats and shortening; tree nuts; infant and junior foods. There are 82 individual food items in these "market baskets". Thirty markets are sampled in 28 different cities. In a report by Corneliussen (1972) which covers the period from June 1969 to April 1970, 2,4-D was analyzed using a test method with a sensitivity of 0.02 ppm. Some trace residues were reported which were below the sensitivity level of analysis. estimates. Such low values are only Chlorophenoxy acids were found four times during this reporting period. However, 2,4-D is listed as occurring only twice with the highest value being 0.125 ppm in the June, 1969 to April, 1970 reporting period. For earlier periods they report phenoxy her- bicides were found seven times in 1967-1968, eight times in 1966-1967, and 13 times in 1965-1966. Shown in a series of reports (Corneliussen, 1972; Oohnson et al., 1976; Manske et al., 1975; Manske et al., 1974) covering the period of June, 1970 through July, 1973 2,4-D was present only infrequently (once or twice in all the yearly samples) and in low amounts (about 0.01 ppm). From August, 1973 through July, 1975, no 2,4-D was found in any of the samples (Johnson et al., 1977; Manske et al., 1977). summary report (Duggan et al., 1969) /\ f 0r the period July 1963 to July 1970 indicates the maximum possible amount of 2,4-D a person might have ingested during this six-year period (1964-1970) is given. These amounts are well below the daily intake quantity regarded as safe by �the FAO-WHO Expert Committee and even further below the "no-effect-level" established by the World Health Organization. An expert group on pesticide residues working for FAO-WHO has established an acceptable daily intake of 2,4-D for humans of 0.3 mg/kg (Anonymous, 1972). day. For a 132 Ib person this would amount to 18 mg per Taking a maximum daily intake of 0.005 mg as calculated from market basket surveys would give a 3600 safety factor before the acceptable daily intake was even reached. This same international group (Anonymous, 1972) has also accepted 31 mg/kg/day of body weight in the rat as being a dose level that causes no toxicological effect. The use of this figure of course would result in an even larger safety factor. When considering this point, it should be noted that the Environmental Protection Agency (EPA) has issued a number of tolerances for 2,4-D. 2,4-D These tolerances are permissible residues of in our food and water that EPA considers would not be harmful to human health. Following are printed notices EPA has presented as guidelines. This was taken from the Code of Federal Regulations Title 40 Protection of Environmental Parts 100 to 399 revised as of July 1, 1979 p. 489 Section 180.142 2,4-D: tolerances for residues, (a) Tolerances are established for residues of the herbicide plant regulator and fungicide 2,4-D (2,4-dichlorophenoxyacetic acid) in or on raw agricultural commodities as follows: 5 parts per million in or on apples, citrus fruits, pears and quinces. 2,4-D The tolerance on citrus fruits also includes residues of (2,4-dichlorophenoxyacetic acid) from the preharvest application �of 2,4-D isopropyl ester and 2,4-D butoxyethyl ester and from the postharvest application of 2,4-D alkanolamine salts to citrus fruits, and from the postharvest application of the 2,4-D isopropyl ester to lemons. 0.2 part per million in or on potatoes, (b) Tolerances are established for residues of 2,4-D (2,4-dichlorophenoxyacetic acid) at: 1000 parts per million in or on grasses (pasture and rangeland): 300 parts per million in or on grass hay: 20 parts per million in or on the forage of barley, oats, rye, and wheat, corn fodder and forage, rice straw, sorghum fodder and forage, and sugar cane forage: 2 parts per million in or on sugarcane: 0.5 part per million in or on the grain of barley, oats, rye, and wheat, corn grain and fresh corn including sweet corn (kernels plus cob with husks removed), cranberries, grapes, and sorghum: 0.1 part per million in or on blueberries and rice from application of 2,4-D in acid form, or in the form several salts or esters. The FDA, according to Pesticide and Toxic Chemical News (Anonymous, 1979), list 2,4-D as being found twice in its food monitoring programs in fiscal 1978 and zero in fiscal 1979. To put this in perspective the insecticides malathion and captan were found 329 and 305 times; and 141 and 185 respectively in the same time period. Thus, it may be concluded that from a health standpoint there is essentially no 2,4-D in our food supply. Measurable amounts of 2,4,5-T were found only in.two food samples 2-29 �in FDA market basket surveys in 1966-67 and in one sample in 1967-68. No 2,4,5-T has been found in food since 1968 in the FDA studies. total of over 2,000 samples was collected and analyzed. The highest 2,4,5-T concentration in the 1966-67 samples was 0.19 ppm. residues of silvex were found. A Only two These occurred in dairy products collected in 1965-66 and were 0.018 ppm and 0.029 ppm (Environmental Protection Agency, 1978). Therefore, based on FDA market basket sur- veys, the amount of phenoxy herbicides in food is virtually undetectable. The most direct exposure of humans to 2,4,5-T through food products is probably via plants; however, research has shown that phenoxy residues in forage and agronomic crops usually disappear rapidly. Since most weeds in crops are treated in early spring, residues disappear by harvest time. Devine (1970) analyzed 27 samples of rough rice fromm Texas, Arkansas, and Louisiana for residues of 2,4,5-T at intervals from 50 to 84 days after application of 2,4,5-T for weed control. No detectable residues (0.01 ppm) were found. Rice straw contained residues that varied from <0.01 ppm to 1 ppm. In the case of pasture and rangeland plants, which may intercept relatively high amounts of phenoxy herbicides (up to 200 ppm), residues can be avoided in meat and milk products by deferring grazing for milk cows on the treated area a few days to a few weeks and removing meat animals from treated pastures 2 weeks before slaughter. These restrictions appear on current product labels. W i l l the meat from slaughtered animals which had been bean fed 2,4-D or 2,4,5-T for four weeks contain residues? Negligible levels 2-30 �(less than 0.1 ppm in muscle and fat) were found when no withdrawal time was employed. If the animals were fed a regular diet for 1 week after the 2,4-D or 2,4,5-T feeding occurred, no detectable phenoxys were found in fat or muscle tissue (Leng, 1972). Even when w i l d l i f e species or livestock graze on pastures immediately after spraying, only small amounts of phenoxy herbicide may appear in meat or m i l k . They disappear after a few days due to rapid loss of the herbicides from forage and by normally rapid excretion from the grazing animal. Klingman et al. (1966), in actual field grazing trials w i t h cattle, found 0.01 ppm to 0.09 ppm of 2,4-D in m i l k the first 2 days after spraying 2,4-D at 2 pounds per acre, and lower amounts thereafter. No residues of 2,4,5-T were found in m i l k from cows put into pastures 4 days after spraying. Bjerke et al. (1972) found no residues of 2,4-D, 2,4,5-T, or MCPA, or their corresponding phenols, greater than 0.05 ppm in m i l k from cows exposed to 30, 300, or 1,000 pm 2,4,5-T in their feed level. silvex were found only at the 1,000 ppm feeding level. Residues of Clark et al. (1975) concluded that residues of phenoxy herbicides or phenolic metabolites in meat of sheep or cattle are unlikely under normal patterns of 2,4,5-T use. In field studies, Newton and N o r r i s (1968) found that b l a c k t a i l deer did not accummulate large amounts of 2,4,5-T when they grazed browse that had been treated with 2 pounds per acre. Concentrations in tissue rarely reached detectable levels, and the ruminant was able to degrade and e l i m i n a t e the herbicide, soon after ingestion. Obviously, game animals may graze in treated areas immediately after 2-31 �spraying, but in most cases spray areas are substantially smaller than the home range of large game animals, and thus exposure is not continuous. Game animals are likely to constitute a vanishing small proportion of the average human diet in the United States, but they may be an important component in the diet of a few individuals. Fish and other aquatic organisms are also important components of the diet of humans. The occurrence of significant amounts of phenoxy herbicides in the FDA market basket survey in fish products was not indicated. Research shows that most fish do not accumulate large amounts of the phenoxy herbicides «1 ppm), even when the herbicide is applied directly to water surrounding the fish. The degradation rate of phenoxys in surface water is quite rapid. Fish also have the capability of eliminating and degrading the phenoxys. None of these reports show food to be a major source of human exposure. Assessment of human exposure to the phenoxy herbicides Man's potential exposure to phenoxy herbicides through the air he breathes, water he drinks or food he eats has been shown to be minimal. Levels that his skin may absorb by coming into casual contact with sprayed vegetation would probably be low since these compounds degrade rapidly on both leaf surfaces and soil. Although much of phenoxy applied to vegetation may find its way to the soil or run-off into water the degradation rate is so rapid that no buildup occurs. Thus, all of avenues for human exposure examined so far have indicated that humans are receiving non-existant to minimal levels. In the next chapter an in-depth look into exposure levels 2-32 �received by those who apply herbicides for their livelihood will be made. It is important to differentiate between man's exposure and his absorbed dose. It was pointed out that humans may be exposed via inhalation, dermal depositions or ingestion. An important item which dictates the degree of risk man encounters is his absorbed dose or how much of the chemical actually enters the body fluids. The rate and amount excreted from the body is also important. The absorbed dose resulting from inhalation must take into consideration how much was present in the air a man breathes in and also how much is in the air he exhaled. The difference would conceivably be the absorbed dose unless a fraction of this was expelled as sputum. Absorbed dose resulting from dermal exposure is that amount penetrating man's dermal membranes. Herbicide contacting the skin does not necessarily penetrate the surface. In the case of water soluble phenoxy compounds a considerable amount can be washed off using soap and water if this is done soon after contact. In the case these compounds are ingested, the opportunities for absorption are much greater. Thus, man or any other organism can not be injured merely by being exposed to a toxin. 2-33 �CHAPTER 2 References Adams, D. F., C. M. Jackson, and W. L. Bomesberger. 1974. Quantitative studies of 2,4-D esters in the air. Weeds 12:280-83. Akesson, N. B. and W. E. Yates. 1978. application handbook. Wildlands aerial herbicide Altom, J. D. and J. F. Stritzke. 1973. Degradation of dicamba, picloram and four phenoxy herbicides in soils. Weed Sci. 21:556-60. American Conference of Government and Industrial Hygienists. 1977. Threshold limit values for chemical substances and physical agents in work room environment with intended changes for 1977. Cincinnati. Anonymous. 1972. WHO Pesticide Residue Series, No. 1. 1971. Evaluations of some pesticide residues in food. Geneva, World Health Organization, 1972. Anonymous. 1975. Part IV. Environmental Protection Agency, Water Programs, National Interim Primary Drinking Water Regulations. Federal Register December 24. p. 59571. Anonymous. 1979. FDA lists pesticides most frequently found in food monitoring programs. Pesticide and Toxic Chemical News. October 24, 1979, p. 16. Ashton, F. M. and T. J. Sheets. 1959. Averitt, W. K. 1967. An evaluation of the persistence of 2,4-D amine in surface waters in the State of Louisiana. Proc. South. Weed Conf. 20:342-47. Bamesberger, W. L. and D. F. Adams. 1966. An atmospheric survey for aerosol and gaseous 2,4-D compounds, pp. 219-227. In: R. F. Gould (ed.) Organic Pesticides in the Environment. Ad. Chem. Series, Am. Chem. Soc. Bjerke, E. L., J. L. Herman, P. W. Miller, and J. H. Wetters. 1972. Residue study of phenoxy herbicides in milk and cream. J_. Agric. Food Chem. 20:963-67. Blackman, G. E., J. D. Fryer, A. Lang and M. Newton. 1974. The effects of herbicide in South Vietnam, Part B. Persistence and disappearance of herbicides in tropical soils. Washington, National Academy of Sciences. Blackman, G. E., J. D. Fryer, A. Lang, and M. Newton. 1974. The effects of herbicides in South Vietnam. Part D. Working Papers. Washington, National Academy of Sciences/National Research Council. Bovey, R. W., F. R. Miller, and J. D. Diaz-Colon. 1968. Growth of crops in soils after herbicidal treatments for brush control in the tropics. Agron. G. 60:678-79. 2-34 �Bovey, R. VI. and J. R. Baur. 1972. Persistence of 2,4,5-T in grasslands of Texas. Bull. Environ. Contam. Toxicol. 8:229-33. Bovey, R. W. and A. L. Young. 1980. The Science of 2,4,5-T and Associated Phenoxy Herbicides. New York, John Wiley and Sons. p. Brown, E. and Y. A. Nishloka. 1967. Pesticides in selected Western streams; A contribution to the national program. Pesticides Monitoring J. 1:38-46. Bruns, V. F., B. L. Carlile, and A. D. Kelley. 1973. Responses and residues in sugarbeets, soybeans, and corn irrigated with 2,4-D or silvex treated water. United States Department of Agriculture, Agric. Res. Ser., Technical Bulletin No. 1476. 32 pp. Clark, D. E., et al. 1975. Residues of chlorophenoxy acid herbicides and their phenolic metabolites in tissues of sheep and cattle. J^. Agric. Food Chem. 23:573-78. Corneliussen, P. E. 1972. Pesticide residues in total diet samples. VI Pesticides Monitoring J. 5:313-30. Craft, A. S. 19:141-58. 1949. Toxicity of 2,4-D in California soils. Hilgardin Crockett, A. B. 1974. Pesticide residue levels in soils and crops, FY-70 National Soils Monitoring Program. II. Pesticides Monitoring J_. 8:69-97. Crosby, Donald G. 1976. The Significance of Light Induced Pesticide Transformations. Pp. 568-76. DeRose, H. R. and A. S. Newman. 1947. The comparison of the persistence of certain plant growth regulators when applied to soil. Soil Sci. Soc. Proc. 12:222-26. Devine, J. M. 1970. Report on 2,4,5-T residues in rough rice and rice straw. Syracuse Univ. Res. Conf. Jan. 26, 1970. Duggan, R. E., G. Q. Lipscombe, E. L. Cox, R. E. Heatwole and R. C. Kling. Residues in food and feed - pesticide residue levels in foods in the United States from July 1, 1963, to June 30, 1969. Pesticides Monitoring J. 5:73-212. Fetisov, M. I. 1966. Occupational hygiene in the application of herbicide of the 2,4-D group. Gig. Sanit. 31:383-86. Gratkowski, H. and R. Stewart. 1973. USDA Forestry Service Gen., Rept. PNW-3 Pac. Northwest For. Range Exp. Stn., Portland, OR. Hansen, J. R. A bi-fluid spray system for application of invert emulsions. Oklahoma Agric. Aerial Appl. Conf. Phenoxy Herbicide Bull. No. 207. Hercules Powder Co., 1965. �Helling, C.S. 1971. Pesticide mobility in soils. Proc. Soil Sci. Soc. Am. 35:732-748. Johnson, J. E. 1971. The public health implications of widespread use of the phenoxy herbicides and picloram. BioScience 21:899-905. Johnson, R. D. and D. D. Manske. 1976. Pesticide residues in total diet samples. IX. Pesticides Monitoring J. 9:157-69. Johnson, R. D. and D. D. Manske. 1977. Pesticide and other chemical residues in total diet samples. Pesticides Monitoring J. 11:116-131. Kearney, P. C. and D. D. Kaufman. Marcel Dekker. Vol. I. 500 pp. 1975. Herbicides. New York, Klingman, 0. L., C. H. Gordon, 6. Yip, and H. P. Burchfield. 1966. Residues in the forage and in milk from cows grazing forage treated with esters of 2,4-D. weeds 14:164-67. Klingman, G. C. 1961. Weed Control as a Science. Wiley and Sons. P. 130. Klingman, Glenn C. and Floyd M. Ashton. York, John Wiley and Sons. P. 449. New York, John 1982. Weed Science. New Kutz, F. W., A. R. Yobs and H. S. C. Yank. 1976. National pesticide monitoring programs. In: Robert E. Lee, Jr., (ed.), Air Pollution from Pesticides and Agriculture Processes, Boca Raton, Florida. CRC Press. Lavy, T. L, 1968. Micromovement mechanisms of s-triazines in soil. Soil Sci. Soc. Amer. Proc. 32:377-80. Leng, M. L. 1972. Residues in milk and meat and safety to livestock from a use of phenoxy herbicides in pasture and rangeland. Down to Earth 28:12-20. Lictenberg, J. J., J. W. Eichelberger, R. C. Dressman and J. E. Longbittom. 1970. Pesticides in surface waters of the United States; A five-year summary, 1964-1968. Pesticides Monitoring J. 4:71-86. Lutz, J. F., G. E. Byers and T. J. Sheets. 1973. The persistence and movement of picloram and 2,4,5-T in soils. vL Environ. Qua!. 2:485-88. Manigold, D. B. and J. A. Schulze. 1969. Pesticides in water pesticides in selected western streams - A progress report. Pesticides Monitoring J. 3:124-35. Manske, D. D. and P. E. Corneliussen. 1974. Pesticide residues in total diet samples. VII. Pesticides Monitoring J. 8:110-24. �Manske, D. D. and R. D. Johnson. 1977. Pesticides and other chemical residues in total diet samples. X. 1977. Pesticides Monitoring J. 10:134-48. Manske, D. D. and R. D. Johnson. 1975. Pesticide residues in total diet samples. VIII. Pesticides Monitoring J. 9:94-105. McKinlay, K. S., S. A. Brant, P. Morse, and R. Ashford. 1972. Droplet size and phytotoxicity of herbicides. Weed Sci. 20:450-52. Newton, M. and L. A. Norris. 1968. Herbicide residues in blacktail deer from forests treated with 2,4,5-T and atrazine. In: Proc. West. Soc. Weed Sci., pp. 32-34. Norris, L. A., M. L. Montgomery and E. R. Johnson. sistence of 2,4,5-T in a Pacific Northwest forest. 25:417-22. 1977. The perWeed Sci. Plumb, T. R., L. A. Norris, and M. L. Montgomery. 1977. Persistence of 2,4-D and 2,4,5-T in chaparral soil and vegetation. Bull. Environ. Contam. Toxicol. 17:1-8. Ramel, C. 1977. Chlorinated Phenoxy Acids and Their Dioxins-Mode of Action, Health Risks and Environmental Effects. Ecological Bulletins/NFR27. Stockholm, Royal Swedish Academy of Sciences. Seymour, K. 6. 1969. Evaluation of Spray Drift Potential in Pesticidal Formulations Research. In: Advances in Chemistry Series 86, Washington, American Chemical Society. Smith, G. E. and B. G. Isom. 1967. Investigation of effects of large scale applications of 2,4-D on aquatic fauna and water quality. Pesticides Monitoring Ji. 1:16-21. Stewart, R. E. and H. Gratkowski. 1976. Aerial application equipment for herbicide drift control reduction. U.S. Dep. Agric. For. Serv. Gen. Tech. Rep. PNW-54 Pac. Northwest For. Range Exp. Stn. Portland, Oregon. Turner, D. J. 1977. The Safety of the Herbicides 2,4-D and 2,4,5-T. Forestry. Forestry Commission Bulletin 57. London, H. M. Stationery Office. U.S.D.A. Technical Bulletin Number 1671, 1982. Economic Assessment of 2.4,5-T. The Biological and U.S. Environmental Protection Agency. 1978. 43 Federal Register Excerpts from Position Document-1 2,4,5-T Working Group 17128, 17139-17141. Walker, Allan and Allan E. Smith. 1979. Persistence of 2,4,5-T in a heavy clay soil. Pesticide Sci. 10:151-57. 2-37 �Warren, G. F. 1954. Rate of leaching and breakdown of several herbicides in different soils. North Central Weed Control Conf. 11:5-6. Wiersma, G. B., H. Tai, and P. F. Sand. 1972. Pesticide residue levels in soils, FY 1969 National Soils Monitoring Program. II. Pesticides Monitoring J. 6:194-228. Young, A. 1., E. L. Arnold, and A. M. Wachinski. 1974. Field studies on the soil persistence and movement of 2,4-D, 2,4,5-T and TCDD. Abstr. No. 226, Weed Sci. Soc. Am. Young, A. I , C. E. Thalken, W. E. Ward, and W. J. Cairney. 1974. . The ecological consequences of massive quantities of 2,4-D and 2,4,5-T herbicides. Summary of a five-year field study. Abstr. No. 164, Weed Sci. Soc. Am. 2-38 �CHAPTER 4 Pharmacokinetic Studies of the Phenoxy Herbicides in Humans and Animals Toxicity can be defined as the ability of a material to cause injury by other than physical (mechanical) means. It can be deter- mined by administering known doses of a test substance to laboratory animals. A relatively simple test for acute (single dose) toxicity is the determination of a substances' acute oral median lethal dose LDsg. Such an LDso (mg/kg) is defined as the single dosage of milligrams per kilogram body weight required to kill 50& of a test population of animals when given by mouth. The lower the number, the greater the toxi- city. Most commonly used pesticides have some degree of toxicity. Table 4-1 provides information on toxicity, terminology, LDso's and estimates for probable lethal dose for a 150 Ib man if exposed to a toxic material. The toxicity ratings in this table will be used to evaluate pesticide toxicity presented in Table 4-2. exposed to concentrated pesticides. Humans are seldom Batchmen mixing chemicals for a spray operation or those in manufacturing plants are most likely to contact the concentrates. Some accidents and suicides have occurred. 4-1 �Table 4-1. dose. Ratings of toxicity, LDso, and estimates of man's lethal (Cleason, et al. 1957). Probable lethal dose for Toxicity rating 2 LDso (mg/kg) 1 _ 150 Ib man3 Extremely toxic less than 5 a taste (less than 7 drops) Very toxic 5-49 7 drops to 1 teaspoonfull Moderately toxic 50-499 1 teaspoon to 1 ounce Slightly toxic 500-4,999 1 ounce to 1 pint (1 pound) Almost non-toxic 5,000-14,999 1 pint to 1 quart Non-toxic 15,000 and above more than 1 quart ^Toxicity rating is from Clinical Toxicology of Commercial Products by Cleason, M.N. , Gosselin, R.E. and Hodge, H.D., and published by Williams and Wilkins Co., Baltimore, MD, 1957. ^These doses refer to a 100% concentrate. Many pesticides are con- tained in solvents which dilute the active ingredient. the solvent may be more toxic than the pesticide.) 4-2 (In some cases �Table 4-2. Relative Toxicity of Some Pesticides to Mammals. (Virginia Pest Management Guide, 1980). Trade Common name LDso1 name Toxicity rating (mg/kg) Herbicides Alachlor Lasso 1,800 slightly toxic Ami no triazole Amitrole-T 5,000 almost non-toxic ASPIRIN (for comparison 750 slightly toxic Atrazine Aatrex 3,080 slightly toxic Bromacil Hyvar-X 5,200 almost non-toxic 4,659-5,431 Butyl ate slightly toxic slightly toxic Chloramben Ami ben 3,500 Cyanazine Bladex 334 Dalapon Dowpon Dicamba Banvel-D 1,028 slightly toxic Dinitramine Cobex 3,700 slightly toxic Dinoseb DNBP Diquat 7,570-9,330 moderately toxic almost non-toxic 58 moderately toxic Diquat 230 moderately toxic Diuron Karmex 3,400 slightly toxic Fluometuron Cotoran 1,550 slightly toxic Glyphosate Round-up 4,320 slightly toxic Linuron Lorox 1,500 slightly toxic MSMA MSMA 1,800 slightly toxic Metolachlor Dual 4,286 slightly toxic Metribuzin Sencor 1,090-1,206 slightly toxic 4-3 �Table 4-2 (cont.) Trade Common name name LD501 Toxicity rating (mg/kg) Monuron Telvar 2,300-3,700 Oxadiazon Ronstar 30,000 Paraquat Paraquat Pencil me thai in Prowl 3,380 slightly toxic Picloram Tordon 8,200 almost non-toxic Profluralin Tolban 2,200 slightly toxic Propachlor Ramrod 710 slightly toxic Si 1 vex Weed-B-Gon 1,070 slightly toxic Simazine Princep 5,000 almost non-toxic Sodium Chlorate Tumbleaf 5,000 almost non-toxic 3,320 slightly toxic 120 slightly toxic non-toxic moderately toxic Table salt (for comparison) Terbacil Sinbar 5,000-7,500 Trifluralin Tref Ian 3,700 2,4-0 2,4-D 2,4-DB 2,4-DB 2,4,5-T 2,4,5-T Hexazinone Velpar 16,900 non-toxic Acephate Orthene 866-945 slightly toxic Aldicarb Temik 1.0 extremely toxic Azinphosmethyl Guthion 13 very toxic 300-1,000 1,960 300 almost non-toxic slightly toxic moderately toxic slightly toxic moderately toxic Insecticides 4-4 �Table 4-2 (cont.) Trade Common name name LDSQ! Toxicity rating (mg/kg) Carbaryl Sevin Carbofuran Furadan Chlordane Ortho-Klor 850 10 335 slightly toxic very toxic moderately toxic Belt Chlordimeform Fundal, 170-220 moderately toxic 97-276 moderately toxic Galecron Chlorpyrifos Dursban, D erne ton Systox 12 DDT Diazinon 113-118 very toxic moderately toxic 108 moderately toxic 80 moderately toxic 215 AG 500 moderately toxic Spectracide Dichlorvos Vapona, DDVP Dimethoate Cygon, DeFend Disulfoton Di-Syston 12 Ethylene dibromide Dowfume 108-70 moderately toxic Fenthion Baytex, 313 moderately toxic 88 moderately toxic very toxic Tiguvon Entex Lindane Malathion Cythion 1,375 s l i g h t l y toxic Methiacarb Mesurol 130-135 moderately toxic 4-5 �Table 4-2 (cont.) Trade Common name name LD501 Toxicity rating (mg/kg) Methomyl 17-24 very toxic 14 Lannate, very toxic Nudrin Methyl Parathion Penncap-M 480 (Microencapsulated MP) moderately toxic Parathion 13 very toxic Pentachlorophenol 27 very toxic Phorate Thimet 2.3 200 Pyrethrum Ronnel Korlan, extremely toxic moderately toxic 906-3,025 slightly toxic Trolene, Ronnel 132 moderately toxic 14,000 almost non-toxic 64 almost non-toxic 9,000 almost non-toxic Rotenone Tetradifon Ted ion Fungicides Benomyl Ben!ate Captan Captan Fensulfothion Dasanit 2-10 extremely toxic Dichlone Dichlone 1,300 slightly toxic Captafol Difolantan Oxydisulfoton Disyston Perbam Ferbam Carbofuran Furadan Maneb Maneb 6,200 almost non-toxic 3.5 extremely toxic 17,000 11 6,750 4-6 non-toxic very toxic almost non-toxic �Table 4-2 (cont.) Trade Common name name LD5Q1 Toxicity rating Captan Orthocide 9,000 almost non-toxic Thiram Thiram Zineb Zineb 780 5,200 slightly toxic almost non-toxic (mg/kg) - The dosage of milligrams per kilogram of body weight required to kill 50% of test animals when given a single dosage by mouth. From: Herbicide Handbook of the Weed Science Society of America, 4th Edition, 1979. A list of several pesticides grouped as families of herbicides, insecticides, or fungicides is presented in Table 4-2. This table suggests that as a family, the toxicity of herbicides presents a minimal risk to human health. Phenoxy herbicides have been specifically designed to kill broadleaf weeds; as such it is not surprising that they have relatively low capability to cause harm to animals (Vaccaro, 1983). Insecticides as a family have a higher potential for harming higher animals. It is interesting to look at information from the Poison Control Center on reported cases of poisoning. In 1979, there were over 144,000 reported cases of poisoning with 70 fatalities. 4-7 �None of the 70 fatalities were due to herbicides. Of the 144,000 reported cases, 5% were attributed to all pesticides and 0.2% to herbicides. The LDso values as listed in Table 4-2 represent acute doses or the amount that an animal can tolerate if all was administered in one dose. These values from test animals are of real assistance when addressing poisoning or attempted suicide cases in humans. Since men are seldom used as guinea pigs in our society to study potential longterm (chronic) effects of repeated exposure to low levels of pesticides, animal feeding experiments are conducted. In these studies several different concentrations of pesticides are routinely incorporated into all of the food ingested by the test animal. The duration of feeding periods may range from 1 week to more than a year for many of the tests. The studies are required to include concentrations that are sufficiently high to at least cause injury to some of the test animals. No observed effect levels (NOELs) are determined by citing the highest pesticide level at which no harmful effects are detected on the test animals. It is relatively easy for the results of this type of study to be misinterpreted since often the information is not presented in a manner that one can easily understand regarding how much of the chemical was required to cause the effect. Information derived from studies of this kind are of importance to man as he seeks to make judgements regarding any exposure to pesticides that he may receive. Conducting a well designed research study is not as difficult as effectively communicating the findings to an interested audience of 4-8 �varying backgrounds. Results from the applicator exposure studies have shown that a portion of the phenoxy herbicides as they are being applied as vegetation control do not reach their target. In fact, at times humans may be inadvertent receptors resulting in an "absorbed dose." The goal of this chapter is to investigate the fate of the phenoxy compounds after they have contacted humans, entering the body by penetrating the skin (the most common route) via the food and water consumed or in the air breathed. The most commonly used phenoxy compounds 2,4-D and 2,4,5-T will be discussed. Since the LDso for 2,4-D ana 2,4,5-T is "upwards of 300 mg/kg" for several animal species (CAST, 1978) it is highly likely that these compounds have some biological activity for humans also. The mere fact that a compound has some moderately toxic properties does not make it dangerous unless one receives an absorbed dose and even then the amount of the dose is the important factor. A philosopher of the Middle Ages, Paracelsus, said "All substances are poison; there is none which is not a poison. The right dose differentiates a poison and a remedy." If the full understanding of the meaning of dose could be clearly commmunicated an informed U.S. public would have less trouble sleeping at night. The user needs to be aware of the toxicity of the compound he is using. As shown in Table 4-2 pesticides vary greatly in their toxicity to man and other animals. The phenoxy compounds have an excellent human health record when considering their widespread use and the lack of complaints or health problems arising from pesticide applicators, the most exposed group. 