Epidemiology in Military and Veteran Populations: Proceedings of the Second Biennial Conference, March 7, 1990 (1991)

Han K. Kang, Kevin K. Watanabe, Joseph Breen, Janet Remmers, Margaret G. Conomos, John Stanley, and Michele Flicker ^*

The primary reason for concern about the adverse effects of exposure to Agent Orange is attributable to its toxic contaminant, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), or dioxin. This study investigated whether 36 Vietnam veterans had significantly higher levels of dioxin in adipose tissue than a similar group of 79 non-Vietnam veterans or 80 civilians. The adipose tissue specimens for the study were selected from the 8,000 archived tissues that had been collected from the non-institutionalized general population by the U.S. Environmental Protection Agency