4-9 �What protective mechanism(s) does man possess which allows him to be in contact with these compounds, sometimes over an extended period, without them causing him noticeable physical harm? To gain a better understanding of how his body is able to excrete or rid itself of these compounds an evaluation of animal feeding studies with the phenoxy compounds is in order. An indication of the extent of animal toxicological testing with the phenoxy compounds is obtained from Diaz-Col en et al. (1977) who indicate that more than 870 such studies have been completed in the past 25 years. In one study residues of the 2,4-D and its breakdown products were determined in muscle, fat, liver and kidney of cattle and sheep fed up to 2000 mg 2,4-D per kg in their feed. Muscle and fat contained the lowest levels while kidneys and liver contained the highest residue level. Withdrawal from treatment for one week resulted in a significant reduction in tissue residue levels. With the exception of the kidneys, 2,4-D residues averages less than I mg/kg in the tissue analyzed. The kidney tissue level averaged 7.82 mg/kg with 0.37 mg/kg present after a 7-day withdrawal period. No 2,4-D was detected in fat or muscle of any animals at a detection limit of 0.05 mg/kg. (Lack of build-up of the phenoxy herbicides in fatty tissues of animals is common; this is in sharp contrast to some of the more persistent chlorinated insecticides which are stored in fatty tissues of mammals.) All treated animals in this study showed some loss of apetite, weight loss or poor weight gain depending on the level 2,4-D present in the feed. During the 7-day withdrawal period, feed consumption in all groups returned to normal (Clark et al., 4-10 �1975). Cattle, which were fed 50 or 100 mg/kg/day phenoxy herbicide (2,4,5-T or Silvex) for as long as 1 year, suffered no ill effects. Livestock can safely eat vegetation freshly treated with phenoxy herbicides, however, milk or meat fromm these animals could contain trace levels of residue. It is recommended that 2 weeks elapse between phe- noxy treatment of vegetation and slaughter of animals grazing the treated area (Leng, 1972). One of the short-comings of some animal testing data as described by the Advisory Panel on Toxic Substances (Beljan et al., 1981) states "One of the common failings of animal toxicity assessments is the use of test doses that are much higher than those that may be encountered by humans. Such massive doses are intended to shorten the time for a response or, in the case of a carcinogenicity test, to increase the probable incidence of tumors within a sample of animals." However, much valuable information can be obtained from animal studies when realistic rates are selected proper A two-year study by Smith et al, (1978) of chronic toxicity and incidence of cancer among rats inge ting diets containing 2,4,5-T revealed few toxicological symptoms (loss of body weight and slight changes in the size or shape of kidneys, livers, and lungs) even at the highest does level (30 mg 2,4,5-T/kg body weight/day). This study also showed no incidence of cancer response in rats even when administration of 2,4,5-T extended over most of their life span at a dosage high enough to induce other toxicity. As for the effects of 2,4,5-T on reproduction, they found in studying three generations of rats that dose levels of 2,4,5-T high enough to cause signs of toxi- 4-11 �city had no effect on the reproductive capacity of rats, except for a tendency to reduce s u r v i v a l of the young at dose levels of 10 and 30 mg/kg/day. Numerous studies on the fate of 2,4-dichlorophenoxy-acetic acid (2,4-D) and related herbicides in animals have shown that these chemicals are absorbed and excreted unchanged in the u r i n e w i t h i n a week after administration (Leng, 1977). Pharmacokinetic studies with 2,4,5-T in rats and dogs (Piper et al., 1973) supported these f i n d i n g s (Pharmacokinetic studies show us how a chemical is treated by the body and how rapidly it is excreted). Repeated doses of 2,4-D or 2,4,5-T in animals do not cause toxic symptoms since they continually eliminate it. This process is protec- tive since it does not allow the concentration to build-up. The study of how animals eliminate foreign chemicals and the rate of elimination is important in understanding the fate of phenoxy compounds in humans. Fate of phenoxy herbicides in man In an experiment to determine the fate of ingested 2,4,5-T in man, f i v e male volunteers (from 31 to 58 years of age) were each given a single dose of 5 mg/kg of body wt of 2,4,5-T (which contained less than 0.05 ppm of TCDO). Blood concentration of 2,4,5-T reached a maximum in 7 hours, and then declined for the f o l l o w i n g 5 days. W i t h i n 5 days, 89% of the ingested 2,4,5-T was excreted unchanged in the u r i n e (Gehring et al., 1973). From these results the investigators predicted that an i n d i v i d u a l ingesting d a i l y doses of 2,4,5-T would have attained his highest level after 3 days. 4-12 �In related research showing 2,4-D excretion Sauerhoff (1977) showed that 2,4-D was also absorbed almost entirely and eliminated from the body in the urine - 95% within 4 days. With repeated doses, a plateau is reached and the body excretes 2,4-D as rapidly as it enters. This research provides evidence of the fate of phenoxy herbicides that are orally ingested by humans; but, as shown in Chapter 3, forest crew members applying phenoxy herbicides appeared to receive their highest dose when these compounds penetrated their bare skin areas. Personnel associated with the "Operation Ranch Hand" program, which involved use of the phenoxy herbicides as defoliants in Vietnam would probably also have received the majority of their exposure by skin absorption. Studies by Sauerhoff et a!., 1976 show that regardless of the path of entry, rapid urinary elimination of 2,4,5-T occurred in rats. Probability of dermal absorption Is it feasible that significant levels of an organic chemical could pass through man's skin? Baranowska-Dutkiewicz (1982) indicates that under industrial conditions the skin is the main route of aniline absorption. To obtain a measure of the absorbed dose of aniline occurring in 30 or 60 minute exposure a 24-h urine sample was analyzed for p-aminopheriol. They cited other research which showed that ben- zene, styrene, xylene and carbon disulphide were even more readily absorbed by human skin. Hosier et al. (1980) analyzed blood plasma and urine to provide indicators for the amount of 10% lindane cream absorbed by human skin in a 12-hr exposure study. 4-13 Jakobson et al. �(1982) indicates that uptake of organic solvents through the skin may pose a risk during occupational handling of these compounds. Warnings are given that considerable uptake may occur when certain solvents contact human skin. Durham et al. (1972) indicates that although absorption by the respiratory route is faster than through the skin, the opportunities for dermal exposure are often much greater. Studies of the dermal absorption of 2,4-D applied to human forearms revealed that 6% of the applied amount was recovered in urine collected during the week after exposure (Feldman and Mai bach, 1974). The findings of the applicator exposure studies presented in Chapter 3 which showed dermal absorption of phenoxys to be the major route of entry into the body are compatable with the results of the research above. Although the odor near an application site often times is quite pronounced, it is interesting that in both of the detailed phenoxy applicator exposure studies reported here and also in other studies the amount of the compound in the air was minimal on a weight basis compared to the dermally absorbed dose. Young et al. (1981) applied the propylene glycol butyl ether esters (PGBE) of 2,4,5-T to the shaved skin of rats. This is the same formulation of 2,4,5-T used in the 2,4,5-T applicator exposure studies described in detail in Chapter 3. When this 2,4,5-T was absorbed through the skin only 2,4,5-T could be detected in the urine. The authors concluded that the application of PGBE ester to the skin of rats resulted in 2,4,5-T itself in the body. Therefore, exposure to esters of 2,4,5-T can be regarded as 4-14 �toxicologically equivalent to the 2,4,5-T itself. Six days after application to the skin they recovered approximately 97% of the 2,4,5-T in the urine. Samples of either blood or urine from exposed individuals should provide a measure of the absorbed dose since the phenoxy compound must enter the blood stream before it is excreted in urine. Total urine collection allows the researcher to recover almost all of the phenoxy herbicide that was absorbed. Young et al. (1981) also observed that adsorption through skin was much slower than absorption of ingested 2,4,5-T. The low rate of absorption of the ester through the skin suggests that removal of the ester from the skin by washing after exposure could substantially reduce the dose of 2,4,5-T received by this route. The relatively slow dermal absorption and excretion rate agrees with human data derived from the 2,4,5-T exposure study presented in Chapter 3. The extraordinary ability of man's body to rid himself of foreign substances and other toxins is vividly demonstrated as one scans the literature. Although the prudent man will protect himself from unne- cessary exposure to toxins in the environment, it should be reassuring to him to learn that he has several built in safety devices. Perhaps in this age when the media almost monthly reveals to us the toxic effects of one of our foods or some other chemical product that we come into contact with it is reassuring to know that our kidneys or other body systems are attempting, on a full-time basis to remove both known and unknown toxins that we encounter. 4-15 �Factors affecting man's absorption and excretion rates of phenoxy compounds One of the more perplexing questions with regard to interpreting results from pesticide applicator exposure data is why the correlation is so low between the amount of pesticide deposited on patches and that excreted in urine. Data presented earlier in this chapter i n d i cates that phenoxy herbicides are nearly quantitatively excreted in m a n ' s urine and that dermal absorption is the major avenue of entry into his body. Thus, i n t u i t i o n suggests that levels derived from quantitatively measuring the herbicide deposited on m a n ' s skin and the amount excreted in his urine would be h i g h l y correlated. In both the 2,4,5-T and the 2,4-D applicator exposure studies (Lavy et a!., 1980; Lavy et al., 1982) poor correlation for these two measurements was found. What are some of the potential reasons for these findings? Figure 4-1 depicts the typical excretion rate of p-aminophenol for an i n d i v i d u a l dermally exposed for 1-hr to l i q u i d a n i l i n e (Baranowska-Dutkiewicz, 1982). When a single dermal exposure occurs over a 1-hr period for a rapidly absorbed compound a smooth uniform excretion pattern occurs. Figure 3-1 on page illustrated the excretion pattern of 2,4-D by one of the more heavily exposed crewmembers in the 2,4-D exposure study who probably did not receive all of his exposure on the day the spray was applied (Lavy, 1980). Since patches were attached to the worker's clothing only on the application day, and urine was collected over a 6-day period it is not surprising to f i n d poor correlation between patch data and urinary excretion data if exposure was not l i m i t e d to the day of application. As discovered 4-16 �Figure 4-1. Excretion rate of p-aminophenol in urine in a subject exposed to 1% aniline solution for 1 h (Baranowska-Dutkiewicz, 1982). �8 £ LxJ en 6 5 o: g h- 3 LJ CC O p X ^ Ld I 8 10 12 14 HOURS 16 18 20 22 24 �in post-study interviews for some workers, the opportunity for cont i n u i n g exposure occurred as they cleaned equipment. Contaminated clothes could also provide prolonged exposure. Another fact that could be contributing to the low correlation between patch and urine data is the physical location of the patches on the test subject. In the 2,4,5-T study (Lavy, 1978) 6-10 x 10 cm paper backed gauze patches were attached to clothing of the worker at the f o l l o w i n g locations: chest, back, upper arms, and upper thighs. In the 2,4-D study the denim patches were attached to worker's clothing near their bare skin areas. The calculations for predicting dermal exposure to humans are dependent on knowing: a) the area of patches retaining the chemical, b) the total weight of the pesticide deposited on the patch, c) the area of bare skin of the crewmember ( t h i s is obtained by viewing photographs of the crewmember in his work clothes and using a human area formula developed by Durham and Wolfe (1962)). Use of patches and consequent interpretation of data derived from their use is based on assumptions that pesticides 1) are uniformly deposited across all parts of the body; 2) that patches capture and retain the same amount of pesticide that the skin retains ( t h i s could hardly be true if a man were perspiring and he was wearing dry patches) and 3) that all portions of a m a n ' s skin are equally penetrable by the pesticide in use. Data presented by Lavy et al. (1980b) showed that the range in weight of 2,4,5-T on s i m i l a r sized patches on the same i n d i v i d u a l ranged from 20 to 8760 micrograms. The anatomical site of application has been shown to affect the percutaneous absorption of compounds in 4-17 �humans (Maibach et al., 1971). The interaction between skin and the specific pesticide will also be important. Some will adhere strongly to the skin while others will more readily pass through. When all of these factors are taken into consideration it is not surprising that low correlations exists between patch and urinary excretion data in phenoxy exposure studies. Gehring et al. (1973) developed an equation to predict a per day excretion curve for humans given a single dose of phenoxy herbicide. Nash (1983) has expanded on this work and devised an equation which allows one to make an estimate of the absorbed dose which occurred on the day of exposure. He indicates that by knowing the elapsed time between the exposure day and the day a 24-hr urine sample was taken, his equation can provide a good estimate of the extent of absorption which occurred. If functional use of the above approaches could be made it would save the field research crews considerable time and the crew participants inconvenience (as contrasted to collecting every drop of urine over a 6-day period). Unfortunately as shown in Figure 3-1 on page , actual field exposure received by some workers does not occur all on the same day. Thus, if the researcher is looking for measurement of the total phenoxy exposure occurring to all crewmembers who work as a part of the field application process he will continuously monitor the phenoxy content of urine from humans during all of the time exposure is potentially occurring and for at least 5 days following the last potential exposure. Even though the exposure to humans applying the phenoxy herbicides was not all limited to the actual application day it was shown in Chapter 3 that even for the 4-18 �i n d i v i d u a l s m i x i n g the concentrate and receiving m u l t i p l e exposures, natural human body processes are capable of keeping the level of the phenoxy herbicides which penetrate the body well below harmful levels. It is vital that we not overlook the primary f i n d i n g s of this monograph - namely although man experiences some exposure to the phenoxy compounds the detection of this exposure is made possible only through advances in analytical equipment which allow pesticide detection capabilities in the parts per b i l l i o n range. The backpack appli- cators and mixers (those receiving the highest absorbed dose) (Lavy, 1978) did not feel i l l or e x h i b i t any symptoms. The absorbed dose for these highest occupationally exposed i n d i v i d u a l s is 300-fold below the no-effect level. Putting this into another context - if one were dependent on surveys and results from questionnaires of workers who mix and apply the phenoxy herbicides, basically we would come to the conclusion that no exposure was occurring. These compounds have been used extensively for over 30 years with no known i l l effects due to their use when used in accordance to the directions provided. Thus, other than through the use of analytical equipment capable of detecting trace levels of pesticide residues, the only evidence that exposure is occurring is by v i s u a l l y observing that sometimes workers get the spray materials on their skin and that a characteristic odor is often present at spray operation sites. Effects of repeated exposures Questions such as the f o l l o w i n g have not been addressed f u l l y : Do repeated exposure to the phenoxy herbicides pose a threat to the health of people who work as pesticide applicators for several months 4-19 �a year? How high a concentration of the phenoxy herbicides will build up in such applicators? The exposure data presented have addressed the question: exposure does a human receive during a one-time exposure? How much Researchers are aware that normally field workers do not apply phenoxy herbicides just one time a year. However, to obtain good replicated data it is important that each of the crewmembers enter the study on an equal exposure basis. If some had been spraying continuously for a week prior to the study, others two days, and others intermitently for 3 or 4 days it would be nearly impossible to attain meaningful replicated data. If the study had been designed to measure the exposure occurring during and after 4 sequential spray days, how would one handle the data collected following a weather pattern which allowed 2 days of application before a 3-day rainstorm prevented continuation of the research? The logistic problems associated with carrying out the research would be greatly increased and the replications would be destroyed unless all workers were at the same location experiencing the identical work conditions for the same period of time. Another reason for selecting a single exposure day was the fact that single dose exposure data could be extrapolated to produce the same results that would have been attained if an individual had been exposed repeatedly. Taking a scientific approach to answering the repeated exposure question Ramsey et al. (1978) made a pharmacokinetic analysis of the 2,4,5-T exposure study reported in Chapter 3 (Lavy, 1978). They indi- cate that the total amount of 2,4,5-T excreted in the urine following 4-20 �exposure represents a minimum estimate of the amount of 2,4,5-T absorbed, since excretion may not be complete at termination of the experiment. However, calculation of the absorbed dose of 2,4,5-T based on pharmacokinetic analysis of urinary excretion data is not dependent on total excretion and can provide a realistic estimate of the absorbed dose. Ramsey et al. (1978) estimated the amount of 2,4,5-T by the workers using knowledge of 2,4,5-T in humans from controlled experiments Gehring et al. (1973). Experiments in laboratory animals have shown that 2,4,5-T is excreted in the urine at the.same rate regardless of how it enters the body. In other words intravenously and orally administered 2,4,5-T are excreted in the urine at the same rate. Since urinary excretion of 2,4,5-T is independent of the route of exposure (Young et al., 1981; Sauerhoff et al., 1977; Piper et al., 1973) these authors used the data of Gehring et al. (1973) to calculate the dose of 2,4,5-T absorbed by forest workers, even though most of it was absorbed through the skin. Ramsey et al. (1978) took this information and prepared the following model for urinary excretion. S(t) k0i B(t) kie E(t) This is a schematic diagram of the pharmacokinetic model for the absorption of 2,4,5-T or its PGBE ester in humans, followed by urinary excretion of 2,4,5-T acid. S(t) = amount of 2,4,5-T in the body at time t. E(t) = amount of 2,4,5-T excreted in urine at time t, and and k^e = first order rate constants for absorption and excretion of 4-21 �2,4,5-T, respectively. B(t) = amount in body at time t. They took urinary excretion data from Lavy (1978) and employed the model. They indicated The amount of 2,4,5-T in the pre-exposure samples fluctuated widely with no apparent pattern, therefore no background corrections were applied. The workers employed as mixers may have been exposed on the day previous to the application date (since the formulations are usually mixed the day before the actual application), but for purposes of pharmacokinetic analysis they were considered to be exposed only at the date of application. The duration of the application procedures ranged from 55 minutes to approximately 4 hours (average 138 minutes). Since this is a short time span relative to the total duration of the experiments (up to seven days), the calculated dose was considered to be a single application to the skin at time zero. Values listed in Table 4-3 represent the percentage of the total absorbed dose of 2,4,5-T which was excreted on each of first seven days following exposure. 4-22 �Table 4-3. Cumulative fraction of the dermally absorbed dose of 2,4,5-T excreted in the urine (Ramsey et a"!., 1978). Days following Cumulative fraction of exposure dose excreted 1 0.1956 2 0.4819 3 0.6974 4 0.8326 5 0..9105 6 0.9532 7 0.9760 Ramsey et al. (1978) commented: The pattern of the daily amount of 2,4,5-T excreted in the urine following exposure is characterized by a maximum on day 2, followed by a steady log-linear decline thereafter. However, an examination of the data shows that, in many cases, there is a significant increase in the amount of 2,4,5-T excreted after the second day following exposure. These data are inconsistent with the excretion pattern expected from a single exposure (Gehring et al., 1973), and thus, probably indicates subsequent exposure to 2,4,5-T or to its ester after the actual application date. The speculation that such exposure might arise from the use of contaminated clothing or footgear should be verified by further experiments and observations. In each case, these increased amounts of urinary 2,4,5-T have the effect of increasing the calculated dose, and therefore result in maximized estimates of the dose absorbed on the application date. Since the data reported by Lavy (1978) indicate clearly that the respiratory route of exposure to 2,4,5-T is virtually negligible, the major portion of the absorbed dose of 2,4,5-T was the result of dermal exposure to ESTERON 245 herbicide formulations. Methods used to calculate the absorbed dose will reflect the total amounts of 2,4,5-T absorbed by all 4-23 �possible routes. Figure 4-2 is a composite of the urine excretion curves from the 21 workers in the 2,4,5-T exposure study. The above detailed interpretation by Ramsey et al. (1978) evaluated each of the worker's excretion data individually. It is interesting that a relatively smooth pattern exists when the whole group is evaluated on a composite basis. This is in contrast to the rather marked variability from individual to individual. The question of maximum attainable body burden is specifically addressed by Ramsey et al. (197$). The pharmacokinetic model describing the absorption and excretion of 2,4,5-T in humans can also be used to predict the accumulated body burden of 2,4,5-T that would result from repeated daily exposures. The results of this mathematical simulation are shown by the solid line in Figure 4-3. These simulated data predict that the maximum accumulated body burden of 2,4,5-T resulting from repeated daily exposures would be 1.4x the daily dose (D0). In other words, if a worker absorbed a dose of 0.05 mg/kg each day, the maximum body burden attained would be 0.07 mg/kg and this maximum would be reached after approximately 7 daily exposures. However, if the 2,4,5-T remaining to be absorbed were removed 6 hr after each exposure (e.g., by washing or changing clothing), the predicted accumulated body burden would be represented by the dotted line in Figure 4-3. In this case, the maximum body burden would be 0.3x the daily dose D0, and this maximum would be reached after approximately 3 daily exposures. In summary, the amount of 2,4,5-T absorbed by forest workers during the application of ESTERON® 245 has been shown to be generally less than 0.1 mg 2,4,5-T per kg of body weight. Since this dose level is far below the no effect level of 20 mg/kg for fetotoxic or teratogenic effects cited by EPA (1978), we conclude that under these conditions the absorption of 2,4,5-T presents a negligible toxic hazard to forest workers. A composite of the urinary excretion of 2,4-D forestry workers is presented in Figure 4-4. Ramsey et al. (1980) attempted to use this information on their pharmacokinetic model: 4-24 �Figure 4-2. Average daily excretion of 2,4,5-T in urine of forest workers applying 2,4,5-T (backpack, mist blower, and helicopter methods)(Lavy, 1978). �MILLIGRAMS 2,4,5-T EXCRETED IN URINE Avg. of 21 workers 0.900 0.600 I ro EU 1 0.300 I I 5 7 DAYS * 9 II SPRAY DATE �Figure 4-3. Simulated body burden of 2,4,5-T following 13 repeated daily exposures in units of the daily dose D0 (Lavy, 1978). �1.0 0.9 0.8 0.7 IN3 -p. tr Body Multiple Exposures 0.6 Burden 05 of 2,4-D 04 (Do) 03 0.2 O.I 0 Single Exposure 6 8 10 12 14 16 No. of Exposures = No. of Days �Figure 4-4. Average daily excretion of 2,4-D in urine of forestry workers applying 2,4-D by helicopter (Lavy, 1980). �MILLIGRAMS 2,4-D EXCRETED IN URINE 0.150 -p> ro o Avg. of 18 workers 0.100 .050 I 6 8 DAYS 10 12 I 14 * SPRAY DATE �In contrast to the 2,4,5-T study cited above, the levels of urinary 2,4-D in the present study could not be measured for a sufficient length of time (because the exposure levels were so low that the concentration of urinary 2,4-D was often below the analytical detection limit) to allow direct determination of the first order rate constant for the dermal absorption of 2,4-D esters in these workers. However, because of the close similarity between the pharmacokinetic profiles of 2,4-D and 2,4,5-T, the rate of dermal absorption of 2,4-D esters in humans may be similar to that for 2,4,5-T esters. This assumption is supported by the observation that the half-life values for the dermal absorption of the PGBE esters of 2,4-D and 2,4,5-T in rats are similar, approximately 20 hours and 24 hours, respectively (Young et al., 1981). In order to calculate the pharmacokinetic fate of dermally absorbed 2,4-D in humans, numerical values for the rate constants of absorption and elimination are required. Based on the considerations above, the dermal absorption rate constant of 2,4-D esters in humans was assigned the same value as that for 2,4,5-T esters (half-life = 18 hours) (Ramsey et al., 1980). The urinary excretion rate constant of 2,4-D and conjugates was assigned the value previously determined in a human study (half-life = 11 hours) (Sauerhoff et al., 1977). These values then allowed the calculation of the amount of dermally absorbed 2,4-D that is excreted in the urine at any time following exposure (Ramsey et al., 1979). This calculation reveals that approximately 99% of a dermally absorbed dose of 2,4-D esters would be excreted in the urine as 2,4-D during the exposure day and for 5 days thereafter. These values for the rate constants were also used to obtain a computer simulation of the body concentration of 2,4-D that would result from repeated daily dermal exposure to 2,4-D esters. The results of prediction are shown in Figure 4-5 in which the body burden of 2,4-D is given in units of the dose that has been absorbed thru the skin. This simulation shows that the maximum body burden attained 4-25 �Figure 4-5. Simulated body burden of 2,4-D in humans after a single dermal exposure and after repeated (daily) dermal exposures. The body burden is expressed in units of the dose level applied to the skin (D0) (Lavy, 1980). �2.8 2.4 2 Body -° Burden of2,4,5-T ^ PO en (Do) Continuously exposed — Washed off after 6 hours �would be approximately 0.7 times the daily dose. In other words, if a dose of 0.050 mg 2,4-D per kg were absorbed through the skin each day the maximum concentration attained in the body after repeated daily exposures would be only 0.035 mg/kg. This bioaccumulation factor is less than might be expected (as calculated based on 1 dose per day) because the absorption rate constant in the pharmacokinetic model is less than the elimination rate constant (i.e., a flip-flop model). Although the primary emphasis of this monograph is devoted to human exposure to phenoxy herbicides an additional point of concern when 2,4,5-T, is discussed, is that of the trace contaminant 2,3,7,8-tetrachloro-2-dibenzo-dioxin (TCDD). Relative toxicities of the phenoxy herbicides and TCDD Considerable confusion exists in the minds of most people with respect to the relative dangers of 2,4,5-T and the 2,3,7,8 isomer of TCDO. In a 2,4,5-T formulation it is not possible for a person, ani- mal, or plant to receive a lethal dose of TCDD before receiving a much much larger and correspondingly lethal dose of 2,4,5-T. The following discussion provides an explanation why people who come into contact with 2,4,5-T are not harmed by the more toxic TCDD. Fisher (1977) reported that the average amount of TCDD in 2,4,5-T is about 0.01 ppm. To compare the toxicities of the two compounds, assume that the NOEL of 2,4,5-T for humans is 20 mg/kg and for TCDD is 0.001 mg/kg (Tschirley, 1979). Dividing the 2,4,5-T NOEL by 0.001 (the TCDD NOEL) gives a value of 20,000. 20,000 times greater than that of 2,4,5-T. The toxicity of TCDD is thus Since 2,4,5-T is more con- centrated than TCDD by a factor of 100 million, however, the effective 4-26 �toxicity can be calculated by d i v i d i n g 100 m i l l i o n by 20,000, which shows that in a 2,4,5-T mixture containing 0.01 ppm of TCDD the 2,4,5-T toxicity is potentially 5,000 times greater than that of TCDD. This same sequence of calculations w i l l be used for evaluating the effective toxicity of 2,4,5-T and TCDD in Agent Orange assuming that Agent Orange contained an average of 2 ppm (Young et al., 1978). For Agent Orange containing 2 ppm TCDD the calculations would be as follows: the composition of Agent Orange is essentially a 50-50 mix- ture of 2,4-D and 2,4,5-T (Young et al., 1978); the no-effect-level for 2,4-D is 24 mg/kg ( H a l l , 1980) and the no-effect level for 2,4,5-T is 20 mg/kg (EPA, 1978); since Agent Orange is a 50-50 mixture of these compounds an average no-effect-level of 22 mg/kg will be used. The herbicide: TCDD ratio would be 1,000,000: or 500,000:1; calcul a t i n g a new toxicity relationship based on the 22 mg/kg no-effectlevel for the herbicide and the 0.001 mg/kg level for TCDD reveals that pure TCDD is 22,000 times as toxic as the herbicides in Agent Orange. Since, there is 500,000 times more herbicide present in Agent Orange the relative toxicity of the mixture was: Relative. toxicity 1,000.000 concentration of herbicide in Agent Orange 2 ' concentration of TCDD in Agent Orange 22 mg/kg HerbicidelOEL 0.001 mg/kg TCDD NOEL 9? 7 "*' Thus, because of the large preponderance of the herbicides in the m i x ture (99.9998% Herbicide and 0.0002% TCDD) the relative toxicity from the herbicides would be 22.7 times greater than the effects from the TCDD. Values for NOEL'S are based on chronic (long-term) feeding trials; 4-27 �some personnel in Vietnam may have had only occasional exposure to low levels of Agent Orange while others may have been exposed for longer periods to higher concentrations. Regardless of the source or dura- tion of exposure, pharmacokinetic studies have shown that the phenoxy compounds are rapidly excreted from the body, and do not build-up to harmful levels. In rat studies TCDD has been shown to be primarily excreted in the feces with 50% eliminated the first day (Piper et al., 1973; Allen et al., 1976). The following studies present information of vital importance to our evaluation of two groups of people who have had some phenoxy exposure. If the assumptions used in the above calculations are valid and if Vietnam veterans have experienced health effects that can be attributed to their exposure to Agent Orange these data indicate that the defoliants themselves were more to blame than TCDD. Ranch Hand personnel The group of U.S. military personnel directly involved with transporting, mixing, and applying the herbicides used as defoliants in South Vietnam was referred to as "Ranch Hands." This group of Vietnam veterans probably received more exposure to Agent Orange than any other group since they handled the concentrated materials. Mechanics and pilots associated with spray missions may also have received an absorbed dose of the herbicides. By studying the health and mortality records of the Ranch Hands and workers in the U.S. who have been, involved in the commercial application of the phenoxy herbicides for 10 years or more we have a large data base from which to 4-28 �draw valid comparisons. A follow-up study on the death rate of ranch hands has been conducted by the Air Force. Of the 1170 individuals who participated in the study (this represents 97% of the total number involved in the spray operation in South Vietnam) no statistically significant difference in the peacetime death rate have found between this group and the comparison group, which was made up of similarly aged American men who served in the armed forces but not in South Vietnam. Both groups / appear to have experienced s i g n i f i c a n t l y less mortality than a similarly aged U.S. white male population ( A i r Force, 1982). Domestic aerial applicators. In 1979 the National Agricultural Aviation Association conducted a study to evaluate the effects of exposure on aerial pesticide applicators. It was designed to deter- mine if continuous or long-term exposure to pesticides had resulted in increased incidences of miscarriage and s t i l l b i r t h in the applicators' families. The study was based on the assumptions that absolute safety exists, i.e., lack of hazard cannot be determined w i t h regard to any chemical in its interactions with Homo sapiens or lower species; there is a dose/response relationship with regard to any chemical and its adverse or beneficial effects since i n d i v i d u a l s or populations receiving higher doses (exposures) w i l l show more effect than comparable i n d i v i d u a l s or populations w i t h lower doses or exposures; and i n d i v i d u a l s or populations who apply pesticides are more exposed as a result of their occupation than the general population. F a m i l i e s of National Agricultural A v i a t i o n Association members and a s i m i l a r p o p u l a t i o n of s i b l i n g s who had had no occupational exposure 4-29 �to pesticides were compared. Questionnaires were developed and distributed to members all across the nation. The results listed in Table 4-4 show that no significant differenTable 4-4. Comparisons among live births, miscarriages and stillbirths between the two basic populations. (Modified from WAA, 1980.) Non-Agricultural Agricultural Aviation Aviation Families Families (no.) Live Births (%) (no.) (*) 373 (92) 360 (88) 29 ( 8) 47 (12) Miscarriages & Stillbirths ces in live births and miscarriages were found among the "exposed families" and the unexposed families. Information on birth defects as related to parents occupation is quite rare. The occurrence of birth defects has neither substantially decreased or increased since pesticides were introduced suggesting that few if any widespread and powerful new teratogens were introduced (WAA, 1980). In studies where hundreds of pesticide applicators have been compound to non-pesticide users no evidence has been found which would indicate that the use of phenoxy herbicides has harmed man. Using present day methods of statistical evaluation there is no evidence of 4-30 �CHAPTER 4 References Air Force. 1982. Health Study - Update, Presented to Veteran's Administration Advisory Committee on Health-Related Effects of Herbicides. Washington, November 30, 1982. Allen, J. R. et al. 1975. Tissue distribution, excretion and biological effect of MC tetrachlorodibenzo-£-dioxin in rats. Food Cosmet. Toxicol. 13:505-15. Baranowska-Dutkiewicz. 1982. Skin absorption of aniline from aqueous solutions in man. Toxicol. Lett. 10:367-72. Beljan, 0. R., N. S. Irey, U. W. Kilgore, K. Kimura, R. R. Suskind, J. J. Vostal, and R. H. Wheater. 1981. The health effects of "Agent Orange" and polychlorinated dioxin contaminants. Technical report. Prepared by the Council of Scientific Affairs. Advisory Panel on Toxic Substances. Chicago, American Medical Association. CAST (Council for Agricultural Science and Technology). 1978. The Phenoxy Herbicides. 2 ed. Ames, Iowa State Univ., 1978. Clark, D. E., J. S. Palmer, H. F. Radeleff, R. D. Crookshank, and F. M. Farr. 1975. Residues of chlorophenoxy acid herbicides and their phenolic metabolites in tissues of sheep and cattle. J_. Agric. Food Chem. 23:573-78. Cleason, M. N., R. 6. Gossehn, and H. D. Hodge. 1957. Clinical Toxicology of Commercial Products. Baltimore, Williams and Wilkins. Diaz-Colon, J. 0. and R. W. Boyey. Selected bibliography of the phenoxy herbicides. 3. Toxicological studies in animals. Bulletin MP-1343. Texas Agricultural Experiment Station, Texas A & M University, College Station, TX. Durham, W. F., H. R. Wolfe, and J. W. Elliott. 1972. Absorption and excretion of parathion by spraymen. Arch. Environ. Health 24:381-87. Durham, W. F. and H. R. Wolfe. 1962. Bull. WHO. 26:75-91. EPA. 1978. Rebuttable presumption against registration and continued registration of products containing 2,4,5-T. 43 Federal Register 17116-57, April 21, 1978. Feldmann, R. J. and H. I. Mai bach. 1974. Percutaneous penetration of some pesticides and herbicides in man. Toxicol. Appl. Pharmacol. 28:126-32. , Fisher, J. R. 1977. Letter-p-G12 In Final E1S Vol. 1. Vegetation management with herbicides. U.S. Department of Agriculture Forest Service. 4-32 �Gehring, P. J., C. 0. Kramer, V. A. Schwetz, J. Q. Rose, and V. K. Rowe. 1973. The fate of 2,4,5-trichlorophenoxyacetic add (2,4,5-T) following oral administration to man. Toxicol. Appl. Pharmocol. 26:352-66. Hall, J. F. 1980. In: T. L. Lavy Project Completion Report to National Forest Products Association, Washington, D.C. Hosier, J., C. Tschanz, C. E. Hignite, and D. L. Azarnoff. 1980. Topical application of lindane cream (Kwell) and antipyrine metabolism. _J. Invest. Dermatol. 74:51-53. Jakobson, I., J. E. Washlberg, B. Holmberg, and 6. Johonsson. 1982. Uptake via the blood and elimination of ten organic solvents following epicutaneous exposure anesthetized guinea pigs. Toxicol. Appl. Pharmacol. 63:181-87. Lavy, T. L., S. J. Shepard, and J. D. Mattice. 1980. Exposure measurements to applicators spraying (2,4,5-trichlorophenoxy)acetic acid. J. Agric. Food Chem. 28:626-30. Lavy, T. L., J. D. Walstad, R. R. Flynn, and J. D. Mattice. 1982. (2,4-dichlorophenoxy)acetic acid exposure received by aerial application crews during forest spray operations. J. Agric. Food Chem. 30:375-81. ~ Lavy, T. L. 1978. Project completion report to National Forest Products Assn. Measurement of 2,4,5-T Exposure of Forest Workers. Washington. Lavy, T. L. 1980. Project completion report to National Forest Products Assn. Determination of 2,4-D Exposure Received by Forestry Applicators. Washington. Lavy, T. L., 0. S. Shepard, and D. C. Bouchard. 1980b. Field worker exposure and helicopter spray pattern of 2,4,5-T. Bull. Environ. Contam. Toxicol. 24:90-96. Leng, M. L. 1972. Residues in milk and meat and safety to livestock from a use of phenoxy herbicides in pasture and rangeland. Down to Earth 28:12-20. Leng, M. L. 1977. Comparative metabolism of phenoxy herbicides in animals in fate of pesticides in large animals. In: 6. W. Ivy and H. W. Dorough (eds.), New York, Academic Press. Pp. 53-76. Maibach, H. I., R. J. Feldmann, T. H. Mil by, and W. F. Serat. 1971. Arch. Environ. Health 23:208-11. Nash, R. G. 1983. Agricultural applicators' exposure to pesticides. American Industrial Hygiene Conference Proceedings 1982, Cincinnati. Pp. 23. 4-33 �Piper, W. N., et al. 1973. In: E. H. Blair (ed.), Chlorodioxins— Origins and Fate Ados. Chem. Ser. 120:85-91. Excretion and tissue distribution of 2,3,7,8-tetrachlorodibenzo-jD-dioxin in the rat. Piper, W. N., J. Q. Rose, M. L. Leng, and P. J. Gehring. 1973. The fate of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) following oral administration to rats and dogs. Toxicol. Appl. Pharmacol. 26:339-51. Ramsey, J. C., T. L. Lavy, and W. H. Braun. 1978. Exposure of forest workers to 2,4,5-T: Calculated Dose Levels in Lavy 1978 Project completion report to National Forest Products Assn. Washington. Ramsey, J. C., F. A. Smith, T. L. Lavy, C. N. Park, and W. H. Braun. 1980. Dose levels of 2,4-D in forest workers. In: T. L. Lavy. 1980 Project Completion Report to National Forest Products Association. Washington. Sauerholf, M. W., W. H, Braun, G. E. Blau, and P. J. Gehring. 1976. The dose-dependent pharmacokinetic profile of 2,4,5-trichlorophenoxy acetic acid following intravenous administration in rats. Toxicol. Appl. Pharmacol. 36:491-501. Sauerhoff, M. U., et al. 1977. The fate of 2,4-dichlorophenoxyacetic acid (2,4-D) following oral administration toman. Toxicology 8:3-11. Smith, F.A., B.A. Schwetz, and K.D. Nitschke. 1976. of 2,3,7,8-tetrachlorodibenzo-p-dioxin in CF-1 mice. Pharmacol. 38:517-523. Teratogenicity Toxicol. Appl. Tschirley, F. H. (ed.). 1979. Scientific dispute resolution conference on 2,4,5-T. Sponsored by the American Farm Bureau Federation, Richmond, Nu Vue Printing Company. Pp. 101. Vaccaro, J. 1983. The toxicity of silviculture herbicides in herbicide safety workshop, Southern Forest Products Assn. New Orleans, Louisiana. Virginia pest management guide 1, 1980. Chemical control of insects, plant diseases and weeds. Virginia Polytechnic Institute and State University. Pp. 317. Weed Science Society of America. Weed Science Society of America. 1979. Herbicide Handbook, 4th. ed., Champaign. Pp. 479. World of Agricultural Aviation (WAA). 1980. Investigation of the possible effects of pesticide exposures on reproductive mortality and morbidity. 7:12-30. Young, A. L., J. A. Calcagni, C. K. Thalken, J. W. Trembloy. 1978. The Toxicology, Environmental Fate, and Human Risk of_ Herbicide Orange and Its Associated Dioxin. USAF OEHL Technical Report 78-92. Aerospace Medical Division (AFSC) Brooks Air Force Base, Texas. Pp. 247. 4-34 �Young, J. D., J. C. Ramsey, and W. H. Braun. 1981. Pharmacokinetics of 2,4,5-T P6BE ester applied dermally to rats. Journal of Toxicology in Environ. Health 8:401-08. 4-35 �CHAPTER 5 Assessment of Harm Due To Phenoxy Herbicide Exposure What research information is needed? The complex nature of the issues surrounding phenoxy herbicides and the TCDO contaminants in 2,4,5-T requires that the concerned interested citizen use common sense as he tries to sort out the facts and keep an open mind to new ideas and new research information. Seemingly, a countless number of studies and volumes of writings have already been devoted to learning about the phenoxy herbicides and their effects on man. Should more be done? As long as pertinent questions remain, scientists have an obligation to attempt to answer them. However, it is prudent to conduct only the studies for which there has been no clear cut answer and those that would appear to have attainable answers. Some areas where additional research would be beneficial include: a) Determination of the bioavailability of TCDD from soil. Use of different solvent systems and other factors which could potentially enhance bioavai1 ability or degradation of TCDD should be fully investigated. Work conducted to date indicates that the low water solubility of the compound makes it nearly inert in a soil environment. The Center for Disease Control (CDC) has indicated that 1 ppb TCDD in soil represents a threshold value for man's safety. Definitive new studies should be conducted which clearly demonstrate this threshold value. These new data could strengthen the CDC position. The TCDD bioavailability issue is important due to the toxicity of the compound, the widespread dispersal of TCDD tainted oil in 5-1 �Missouri, and the fact that several sources of TCDD in the environment have been documented. These sources include: fly ash from municipal incinerators (01ie, 1977); home fireplace soot, commercial sludge fertilizer and diesel truck mufflers (Bumb et al., 1980). b) Origin and fate of dioxins in the environment. It has been shown that production of herbicides and other commercial chlorinated phenoxy compounds is not the only source of TCDD. If major sources of these toxic compounds exist in our environment we need to become more aware of them and take the proper precautions. c) Assessing the bioaccumulation of TCDD in animals living in the areas of known TCDD contamination. Experiments could be patterned after those of Newton and Snyder (1978) who sacrificed a few test animals which have had most opportunity for TCDD exposure. By analyzing livers of animals living in the Times Beach area it should be possible to determine if a TCDD build-up is occurring. Animals living in or near this TCDD contaminated area should accumulate higher levels than humans. d) A systematic collection and analytical scheme should be initiated which would allow the procurring of fat samples from human cadavors, some known to have had high exposure to phenoxy herbicides and others with no known exposure. Comparison of the extent of build- up of TCDD in human adipose tissue, if any, could then be made. These studies would be similar to, but much more extensive than those reported by Kutz (1981). e) Improved methods for accurately determining rates of degradation of toxic compounds in our environment. Errors in evaluating 5-2 �or calculating disappearance rates of chemicals highly toxic to man could result in decisions harmful to his health. f) Determining source of the toxic components of a pesticide. In the case of mammalian toxicity to pesticides, the effects of the active ingredient should be separated from those of the carrier or other additives. Some organic solvents may possess more toxic qualities than the "active" compound. It is important to be able to iden- tify the source of any toxic effect observed. g) Soft tissue sarcoma. Reports by Hardell (1981), but unsupported by several researchers, attempt to link phenoxy exposure and soft tissue sarcomas in humans. This possibility should be fully evaluated before the issue is put to rest. If a valid correlation exists this issue should be widely publicized and steps taken to remedy the situation, h) Continuing in-depth assessment of the health characteristics of personnel in the Ranch Hand operation and commercial pesticide applicators. Evaluation of health records of these exposed groups offer valuable opportunities to assemble meaningful health hazard information due to the large numbers of personnel involved and the fact that this was the most highly exposed group in Vietnam. i) Adequate testing of compounds being developed to replace the herbicide 2,4,5-T. An extensive amount of time and effort had been devoted to researching nearly all aspects of 2,4,5-T; it is possible that more environmentally troublesome compounds which have been studied far less extensively may be being used as a substitute for the readily decomposable, much-studied 2,4,5-T molecule. 5-3 �What does the majority of scientific information tell us about man's potential for harm from being exposed to phenoxy herbicides? A large preponderance of the scientific publications concerning the possible effects of the phenoxy herbicides and the low levels of TCDD present show that these compounds do not pose a threat to man's health. Following are a few examples of research, in addition to the author's studies on human exposure to phenoxy herbicides in the field which supports the safety of the phenoxy compounds. USDA R.C. Nash Study In a study by R.C. Nash of the United States Department of Agriculture, it was calculated that if a 175 Ib farmer applied 2,4-D 30 days a year for 30 years he would receive and excrete less than one gram (about the weight of one-half of a dime). To put this into perspective this is several hundred times less than the acceptable daily intake (ADI) for a person over that time period (Nash, 1983). (The ADI is the amount of a chemical, based on all the known facts, which can be consumed every day for an individual's entire lifetime with no harm occurring. These values are expressed on a mg chemical per kilogram body weight basis. The ADI is based on the no effect level in the most sensitive animal species or, if the data are available, in man.) Oregon researchers (Newton et a!., 1975) The registered uses of 2,4,5-T in the forest were effective and did not present a significant threat to the quality of the environment. Data from experimental animals indicate that immediately after spraying 2,4,5-T at an accepted treatment rate, a 100 Ib human could absorb all the TCDD on 1 acre of forest and suffer no toxic effect. 5-4 �TCDD is so tightly bound by vegetation and soil that the probability of significant intake by humans is remote. It is calculated that a 1 Ib/acre application of 2,4,5-T containing 0.1 ppm TCDD applied directly to the soil could result in a maximum of 0.1 parts per trillion TCDD in soil. TCDD does not appear in the seeds of plants grown in soil containing even large amounts of TCDD. Isensee et al. (1971), after applying TCDD to a Lakeland sandy loam and growing oats (Avena sativa) and soybeans (Glycine max), concluded that soil uptake of TCDD by plants is highly unlikely. J.M. Witt American Chemical Society Symposium, Seattle 1983 To evaluate man's risks from another approach, a comparison of the health effects of 2,4-D has been made with aspirin. A man swallowing 2 tablets of aspirin every 4 hours has a safety factor of 25; i.e. if he consumed 50 tablets in that time period, death is expected. By comparison, Witt indicates an individual who was directly under a spray boom applying 2 Ib of 2,4-D per acre has a safety factor of 625. He would need to be sprayed 625 times on the same day before he could receive sufficient 2,4-D to cause death. Data on birth defects for these compounds show a similar pattern. Following is a calculation for the risk of birth defects to the offspring of pregnant female sprayed with 2,4-D as prepared by Witt (1984). 1) A female standing directly under a spray plane applying 2 Ib 2,4-D per acre would receive an exposure level of 20 mg 2,4-D per square foot of body area. 2) 5% of the exposure would be absorbed through her skin. 3) the half-life of 2,4-D: on plants, litter or soil 2 weeks to 2 5-5 �months; in water - 1 week; in mammals 1 day. 4) pregnant females have approximately a 20 sq ft. skin surface. The 2,4-D risk of birth defects for a pregnant female who was sprayed with a spray-mix of 2 Ib/acre (=20 mg/ft 2 ); 50% of her skin would be exposed (10 ft 2 ); weight 110 Ibs (=50 kg). Calculations: 20 mg/ft 2 x 10 ft 2 (exposure level) (50% of skin area exposed) 50 kg (body weight) 200 mg = 4 mg/kg on bare skin areas x 0.05 percent absorbed 50 kg = 0 . 2 mg/kg absorbed dose 20 mg/kg/dayl = No effect 1e,ve1 = 100 = margin of safety 0.2 mg/kg = absorbed dose These data indicate the pregnant female could be sprayed 100 times daily before she would receive an absorbed dose of 2,4-D which is equivalent to the amount of the "no effect level" for birth defects. It is highly unlikely that anyone would be sprayed 100 times in one day. United Kingdoms; Review of the safety of the herbicide 2.4.5-T Following are conclusions from a comprehensive evaluation of the phenoxy-TCDD question by a major research team in the United Kingdom: (Kilpatrick et al., 1980) Given the acute toxic effects of TCDD any significant exposure to this chemical would have been reflected in the health of operators. Against this background of data on exposure to 2,4,5-T and bearing in mind the low levels of TCDD present in l$ince Witt (personal communication) believes the no effect level for 2,4-D is between 20 and 25 mg/kg/day he has chosen to use the more conservative value for these calculations. 5-6 �is M three surveys „„ ,B ,Molved 1100> a, 1970, 1975 Md "76, ^ revealecj any disorders. ted ,in9dom for «.r §i to t, ^ . use of 2.4.5-. wy TCDO eo. of 2>4>5 . T 1983) iysle«1 c o TO n 1980; TS , sen,nwe ^cator of ° ; ::1- ^ "" reactions to reactions are rare, TCOO exposure has 1 - - - ,»» -- -• - - * 5-7 �Relative risks of spraying 2.4,5-T An oncogenicity (tumor causing) study reported by Kociba in 1978 showed that TCDO caused cancer in rats. Assuming the effective rate of 0.1 ug/kg body weight per day is comparable to that of humans, it is possible to calculate a risk factor for workers applying 2,4,5-T containing (assuming worst case) 0.1 ppm TCDD. Leng (1978) discussed the "one hit" model which assumes that pro- bability of tumor production is random and proportional to the total chemical contact. Under worst-case conditions for a female applicator weighing 60 kg applying 2,4,5-T containing 0.1 ppm TCDD, 5 days per week, 4 months per year, for 30 years, the average lifetime dose of TCDD would be one millionth of a microgram per kilogram of body weight per day (1 x 10~& ug/kg/day). Based on a 99% upper confidence limit for the incidence of tumors in female rats given 0.1 ug/kg/day for 2 years in the Kociba study, and adjustment for differences in body surface area (more conservative than relative body weights), and in lifespan (70 vs 2 years) for humans compared to rats, she would have a calculated potential risk of 3.05 x 10"5. Male applicators exposed in the same way would have a risk only half as large since male rats in Kociba's study had a lower incidence of tumors. Table 5-1 indicates that cancer risk to a male smoker is 6000 times greater than to a backpack sprayer applying 2,4,5-T. The relative risk of a worker developing tumors from exposure to the TCDD in 2,4,5-T is far less than for people who drink water in Miami or New Orleans, or even fly cross-country once a year. 5-8 �Table 5-1. Relative Annual Risk of Cancer Per Million People from Exposure to Various Causative Agents (Adapted from Wilson, I960 and Leng, 1978). R i s k x 106 per year TCDD from Application of 2,4,5-T Female backpack sprayer • Male backpack sprayer 0.4 0.2 Eating and Drinking One diet soda per day (saccharin) 10 Average U.S. consumption of saccharin 2 Peanut butter, 4 tablespoons per day (aflatoxin) 40 Milk, 1 pint per day or cheese equiv. (aflatoxin) 10 Drinking water in Miami or New Orleans 1.2 Steak, 1/2 Ib charcoal-broiled, 1 per week 0.4 Alcohol, light drinker (1 beer per day) 20 Alcohol in smokers vs. non-smoker 50 Tobacco Smoker 1200 Being in room with a smoker 10 Concluding remarks The two largest data bases available for evaluating man's exposure to the phenoxy compounds are 1) the health records of hundreds of pesticide applicators, and 2) the health records of the Ranch Hand 5-9 �personnel who were active in the transport, mixing and application of Agent Orange in Vietnam. By continuing to monitor the health of these groups we are studying those individuals who have received the highest dose of phenoxy herbicides. If adverse health effects are caused by exposure to these chemicals the evidence should become apparent when the health records of these groups are compared to those of a nonexposed population. The amount of phenoxy herbicide present in man's air, food and water has been shown to be very low. Field workers who mix and apply these compounds receive a higher absorbed dose than others in our society with the possible exception of workers in phenoxy manufacturing plants. Some workers at these plants using chlorinated phenoxy compounds where explosions have occurred have received higher doses. Although the dose of herbicides received by those transporting, mixing, or applying these chemicals have not appeared to reach health threatening levels, those interested in reducing their exposure levels may easily do so by following simple suggestions listed in Chapter 3. It is recognized that some who have served in Vietnam have later suffered health problems. While probing literature in preparation for this monograph no scientific evidence has been found which would link those health problems to phenoxy or TCDD exposure. To answer questions on health related problems other avenues should be pursued. The ultimate answers to these problems are probably more complex than attributing them to phenoxy herbicide exposure. After reviewing hundreds of research articles on the pros and cons of 2,4,5-T, scientific evidence strongly supports the safety of this 5-10 �compound to man and the environment. Findings from agriculture show that in over 30 years of use no known health problems have been found when this herbicide is used according to label instructions. 5-11 �Bibliography Bradley, E. 1983. Landmark court decision rules in favor of 2,4-D and 2,4,5-T herbicides. In: News Release, National Coalition for a Reasonable 2,4-D Policy. Bumb, R.R., W.B. Crummett, S.S. Cutie, J.R. Gledhill, R.H. Hummel, R.O. Kagel, L.L. Lamparski, E.V. Luoma, D.L. Miller, T.J. Nestrick, L.A. Shadoff, R.H. Stehl, and J.S. Woods. 1980. Trace Chemistries of Fire: A source of Chlorinated Dioxins. Science 210:385-390. Hardell, L. 1981. Relation of soft-tissue sarcoma, malignant lymphoma and colon cancer to phenoxy acids, chlorophenols and other agents. Scand. J. Work Environ. Health 7:119-130. Holmstedt, B. Seveso. 1980. Arch. Toxicol. 44:211-230. Prolegomena to Homberger, E., et al. 1979. Ann. occup. Hyg. 22:327-370. accident; its nature, extent and consequences. The Seveso Kilpatrick, R. 1980. Further review of the safety for use in the U.K. of the herbicide 2,4,5-T. Advisory Committee on Pesticides. London. Kociba, R.J., D.G. Keyes, J.E. Beyer, R.M. Carreon, C.E. Wade, D.A. Kittenber, R.P. Kalnins, L.E. Frauson, C.N. Park, S.D. Barnard, R.A. Hummel, and C.G. Humiston. 1978. Results of a two-year chronic toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats. Toxicol. Appl. Pharmacol. 46:279. Kutz, F.W. 1981. Chemical exposure monitoring in the EPA office of pesticides and toxic substances. Presented at the Veterans Administration Advisory Committee on Health Effects of Herbicides. Nov. 19, 1981. Leng, M.L. 1978. 2,4,5-T/TCDD controversy-Current status. North Central Weed Control Conference 33:164-168. Proc. Nash, R.C. 1983. Background Information: from G.W. Bostrom, E.G. National Coalition for a Reasonable 2,4-D Policy. Newton, M. and S.P. Snyder. 1978. Exposure of Forest Herbivores to 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) in Areas Sprayed with 2,4,5-T. Bull. Environ. Contam. Toxicology 20:743-750. Newton, M. and A. Norris. 1975. A discussion on herbicides. Statement on 2,4,5-T and TCDD. Reproduced from Journal of Forestry 73:410-412. Olie, K.P.L. 1977. Vermeulen 0. Hutzinger, Chemosphere. 5-12 6:455. �APPENDIX Significance of TCDD in Man's Exposure to Phenoxy Herbicides Due to the interest in this subject area and widespread national attention focused on it, many questions arise. Although some items may be somewhat redundant with earlier discussions, they are presented here for convenience. Following is an index of questions addressed in this appendix: Page A-4 What is dioxin? Page A-4 Is there more than one dioxin? Page A-5 Why does TCDD exist? Page A-6 How do we learn about TCDD? Page A-6 What is being done to learn the significance of man's exposure to TCDD? Page A-7 Is TCDD harmful to humans? Page A-ll Were phenoxy herbicides involved at Times Beach? Page A-ll Are people being harmed by the TCDD present at Times Beach? Page A-12 Since the discovery that soils removed from the horse arena and deposited in other Missouri communities still contain detectable TCDD levels, what is being done? Page A-12 What can be done to clean-up or remedy the current TCDD containing areas? Page A-13 HOW do humarrscome into contact with TCDD and other dioxins? Page A-17 Do phenoxy herbicides contain TCDD? Page A-17 Is the TCDD present in 2,4,5-T harmful to animals eating vegetation sprayed with the herbicide? Page A-18 What is Agent Orange? A-l �Page A-20 Why were defoliants used in Vietnam? Page A-20 Were defoliants other than Agent Orange used in Vietnam? Page A-21 Were other pesticides used in Vietnam? Page A-22 Is Agent Orange as toxic as TCDD? Page A-22 How long does TCDD persist in the environment? Page A-24 Do green plants take up TCDD from the soil? Page A-24 Does TCDD magnify in the food chain? Page A-26 Is TCDD in the environment harmful to animals? Page A-27 What are the short- and long-term effects of TCDD? Page A-30 If TCDD entered a human would it accumulate? Page A-31 Why is the Seveso, Italy accident important to the TCDD exposure questions? Page A-33 Since it has been shown that TCDD is widely distributed in the environment, does that mean that it is present in the food we eat? Page A-34 Can man be harmed from ingesting 2,4-D or 2,4,5-T? Page A-35 How toxic is 2,4-D to workers applying this herbicide in the forest? Page A-36 Should a concerned pesticide applicator individual quit worrying about exposure to pesticides since the scientific facts indicate no imminent danger is present? Page A-37 If ones chances of being harmed by phenoxy herbicides and TCDD are so low, why do we keep hearing about the danger on TV and reading about it in the newspapers? Page A-40 Other than in production plants what groups of humans are exposed to the highest amounts of pesticides? A-2 �Page A-40 Do phenoxy herbicides cause soft tissue sarcomas in man? Page A-41 Were there dangerous components of Agent Orange besides TCDD? Page A-41 What happens if 2,4,5-T, which contains a trace level of TCDD, comes into contact with a human? Page A-43 Were U.S. ground troops in Vietnam "at risk from Agent Orange since they could not shower daily? Page A-44 Could Agent Orange harm soldiers exposed to it? Page A-46 How could one know if he had been exposed to significant levels of Agent Orange? Page A-47 Chloracne has been described as the primary symptom for men exposed to chlorinated organic compounds, are there other symptoms that would accompany these symptoms? Page A-47 What has happened to the health of those in Vietnam who mixed and applied Agent Orange? Page A-47 Could Agent Orange affect the fertility of males? Page A-48 Can birth defects be caused by pregnant women being exposed to the phenoxy herbicides and TCDD? Page A-53 Does human mothers milk contain TCDD? Page A-63 How do the teratogenic effects of TCDD and 2,4,5-T compare? Page A-54 Were U.S. military personnel who used Agent Orange well informed about its properties at the time of its use in Vietnam? Page A-55 What evidence do we have which indicates that use of defoliants in South Vietnam has not caused man health problems? A-3 �Page A-55 Why do the Veteran's Administration and other government agencies continue to study Vietnam veteran health problems if the government is not admitting that Agent Orange exposure has caused the problems? What is dioxin? Dioxin is a general term which describes an organic molecule essentially composed of two benzene rings joined by a double oxygen bond. When a chlorine atom is attached to any or all numbered positions of it, -2-dioxin. tions. the molecule is referred to as a chlorinated dibenzo Chlorine atoms may be attached in the 2, 3, 7, and 8 posi- The molecule shown in Figure 1.1 is commonly referred to as dioxin or TCDO. Is there more than one dioxin? As many as 75 different dioxin molecules can be formed by substituting different chloride atoms at the numbered positions given (Kociba et al., 1982). Types of chlori- nated dibenzo-p-dioxins that can be found in the environment include: tetra, hexa, hepta, and octa. Several of these chlorinated dibenzo-p- dioxins are biologically active. The most active one known is the 2,3,7,8 tetra chloro dibenzo-£-dioxin. It is the most toxic of the 22 tetrachlorinated dioxins; this toxic molecule will be referred to herein as TCDD. Table A-l lists some of the properties of TCDD. A-4 �Table A-l. Some TCDD Physical Properties (Bovey and Young, 1980). Molecular weight 322 Melting point, °C 303-305 Decomposition point, °C 980-1000 Solubility in ppm Orange herbicide 580 benzene 570 chloroform 370 acetone 110 methanol 10 water 0.0002 TCDD is quite soluble in fats and very insoluble in water. In the direct presence of sunlight and a hydrogen donor it is degraded readily on leaves, water or soil surfaces. It has a low vapor pressure (1 x 10~7 Ton) and tends not to vaporize in the atmosphere thus its vapors can not be carried away by the wind. It is strongly bound to soil, consequently is very resistant to leaching and to entering the groundwater (Reggiani, 1981). Why does TCDD exist? TCDD is formed as a low-level unwanted by- product during the manufacture of trichlorophenol from 1,2,4,5-tetrachloro-benzene (Sparschu et al., 1971). The production of trichlorophenol is the initial step in producing hexachlorophene and trichlorophenol, common antibacterial agents that have been incor- A-5 �porated into a number of soaps and cosmetics (Rawls et al., 1976). Trichlorophenol is also used in the production of 2,4,5-T. How do we learn about TCDD? Information on the effects of TCDO exposure comes from at least three different sources. Industrial accidents have resulted in severe human exposure. Use of synthetic compounds containing TCDD, particularly the herbicide 2,4,5-T has initiated studies on humans and animals exposure exposed to the compound (Rose et al., 1972). Furthermore, several laboratory investigations have been completed during the past 10 years in an attempt to determine the toxic effects of TCDD (Zabik and Zabik, 1980). What is being done to learn the significance of man's exposure to TCDD? Dr. Frederick Kutz from the Field Studies Branch of the Exposure Evaluation Division of the Office of Toxic Substances in the U.S. Environmental Protection Agency offers some information on the philosophy and scope of their programs: The major orientation of the monitoring programs in the Office of Pesticides and Toxic Substances is toward the assessment of human exposure. Therefore, biological monitoring of human tissues and fluids assumes primary importance. Environmental components, such as air, drinking water, food, etc., which are intimately associated with human life, are considered secondarily. This prioritization scheme focuses monitoring programs on the protection of public health. Current ambient chemical monitoring responsibilities within OPTS include monitoring soils, raw agricultural crops, estuarine and marine organisms, water, human tissue, and air. Many of the programs have operated in cooperation with other Federal agencies, i.e., the National Center for Health . Statistics, the U.S. Geological Survey, and the National Oceanic and Atmospheric Administration. These agencies collect selected specimens for chemical residue analysis by EPA-directed laboratories. Samples of the above human and environmental media are collected on a periodic basis from locations defined by statistically representative designs (Kutz, 1981). In addition to the EPA, scientists in several other governmental agencies including FDA, USDA, and CDC and others are studying TCDD. A-6 �Related research is also being conducted by universities and chemical companies. An in-depth discussion of an applicator exposure study to 2,4,5-T containing trace levels of TCDD was presented in Chapter 3. Is TCDD harmful to humans? It is a widely accepted fact that TCDD is one of the most toxic low molecular weight synthetic molecules known to man. As a result of its toxicity and occassional presence in the environment TCDD has emerged as a compound for which there is great public concern. Since humans and test animals do not always have the same tolerance or sensitivity to a chemical it is difficult to establish a sound basis for exact comparisons. The question is: What is the relative toxicity of TCDD to humans and other mammals? In view of the extreme toxicity of TCDD this question becomes quite alarming when one reviews the facts on thalidomide. The lowest dose of thalidomide to produce teratogenic effects in mammals are: 30, 50, 100, and 350 mg/kg/day for mouse, rat, dog and hamster respectively. For humans this value is 0.5 mg/kg/day. Thus, humans are 60 to 700 times more sensitive to thalidomide than were the animals used to test.it. This relationship is not necessarily true of all compounds, however. In fact, it appears that humans are less sensitive to TCDD than several other animal species. For TCDD a wide range in mammal toxicity exists as expressed as an acute and 1059 (Table A-2). A-7 �Table A-2. Acute oral LDso of TCDD. Single-dose 1050 Literature citation (Micrograms per kilogram body weight) Guinea pig Rabbit 0.6-2.0 Schwetz et al., 1973 115 Schwetz et al., 1973 Rat 22-45 Greig et al., 1973 Dog 100-200 Greig et al., 1973 25-50 Greig et al., 1973 Chicken Bullfrog Hamster over 1000 1157-5051 Beatty et al., 1976 Olson et al., 1980; Henck et al., 1981 Since humans are not routinely used to test compounds as toxic as TCDD there is little information available other than information which arises after an accident or some other "happening." Literature describes in the early 1970's a "happening" known as the Missouri Horse Arena Incident; today it has been extended and now includes Times Beach, Missouri. Following the spraying in 1971 of three horse arenas in Lincoln County, Missouri, with salvage oil contaminated with TCDD, a by-product of hexachlorophene production, a number of people who worked or played in the arenas developed medical disorders (Beale et al., 1977; Kimbrough et al., 1977; Lobes et al., 1972; Young et al., 1978). For a more detailed accounting of this "happening" see A-8 �Chapter 1. The most serious health effects were shown by a six-year- old girl who developed hemorrhagic cystitis and focal pyelonephritis. The urinary tract symptoms were preceded by headache, epistaxis, diarrhea and a general malaise. after a few days. Her urinary tract symptoms cleared An examination three months later revealed she was normal except for punctate hemorrhages of the bladder seen on cystoscopy. The father of the child had developed headaches and nausea while working in the arena. Her mother reported severe headache, nausea, diarrhea and abdominal pains, and arthralgia. A ten-year-old sister developed easy fatigability, epistaxis, headaches, abdominal pain, and diarrhea. lesions (Reggiani, 1978). All developed at least mild acne Follow-up studies on the mother and two children performed five years later were normal (Beale et al., 1977). The symptoms for all of the humans were relatively mild with the most severe being the hemorrhagic cystitis and focal pyelonephritis seen in the six-year-old girl. several years later. The symptoms had cleared on re-examination Exposure, at least to the children must have been significant in that they regularly played in the soil of the arenas. These symptoms can be contrasted with fatalities observed for horses and other animals. In total 62 horses, hundreds of birds, many cats, and a few dogs died (Carter et al., 1975). Thus, the answer to our original question "Are dioxins toxic to humans?" is yes; but in contrast to thalidomide, humans appear to be much less sensitive to TCOD than are many test animals. This fact does not detract from the extreme toxicity of the TCDO molecule but does suggest that well designed experiments using sensitive test ani- A-9 �mals can perhaps provide meaningful information for humans. Center for Disease Control (CDC) has indicated that soils containing 1 ppb could theoretically cause health problems. Since the Center for Disease Control was forced to make a quick decision before all of the scientific data could be thoroughly evaluated it is possible that they have set limits on TCDD concentrations that are too conservative. Soil particles strongly bind TCDD. This reduces the uptake of this compound by the body of test animals according to experiments conducted by Poiger and Schlatter (1980). They also showed that the longer TCDD remains in contact with soil the less available it becomes for uptake. Since humans ingest little TCDD, they have little direct external body exposure to it. Cast (1983) states that implications by Burke (1983) that "dioxin, described by a respected health research as a "doomsday chemical" that can be harmful to humans in microscopic amounts as low as one part per billion" can be equated to the effect of one part of dioxin per billion of soil is without substance. As mentioned several times earlier, it is easy to confuse "exposure" and "absorbed dose." Chemical tests for amounts of TCDD in soil provides a measure of potential exposure but does little to provide a meaningful answer to how much does man absorb. As suggested by CAST, 1983 examination of the liver of wild animals inhabiting the high exposure areas can provide much information from which man can better evaluate the level of risk present. Agencies faced with njany human health decisions have difficult decisions to make. If they are to error, they should error on the A-10 �side of safety. As responsible servants they are also expected to make decisions based on all of the best scientific information available. Due to the worldwide impact these decisions may have these agencies should be permitted the luxury of continually reweighing the facts and the opportunity to alter their recommendations - be they for more stringency or less stringency. Were phenoxy herbicides involved at Times Beach? No, the TCDD by- product occurred during the synthesis of hexachlorophene a bactericidal agent. Are people being harmed by the TCDD present at Times Beach? The TCDD being found at Times Beach is present in soil which was contaminated with oil containing TCDD more than 10 years ago. The low water solu- bility (0.2 ppb) and adsorbing properties of TCDD prevents detectable levels from moving into man's food supply via growing plants. a grease or oil Without solubilizer the TCDD present today has little oppor- tunity to move through the skin of man or other animals. Bovey and Young (1980) have discussed that in highly TCDD contaminated soils the burrowing animals, beach mice and hispid cotton rat, TCDD level in their liver. 10 to 1500 ppt TCDD. had the highest Liver from healthy animals contained from No current health problems attributable to TCDD in the Times Beach area have been reported. Only by carefully extracting the TCDD with special organic solvents and then using the best analytical gas chromatograph-mass spectrometry instrumentation is it possible to identify TCDD. The normal health patterns of humans who have lived in these contaminated areas for the past 12 years in conjunction with the low A-ll �bioavailability of TCDD strongly suggests that TCDD is not available in the environment at levels harmful to plants, humans or other animals. Since the discovery that soils removed from the horse arena and deposited in other Missouri communities still contain detectable TCDD levels, what is being done? Following the Center for Disease Control's announcement that TCDD of 1 ppb and above constitutes a danger to man, EPA has indicated that dollars in its "Superfund" will be used to purchase the areas and relocate the people involved. As of May, 1983 residents of two areas of Times Beach and Imperial, MO had been cited by EPA as communities qualifying for use of the Superfund for relocation purposes. What can be done to clean-up or remedy the current TCDD containing areas? Five possible options are available as listed in Table A-3. A-12 �Table A-3. Possible waste stabilizing and permanent clean-up possibil- ities for Imperial, Missouri (Garmon, 1983). Option Years (est.) Construction of a grout curtain (This involves Millions (est.) 3-4 $29 digging a ditch around the contaminated area and filling it with a clay-like substance in order to prevent migration of the toxic material.) Removal of dirt and placement in an on-site vault 3-4 $9-11 Removal of dirt and placement in an off-site landfill or storage facility 4 8 Incineration of dirt Detoxification of dirt via application of $5-8 $29+ 11 $58-77 chemicals (Use of a sodium hydroxide and polyethylene glycol mixture to dechlorinate remove the chlorine atoms from - the dioxin now is being tested. Estimates of time and cost include that needed for further development of this technique.) How do humans come into contact with TCDD and other dioxins? In addi- tion to trace levels in the phenoxy herbicide 2,4,5-T, Agent Orange and by-products of trichlorophenol production other sources are present. Dr. Kutz of the EPA reports some of their investigations: A-13 �Because of the growing concern for the possibility of human exposure to toxic substances as a result of combustion, a study to provide statistically valid estimates of the level of organic emissions from combustion sources was begun. Among the main categories of concern are coal/refuse-derived fuel combustion, refuse combustion, and residential wood combustion. The compounds of interest are included in a broad category of substances termed "Polycyclic Organic Matter" (POM). These include polychlorinated biphenyls (PCBs), polychlorodibenzo dioxins (PCDDs), polychlorodibenzo furans (PCDFs), phenols, pyrenes, other polynuclear aromatic hydrocarbons (PAHs), and other organic compounds. In order to make a statistically valid estimate of national emissions, it is important to have information on emissions variabil-ity within any one facility. Therefore, a pilot study at two facili-ties was conducted in order to describe emissions variability. This variability was used to design the national study. One of the facilities sampled in the pilot burned 85% coal and 15% refuse-derived fuel (RDF), whereas the other burned raw municipal refuse. These facilities were sampled for 9 and 10 days, respectively, and samples of fuel, ambient air; water, bottom ash, fly ash, and flue gas were taken (Kutz, 1981). Results of these gas chromatography-high resolution mass spectremetry analyses are presented in Table A-4. A-14 Research by Reggiani �Table A-4. Residues of dioxins and furans observed in flue gas of a municipal waste combustion facility.! (Kutz, 1981). Mean^ Mean^ quantities concentrations emitted (ng/cubic m) Isomer groups (ug/hr) Tri-CDD 3 13 1,100 CDF4 300 26,000 Tetra-CDD 6.3 540 2,3,7,8-TCDD 0.4 34 90 7,600 Hexa-CDD 16 1,400 -CDF 62 5,200 -CDF Hepta-CDD 7.6 640 -CDF 7.5 640 Octa-CDD 2.5 220 0.6 52 -CDF corrected for recoveries. 2 Mean of 3 data points. refers to chlorodibenzo dioxins. refers to chlorodibenzo furans. (1981) indicates that various polychlorinated dioxins, sometimes including TCDD, may be formed during various combustion processes, A-15 �such as municipal incinerators, fossil-fueled power plants, internal combustion engines, home fireplaces and cigarette smoke. Knowledge that the toxic TCDO is introduced into man's environment through several sources does not necessarily make this a safer world in which to live but it does allow us to redirect the thinking of some of our environmental concerns. Since the presence and significance of TCDD as a contaminant in herbicides at levels from 1 to 70 ppm was found by Higginbotham (1968) the concentrations have been markedly reduced. Today's production methods routinely produce 2,4,5-T con- taining less than 0.1 ppm which is the legal level. Another source of exposure could be through the use of cosmetics that have as one of their ingredients compounds derived from trichlorophenol (Rawls et al., 1976). Although no research has been reported on the absorbed dose of TCDD arising from these uses one can infer that due to the lipophyllic nature of TCDD its percutaneous absorption would be expected when present in creams applied to the skin. A report by Schmittle et al., 1958 indicates that human and animal exposure to TCDD has also occurred from the improper disposal of distillate residues of chlorinated benzenes which are used as a starting material in the production of a variety of chemical products. They indicate that perhaps the most extensive occurrence of animal exposure to the dioxins (including dichloro-, trichloro-, tetrachloro-, pentachloro-, hexachloro-, heptachloro-, and octachlorodibenzodioxin) occurred in the United States in 1957. At that time, in order to increase the caloric content of poultry rations, feed manufacturers were using a wide variety of low cost fats. These fats were A-16 �collected from various sources and pooled. Some of the fats were contaminated with oily residues containing dioxins. These mixtures were sold to the feed manufacturers and incorporated in poultry feeds. As a result of the ingestion of these contaminated fats, millions of chickens died. Hundreds of thousands of the surviving contaminated chickens were processed for human consumption before the origin of the problem was discovered and steps initiated to prevent the use of this food. There are no data which suggests that humans were affected by consuming adulterded poultry products (Allen et a!., 1978). It is not surprising that there are no accounts of humans becoming ill from eating TCDD contaminated chicken because the high toxicity of the compound would probably kill the chicken. Thus, there is a built in safety factor since humans seldom eat meat from animals that they or the slaughter house has not killed. Do phenoxy herbicides contain TCDD? 2,4,5-T and si 1 vex contain trace levels of the contaminant TCDD. Although it has not been feasible to eliminate this contaminant entirely, present production methods are able to reduce the dioxin level routinely to less than 0.01 ppm in commercial 2,4,5-T with occasional batch containing as much as 0.05 ppm (McQueen, 1977). The legal limit for TCDD in 2,4,5-T is 0.1 ppm. Is the TCDD present in 2,4,5-T harmful to animals which eat vegetation sprayed with the herbicide? As previously stated, TCDD is present in 2,4,5-T in only trace amounts but the extremely toxicity of even small amounts of TCDD has also been stated. The quotation from CAST (1978) will be used to answer the question: The amount of TCDD distributed in the United States in 2,4,5-T and silvex is probably no more than 1 ounce (28 g) annually. This material is distributed over approximately 5 A-17 �million acres at the rate of 5 micrograms per acre. The most sensitive species known is the guinea pig, which has an LDso of 0.6 microgram per kilogram (Johnson, 1971). If we assume that we have a grazing animal about the size of a sheep or deer (175 Ib or 80 kg) and the sensitivity of a guinea pig, this animal would have to consume, without excretion, all of the treated vegetation on more than 9 acres of land to get a lethal dose. If the animal had the sensitivity of a rat, it would have to consume, without excretion, all of the treated vegetation on more than 400 acres. There are no cases where livestock has been injured due to ingesting forage treated with 2,,4,5-T containing trace levels of TCOD. Perhaps the fact that phenoxy compounds (Lavy, 1980a) and TCDD (Jensen et al., 1981) are rapidly excreted from animals allow these chemicals to be absorbed without apparent harmful effects. What is Agent Orange? Herbicide Orange, later Agent Orange, was a label given by the U.S. military forces to a 50:50 mixture of the nbutyl esters of 2,4-dichlorophenoxyacetic acid (2,4-D) and 2,4,5-trichlorophenoxy-acetic acid (2,4,5-T). One gallon of Herbicide Orange theoretically contained 4.21 Ib of the active ingredient of 2,4-D and 4.41 Ib of the active ingredient of 2,4,5-T. Orange was formulated to contain a 50:50 mixture of the n-butyl esters of 2,4-D and 2,4,5-T. The percentages of the formulation typically are shown in Table A-5. A-18 �Table A-5. Chemical composition of Agent Orange (Shepard, 1980) Materi al % n-butyl ester of 2,4-D 49.49 free acid of 2,4-D 0.13 n-butyl ester of 2,4,5-T 48.75 Free acid of 2,4,5-T 1.00 inert ingredients (e.g., 0.62 butyl alcohol and ester moieties) Agent Orange also included amounts (from 0.02 to 15 ppm, or a weighted mean of 1.98 ppm) of the contaminant 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). According to Young et al. (1978), about 25 million kilograms (55,940,150 Ib) of 2,4-D and 20 million kilograms (44,232,600 Ib) of 2,4,5-T - for a total of 166 kg (368 Ib) of TCDD - were distributed over approximately 1.4 million hectares (3.5 million acres) of South Vietnam from January 1962 to February 1971. It is significant to note that almost 40% of the entire amount of TCDD that was released in South Vietnam - or 64 kg (143 Ib) was contained in Agents Green, Pink, and Purple, which materials were used on about 36.4 thousand hectares (90 thousand acres) from 1962 thru 1964. During this time few American troops were assigned to South Vietnam. The name Herbicide Orange was a code name by the military for herbicide combination given. Later the phrase "Agent Orange" was A-19 the �coined. Barrels containing this plant defoliant were painted with an orange stripe to differentiate it from other compounds being used in Vietnam. Why were defoliants used in Vietnam? The use of chemicals (herbicides) to control vegetation has been one of the most controversial subjects arising from the Vietnam conflict. The U.S. Air Force applied most of these herbicides in jungle areas to clear vegetation from the perimeters of military bases and camps, along lines of communication, and in enemy staging areas. The objective was to provide defoliated zones that would reduce ambushes and disrupt enemy tactics (Young, 1981). Ninety percent of all Herbicide Orange was used to defoliate forests; it was especially effective in defoliating mangrove forests. Eight percent of Herbicide Orange was used in the destruction of broadleaf crops (beans, peanuts, ramie, and root or tuber crops). The remaining 2 percent was used around base perimeters, cache sites, waterways, and comunication lines (Shepard, 1980). The use of defoliants in Vietnam saved hundreds of U.S. military lives. Were defoliants other than Agent Orange used in Vietnam? Several others were employed; their code names are White, Blue, Purple, Pink and Green. The latter three which all contained 2,4,5-T were used in (Table A-6). A-20 �Table A-6. Number of gallons of military herbicide procured by the U.S. Department of Defense and disseminated in South Vietnam during January, 1962 - October, 1971 (Young et al., 1978). Period of Code name Herbicide Quantity Orange 2,4-D; 2,4,5-T 10,646,000 1965-1970* White 2,4-D; Picloram 5,633,000 1965-1971** Blue Cacodylic Acid 1,150,000 1962-1971** Purple 2,4-D; 2,4,5-T 145,000 1962-1965 Pink 2,4,5-T 123,000 1962-1965 Green 2,4,5-T 8,200 1962-1965 Total use 17,705,200 *Last fixed-wing mission of Orange 16 April 1970; last helicopter mission of Orange 6 June 1970. **Last fixed-wing mission 9 January 1971; all herbicides under U.S. control stopped 31 October 1971. Were other pesticides used in Vietnam? The control of malaria and other mosquito-borne diseases in South Vietnam necessitated an extensive aerial insecticide application program in order to control these vector insects. From 1966 through 1972, three C-123 aircraft were used to spray Malathion, an organo-phosphate insecticide. These aircraft could be distinguished from the herbicide-spraying aircraft because they were not camouflaged. These aircraft routinely sprayed insecticide adjacent to military operations were in progress, or installations, as well as in areas where military operations were in progress, or about to commence." (Shepard, 1980). A-21 �Is Agent Orange as toxic as TCDD? As indicated above, TCDD was present in Agent Orange at an average concentration of nearly 2 ppm. This translates to a 50:50 mixture of 2,4-D and 2,4,5-T; the final mixture contained 99.9998% phenoxy herbicide and 0.0002% TCDD. Information taken from Cast (1978) states The evidence shows that TCDD is from 5,000 to 500,000 times more toxic to mammals than 2,4,5-T, depending on species. Under current standards set by the Environmental Protection Agency, the content of TCDD in commercial 2,4,5-T must be less than 1 part in 10 million parts of 2,4,5-T. In practice the content is about 1 part in 100 million (Fisher, 1977). On this basis, a single toxic dose of TCDD would be contained in from 200 to 20,000 toxic doses of 2,4,5-T, depending upon the test species. Thus, the current level of TCDD in 2,4,5-T does not contribute significantly to the toxicity of herbicidal preparations of 2,4,5-T. A more detailed calculation of the relative toxicity of TCDD and the phenoxy herbicides in the mixture is presented in Chapter 4. How long does TCDD persist in the environment? Crosby and Wong (1977) analyzed the persistence of TCDD in actual herbicide formulations on leaves, soil, or glass plates. When leaves were exposed to natural sunlight, most of the TCDD was lost from the leaves in less than 6 hours. This loss was due principally to "photochemical dechlorination." The herbicide formulation provides a hydrogen donor, which allows rapid photolysis to occur. excellent source of hydrogen donors. markedly decreased. Since Agent Orange has an TCDD persistence should be Thus, despite the known persistence of pure TCDD, it is not stable in thin films of formulated herbicide when exposed to outdoor light. Plant uptake of TCDD from soils does not appear to be significant. Soybean and oat plants took up only trace amounts of TCDD in the first A-22 �10 to 14 days after exposure to sandy soil containing 200,000 times the amount of TCDO contained in an application rate of 2 pounds per acre 2,4,5-T (with 0.1 ppm TCDD). There was no detectable TCDD in the grain or beans at maturity, probably because of normal dilution by plant growth, volatilization, or photodecomposition on the leaf surface, or as a result of metabolism. TCDD is not translocated from the point of application on the leaf surface (Isensee, 1971). In five soils with widely varying properties, TCDD was found to be immobile even when subjected to leaching (Helling et al., 1973). The possibi- lity of TCDD entering ground water is remote (Tschirley, 1971). TCDD is incorporated into soil, it disappears slowly. If It seems unli- kely, however, that TCDD would be incorporated in soils under most conditions of use since it does not leach into the soil. TCDD is not produced from breakdown products of 2,4,5-T in soils or in sunlight (Kearney et al., 1972). TCDD is nearly insoluble in water - 0.2 ppb. For this reason, it would be expected to remain on the surface of plants and soil at the application site. Because it is immobile in soils, Kearney et al. (1973) also concluded that there would be "no ground water contamination problems." Recent studies on half life determinations of TCDD in soils indicate that 10-12 years are required for 50% of this compound to degrade. This very persistent molecule has earlier been shown to be quite toxic. Other factors of importance in assessing its hazard to humans is how much is present? how bioavailable is it? and ultimately how much is absorbed by humans? A-23 �Do green plants take up TCDD from the soil? Isensee et al. (1971) determined the uptake of TCDD from a Lakeland sandy loam soil using two crop species. Oats (Avena sativa) and soybeans (Glycine max) accumulated less than 0.15% of the TCDD applied to the soil. In a similar study Cuppelo et al. (1976) using sorghum (Sorghum bicolor) found that the amount of TCDD taken up from a Ulysses sandy loam soil was approximately one millionth of one percent of the amount of TCDD in the soil. One of the major contributing factors for the low % uptake is the low 0.2 ppb water solubility of TCDD. As a general rule, in the case of herbicides, those with low water solubilities tend also to be more tightly adsorbed to soil colloids thus further decreasing their potential for plant uptake. In U.S. Air Force Experimental plots at Eglin AFB, Florida where massive quantities of 2,4,5-T were applied to soil, TCDD concentrations were 150, 160, 700, and 44 ppt at soil depths of 0-2.5 cm, 2.5-5 cm, 5-10 cm and 10-15 cm, respectively. In an ecological survey conducted over a 5-year period in that area more than 123 different plant species were present. No TCDD was found in any of the plant species ana- lyzed (Young et al., 1975). Due to the extremely low levels of TCDD present in most soil and its low water solubility, TCDD plant uptake from soil is virtually nil thus escaping detection sensitive gas chromatographic determinations with capabilities for detecting TCDD in the low ppt range. Does TCDD magnify in the food chain? The term bioaccumulation means the uptake of a chemical and some storage of it by an organism. Since TCDD is present in such low levels in the environment ingestion, inha- A-24 �lation or dermal absorption is minimal (Norris et al., 1977). An EPA study monitoring fat from beef animals indicated that bioaccumulation of TCDD in grazing animals could occur, but levels were not high enough to be regularly detected by instruments with a 10 ppt sensitivity. The results of these various tests indicate that, if TCDD is present in the environment in a form that is available to organisms, then bioaccumulation will occur if organisms are exposed. This concept is supported, both from an examination of the physical-chemical properties of TCDD, as well as by studies of its behavior in animals exposed through feeding studies or in laboratory model aquatic ecosystems. The degree to which bioaccumulation of TCDD occurs in the field depends not only on the physical-chemical properties of the compound, but also on the persistence and availability of TCDD in the environment. Mechanisms of degradation and dilution that operate in the natural environment reduce the opportunities for organisms to be exposed, and thereby reduce the degree to which bioaccumulation might occur. A wide-scale monitoring of water, sediment, fish, beef, and human milk from areas in the Midwestern United STates where 2,4,5-T has been applied also showed no detectable TCDD residues at minimum detection levels that averaged 10 ppt. These monitoring efforts indicate that substantial bioaccumulation of TCDD (sufficient to produce residue levels in excess of 10 ppt TCDD in the majority of the population) is not occurring in animals in or near areas treated with 2,4,5-T in current operational programs (USDA, 1982). A-25 �Is TCDD in the environment harmful to animals? As shown in Table A-2 several animal studies have been completed to evaluate the toxicity of the TCDD molecule. Thus guinea pigs are quite sensitive to the molecule while hamsters may be as much as 8000 times less sensitive. The question to be answered is - is there enough TCOD in the environment to harm animals? Several animal studies have completed but the only formal study designed to test the accumulation of TCDD in animals in the wild Newton et al. (1978), could detect no TCDD in tissues of mountain beaver at the end of 45 to 60 days of feeding on 2,4,5-T treated vegetation. The area had been sprayed with 2 Ib 2,4,5-T per acre at the beginning of the feeding period. Shadoff et al. (1971) analyzed animals for accumulation of TCDD from regions in Midwestern United States where 2,4,5-T is commonly used. They did not detect any TCDD in samples of fish, water, mud and human milk from Arkansas and Texas. Their detection limit was 10 PPt. Jensen et al., 1981 fed artificially exaggerated levels of 24 ppt TCDD to seven beef cattle for 28 days. No adverse effects were observed in the animals during the 50 weeks of the feeding study. They consumed approximately 0.02 ug/kg in 4 weeks. Beef fat which contained the highest level of TCDD, contained 61 to 95 ppt. These levels dissipated when the cattle were put on non-treated feed. These data represent residues obtained following an artificially exaggerated TCDD ingestion rate when compared to the intake that might be obtained grazing grass sprayed with 2,4,5-T. Typically, the ini- A-26 �tial deposit of 2,4,5-T on grass is 100 ppm for each pound of herbicide applied per acre (Morton et al., 1967). Since 2,4,5-T presently contains <0.05 ppm of TCDD, the treated grass could have a maximum initial TCDD level of 5 ppt. This residue decreases with time, having a half life on grass somewhere between the 4-h value found on excised leaf surfaces (Crosby and Wong, 1977) and a value of 1 week in a field experiment where grass was sprayed with a 2,4,5-T formulation (Getzendaner and Hummel, 1975). Assuming these halflives, the TCDD level would decline to between <0.001 and 0.3 ppt in 1 month. TCDD residue levels obtained in the Jensen study in which a continuous 24-ppt diet level was fed should not be confused with the greatly reduced potential for TCDD residues in normal herbicide use. The results of the Jensen feeding study provide a foundation of ruminant animal data supporting the conclusion that residues of TCDD would be detected in beef fat if range animals had been exposed to measurable levels of TCDD. However, no TCDD has been found at levels above the detection limit in beef samples taken from areas which would maximize exposure to TCDD through treatment with 2,4,5-T, except for 3 samples out of 89 from treatment areas (Shadoff et al., 1977) which came from an area in Missouri where highly TCDD contaminated soil from Missouri Horse Arena incident had been deposited (Pesticide Chemical News, 1976). Rabbits force fed with TCDD contained in soil or organic solvents absorbed higher levels from the organic solvents than from the soil which tightly binds TCDD (Bonaccorsi et al., 1983). What are the short- and long-term effects of TCDD? We could answer this question quite simply and correctly by indicating we do not yet A-27 �have all of the answers. However, this same answer could be given for most all compounds that have biological activity. Since some information is available, portions of this will be discussed. Much of this information is based on animal feeding studies. These effects could be more severe or less severe in man since TCDD has been shown to be to some extent species-specific, however indications are that man is less sensitive to TCDD than are several other animal species (Stevens, 1981). TCDD appears to be excreted primarily in the feces of the rat (Piper et al., 1973) and is assumed to be associated with unabsorbed compound. Calculated half-lives of 10-30 days were markedly longer than those from the phenoxy herbicides. The extreme toxicity as exhibited by the very low LDso's (Table 5-2) for TCDD in combination with a relatively long residence time in animals provide this compound with an enhanced potential for causing harmful effects in animals. Due to the extreme toxicity of this compound man should take precautions to insure that he is not being exposing himself to dangerous levels. Whether it is possible for him to expose himself to health threatening levels in the environment is debateable. Nevertheless, he also owes it to himself to make a rational attempt to become as well informed as possible to the actual levels of exposure he can conceivably be expected to receive. Workers involved in accidental exposure to TCDD as a contaminant of trichlorophenol or in factories synthesizing the compound have reported a number of adverse health reactions (May, 1973; Poland et al., 1971). Besides chloracne, which is the best known and most sen- A-28 �sitive sign of exposure to TCDD, other reported toxic manifestations include conjunctivitis or blepharoconjuctivitis, hepatotoxicity, gastrointestinal and kidney disturbances, anorexia, weight loss, fatigue, hyperpigmentation, hirsutism, porphyria cutanea tardia, neuromuscular problems, and psychological changes. However, it has not been established with certainty that all these clinical symptoms were related to TCDD exposure. In their section on "Public Health Hazards" the National Research Council of Canada (NRCC, 1978) indicates that exposure to high concentrations of phenoxy herbicides as residues in food or water is relatively unlikely, the primary risk to human health may be expected to be from occupational exposure during manufacture and application. Bleiberg et al. (1964) and Poland et al. (1971) reported surveys of workers in a plant producing 2,4-D, 2,4,5-T, and the di- and trichlorophenols. Chloracne was found in 13 of 73 employees examined and, in the earlier survey, several cases of porphyria and uroporphyrinuria were described.. These effects were observed before and just after major steps were taken to decrease the TCDD content of the products produced in this plant. These are signs of TCDD poisoning although the exact nature of the chemicals to which the workers were exposed is unknown. In another study no differences were observed in 220 men exposed to 2-8 mg 2,4,5-T/day for 1 to 960+ days (Johnson, 1971). The control group contained 4600 men. "The workmen were given extensive physical examinations including a battery of at least 20 laboratory tests." No chromosomal effects were reported. sures were not specified. A-29 Their level of expo- �Based on dose response studies in animals and limited epidemiologic data at this time there is no conclusive evidence that TCDD is mutangenic, teratogenic, or carcinogenic to man (Bovey and Young, 1980). If TCDD enters a human would it accumulate? In animal studies in which l^C-labeled TCDD was given orally to rats the largest accumulation occurred in liver and fat. These tissues contained 10 to 50 times as much as other body tissues. This fact does not answer the question of accumulation. TCDD is eliminated primarily through the feces for guinea pigs, rats, and mice. The total excreted through this route was 94, 99, and 74%, respectively for the above animals. Urinary excretion accounted for the difference. The half life for elimination was 30, 31, and 17 days respectively for guinea pigs, rats, and mice (Neal et al., 1982). Dr. Frederick Kutz (1981) of the EPA has analyzed human adipose tissue for TCDD content. His statements follow. Several investigators have indicated that minute quantities of 2,3,7,8-TCDD are present (in the low parts per trillion range) in specimens of adipose tissue collected from members of the general population. We also have conducted a very limited number of analyses of this type. As control specimens for some analytical studies done by the EPA Dioxin Monitoring Program in early 1980, six specimens of human adipose tissue were collected from residents of an urban Ohio county. These specimens were exised during post-mortem examinations from individuals with m) recorded or known exposure to 2,4,5-T or silvex. The source of TCDD in these specimens was not phenoxy herbicides. Subsequently, the samples were analyzed in duplicate following the EPA Dioxin Monitoring Program protocol. Instrumental determinations were conducted at two independent laboratories. The results demonstrated that all specimens contained residues of 2,3,7,8-TCDD. Levels ranged between 5 and 12 parts per trillion, with a detection limit below 5 ppt. It should be emphasized that all studies conducted to date, including this one, have been accomplished utilizing small A-30 �sample sizes and deliberate specimen selection criteria. Consequently, these data cannot be construed as being representative of the general population. Why is the Seveso. Italy accident important to the TCDD exposure questions? By evaluating health problems suffered by humans which have been exposed to high levels of TCDD, we will be in a better posv tion to attempt to quantify exposure to the Vietnam veteran and Vietnamese living in the defoliated areas. Stevens (1981) provides a quantitative approach to the animal deaths and human health factors. About noon on July 10, 1976, a factory north of Milan which produced trichlorophenol had an explosion which released boiling reagents into the atmosphere. Most of the cloud settled on a 110 hectare (1 ha = 2.47 acres) area in the town of Seveso where it deposited about 2 kg of TCDD. This area was designated Zone A and a larger, more distant area of 269 ha which received only 20 g of TCDD was designated Zone 8. The 733 residents of Zone A were evacuated 2 weeks after the explosion. The 4800 residents of Zone B were not evacuated but were prohibited from raising or using produce from either Zone B or the surrounding 1430 ha Zone R. These amounts of TCDD would produce average concentrations of 1800 and 7 ug of TCDD per square metre in Zones A and B respectively. Analyses carried out days to weeks after the explosion showed a very spotty distribution with a range in Zone A from 5447 ug per square meter to not detectable (Pocchiari, Silano & Zampieri, 1979). Actual concentrations on vegetation and soil shortly after the explosion were undoubtedly higher than these recorded values. Birds began dying in Zone A a few days after the explosion. Birds preen their feathers regularly. The surface area of the feathers is many times that of the skin surface so even a small bird must have at least 0.1 m2 of feather surface. 2 If the feathers held the average concentration of 1800 ugnr , then 180 ug would be present on a bird weighing only 30 to 100 g. Thus the bird could ingest 1000 ugkg'1, about ten times the acute LD5Q for birds (Young et al., 1978). Rabbits which ate fresh vegetation from local fields died next. A 1 kg rabbit consuming vegetation from an area of one square meter would take in at least 1800 ug, over ten times the 1059 for rabbits. More cats than dogs died since cats both groom frequently and eat birds. Later a number of sheep.and cattle died; most livestock in Zone A was destroyed. In Zone B, however, birds and rabbits would consume TCDD in the order of 7 ugkg-1 (below the actue toxic dose) and here few of these animals were found dead. A-31 �The early toxic effects in man were not due to TCDD but to other chemicals, for example sodium hydroxide. Chloracne has been the only prominent effect found to be caused by TCDD; 193 cases have been diagnosed with 50 cases in residents of Zone A. The diagnosis in the 143 cases outsize Zone A is confused because these were found by screening 49,000 people for an incidence of 0.3% and this percentage is reported to be the incidence of chloracne in more northern industrial Italian cities not involved in the accident. However, in Zone A the incidence of chloracne was 7%: 23 times this base rate. In addition, about 10% of the Zone A population showed either minor clinical signs of polyneuropathy or laboratory evidence of slowed nerve conduction times. Transient changes in liver enzymes were noted in a few cases. There have been no deaths, no serious disease, no immunological changes, no birth defects, no increase in spontaneous abortions; and no chromosomal changes attributable to TCDD (Pocchiari et al., 1979). Cases of chloracne outside Zone A were probably due to external contact with transported TCDD. The overall pattern indicates that the mean concentration of TCDD in Zone A was below the minimum toxic dose but the spotty distribution produced minimal signs and symptoms in about 10%. Since most clinical findings were in children under age 12, 30 kg will approximate the average weight. To reach the MTD of TCDD of 0.1 ugkg"^ would require an intake of 3 ug. The mean concentration of TCDD was 1800 ugm~2 which must have been considerably lower than the level needed to provide the MTD. It will therefore be assumed that the 10% affected were exposed to about three times this mean value or 5400 ugm~2. The intake transfer factor for Seveso would then be 5400/3 or 1:1800. TCDD is most toxic for guinea pigs and least for dogs. The acute toxicity for chickens, mice, rats, cats and monkeys are all within a proportional order of magnitude, and chronic toxicity in man, monkeys and rats is not very different (Young et al., 1978). Hence the reason that human toxicity was low in Missouri and Seveso while animal toxicity was high, related not to innate differences in sensitivity but to a very much lower intake of the environmental TCDD. Humans wear shoes and clothing, they do not groom with their mouths, they wash themselves and their food and they do not eat vegetation in bulk. These features account for the intake transfer factors of 1:2300 calculated from a single individual for whom accurate measurements of environmental TCDD and of the duration and type of exposure were known, and of 1:1800 calculated from the signs and symptoms of about 75 individuals with less accurate TCDD measurements and of their exposure. The agreement is good and the average value of 1:2050 will be used for the intake transfer factor. In the Seveso accident TCDD was released in a village at rates per acre that were millions of times greater than those that occur from A-32 �2,4,5-T treatments (Reggiani, 1978). Beljan et al. (1981) discussing the Seveso accident states Thus far, one of the most extensive "human experiments" is being conducted on the residents of Seveso, Italy. The whole story may not be known for another five years; but preliminarily, it appears that the dermatological effects were reversible; there has been no progression of neurological damage; differences in immune response between exposed and nonexposed persons have not been significant; and there has been no notable increase in infectious diseases. There may never be an entirely clear picture from Seveso of the effect of TCDO on reproduction or the fetus, because it is suspected that a number of women elected to have therapeutic abortions. No major pathology, however, was reported by the area physicians. May (1982) reported an accident involving TCDD which occurred in the United Kingdom in 1968. During the manufacture of 2,4,5 trichlorophenol (TCP) an accident led to the release of crude TCP containing a higher than normal concentration of 2,3,7,8-TCDD. A total of 79 cases of chloracne was reported, but there was no clinical evidence of other dioxin-related diseases. responded favorably to simple treatments. The majority of the cases Ten years later, symptoms had disappeared completely from half the cases under observation while in the remainder symptoms were insignificant. Since it has been shown that TCDD is widely distributed in the environment, does that mean that it is present in the food we eat? If normally present it is below our level of detection as shown by several food surveys. In a study designed to favor the conditions for a positive identification of TCDD the following action was taken. Beef fat containing residues of ronnel and its metabolite 2,4,5-trichlorophenol at 46.0 and 2.8 ppm, respectively, was examined before and after cooking for presence of 2,3,7,8-tetrachlorodibenzo- A-33 �p-dioxin. Fat samples fortified at 2000 ppm with 2,4,5-trichlorophenol or its sodium salt were cooked at 500°F for periods of 6 to 22 hr. No tetra-dioxin was found in any of the samples, using a method with a sensitivity limit of 0.05 ppm (Watts et al., 1973). Can man be harmed from absorbing or ingesting 2,4-D or 2,4,5-T? Berwick (1970) reports an incident where a man accidentally ingested a 110 mg/kg dose of 2,4-D (isooctyl ester). Besides depression and irritation of the central nervous system, the victim suffered from myopathy, 61 irrigation, nephroprethy and toxicity to the liver. individual recovered fully. This Several other cases give us additional information on man's susceptibility to phenoxy herbicides. Symptoms of poisoning occur at an oral dose of around 3 to 4 grams (Hayes, 1963). In an experiment conducted on himself, a man consumed 500 mg of purified 2,4-D each day for 21 days without ill 1951). effects (Assouly, In another case, 18 intravenous doses - the last 12 of which were 800 milligrams or more - were delivered to a patient over the course of 33 days for the treatment of coccidioidomycosis; there were no side effects even after the 18th dose of 2,000 mg. The last injection of 3,600 mg produced illness - but, by the end of 48 hours, the subject had completely recovered (Beljan et al., 1981). Some information on phenoxy toxicity to man has been gained by evaluating data gathered from suicide victims. The lethal oral dose for 2,4-D as based on findings from a 23-yr-old male is slightly over 80 mg/kg; a man weighing 80 kg would thus require at least 6.4 grams of the material to kill him. Nielsen et al. (1965) reports this A-34 �suicide with an estimated 6.5 g of 2,4-D. death was 18 hours. The time from ingestion to This 80 mg/kg dose was received all at one time. If low doses are received over a long time period man's body effectively rids itself of phenoxy compounds. Rats, when fed levels of 2,4-D at or above the NOEL for teratogenicity, developed peripheral neuropathy (impaired functioning of outermost joints of forearms and hindlimbs (Squibb et al., 1983). In all cases grip strength of the rats returned to normal 6 weeks after daily dosing ended. Peripheral neuropathy was reported for two individuals who had significant dermal contact with 2,4-D (Berkley et al., 1963; Todd, 1962). No evidence has been shown that indicates that chronic expo- sure to 2,4-D in man will cause peripheral neuropathy. How toxic is 2,4-D to workers applying this herbicide in the forest? Using the above information on acute poisoning in the suicide case let us now make a calculation from data on applicator exposure to 2,4-D provided in Chapter 3 (Lavy et al., 1982). The group of workers receiving the highest dose were the pilots and the batchmen, both of whom received 0.0198 mg/kg. If we assume similar toxicity between the dose received orally by the suicide victim and that received as a consequence of dermal absorption by the field workers the following calculation can be made. The 80 kg suicide victim's level of ingestion was 80 mg/kg; the group with the highest absorbed dose in the 2,4-D applicator exposure study (Lavy et al., 1982) was 0.0198 mg/kg. The calculation: A-35 �Lethal acute dose 80 mg/kg = 4040 One time applicator exposure dose 0.0198 mg/kg reveals that the workers could have absorbed 4040 times more 2,4-D than they did before they would have received a lethal dose. Thus, a 4040 fold margin of safety was involved for the most highly exposed workers in the study. In reality the margin of safety would be much higher since the approximately 0.0198 mg/kg was the total absorbed dose of 2,4-D occurring over the 5-day observation period. As discussed previously, absorption through the skin in a relatively slow process, thus the body has the ability to excrete it at a faster rate than it can enter through the skin. This rapid elimination of the phenoxy compounds does not allow them to build-up in the body. The fact that over 9556 of the absorbed phenoxy dose is excreted from the worker's bodies within 6 days probably accounts for the fact that many aerial applicators have made a career of applying phenoxy herbicides with no apparent detrimental health effects. Recent applicator exposure studies from New Zealand indicate that calculated results when compared to epidemological data show that chemical applicators are at no greater risk from teratogenic or carcinogenic hazard that the general population (Ferry et a!., 1982). Should a concerned pesticide applicator quit worrying about exposure to pesticides since the scientific facts indicate no imminent danger is present? All pesticides have biological activity; as such they should be handled with due caution as prescribed on the label. Nonapplicators receive much, much less exposure than others in our society. Education, protective clothing, and a desire to work in a A-36 �safe manner are prerequisites for concerned pesticide applicators who wish to limit their absorbed dose of the pesticide. To put into perspective the potential dermal exposure received by workers applying 2,4,5-T by backpack sprayers at a 2 Ib/acre rate compared to the TCDD levels discussed in other examples in this monograph (see Table A-7). Table A-7. Comparisons of TCDD levels as reported in different studies. Commodity Concentration reported (ppb). Soil from Missouri horse arenas 33,000 Soil from near Seveso explosion site Soil from Times Beach, Missouri 5,500 300 CDC expressed concern level 1 Soil concentration present when top 3 inches is treated with 2#/acre 2,4,5-T (containing 0.04 ppm TCDD) 0.00008 Present on humans applying 2,4,5-T (containing 0.04 ppm TCDD) at a rate of 1.6 Ibs/acre in the forest 0.00005 If one's chances of being harmed by phenoxy herbicides and TCDD are so low, why do we keep hearing about the danger on TV and reading about it in the newspapers? Several different answers are possible: A-37 �1) The issues being discussed are complex; assuming the news story writer heeds all of the facts available to him, it would not be a simple task to convey the message to a broad audience. Results of an extensive literature search and discussions with scores of concerned laymen reveals that the U.S. public has heard little factual information pertaining to the man's exposure to phenoxy herbicides and TCDD. The objectives of the news media and of reporters of scientific data are often at opposite poles. The journalist does not provide details on the millions of cars traveling the highways each day. But a car which crashes, bursts into flame, and kills three people makes front page news. Likewise, the public is seldom told of the major role that pesticides play in allowing the U.S. farmer to achieve production levels which provide us an opportunity to ship our abundant food supplies to millions of starving people around the world. But if a child inadvertently drinks concentrated gasoline or insecticide and dies this story makes news. The scientist on the other hand collects data to determine how safe cars are, or how to make them safer; he seeks to find new means to feed people; he tests to determine the danger levels of cars or pesticides; realistically, these facts do not make good front page stories. Evidence overwhelmingly supports the concept that phenoxys and TCDO when in a concentrated form, can harm man and animals. The news media responds to this side of a complex story. The evidence also convincingly shows that man can not, through his occupational use of the herbicides or through his food supply, attain a dose of the phenoxy herbicides or TCDD approaching a health threatening level. The A-38 �scientist bases his endorsement on the current use of these materials after assessing the issues and interpreting them to the best of his ability. 2) The public hears news stories; most do not pursue scientific literature. The audience is more accustomed and perhaps more receptive to listening to catastrophic happenings than to statements such as "these pesticide levels do not constitute a threat to man's health." 3) It is much easier to speculate that something bad may happen in the future than it is to prove unequivocally that 20, 40, or 80 years from now a compound will not present a danger to man or his grandchildren. 4) Politically, it is more efficacious to vote on the side of safety than it is to invest the time and determination to search out the facts. The masses informed by the media represent more votes than those reading the scientific journals. 5) Some Vietnam veterans have health related problems; for many individuals it will not be possible to arrive at the exact source of the problem due to complex world in which we live. In fairness to this group of individuals only factual scientifically validated information should be offered as the cause. 6) As discussed previously, on a per gram basis, the toxicity of the phenoxy herbicides is much, much less than that of TCDD. This monograph has presented information which shows that phenoxy herbicides are not a major source of TCDD in our environment today. Dioxin (TCDD) will continue to be of keen interest to man due to its extreme toxicity and its persistence in the environment. A-39 �Other than in production plants what groups of humans are exposed to the highest amounts of pesticides? Commercial pesticide applicators and mixers receive doses that are hundreds to hundreds of thousands times greater than others in the vicinity. Do phenoxy herbicides cause soft tissue sarcomas in man? Reports from Sweden (Harden, 1981) suggest that man's exposure to phenoxy herbicides used in forestry and agriculture results in an increased rate of soft tissue sarcomas. Many researchers have been critical of the work charging that he failed to accurately separate other potential tumor causing factors. Workers are unlikely to recall with any accuracy the chemicals being used some years in the past and it seems somewhat simplistic to differentiate between workers exposed to phenoxy acetic acid herbicides and those not exposed to this group of herbicides (Kilpatrick, 1980). Estimates of the level of exposure occuring or length of time involved are obscure. The fact that several different sarcoma types were found also make the findings suspect. A study in the Netherlands was conducted by the State Occupational Health Service to evaluate claims by newspapers that personnel involved with 2,4,5-T work in a forested community were encountering a higher prevalance of tumors. A total of 54 of their most highly exposed personnel were selected from a group of 400 workers who had had some exposure. This 54 man group was matched with 54 non-exposed workers (matched by sex, age and residence). Acne and liver dysfunc- tion data were used as the criteria for evaluating the workers. No differences in acne or urinary porphyrin patterns were shown for the forestry workers exposed to 2,4,5-T and the non-exposed group (Van A-40 �Houdt et al.t 1983). The question of soft-tissue sarcoma is important and must not be avoided. Current research has failed to show a cause and effect relationship between exposure to phenoxy herbicides and soft tissue sarcomas. This subject is receiving continuing attention by research being conducted by the National Institute for Occupational Safety and Health (NIOSH). Were there dangerous components of Agent Orange besides TCDD? This herbicide was a 50-50 mixture of the n-butyl esters of 2,4-D and 2,4,5-T. The 2,4,5-T component contained a trace level of TCDD. In normal use patterns such as these employed in U.S. agriculture and in South Vietnam defoliation efforts levels harmful to man would not be encountered. As discussed earlier (see page ) 2,4,5-T represents more of a health hazard to man than does TCDD which is present as a trace component. What happens if 2.4,5-T, which contains a trace level of TCPD, comes into contact with a human? An in-depth discussion of an applicator exposure study where 2,4,5-T was applied by helicopter crews, mist blower crews and backpack spray crews was presented in Chapter 3. Nearly all crewmembers were exposed. None became ill. These results agree with findings discussed by Kilpatrick (1980). Dermal absorption was cited as the primary entry mechanism into the body. Excretion in urine over a 6-day period accounts for over 95% of the absorbed dose. Individuals who mixed the concentrate received the highest absorbed dose. The most highly exposed group, the mixers, could have received an absorbed dose 300 times greater than their actual dose before they A-41 �would have attained a no-adverse-effect dosage for teratogenicity. Another way of expressing this fact is that a safety factor or "A Margin of Safety" of 300 was present. The 20 mg/kg no-adverse-effect dosage for 2,4,5-T teratogenicity as established by the working group in EPA Position Document 1 is a conservative value when employed for other human health effects. Since the developing fetus is quite sen- sitive to toxins this value of 20 mg/kg represents a realistic level for evaluating exposure of females of childbearing age. Thus, the results from the monitored human applicator exposure studies to the phenoxy herbicides showed that the dose the workers received was well below any level which could harm their health. Of the personnel exposed to Agent Orange in Vietnam, only members of Ranch Hand operations could have received exposure levels as high as the field crews discussed here. Ground troops walking through vegetation sprayed with Agent Orange probably received much lower exposure than the Ranch Hand personnel based on evaluation of the following limited study. A worker walking through a field, which 2 hours earlier had been treated with an ester formulation of 2 Ib/acre 2,4,5-T was outfitted with 6 gauze patches attached to his clothing allow to measurement of the extent of dislodgeable 2,4,5-T residues occurring. Analysis of the patches revealed no detectable 2,4,5-T in spite of the fact that the worker walked through chest-high to 15-feet-tall sprayed vegetation. The gas chromatographic technique could have detected levels as low as 10 micrograms. Calculations presented in Chapter 4 indicate that 2,4,5-T, a pesticide relatively non-toxic to man, represents a greater threat to A-42 �man's health than does the highly toxic TCDD - due to the trace amount of TCDD present in 2,4,5-T. As also shown in Chapter 4 some of our commonly used insecticides have toxicity properties that are 60 times greater than 2,4,5-T. Aerial application of some insecticides, classified as very toxic, can be accomplished in a manner that does not appear to pose a health threat to the crewmembers applying them. Were U.S. ground troops in Vietnam "at risk" from Agent Orange since they could not shower daily? Records of 2,4,5-T excretion in urine from commercial applicator exposure studies indicate that although the backpack crewmembers sprayed only 1 day per test some of them received an absorbed dose 1 or 2 additional days during the 6 days following the application day. Observations of the spray operation and crewmember habits suggest that not all of these people changed clothing and bathed daily. Contaminated boots, gloves, or other clothing probably contributed to additional exposure they received 3, 4, or 5 days after the application was completed. As presented in the Lavy (1978) National Forest Product Association report the mixers received the highest exposure, 0.069 mg/kg while the backpack sprayers received 0.048 mg/kg and the flagger received 0.003 mg/kg. An observer in the forest who was sprayed 2 or 3 times with 2,4-D received 0.00056 mg/kg. Following are the margins of safety* for each: Mixers 299 Backpack sprayers 417 Flagmen 6,666 Observer (sprayed 5,333 with helicopter 2 or 3 times) A-43 �NOTE: The actual margin of safety is probably much higher than these values since the NOEL values were obtained from sensitive test animals who daily consumed these compounds in their food; most crewmembers are not exposed daily. These data indicate that workers in the Ranch Hand Operation would have been exposed much more heavily than ground troops; thus, if health problems were to arise due to Agent Orange exposure it should be expressed in health records from the Ranch Hand personnel. If U.S. military ground troops who were exposed to Agent Orange did not take daily showers, they too had the potential for continuing exposure from exposed skin or clothing. Even though the exposure for ground troops may have been prolonged due to the clothes they were wearing these levels would not build up in their bodies since the phenoxy herbicides are readily excreted in urine. The maximum exposure levels that the ground troops could have received were less than that received by crewmembers of backpack spray crews in the U.S. who did not receive health threatening exposure application levels. A survey of long-term health studies of aerial applicators of phenoxy compounds has not revealed any health associated problems (WAA, 1980). Could Agent Orange harm soldiers exposed to it? Information obtained by calculating the trace levels of TCDD in 2,4,5-T and interpreting the human symptomology exhibited by the large TCDD exposed population in Seveso, Italy suggest that humans exposed to 2,4,5-T or Agent Orange have not encountered either serious acute or any chronic health threatening doses of TCDD. Stevens (1981) uses the TCDD exposure fin- A-44 �dings derived from the Missouri Horse Arena Incident and the Seveso, Italy accident to assess the hazards to humans in Vietnam: Agent Orange is an oily liquid composed of equal volumes of the ji-butyl esters of 2,4-D and 2,4,5-T. TCDD, present in Agent Orange in an average concentration of 2 ppm, is extremely insol-uble in water but readily soluble in Agent Orange. The herbicide was delivered as a coarse spray with droplets of 320-350 urn diameter (Young et a!., 1978). While gases readily enter the lungs, only droplets which are less than 10 urn can enter (Walton, 1971), thus Agent Orange could not be absorbed from the lungs. Each gallon of Agent Orange contained 10.7 Ibs (4.9 kg) of total herbicide (Young et al., 1978). One acre is 3660 m2 so the spraying rate of three gallons per acre converts to metric mass units of 4.1 gm~2. Since the mean TCDO concentration in Agent Orange was 2 ppm, the delivered amount of TCDD would be 8 ugm"2. The average concentration of TCDD in Zone B in Seveso, where the people were not evacuated, was 7 ugm"2 (Pocchiari et al., 1979). The conditions for destruction of TCDD in Vietnam were almost ideal. TCDD is very stable in the dark but is photodechlorinated in position eight. The resulting 2,3,7-TriCDD is 10,000 times less toxic (McConnell & Moore, 1979). Photodechlorination requires light, an organic hydrogen donor (Agent Orange works well) and the reaction proceeds three times faster at 30°C, the mean annual daytime temperature in Saigon, than at 23°C (Liberti et al., 1978). When Agent Orange is applied to leaves in full sunlight, the half-life of the contaminant TCDD is only 2 h (Crosby & Wong, 1977). Grass receives less light and some TCDD reaches the soil; the half-life of TCDD applied to turf is about 6 days (Young et al., 1978). Since spraying was done in early morning (Young et al., 1978), a single day of full sun would decrease the TCDD on the leaves from 8 to 0.3 ugrrr2. About 90% of Agent Orange was sprayed on forests and only 6% of that reached the forest floor (Young et al., 1978) to yield 0.5 ugnr2 of TCDD. Hence under forest cover, troops passing through, even under direct spraying, would receive less than 1 ugrrr2 in their environment. For the 10% not sprayed on forests, the early morning concentration of 8 ugrrr2 would be less than half of that by evening. For most soldiers who had any exposure to Agent Orange, that exposure would consist of walking through vegetation which contained no more than 1 ugm"2 of TCDD. Using the transfer factor of 1:2050, the soldier would take in 5 x 10~4 ug of TCDD. For a 70 kg soldier this represents 7 x 10"" ugkg~* or 1/14,000 of the minimum toxic dose of 0.1 ugkg'-"-. If soldiers were exposed to direct spraying from airplanes the intake would still be only 1/1750 of the MTD. Even if a soldier were exposed to 8 ugm"2 each day of his one year tour in Vietnam and each day was treated as a new exposure, his cumulative intake would be only 1.4 ug or 0.02 ugkg"1 of TCDD. A-45 �Since it would take about 5 years of daily close with Agent Orange to reach toxic levels of TCDD, illnesses in Vietnamese (Tung et al., 1971, 1973) Americans (Bogan, 1979) were due to Agent Orange merit" (Stevens, 1981). contact claims that and are without Stevens' description above of the prospects for significant TCDD exposure of U.S. military personnel in Vietnam were made with no reference to the fact that a portion of the TCDD absorbed during the exposure period would also be excreted. Assuming the other con- siderations to be valid, then an even greater margin of safety could be projected. There is another important but difficult fact to evaluate regarding the U.S. spraying program in Vietnam. Many soldiers were accustomed to seeing U.S. planes routinely spray jungle and base perimeter areas; for many it would have been impossible to distinguish between the spraying of malthion, an insecticide, or one of the several defoliants. How could one know if he had been exposed to significant levels of Agent Orange? From the studies on humans accidently exposed to large doses of TCDD in Seveso, Italy in 1976 the most common symptom was the development of chloracne. However, not all of the exposed population developed chloracne. This particular skin disease is, in fact, regarded as the clinical marker for TCDD exposure. Systemic disorders in man from exposures to TCDD are not likely to occur if chloracne is not present. Such symp- toms as impaired liver function, nephropathy, GI irritation, myopathy and neuropathy, including depression and irritation of the central nervous system, have been reported after exposure to large amounts of TCDD in industrial accidents; however, these symptoms have not been A-46 �progressive, and they have always cleared with time (Beljan, et al., 1981). In the Seveso exposed population evidence of chloracne was present 2 weeks after exposure,however some cases took as long as 7 weeks to develop. A total of 175 cases (Pocchiari, 1978) was reported; this accounted for 0.6 to 1.256 of the population exposed. It is highly unlikely that U.S. military personnel in Vietnam were exposed to TCDO levels high enough to cause chloracne, however, if it did occur evidence of chloracne should have developed within 2 to 7 weeks of the time exposure occurred. Chloracne has been described as the primary symptom for men exposed to chlorinated organic compounds, are there other symptoms that would accompany these symptoms? Chloracne development commonly follows industrial accidents involving man's exposure to TCDD. Evaluation of health records of 174 humans, who developed chloracne after the explosion or occupational exposure to some form of dioxin since 1960, revealed no obvious effects of other systemic disorders (Crow, 1981). What has happened to the health of those in Vietnam who mixed and applied Agent Orange? These individuals are a part of the Ranch Hand group where ongoing tests are being conducted. To date this group has not reported health related problems due to their job assignment. If one could accurately extrapolate from exposure data on humans who mix and apply phenoxy herbicides as a vocation in the U.S., no deleterious health effects would be expected. Could Agent Orange affect the fertility of males? Male mice were fed a replication of substances similar to Agent Orange (Lamb et al., A-47 �1980). In previous studies some female mice when exposed to 2,4,5-T at high doses during pregnancy delivered offspring with cleft palate, cystic kidney or retarded growth; no other laboratory animals had shown this effect. Although the dose levels in this study of male mice were sufficient to affect the liver and the thymus, and to reduce growth rates, there was no alteration in the concentration or motility of sperm or in the proportion of abnormal sperm; no significant decrease in fertility was shown and the development of the offspring of treated males appeared to be normal. Can birth defects be caused by pregnant women being exposed to the phenoxy herbicides and TCDD? The human event most sensitive to harm is that of possible birth defects in the developins fetus of pregnant women. The level of any compound required to cause this effect in test animals is much less than the levels required to show other health problems (Hunter et al., 1981). Much of this section has also been taken from Hunter et al., 1981. Experiments with test animals can be used to determine a "no-effect" level (NEL), which describes a definite amount of applied chemical at which no risk exists and birth defects do not occur. The level for the most sensitive test species can then be extrapolated to human exposure. The likelihood of contacting this level of exposure during field use of the chemical will determine its hazard. The NEL's1 for 2,4-D, 2,4,5-T and TCDD are 24 mg/kg/day, 20 ^•Although the term NEL, "No effect level", is used here there remains a dose-response curve for teratogenicity at levels below the NEL's listed. For phenoxy compounds teratogenicity is not a problem because the chemical clears the body so rapidly. A-48 �mg/kg/day and 0.03 ug/kg/day respectively. (Figures for 2,4,5-T and TCDO are in accordance with the U.S. Environmental Protection Agency.) To attain this much exposure a 60 kg pregnant woman would have to ingest 1.6 g/day of 2,4-D, 1.2 g/day of 2,4,5-T or 1.8 ug/day of TCDD (equivalent to 19 g/day of 2,4,5-T containing the maximum permissible level of TCDD) during the critical period of fetal development (Australian Consultative Council, 1978). To help evaluate the possi- bility of this much exposure occurring consider 1) pesticide applicators are among the most highly exposed group of humans, and 2) applicator exposure to 2,4,5-T as discussed in Chapter 3 revealed that the most exposed crewmembers received less than 0.1 mg/kg/d (Lavy, 1978). Exposure could occur by either: a. direct contact with the sprays; or b. residues of chemical in food or water. a. skin absorption and inhalation of sprays Calculations of direct spraying onto 1 square metre of a woman's skin (approximately 2/3 of total area) and allowing for a 10% absorption rate (based on EPA data) would give safety factors of 40 for 2,4,5-T; 600 for TCOD and 60 for 2,4-D. Thus the NEL will be reached for each of these three chemicals by a woman standing in an open paddock, with two-thirds of her skin exposed, while a low-flying aircraft spraying 2,4-D or 2,4,5-T at the normal rate makes between 40 and 600 passes directly over her each day for many consecutive days during the sensitive period of fetal development. Also the woman makes no attempt to wipe or wash the spray A-49 �off for many days. It is highly unlikely that the above scenario could occur; however, it is desirable to err on the safe side, but there are limits beyond which it becomes unrealistic. Assuming 100% of absorption of inhaled chemical, a 10 minute exposure would give safety factors of 15,000; 225,000; and 31,000 for 2,4,5-T, TCDD and 2,4-D respectively (Australian Consultative Council, 1978). b. residues in food and water Table A-8 shows the amounts of spray or foodstuffs which must be ingested by a 60 kg woman in order to obtain the "no-effect" level. A-50 �Table A-8. Amounts of food or spray mix to be ingested per day to reach the NEL of 2,4-D, 2,4,5-T, and TCDD with respect to a 60 kg pregnant woman, as determined from experiments with laboratory animals (Modified from Australian Consultative Council, 1978). Source 2,4-D 2,4,5-T TCDD Spray mix 3 .45 1 0.6 1 9.0 1 Water b 72,000 60,000 1 900,000 1 Milk c 14,500 12,000 1 180,000 1 Meatd 7,200 6,000 kg 90,000 kg a 2,4,5-T is diluted approximately 400-fold for conventional spraying. ^Assuming the 2,4,5-T concentration is at the maximum recommended residue level of 0.02 ppm. c Assuming residue level of 0.1 ppm in milk. d Assuming residue level of 0.2 ppm in meat. NOTE: With respect to the residue levels assumed for water, milk, and meat, these levels are virtually never obtained with normal field use. The calculations involved in determining the values in Table A-8 represent an exaggeration since: 1. The 0.02 ppm maximum recommended residue level for 2,4,5-T in water is based on potential run-off, or direct application to surface waters in catchment areas; not to the residue occurring in normal field use which, in the bulk of instances, is below the A-51 �level of detection; and 2. The residue levels chosen for milk and meat are taken from experimental work in which animals were deliberately fed 2,4,5-T at much higher levels than would be expected to occur through the ingestion of treated pasture. Under normal agricultural use resi- dues of 2,4j5-T in meat and milk are extremely unlikely to occur. In addition to the facts mentioned above, the use patterns of 2,4,5-T are such that potential accumulation of the herbicides would be unlikely. They are used only in areas where weed pressure is high and very few individual areas would be treated more often than once in any year. As with 2,4,5-T general residue surveys have shown that 2,4-D is rarely found in food, and when found the levels are well below the maximum permissible residue levels. The latter values have been set to cover relatively unusual circumstances. Domestic users of 2,4-D and 2,4,5-T can be assured that provided the label recommendations are followed, no harmful effects will be caused to any unborn child, unlike other well documented risks associated with cigarette smoking and alcohol (Chlorinated Dioxin Task Force, 1978; U.S. Dept. of Health Educ. and Wei., 1978). Studies were designed to determine whether paternal exposure to TCDD or other dioxins might be associated with adverse pregnancy outcomes. Wives of husbands potentially exposed were compared to a group whose husbands had no known exposure to any dioxin. During the study in groups over 2700 conceptions and 2400 live birth resulted; overall no statistically significant associations were found between any expo- A-52 �sure and pregnancy outcomes (Townsend et al., 1982). Does human mother's milk contain TCDD? This question is of interest since if mother's milk did contain TCDD, a large proportion of her infants nourishment could contain the toxin. If a nursing mother's body .contained TCDD the fat content of the milk would most likely contain some TCDD. Before TCDD could be found in human mother's milk it must be present in her body at a concentration detectable by modern methodology. Techniques are now available which permit the scientist to be able to detect TCDD at, or below, the 1 ppt level. In studies reported by Shadoff (1980) of 4 correctly analyzed human milk samples none contained levels of TCDD above the 1 ppt detection limit. This finding is not surprising since man's food supply seldom contain detectable levels of TCDD. Another report in 1980 states "All 103 mothers' milk samples from the 2,4,5-T spray area were negative for dioxin, EPA noted this week. The Agency said the samples were taken from 2,4,5-T spray areas in California, Oregon and Washington in 1977. Samples were also collected in control areas of California and Alaska not sprayed with dioxin-containing pesticides, the agency said. It noted that the samples were checked on equipment capable of measuring residues down to 1 to 4 parts per trillion" (Pesticide and Toxic Chemical News, 1980). How do the teratogenic effects of TCDD and 2,4,5-T compare? much more toxic compound than 2,4,5-T. TCDD is a Cranmer (1978) has calculated that TCDD is 10,000 times more teratogenic than 2,4,5-T. However, since 2,4,5-T is present at levels which are 10,000,000 times greater A-53 �than TCDD, the TCDD would be required to bioaccumulate by at least 1000 times before it would equal the toxic effect that the larger amount of 2,4,5-T possesses. As discussed earlier the low application rates of 2,4,5-T (approximately 2 Ib/acre) coupled with its short half-life (20-40 days) in the environment make it impossible for the agricultural use of 2,4,5-T to represent a threat to man's health. Were U.S. military personnel who used Agent Orange informed about its properties at the time of its use in Vietnam? Facts regarding how to handle these herbicides were printed in a Department of Army information manual in 1969 (Irish et al., 1969). The publication indicated that spillage onto skin or clothing should be removed promptly to avoid possible irritation through prolonged skin contact. This same report also stated "These chemicals present no hazard to man or animals in areas subjected to defoliation at the time of or following spray application." Results from experiments conducted later, which were designed to measure the amount of 2,4,5-T and 2,4-D exposure received by field workers routinely mixing and applying these chemicals, appear to substantiate the statements presented by Irish et al. in 1969. Results from the most highly exposed 2,4-D and 2,4,5-T applicators revealed that the amount of exposure occurring did not result in absorbed doses approaching health threatening levels (Lavy et al., 1980; Lavy et al., 1982; Nash et al., 1982; Nigg, 1983). A-54 �What evidence do we have which indicates that use of defoliants In South Vietnam has not caused man health problems? Ground troops would have been exposed to that material which penetrated the canopy. Using phenoxy herbicides similar to Agent Orange Tschirley (1969) found that about 6% of the material applied actually reached ground level. The percentage falling to the ground was lower in areas where forest canopy was the thickest. From studies on commercial applicator exposure studies we have learned that those standing under aircraft which was applying phenoxy herbicides do not receive nearly as much exposure as those who mix and prepare the concentrated materials (Lavy et al., 1980; Lavy et al., 1982). Personnel in Vietnam who routinely mixed the concentrates and applied the defoliants to the jungle (the Ranch Hand personnel) would have received a much higher absorbed dose than ground troops. Ongoing monitoring of the health effects of this group has not indicated an increased incidence in death rate or in health problems. Possibly one of the reasons that a man is not injured when exposed to the phenoxy compounds is that his absorbed dose of phenoxy compounds is rapidly and nearly quantitatively excreted in his urine, thus, no build-up occurs in his body. Since these materials are not stored in fatty tissues of humans any health problems arising from their use should occur near the time of exposure. Why do the Veteran's Administration and other government agencies continue to study Vietnam veteran health problems if the government is not admitting that Agent Orange exposure has caused the problems? As responsible interested agencies, the VA, CDC et al. are continuing to A-55 �search for possible scientific evidence which could link Agent Orange exposure to the health effects being reported. The search for new facts should not be interpreted to mean that these agencies believe there is a positive relation between exposure to Agent Orange and consequent health problems; it is their duty to explore all of the factors which could contribute to the health problems reported. A-56 �APPENDIX References Allen, J. R. and J. P. van Miller. 2,3,7,8-Tetrachlorodibenzo-p-dioxin K. R. Rao (ed.), Penachlorophenol; Environmental Toxicology, New York, 1978. Health implications of exposure in primates. In: Chemical Pharmacology and Plenum Press. Assouly, M. 1951. Selective herbicides and growth substances. Pathologic effects on man during the manufacture of the ester of 2,4-D. Arch. Mol. Profess. 12:26-30. Australia. 1978. Consultive Council on Congenital Abnormalities in the Yarram District, Victoria, Australia, Department of Primary Industry, Canberra, Report. Beale, M. G., W. T. Shearer, M. M. Karl, and A. M. Robson. 1977. Long-term effects of dioxin exposure. Lancet 1:748. Beatty, P. W., M. A. Holscher, and R. A. Neal. Contarn. Toxicol. 5:578. 1976. Bull. Environ. Beljan, J. R., N. S. Irey, U. W. Kilgore, K. Kimura, R. R. Suskind, J. J. Vostal, and R. H. Wheater. 1981. The health effects of Agent Orange and polychlorinated dioxin contaminants. Technical report prepared by the Council on Scientific Affairs, American Medical Association, Chicago. Pp. 39. Berkley, M.C. and K.R. Magee. a diethylamine salt of 2,4-D. 1963. Neuropathy following exposure to Arch. Intern. Med. 111:351-352. Berwick, P. 1970. 2,4-Dichlorophenoxyacetic acid poisoning in man. JAMA 214:1114-1117. Bleiburg, J., M. Wallen, R. Brodkin, and I. L. Applebaum. 1964. Industrially acquired pophyria. Arch. Dermitae 89:793-97. Bonaccorsi, A., A. di Domenico, R. Panel!i, F. Merli, R. Motta, R. Vanzati, and 6. Zapponi. 1983. Peer Review Scientific Panel on Dioxin Consensus Statement, 7/29/83. Bogan, 6. 1979. Symptoms in Vietnam veterans exposed to Agent Orange. JAMA 24:23-91. Bovey, R. W. and A. L. Young. 1980. The Science of 2,4,5-T and Associated Phenoxy Herbicides. New York, John Wiley and Sons. Burke, C. 1983. U.S. Checking 6 Illinois Sites for "Doomsday Chemical" Dioxin. Chicago Tribune, January 9. Carter, C. D., R. D. Kimbrough, J. A. Liddle, R. E. Cline, M. M. Zack, Jr., and U. F. Barthel. 1975. Tetrachlorodibenzodioxin: An accidental poisoning episode in horse arenas. Science 188:738-40. A-57 �Case, A. A. 1973. Vet, din. North Am. 3:273. CAST (Council for Agricultural Science and Technology). 1978. Phenoxy Herbicides. 2 ed. Ames, Iowa State Univ., 1978. The CAST. 1983. The Missouri Dioxin Controversy: Scientific Overview Council for Agricultural Science and Technology. April, 1983. Chlorinated Dioxin Task Force. 1978. The Trace Chemistries of Fire. The Source and Routes for the Entry of Chlorinated Dioxins into the Environment. Midland, Michigan, Dow Chemical Company. Cranmer, M. USDA/EPA Symposium. Arlington, Virginia, 1978. The Use of Herbicides in Forestry. , Crosby, D. 6. and A. S. Wong. 1977. Environmental degradation of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Science 195:1337-38. Crow, K. D. 1981. Chloracne and its potential clinical implications. Clin. Exper. Dermatol. 6:243-57. Cupello, J. M. and A. L. Young. 1976. Radiochemistry Bioassay of TCDD Uptake in Plant Material. Extract Annual Research Progress Report No. 11, Dean of Faculty, U.S. Air Force Academy, p. 22. Ferry, D. G., L. R. Gazeley, and 0. R. Richards. 1982. 2,4,5-T absorption in chemical applicators. Proc. Univ. Otago Med. Sch. 60(2):31-33. Fisher, J. R. 1977. Letter PG-12 in Final EIS Vol. 1. Vegetation Management with Herbicides. U. S. Department of Agriculture Forest Service. Gannon, L. 1983. The buying of Times Beach: beings. Science News 123:132-33. Getzendaner, N. E. and R. A. Hummel. Unpublished Data. 1975. A town unfit for human Dow Chemical USA Greig, J. B., G. Jones, W. H. Butler, and J. M. Barnes. Cosmet. Toxicol. 11:585. 1973. Food Hardell, L. 1981. Relation of soft tissue sarcoma, miligant, lymphoma, and colon cancer to phenoxy acids, chlorophenols and other agents scanned. J_. Work Environ. Health 7:119-30. Hayes, W. J. 1963. Chlorophenoxy herbicides. In Clinical Handbook on Economic Poisons. Emergency Information for Treating Poisoning. U.S. Department of Health, Ed Wolf, Public Health Service Public. No. 476. Pp. 106-09. Helling, C. S., A. R. Isansee, and E. A. Wollson. 1973. Chlorodioxins in pesticides soils and plants, J_. Environ. Qua!. 2:171-78. A-58 �Henck, J. W., M. A. New, R. J. Kociba, and K. S. Rao. 1981. Appl. Pharmacol. 59:405. Toxicol. Higginbothatn, G. R., A. Huang, D. Firestone, J. Berrett, J. Ress, and A. D. Campbell. 1968. Chemical and toxicological evaluations of isolated and synthetic chloro-derivatives of dibenzo-p-dioxin. Nature 220:702-12. Hunter, H. C. et al. 1981. A Report on 2,4-D, 2,4,5-T and human health. By an Interdepartmental Committee Appointed by Queensland Cabinet. Pp. 38. House, W. B., L. H. Goodson, H. M. Gadberry, and K. W. Docter. 1967. Assessment of ecological effects of extensive or repeated use of herbicides. Final Report on Contract DAHC15-68-C-0119. Kansas City, Midwest Research Institute. Irish, K. R., R. A. Darrow, and C. F. Minarik. 1969. Information manual for vegetation control in southeast Asia. Dep. of Army. Miscellaneous Publication 33, Fort Detrick, Maryland. Isensee, A. R. and G. E. Jones. 1971. Absorption and translocation of root and foliage to 2,4-dichlorophenol, 2,7-dicholorodibenzo-pdioxin and 2,3,7,8-tetrachlorodibenzo-p-dioxin. J. Agric. Food Chem. 19:1210-14. ~ Jensen, D. J., R. A. Hummel, N. H. Mahle, C. W. Kocher, and H. S. Higgins. 1981. A residue study on beef cattle consuming 2,3,7,8tetrachlorodibenzo-p-dioxin. JL Agric. Food Chem. 29:265-68. Johnson, J. E. 1971. The public health implications of widespread use of phenoxy herbicides and picloram. BioScience 21:899-903. Kalter, H. 1960. Teratology of the Central Nervous System. Univ. of Chicago Press. Chicago, Kearney, K. C., E. A. Wollson, and C. P. Ellington. 1972. Persistence and metabolism of chlorodioxins in soils. Environ. Sci. Techno!. 69:1017-19. Kearney, K. C., A. R. Isensee, and C. S. Helling. 1973. Environmental significance of chlorodioxins in soils. In: Ectyl Blaired Chlorodioxins - Origin and Fate. Advances in Chemistry Series. Vol. 120 Am. Chem. Soc. Wash., D.C. 1973. Pp. 105-11. Kilpatrick, R. 1980. Further Review of the Safety for Use in the U.K. of the Herbicide 2,4,5-T. Advisory Committee on Pesticides. United Kingdom. Kimbrough, R. D., C. D. Carter, J. A. Liddle, R. E. Cline, and P. E. Phillips. 1977. Epidemiology and pathology of a tetrachlorodibenzodioxin poisoning episode. Arch. Environ. Health 32:77-85. A-59 �King, C. 1965. J. Pharmacol. Exp. Ther. 147:391. Kociba, R. J. and B. A. Schwetz. 1982. Toxicity of 2,3,7,8tetrachlorodibenzo-p-dioxin (TCDD). Drug Metab. Rev. 13:387-406. Kutz, T. W. 1981. Chemical exposure monitoring in the EPA Office of Pesticides and Toxic Substances Presentation to the Veterans Administration Advisory Committee on the Health Effects of Herbicides. Washington, November 19, 1981. Lamb, J. C. et al. 1980. U.S.A. National Toxicology Program. Report No. NTP-80-44 "Evaluation of 2,4-D Chlorophenoxy acetic acid (2,4-D) 2,4,5-Trichlorophenoxyacetic acid (2,4,5-T) and 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) toxicity in C57BL/6 mice; reproduction and fertility in treated male mice and evaluation of congenital malformations in their offspring." Lavy, T. L. 1978. Measurement of 2,4,5-T Exposure of Forest Workers. Project Completion Report to National Forest Products Association. Lavy, T. L., J. S. Shepard, and J. D. Mattice. 1980a. Exposure measurements of applicators spraying (2,4,5-trichlorophenoxy) acetic acid the forest, vh Agric. Food Chem. 28:626-30. Lavy, T. L., J. S. Shepard, and D. C. Bouchard. 1980b. Field worker exposure and helicopter spray pattern of 2,4,5-T. Bull. Environ. Contain. Toxicol. 24:90-96. Lavy, T. L., et al. 1982. (2,4-dichlorophenoxy)acetic acid exposure received by aerial application crews during forest spray operations. J_. Agric. Food Chem. 30:375-81. Liberti, A., D. Bracco, I. Allegrini, A. Cecinato, and M. Possanzini. 1978. Solar and ultroyiolet photodecomposition of 2,3,7,8tetrachlorodibenzo-p-dioxin in the environment. Sci. Total Environ. 10:97-104. Lobes, L.A., R.E. Koehler, W.F. Barthel, R.A. Feldman, and J.V. Bennett. 1972. Administrative report to the director of the Center for Disease Control (CDC) on toxic illness, Lincoln County, Missouri, CDC No. EPI-72-13-2, U.S. Public Health Service, CDC, Atlanta, Georgia. May, G. 1973. Chloracne from the accidental production tetrachlorodibenzo-p-dioxin. Br. J_. Ind. Med. 30:276-83. May, G. 1982. Tetrachlorodibenzodioxin: a survey of subjects ten years after exposure. Br. J_. Ind. Med. 39:128-35. McConnell, E. E. and J. A. Moore. 1979. Toxico pathology characteristics of the halogenated aromatics. Ann. NY Acad. Sci. 320:138-50. A-60 �McQueen, E. 6., et al. 1977. 2,4,5-T and Human Birth Defects. Report Prepared in the Discussion of Public Health Dept. of Health, New Zealand. Morton, H. L., E. D. Robison, and R. E. Meyer. 1967. Weeds 15:268. Nash, R. 6,, P. C. Kearney, J. C. Maitlen, C. R. Sell, and S. N. Fertig. 1982. Agricultural applicator exposure to 2,4-dichlorophenoxy acetic acid. ACS Symposium Series 182, American Chemical Society. National Research Council Canada. effects on environmental quality. 1978. Phenoxy herbicides - Their NRCC No. 16075, Ottawa, Canada. Neal, R. A., J. R. Olson, T. A. Jasiewicz, and L. E. Geiger. 1982. The toxicokinetics of 2,3,7,8-tetrachlorodibenzo-p-dioxin in mammalian systems. Drug Metab. Rev. 13:355-85. Neil sen, K., B. Kaempe, and J. Jensen-Holm. 1965. Fatal poisoning in man by 2,4-dichlorophenoxyacetic acid (2,4-D): Determination of the agent in forensic materials. Acta. Pharmocol. Toxicol. 22:224-34. Newton, M. and S. P. Snyder. 1978. Exposure of forest herbivores to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in areas sprayed with 2,4,5-T. Bull. Environ. Contam. Toxicol. Nigg, H. N. and 0. H. Stamper. 1983. Exposure of Florida airboat aquatic weed applicators to 2,4-dichlorophenoxy acetic acid (2,4-D). Chemosphere 12:209-15. Norris, L. A., M. L. Montgomery, and E. R. Johnson. 1977. The persistence of 2,4,5-T in a Pacific Northwest Forest. Weed Sci. 25:417-22. Olson, J. R., M. A. Holscher, and R. A. Neal. Pharmacol. 55:67. Pesticide Chemical News. 1980. Toxicol. Appl. 1976. Washington, June 23, 1976. Pesticide and Toxic Chemical News. Volume No. 8, Jan. 16, 1980. 1980. P. 17. Mothers milk samples. Piper, W. N., J. L. Rose, and P. J. Gehring. 1973. Excretion and tissue distribution of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the rat. Environ. Health Perspect. 5:241-44. Pocchiari, F., B. Silano, and A. Zampieri. 1979. Human health effects from accidental release of tetrachlorodibenzol-p-dioxin (TCDD) at Seveso, Italy. Ann. NY Acad. Sci. 320:311-20. Poiger, H. and C. H. Schlatter. 1979. A-61 Nature 281:706. �Poiger, H. and C. Schlatter. 1980. Influence of solvents and adsorbents on dermal and intestinal absorption of TCDD. Food Cosmet. Toxicol. 18:477-81. Poland, A. P., D. Smith, G. Wetter, and P. Possick. 1971. A health survey of workers in a 2,4-D and 2,4,5-T plant. Arch. Environ. Health 22:316-27. Rawls, R. 1. and D. A. 0'Sullivan. 1976. Italy seeks answers following toxic release. Chem. Eng. News. Aug. 23, 1976. Pp. 27-35. Reggiani, G. 1981. Toxicology of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD): Short review of it's formulation occurrence toxicology and kinetics, discussing human health effects, safety measures, and disposal. Regulatory Toxicol. Pharmacol. 1:211-43. Reggiani, G. 1978. Medical problems raised by TCDD contamination in Seveso, Italy. Arch. Toxicol. 40:161-88. Rose, H. R. and S. P. R. Rose. 1972. Chemical spraying as reported by refugees from South Vietnam. Science 177:710-12. Schmittle, S. C., A. M. Edwards, and D. Morris. 1958. A disorder of chickens probably due to a toxic feed—preliminary report. J_. Am. Vet. Med. Assoc. 143:216-19. Schwetz, B. A., J. M. Norris, G. L. Sparschu, V. K. Rowe, P. J. Gehring, J. L. Emerson, and C. G. Gerbig. 1973. Environ. Health Perspect. Exp. 5:87. Shadoff, L. A., R. A. Hummel, L. Lampachki, and J. H. Danidson. 1977. A search for 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in an environment exposed annually to 2,4,5-trichlorophenoxyacetic acid esther (2,4,5-T) herbicides. Bui 1. Environ. Contam. Toxicol. 18:478-85. Shadoff, L. A. 1980. The determination of 2,3,7,8-tetrachlorodibenzop-dioxin in Human Milk. Am. Chem. Soc. Shepard, B. M. 1980. Proceedings from the Second Continuing Education Conference on Herbicide Orange. Washington, May 1980. Smith, F. A., F. J. Murray, J. A. John, K. D. Nitschke, R. J. Kociba, and B. A. Schwetz. 1981. Three-generation reproduction study of rats ingesting 2,4,5-trichlorophenoxyacetic acid in the diet. Food Cosmet. Toxicol. 19:41-45. Sparschu, G. L., F. L. Dunn, and V. K. Rowe. 1971. Study of the teratogenicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the rat food. Food Cosmet. Toxicol. 9:405-12. Squibb, R.E., H.A. Talson, and C.L. Mitchell. 1983. Neuro-behavioral assessment of 2,4-dichlorophenoxyacetic acid (2,4-D) in rats. Neurobeh. Toxicol. Teratol. 5:331-335. A-62 �Stevens, Kingsley M. 1981. Agent Orange toxicity: A quantitative perspective. Human Toxicol. 1:31-39. Terry, D. 6., L. R. Gazeley, and I. R. Edwards. 1982. 2,4,5-T absorption in chemical applications. Proc. Univ. Otago Med. Sch. 60:31-32. Todd, R.L. 1962. A case of 2,4-D intoxication. J. Iowa Med. Sco. 52:663-664. Townsend, J. C., K. M. Bodner, P. F. D. Vanpeenen, R. D. Olson, and R. R. Cook. 1982. Survey of reproductive events of wives of employees exposed to chlorinated dioxins. Am. J. Epidemiol. 115:695-713. Tshirley, F. H. 1969. Research Report - Response of tropical and subtropical woody plants to chemical treatment. Agric. Res. Service, U.S. Dept. of Agriculture, ARPA Order No. 424. Tshirley, F. H. 1971. Report on status of knowledge regarding 2,4,5-T. Submitted by USDA to EPA. March 5, 1971. 2,4,5-T Advisory Commi ttee. U.S.D.A. 1982. The biologic and economic assessment of 2,4,5-T. Technical Bulletin 1671 U.S.D.A. U.S. Dept. of Health, Education, and Welfare. 1978. Smoking and Health - A Report to the Surgeon General - Pregnancy and Infant Health. Publication Number PHS 79-50066. Van Houdt, J.J., L.G. Fransman, and J.J.T.W.A. 1983. Epidemiological case control study in personnel exposed to 2,4,5-T. Chemosphere 12:575. Walton, W.H. (ed.). Greshan Press. 1971. Inhaled particles. Old working, surrey. Watts, R. R. and R. W. Storherr. 1973. Negative finding of 2,3,7,8-tetrachlorodibenzo-p-dioxin in cooked fat containing actual and fortified residues of ronnel and/or 2,4,5-T trichlorophenol. J^. Assoc. Off. Anal. Chem. 56:1026-27. Wipf, H. K., E. Homberger, N. Neuner, and F. Schenker. 1978. Field trials on photodegradation of TCDD on vegetation after spraying with vegetable oil. In: Dipxin: Toxicological and Chemical Aspects. New York, Spectrum. Pp. 201-17. World of Agricultural Aviation (WAA). 1980. Investigation of the possible effects of pesticide exposures on reproductive mortality and morbidity. 7:12-30. Young, A. L., C. E. Thai ken, and W. E. Ward. 1975. Studies of the ecological impact of repetitive aerial applications of herbicides on the ecosystem of test area C-52A, Eg!in Air Force Base, Florida Technical Report AFATL-TR-75-142 Air Force Armament Lab. Pp. 127. �Young, A. L., et al. 1978. The Toxicology Environmental Fate and Human Risk of Herbicide Orange and Its Associated Dioxins. USAF Occupational and Environmental Health Laboratory Technical Report Number TR-78-92, Final Report Brooks Air Force Base, Texas. Young, A. L. 1981. Agent Orange at the Cross-roads of Science and Social Concern. Air Command and Staff College Student Research Report, Maxwell Air Force Base, Alabama. Zabik, M. E. and M. J. Zabik. 1980. Dioxin levels in raw and cooked liver, loin steaks, round, and patties from beef fed technical grade pentachlorophenol. Bull. Environ. Contam. Toxicol. 24:344-49. A-6J. � 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 207 Folder The folder containing the original item. 5827 Series The series number of the original item. Series VIII 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 Lavy, T. L. Date A point or period of time associated with an event in the lifecycle of the resource 1984 Title A name given to the resource Typescript: Human Exposure to Phenoxy Herbicides ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/71ffbcbbad4ac1cfa431d23b00c5f26a.pdf ca7e266c8bcc9d9bd27ef2466f812d0a PDF Text Text item D Number °5241 Leng, Marguerite L. Corporate Author RBpOrt/ArtiClB Title Letter to Federal Register Section, Technical Services Division, Office of Pesticide Programs, EPA from Marguerite L. Leng, July 7, 1978 Journal/Book Title Year 000 ° Month/Day Color D Number of Images ° DBSCrlptOn NOtBS Letter has subject line 2,4,5-T Rebuttable Presumption Against Registration OPP 30000/26. Leng is responding to the RPAR and encloses two of her papers. First is "Comparative Metabolism of Phenoxy Herbicides in Animals" in Fate of Pesticides in Large Animals, 1977 (not scanned), and the second is a draft of "Government Requirements for Pesticide Residue Analyses and Monitoring Studies" Chapter 11 in Pesticide Residue Analysis, January 31, 1977. Friday, March 01, 2002 Page 5241 of 5263 �l>1*r* 1714 Sylvan Lane Midland, Michigan 48640 July 7, 1978 Federal Register Section Technical Services Division (WH-569) Office of Pesticide Programs, EPA Room 401, East Tower 401 M Street SW Washington, D. C. Subject: 20460 2,4,5-T Rebuttable Presumption.Against Registration OPP 30000/26 This response to the 2,4,5-T RPAR is prompted by my dedication as a scientist who is continually seeking knowledge about the effects of chemicals on man and his environment. I am submitting it as an individual who is an active member in the Division of Pesticide Chemistry of the American Chemical Society, rather than as a person who is currently employed by Dow Chemical, one of the principals in the continuing debate on the safety of this useful herbicide. Enclosed are pertinent excerpts of two documents dealing with studies on phenoxy herbicides in humans, and with the reliability of data on residues of 2,4,5-T and its trace contaminant TCDD in foods and environmental samples. The first paper summarizes three studies with phenoxy herbicides in humans. It includes a discussion about the fallacy of using data from studies at highly exaggerated dosage levels to predict what might happen at levels likely to be encountered from recommended uses of 2,4,5-T. (This paper was cited as reference 79 in EPA's 2,4,5-T Position Document No. 1, but only as a basis for calculating highly exaggerated residues for 2,4,5-T in meat and milk.) The second enclosure is from my draft chapter on government requirements for pesticide residue data and monitoring studies, scheduled for publication in 1978. As indicated in this excerpt, many of the allegations of hazard from use of 2,4,5-T are based on data from studies which do not meet even minimum requirements for such investigations. Sincerely, / ' Marguerite L. Leng, Ph.D. abc cc to attached list. �M. L. Leng letter dated 7/7/78 2,4,5-T Rebuttable Presumption Against Registration OPP 30000/26 Copies to: Federal Register Section Technical Services Division (WH-569) Office of Pesticide Programs, EPA Room 401, East Tower 401 M Street SW Washington, D. C. 20460 Dr. Gerald G. Still, Chairman Division of Pesticide Chemistry American Chemical Society USDA, ARS West Building 005 Beltsville, MD 20705 Mr. Edwin L. Johnson Deputy Assistant Administrator (WH-566) Office of Pesticide Programs, EPA 401 M Street, SW Waterside Mall, East Tower Washington, D. C. 20460 Dr. Philip C. Kearney Chairman, Technical Advisory Group National Agricultural Pesticide Impact Assessment Program USDA, Office of Environmental Activities Office of the Secretary 14th and Independence Ave., SW Washington, D. C. 20250 Dr. William A. Wells Acting Administrator Special Pesticide Reviews Div. (WH-566) U.S. Environmental Protection Agency Office of Toxic Substances 401 M Street, SW Waterside Mall, East Tower Washington, D. C. 20460 Representative James G. Martin (North Carolina) 115 Cannon House Office Building Washington, D. C. 20515 Mr. Douglas M. Costle Administrator (A-100) Environmental Protection Agency Room 1200 Waterside Mall, West Tower 401 M Street SW Washington, D. C. 20460 Harvey L. Warnick Office of Special Pesticide Review (WH-566) Environmental Protection Agency Waterside Mall, East Tower 401 M Street SW Washington, D. C. 20460 Dr. Anna Harrison, President American Chemical Society Carr Laboratory Mount Holyoke College South Hadley, Massachusetts 01075 �GOVERNMENT REQUIREMENTS FOR PESTICIDE RESIDUE ANALYSES AND MONITORING STUDIES by Marguerite L. Leng Government Registration Dow Chemical U.S.A. Midland, Michigan Chapter 11 in PESTICIDE RESIDUE ANALYSIS edited by H. Anson Moye No. 43 in Wiley Interscience Series of Monographs on Analytical Chemistry and Its Applications Philip Ebving and James D. Winefordner, Editors January 31/ 1977 �TABLE OF CONTENTS 1. INTRODUCTION 2. REGULATION OF PESTICIDES 2.1 Government Agencies 2.2 Laws Regulating Pesticides 2.2.1 The Federal Insecticide/ Fungicide, and Rodenticide Act 2.2.2 The Federal Food, Drug, and Cosmetic Act 2.2.3 Implementation of Laws on Pesticide Residues 2.2.4 3. History of Tolerances for 2,4-D in Food Crops RESIDUE DATA REQUIREMENTS 3.1 Analytical Methodology 3.1.1 Validation of the Method 3.1.2 Extraction Efficiency 3.1.3 3.1.4 Total Residue Including Metabolites Regulatory Method 3.2 Residue Studies in Plants 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.2.6 3.2.7 3.3 Field Experiments Sampling Residues in Crops Residues in Crop Byproducts Crop Groupings for Residue Data Negligible or Low Level Residues Temporary Tolerances Residue Studies in Animals 3.3.1 3.3.2 3.3.3 Residues from Treated Plants Residues from Treatment of Animals Residues from Treatment of Agricultural Premises 3.4 Miscellaneous Residue Studies 3.4.1 Aquatic Sites 3.4.2 Stored Commodities 3.4.3 3.4.4 3.4.5 3.4.6 3.4.7 Food Handling Establishments Seed Crops and Seeds Postharvest Treatments Minor Changes in Use or Formulation Imported Foods 3.4.8 Environmental Studies �3.5 Reporting of Residue Data 3.5.1 3.5.2 3.5.3 3.5.4 4. Identification of Report Residue Summary Body of Report Tables and Figures MONITORING STUDIES 4.1 National Pesticides Monitoring Program 4.1.1 Working Group and Monitoring Panel 4.1.2 Pesticides Monitoring Journal 4.1.3 Pesticides in National Monitoring Program 4.2 National Food and Feed Monitoring Program 4.2.1 Surveillance of Unprocessed Food and Feed 4.2.2 Surveillance of Meat and Poultry 4.2.3 Total Diet Studies on Market Basket Samples 4.3 Pesticide Residues in Environmental Samples 4.3.1 Monitoring Studies in Fish/ Wildlife and Estuaries 4.3.2 Monitoring for Pesticides in Soil/ Water and Air —— 4.4 Monitoring in Humans -—— 4.5 Special Monitoring Programs 5. INTERPRETATION OF RESIDUE DATA *—"— 5.1 Variability among Samples 5.2 Sensitivity vs Limit of Detection —— 5.3 Interpretation of Monitoring Data —— 5.4 Confirmation of Residue Identity — "• 5.5 Statistical Evaluation of Results 6. CONCLUSION 7. ACKNOWLEDGEMENT 8. REFERENCES �-4- 1. INTRODUCTION Requirements for analysis of pesticide residues have increased vastly since the 1950's when regulations were first implemented to require tolerances (maximum residue limits) for these useful chemicals in food crops intended for human and animal consumption. The evolution of changes in such requirements can be traced in reviews published between 1955 and 1971 by Gunther and Blinn (1), Pogelman (2), Harris (3), Vorhes (4), Frehse (5), Harris and Cummings (6), and Bevenue and Kawano (7). In 1968/ the U.S. Food and Drug Administration issued FDA Guidelines for Chemistry and Residue Data Requirements of Pesticide Petitions (8). These official guidelines described the studies needed at that time for obtaining analytical data suitable for establishment of tolerances for pesticides in food and feed crops and their byproducts, as well as in meat, milk, poultry and eggs. During the early 1970's bills were passed in the U.S. House and Senate proposing major revisions in the Federal Insecticide, Fungicide/ and Rodenticide Act (FIFRA), resulting in passage of the Federal Environmental Pesticide Control Act (FEPCA) on October 8, 1972. Detailed regulations for implementation of Section 3 on Registration of Pesticides were finalized by publication in the Federal Register on July 3, 1975 (9). �-5- At that time certain sections were "Reserved" for later issuance, Registration Procedures were published as regulations on September 9, 1975 (11). A preliminary draft of sections on Petitions for Tolerances (12) was made available by EPA in 1975 for comment by outside experts, but a revised version has not been issued as of this writing. Review of the draft indicated that requirements for residue data by the Environmental Protection Agency (EPA) under FIPRA as amended will be similar but more extensive than those outlined by FDA under the Food, Drug ( and Cosmetic Act in' 1968 (8) . Proposed guidelines for studies to meet Section 3 Regulations were published in the Federal Register on June 25, 1975 along with extensive appendices outlining how the studies should be conducted (10). Final guidelines are scheduled for publication in 1977 incorporating revisions based on comments by other government agencies, industry, and any interested parties. This chapter presents a brief outline of how certain government agencies control.pesticides, of the laws under which they operate, of regulations implemented to carry out this task, and of how tolerances were established for the herbicide 2,4dichlorophenoxyacetic acid (2,4-D) over a period of 25 years under these laws and regulations. It also provides a detailed description of current requirements for residue data based on �proposed EPA Guidelines, based in part on personal experience during more than ten years of compiling data in petitions for tolerances, including many volumes on residue studies with 2,4-D. Finally, it discusses monitoring studies for pesticide residues in raw agricultural commodities, total diet samples, and environmental samples, as well as pitfalls in interpretation of "positive" findings. �-78- 4. MONITORING STUDIES During the 1960's, concern grew about potential long-term effects of persistent pesticides in the environment. This concern was fed in part by reports of finding residues of DDT in samples such as pelicans, walruses, and even soil far removed from where the insecticide had been used. The impact of such reports has continued long after many of the results of early studies have been refuted by more specific analytical data. Positive findings originally attributed to DDT have often been shown to be due to the presence of PCB's (polychlorinated biphenyls) (48). More recently, environmental and laboratory contamination by phthalate esters has been shown to cause responses frequently mistaken for DDT residues (49). Thus, it is imperative that analytical methods used in monitoring for pesticides are specific for the chemicals being sought and that positive findings are confirmed by secondary methods based on different principles of analysis. 4.1 NATIONAL PESTICIDES MONITORING PROGRAM The National Pesticide Monitfsring Program (NPMP) was initially designed on the basis of the minimum monitoring needed to establish baseline levels of pesticideSvin substrates of food and feed, humans, soil, water, air, wildlirX, fish and estuaries, �-98- Variables such as wihdjrain, temperature, snow, and inversion are unpredictable and sfr®nl.d be recorded during the sampling period. 4-4 MONITORING IN HUMANS Concern about potential long-term effects of persistent chemicals led to the National Human Monitoring Program for Pesticides. In 1967, the Division of Pesticide Community Studies initiated a program in which adipose tissue (fat) was to be collected from postmortem and surgical specimens (50). The design provided for sampling which would yield valid data concerning pesticide incidence variation with geographic distribution, and with age, sex, and race variation of the donors. The sampling population was drawn from sources in cities of 25,000 or more population (1960 figures) distributed among the four census regions (Northeast, North Central, South, and West), without regard for the donor's occupation. The total number of collection points (or pathologists) was 39 and each was to provide a minimum of 50 samples per year, for an annual total of at least 1950 random samples. For each collection point, the sample quota was further broken down by age and sex: 14 years and under, male and female; 15 through 49 years, male and female; and 50 years and over, male and female. Since the program objective was to reflect �-99- the incidences in the normal (man-on-the-street) general population, samples were excluded from cases of known or suspected acute pesticide poisoning or chronic debilitating illness, and from patients who had been institutionalized for long periods. Analyses of human adipose tissue were initially limited to the measurement of certain chlorinated hydrocarbons due to sensitivity limitations of available analytical methodology. The program originally included only DDT and its metabolites, dieldrin, heptachlor epoxide, and the isomers of BHC. As refinements of methodology progressed, additional pesticide chemicals or metabolites were added such as oxichlordane and mirex. Other pollutants such as PCB's (polychlorinated biphenyls) were added as they became identifiable. 4.5 SPECIAL MONITORING PROGRAMS Special programs are instituted as problems arise or are suspected. Many of these are a result of industrial pollution or accidental contamination rather than from direct use of pesticides. One such example was a monitoring program for mercury, or more specifically methyl mercury, in fish and bottom sediments in the Great Lakes attributed to dumping of wastes from giant electrolytic processes. A more localized �-100- problem was contamination of cattle and poultry in Michigan beginning in 1973 from ingestion of feed containing polybrominated biphenyls (PBB's). Intensified programs are also put into operation when emergency use of cancelled pesticides is authorized. During recent years, EPA has taken stepwise action against DDT, mirex, aldrin and dieldrin, heptachlor and chlordane, and other chlorinated hydrocarbons are expected to follow. However, in the absence of adequate substitutes, EPA occasionally approves limited use of DDT to prevent spread of a specific pest. Similarly, EPA issued a permit in 1976 for aerial application of mirex to control fire ants on six million acres in Georgia, Mississippi, Louisiana, and Arkansas. Scientists from APHIS will monitor water, soil, sediment, invertebrates (crabs and crayfish), and vertebrates (mostly birds) to determine the fate of residues in the non-target environment. Such monitoring programs using specific method- ology should provide more reliable information on the persistence of these useful pesticides in the environment. Expansion of special programs is anticipated as EPA continues to review data for other pesticides on their list of candidates for Rebuttable Presumption Against Registration (RPAR). Several chlorinated phenols and phenoxy herbicides are included �-101- in the list, due in part to the presence of dioxin contaminants. The most widely publicized of these is the herbicide 2,4,5trichlorophenoxyacetic acid (2,4,5-T), originally implicated as a teratogen in screening studies at high doses in susceptible strains of mice. The 2,4,5-T used in the first teratogenic studies by Bionetics (60) contained about 30 ppm of the highly toxic 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). This dioxin is formed during the manufacture of 2,4,5-trichlorophenol from tetrachlorobenzene at high temperatures under alkaline conditions. Current specifications are for less than 0.1 ppm TCDD in 2,4,5-T and other pesticides derived from 2,4,5-trichlorophenol, such as the herbicide silvex (2,4,5-trichlorophenoxypropionic acid) and the fungicide hexachlorophene (2,2'-methylene bis(3,4,6-trichlorophenol)). Cancellation proceedings against 2,4,5-T were withdrawn by EPA in June 1974 pending results of extensive monitoring studies for TCDD. Analytical methods have been developed capable of detecting residues down to 10 parts per trillion (0.00001 ppm) or less in milk, fat, .and environmental samples (61, 62). However, caution is needed in attributing positive responses at this level to the actual presence of TCDD residues due to interferences from unrelated substances in the same samples. The TCDD monitoring program initially called for sampling of many substrates including human milk, fat, and liver. �-102- Substrates most likely to contain residues were samples taken from range cattle maintained in areas with a confirmed history of repeated treatment with 2,4,5-T for many years, and slaughtered without the usual fattening period in a feed lot. Samples of fat from these animals were distributed to several cooperating laboratories including EPA and The Dow Chemical Company. Preliminary results were erroneously reported in 1975 to indicate "positive" results for about 50% of the initial 34 fat samples analyzed, but the findings were not consistent among the laboratories conducting the analyses. As reported by Dr. Ralph T. Ross, Chairman of the EPA Dioxin Project in June 1976 (63), only one of the 85 beef fat samples analyzed showed a positive TCDD level at 60 ppt, and two samples appeared to have TCDD levels at 20 ppt. He also stated that the analytical method is not valid below 10 ppt. The method used in the above analyses involves use of gas chromatographic separation to remove interfering substances, in combination with high resolution mass spectrometry for quantitation. This method has also been used to analyze other monitoring samples collected by Dow. As reported by Shadoff in 1975 (64), no residues of TCDD were detected in samples of fish, water and mud from two locations in Arkansas and Texas that have a long history of treatment with 2,4,5-T �-103- containing considerably more than the current limit of 0.1 ppm TCDD. Thus, it seems unlikely that continued use of 2,4,5-T would result in significant residues of TCDD in the environment. Addendum, July 1978 Reports of the study by Shadoff et al. (above) and of two studies on surveillance samples of milk and beef fat were published in 1977 and 1978. Summaries of these studies are quoted below. Summary -' As part of a broad study to determine whether 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) is accumulating in the environment due to approved uses of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) based herbicides, samples of fish, water, mud and human milk were collected from areas in Arkansas and Texas where 2,4,5-T herbicides are used and were analyzed for TCDD. No TCDD was detected by a GC-MS procedure with a detection limit which averaged less than 10 ppt. 2/ Conclusion •=* Surveillance samples of milk from the states of Oklahoma, Arkansas, and Missouri were collected from cows grazing on pasture or rangeland treated with normal applications of 2,4,5-T. These samples and control samples were analyzed for TCDD by GC/MS. A detection limit of 1 ppt was achieved. With this sensitivity the control samples were indistinguishable from those from treated areas. Hence, TCDD was not found. 3/ Summary — Specimens of fat taken from steers which had grazed on rangeland previously treated with 2,4,5-T herbicides were analyzed for the presence of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). A cleanup procedure resulting in a 500-fold concentration was followed by a gas chromatographymass spectrometry detection technique. The limit of detection of TCDD (2.5 times peak to peak noise) was found to be in the 30-60 picogram range (3-6 ppt in beef fat using 10 gram samples). None of the sixteen samples comprising two of the three studies showed any response for TCDD. In the third study, in which the animals were confined to a fenced pasture sprayed in its entirety with a 2,4,5-T herbicide, samples from three of the seven animals gave a positive response at the extremely low level of 3 to 4 ppt TCDD. which is'at the detection limit. — Shadoff, L.A., Hummel, R.A., Lamparski, L., and Davidson, J.H. A Search for 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) in an Environment Exposed Annually to 2,4,5-Trichlorophenoxyacetic Acid Ester (2,4,5-T) Herbicides. Bull. Environ. Contam. Toxicology 18_, 478-485, 1977. 2/ — Mahle, N.H., Higgins, H.S., and Getzendaner, M.E. Search for the Presence of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Bovine Milk. Bull. Environ. Contam. Toxicology 18., 123-130, 1977. -^Kocher, C.W., Mahle, N.H., Hummel, R.A., Shadoff, L.A., and Getzendaner, M.E. A Search for the Presence of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Beef Fat. Bull. Environ. Contam, Toxicology, ]J9, 229-236, 1978. �-104- 5. INTERPRETATION OF RESIDUE DATA Several basic principles should be applied in the interpretation of positive responses obtained in pesticide residue analyses. For example, a nanogram quantity of a pesticide in a solvent may cause a measurable peak in a gas chromatogram. However, the same quantity may not be distinguishable when superimposed on the "noisy" background obtained from extracts of a substrate. Similarly a small peak from a treated sample should not be compared directly to a large peak from a sample spiked at a considerably higher level. Recovery studies should include complete analyses of several control samples fortified at the claimed sensitivity of the method. Any "positive" responses below that should be reported as "less than" or "traces". A response below the limit of detection should be reported as "none-detected" since it is not possible to demonstrate a zero residue. For example, a residue of only 0.01 ppm DDT still represents 1013 molecules per gram of foodstuff (65). Above all, confirmation of residue identity is needed before instrument responses can be translated into positive values. 5.1 VARIABILITY AMONG SAMPLES The inherent biological variation among individuals of the same species can cause differences in response during analysis which might be interpreted as positive residue �-105- values unless numerous samples were analyzed. For example, in a recent study reported by Clark et al, (66), residues of several chlorinated hydrocarbon insecticides in fat of groups of feedlot cattle were compared to the residues in the feed given to them for 112 days prior to slaughter. Both feed and fat samples contained detectable levels of lindane, dieldrin, DDT, DDE, and DDD. Except for dieldrin, each insecticide was found at higher levels in fat than in feed. However, the levels were consistently low except for DDT and its degradation products in fat. Average values for lindane were 0.002 ppm in feed and ranged from 0.014 ppm to 0.032 ppm in fat of five groups of ten calves each given various levels of ammonium salts in addition to the feed. Average levels found for DDT were 0.083 ppm in feed and ranged from 0.324 ± 0.256 ppm in fat from Group II to 0.518 ± 0.104 ppm in Group IV. If only average values for DDT are compared, a significant difference appears to exist between Groups II and IV. However, overlap of values for individuals in the two groups shows that they are not different (i.e., 0.324 +0.256 = 0.580 which is higher than 0.518 - 0.104 = 0.414) .' 5.2 SENSITIVITY VS LIMIT OF DETECTION As discussed recently by the Federal Working Group on Pest Management, the terms sensitivity and limit of detection do �-106- not have the same meaning (52). Sensitivity depends on the response of an instrument such as a gas chromatograph under a specific set of conditions. However, as the instrument sensitivity is adjusted to a maximum, the baseline "noise" level generally increases too. The response due to a pesti- cide residue can be considered significant only if the peak height at the maximum usable sensitivity is twice the peakto-peak height of the noise on the chart. Operation of the instrument at less than maximum sensitivity may be preferable, particularly for comparative analyses in different laboratories, For example, FDA analysts set instruments to give 50% of full scale recorder deflection on injection of one nanogram of heptachlor epoxide or two nanograms of parathion to assure that each laboratory operates at the same sensitivity. On the other hand, the limit o_f detection of an analytical method may be defined as the concentration of pesticide above which a given sample of material can be said, with a high degree of confidence, to actually contain the chemical analyzed. This limit of detection depends upon a number of factors in addition to the instrument sensitivity, including sample size, adequacy of extraction and cleanup, size of aliquot injected, and knowledge and experience of the analyst. Deviations in peaks or baselines caused by inade- quate cleanup are more evident when the instrument is operated at maximum sensitivity. These deviations may �-107- cast doubt on the validity of the peaks of interest in assigning positive values for specific residues. Sutherland has reviewed the analytical limit of detectability for residues (67) . For all practical purposes, the response of control samples to a specified analytical method is the sole determinant of the limit of detectability. However, these control samples must be taken from plants or animals exactly comparable to the treated samples. They must be sampled at the same time, from identical varieties of the same degree of maturity, and must have been treated in all respects the same as the treated samples except that the chemical to be determined must not have been used on them. Furthermore, the samples must be independent and normally distributed according to statistical definitions. The responses given by, or apparent residues found in these control samples will usually follow a bell-shaped distribution curve. average of all samples analyzed will be the true mean. The An old rule-of-thumb was that any uncorrected residue value greater than twice the average control represented a real residue (65) . However, this is not necessarily so depending on the spread of the frequency distribution around the mean for that substrate. Hahn has recently discussed the conse- quences of incorrectly assuming that distribution is normal about the mean (68). �-108- Even greater care must be taken in interpreting data for extremely low residues of highly toxic chemicals in the environment. This was demonstrated following collaborative studies on TCDD conducted as part of the EPA Dioxin Implementation Plan in 1976. Agreement was reached among scientists from EPA, USDA, Harvard, University of Utah and Dow that it takes a response of 2.5 times background noise to indicate a positive finding. Furthermore, it takes a response of 10 times background noise before quantitative values at low parts per trillion can be assigned. 5.3 INTERPRETATION OF MONITORING DATA Unfortunately, analysts conducting monitoring studies do not have the benefit of comparisons to control samples because samples reliably free from pesticides are usually not available. In such cases, the limit of detection is based on the analytical variability for each pesticide on each substrate, generally by carrying through the entire method at least six replicate samples containing levels near the estimated limit of detection of the pesticide (52). Under these conditions, the limit of detection is considered equal to two standard deviations (approximately 95% confidence level) calculated from the replicated results. The identity of any peak larger than this should be subject to confirmation �-109- as discussed in Section 5.4, but practical limitations preclude confirmation of most "positive" values. Although variability among analyses of six replicates may be adequate for samples of certain species of plants or animal tissues, it may not be adequate to establish a reliable baseline for samples of soil or water from widely divergent geographical locations subjected to contamination by a great variety of chemicals. Unfortunately, some chemists do not recognize such limitations and interpret any positive response similar to that produced by a pesticide as proof that a residue is present. Furthermore, reports of moni- toring studies tend to emphasize the frequency of positive responses without regard to the significance of the "residues" found. One such report issued in 1973 was entitled Pesticides in Selected Western Streams — 1968-1971 (69). It gave results of analyses by the U.S. Geological Survey for nine chlorinated hydrocarbon insecticides and three phenoxy herbicides in water from twenty locations in fourteen states. The authors reported "total occurrence (of phenoxy residues) reached a peak of 106 during 1968-69 and declined sharply to 54 during 1970-71. This decrease of about 50% was due to 'reduced occurrences of 2,4-D and 2,4,5-T1." Examination of the lengthy data tables revealed that traces of phenoxy herbicides �-110- amounting to less than 0.0005 ppm (0.5 part per billion) appeared to be present in about 16% of the samples for 2,4-D and 2,4,5-T and 7% of the samples for silvex. Furthermore, four-fifths of the positive findings were at levels below 0.1 part per billion (ppb). The detected limits were listed at 0.005 yg/liter for silvex and 2,4,5-T and 0.02 yg/liter for 2,4-D (i.e., a sensitivity of 5 to 20 parts per trillion). Recoveries of 80 to 100% were claimed for validation studies but the spiking levels were not specified and may not have approached the claimed sensitivity. Furthermore, confirmatory analyses for phenoxy compounds were" not mentioned although other methods or other columns were discussed for confirmation of the insecticides found. The report also stated that 64% of the silvex occurrences were at one station in Nevada and 47% of the 2,4,5-T occurrences were in two associated sites in Arkansas and Oklahoma. In view of the localization of positive findings, it is possible that the parts per trillion residues "found" may have been due to the presence of interfering substances in samples from these locations rather than to actual residues of silvex or 2,4,5-T in the water. Unfortunately, the report has been cited as evidence of widespread contamination of water by these controversial herbicides in Western States. �-111Another example is a paper in the September 1974 issue of Bulletin of Environmental Contamination and Toxicology (70). A total of 13 authors summarize studies on the distribution of pesticides in randomly selected soil samples taken from diverse environments in five west Alabama counties. Pesti- cide residues were found in all samples examined/ even soil samples from areas where records indicated that no pesticides had been applied directly. The pesticides included eleven chlorinated hydrocarbon insecticides, six nonchlorinated insecticides, and two herbicides (atrazine and 2,4,5-T). They reported finding 2,4,5-T in 24% of the samples of top soil analyzed. Examination of the data revealed that 37 samples of top soil contained no detectable residues (sensitivity not specified) and only one sample containing more than 0.1 ppm 2,4,5-T. The samples were analyzed twice by GLC using two columns. No analytical data were provided to indicate reliability for the eight samples reported as "positive" at <0.1 ppm. Unfortunately, such data are used as evidence of persistence of pesticides in the environment. The above examples also demonstrate the pitfalls in drawing conclusions from reports which emphasize the maximum residue found or the percent "positive" samples without adequate allowance for variations in background "noise". This becomes even more important in special monitoring programs for highly �-112- toxic chemicals at extremely low levels. An example dis- cussed previously was the premature release of preliminary data from, analyses for TCDD at levels around 10 parts per trillion in beef fat which could be construed by some as sufficient evidence that 2,4,5-T and related pesticides should be banned (63). 5.4 CONFIRMATION OF RESIDUE IDENTITY Residue analyses without confirmatory tests provide only qualitative information. It is not safe to assume that a response similar to that produced by a known pesticide means the pesticide is actually present. Many instances have been noted in which one material masquerades as another in any one test (e.g. PCB's, phthalate esters, etc.). The certainty of identification is increased when the behavior of the unknown and the standard is the same in a number of tests. Subscribers to Pesticides Monitoring Journal received a June 1975 issue entitled Guidelines on Analytical Methodology for Pesticide Residue Monitoring (52). The following specific guidelines on confirmation of residue identity were listed: Residues reported should be confirmed by tests such as TLC, element specific GCf p-value determinations, derivatization, ultraviolet photolysis, etc., in addition to EC-GC. �-113- PCB's or other suspected complex interfering substances must be separated and confirmed. Combinations of confirmatory tests that measure the same physical property parameters as tests that measure the same parameters as the initial analysis, although they may help to support the identification, are not the best means of doing so. If sufficient material is available/ excellent identifications can be made by infrared spectrometry, particularly with the use of microcell techniques or Attenuated Total Reflectance. Mass spectrometry (MS) or combined GC-MS is recommended for identifying or confirming identity of important residues not adequately identified by other techniques. The methods used for confirming residue identity should always be described in the report of residue results. Although some laboratories may not have time to confirm every residue tentatively identified by <3C, confirmatory analyses should be conducted at intervals such as one in every five samples when the same residues are apparently present throughout the group. When insufficient sample is available, con- firmatory analyses may be conducted on a pool of several cleaned up sample extracts. �-114- 5.5 STATISTICAL EVALUATION OF RESULTS The Federal Working Group on Pest Management also differentiates between accuracy and precision of pesticide residue analyses (52). Accuracy involves statistical measurements that relate to the differences between the test results and the true result when the latter is known or assumed, and may be expressed as the mean error or relative error. Preci- sion involves statistical measurements that relate to the variation among test results themselves — i.e., the scatter or dispersion of a series of test results without assumption of any prior information as to the true result. Precision may be expressed in terms of the variance, standard deviation, relative standard deviation, range, etc. Values from single or duplicate injections of extracts from duplicate samples are more meaningful than values from multiple injections from single samples. These should be differentiated in evaluating differences among groups of data. As noted in Section 3.1.1, pesticide residue methods are often subjected to collaborative studies through the Association of Official Analytical Chemists (AOAC). These studies evaluate specificity, repeatability, and reproducibility of independent analyses for various samples. Repeatabi1ity is a measure of how well an analyst can expect to agree with �-115- himself from day to day, whereas reproducib i 1i ty is how well analyses in one laboratory are likely to agree with results from other laboratories. The Statistical Manual of the AOAC presents excellent information on formulating and statistically analyzing collaborative studies (71). �-116- 6. CONCLUSION The regulation of pesticides involves .a complex maze of rules implemented by various government agencies in their everchanging interpretation of laws enacted and amended by the U.S. congress to control the use of such chemicals. Two basic elements in these laws are that all pesticide products must be registered and that tolerances must be set for safe residue limits of all pesticides used in food or feed crops. Requirements for residue data in support of these tolerances are discussed in Section 3. Monitoring studies are conducted by groups within several government agencies to assure that excessive residues are not present in foods or accumulating in the environment, as discussed in Section 4. The importance of careful interpretation of residue data is discussed in Section 5 of this chapter. Concern about the significance of monitoring data was expressed at a 1973 hearing on EPA studies of specific chemicals in estuarine and freshwater environments. At that time the agency's Hazardous Materials Advisory Committee was told that virtually all measurements of DDT in the environment were suspect due to interferences in the analyses (72) . Early data did not distinguish between DDT and residues of PCB's used extensively in transformer fluids, etc. (48, 52). �-117- Furthermore, phthalate esters used widely as plasticizers accompany DDT through virtually all clean-up procedures according to D. G. Crosby of the University of California (49), Crosby also reported that interferences have been a major factor in chemical analyses for arsenic residues, and for mercury and phenoxy herbicides such as 2,4-D and 2,4,5-T. Responsibilities in the use and misuse of scientific data were discussed during a symposium of the American Association for the Advancement of Science (AAAS) at their national meeting in New York in 1975. It was concluded that responsi- bility for proper reporting of data lies with individual scientists/ scientific societies, and publishers. One of the speakers, Dr. Richard W. Roberts, director of the National Bureau of Standards, estimated that at least half of the data reported in scientific literature are unusable, often because too little information has been provided for inde. pendent evaluations (73) . Dr. Mary L. Good, a director of the American Chemical Society, cited four areas in need of responsible attention: determination of permissible levels of potentially dangerous materials in the biosphere, lack of information on error limits and credibility of published data, overmassage or excessive manipulation of raw data by computer techniques, and evaluation and use of computerized data banks. Also, care must be taken to �-118- delineate between those parts of reports that are factual and can be repeated by other workers and those reflecting interpretation of the data (74). Dr. Bernard L. Oser entitled his presentation Exaggerated Conclusions Based on Inadequate Data (75). He defined scientific data as observations and findings generally expressed in numerical or descriptive terms and stated that, even when correctly reported/ "data" are not necessarily equatable with "facts". Implicit in the term "facts" are the accuracy and reproducibility of findings and the competence and integrity of those responsible for the design, execution, and interpretation of the studies. Data are often statistically manipulated or "staticulated", such as by use of average values without expressing the range or distribution around the mean. Scientists may also draw false or misleading conclusions from their own data or from that of others. This may take the form of overestimating the importance of uncorroborated findings, or as one writer expressed it, extrapolation of unproven speculations. Thus, problems arise from misuse of scientific data as well as from use of unscientific data. Emphasis in this chapter has been on the care needed in conducting analyses for pesticide residues at sub-part-per-million �•119- levels in a variety of substrates, of the need for adequate reporting of such residue data, and of pitfalls in interpretation of data at levels below the validated sensitivity of the method. More specificity is needed rather than "picogram" measurements as often pushed by the scientific community. Premature or unwise interpretation of data can lead to precipitous action by regulatory agencies who have to respond to public pressures from individuals or groups of persons. Unfortunately, most pressure comes from lawyers and politicians who do not comprehend the limitations of pesticide residue analyses, particularly when the identity of the residues "found" has not been adequately confirmed. �-121- 8. 1. REFERENCES F. A. Gunther and R. C. Blinn, Analysis of Insecticides * and Acaricides-. Vol. VI in'Chemical Analysis Series, Interscience, New York-London, 1955. 2. R. W. Fogelman, J. Agr. Food Chem. £, 410 (1956). 3. T. H. Harris, J. Agr. Food Chem. 4,, 413 (1956). 4. F. A. Vorhes, Jr., J. Agr. Food Chem. £, 415 (1956). 5. H. Frehse, Residue Reviews J5, 1 (1964) . 6. T. H. Harris and J. G. Cummings, Residue Reviews j>, 104 (1964). 7. A. Bevenue and Y. Kawano, Residue Reviews 35, 103 (1971) . 8. FDA Guidelines for Chemistry and Residue Data Requirements of Pesticide Petitions, Bureau of Science, Food and Drug Administration, Dept. of Health, Education, and Welfare, Washington, D.C. 20204, March 1968. �-128- 47. Handbook for Authors of Papers in the Journals of the American chemical Society, American Chemical Society Publications, Washington, D.C. (1967). 48. J. H. Ruzicka, "Methods and Problems in Analyzing for Pesticide Residues in the Environment" in Environmental Pollution by Pesticides, ed. by C. A. Edwards, Plenum Press, New York, 1973. 49. D. G. Crosby - Statement before EPA's Hazardous Materials Advisory committee, Pesticide Chemical News 1 (45), 6 (October 17, 1973). 50. National Pesticides Monitoring Program (Revised), Pest. Monit. J. 5, 35 (1971). 51. Federal Working Group on Pest Management, Guidelines on Sampling and Statistical Methodologies for Ambient Pesticide Monitoring, Washington, B.C., October 1974. (U.S. Government Printing Office 1975-626-592/216 3-1). 52. Federal Working Group on Pest Management, Guidelines on Analytical Methodology for Pesticide Monitoring, Washington, D.C., June 1975. �-129- 53. S. D. Fine - Statement before Subcommitee on Agricultural Research and General Legislation, Committee on Agriculture and Forestry, U.S. Senate, January 27, 1976. 54. Bureau of Veterinary Medicine Memo 19: Monitoring for Residues in Food - Animal Tissues, DREW Pub. No. (FDA) 76-6001, July 1975. 55. R. D. Johnson and D. D. Manske, Pest. Monit. J. J9/ (1976) . 56. F. C. Lu - Toxico logical evaluation of food additives and pesticide residues and their "acceptable daily intakes" for man; the role of WHO, in conjunction with FAQ. Residue Reviews 4J5, 81 (1973) . 57. DHEW comments on draft GAO report to Congress entitled "Federal Pesticide Registration Program: Is It Adequately Protecting the Public and the Environment from Pesticide Hazards?" (December 1975). 58. National Academy of Sciences, Principles for Evaluating Chemicals in the Environment, a report of the Committee for the Working Conferences on Principles of Protocols for Evaluating Chemicals in the Environment, Environmental Studies Board, National Academy of Sciences - National �-130- Academy of Engineering and Committee on Toxicology, National Research Council, Washington, D.C., 1975. 59. L. P. vanDyk and K. Visweswariah, Residue Reviews 55, 91 (1975) . 60. K. D. Courtney, et al., Science 168, 864 (1970). 61. R. Baughman and M. Meselson, Environmental Health Perspectives, Experimental Issue No. 5, September 27, 1973. 62. (DHEW Publication No. (NIH) 74-218). W. B. Crummett and R. H. Stehl, Environmental Health Perspectives, Experimental Issue No. 5, September 15, 1973 (DHEW Publication No. (NIH) 74-218). 63. R. T. Ross, Statement on U.S. Environmental Protection Agency's Dioxin Implementation Plan (2,4,5-T/2,3,7,8Tetrachlorodibenzo-p-dioxin), June 21, 1976. 64. L. A. Shadoff and R. A. Hummel, Paper No. 80, Division of Analytical Chemistry, 170th ACS National Meeting, Chicago IL, August 27, 1975. 65. F. A. Gunther, Adv. Pest. Control Research 5, 222 (1962). �-131- 66. D. E. Clark, H. E. Smalley, H. R. Crookshank, and F. M. Farr, Pest. Monit. J. 8, 180 (1974). 67. G. L. Sutherland, Residue Reviews 10, 85 (1965). 68. G. J. Hahn, Chemtech, August 1976 (p. 530). 69. J. A. Schulze, D. B. Manigold, and F. L. Andrews, Pest. Monit. J. 7_, 73 (1973) . 70. R. Albright, et al., gull Environ. Contain. Toxicology 1,2, 378 (1974) . 71. W. J. Youden, E. H. Steiner, Statistical Manual of the AQAC, Assoc. Offic. Anal. Chemists, 1975. 72. Pesticide Chemical News !_ (45) , 6, October 17, 1973. 73. Chemical & Engineering News, February 10, 1975 (p. 17). 74. Chemical & Engineering News, March 3, 1975 (p. 4) . 75. B. L. Oser, International Flavours and Food Additives j5, 163 (1975) . *(No.) CRF is Title (No.) of the Code of Federal Regulations, (No.) FR is Volume (No.) of the Federal Register. � 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 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. For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a> 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. 178 Folder The folder containing the original item. 5241 Series The series number of the original item. Series VIII 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 Leng, Marguerite L. Title A name given to the resource Letter to Federal Register Section, Technical Services Division, Office of Pesticide Programs, EPA from Marguerite L. Leng, July 7, 1978 ao_seriesVIII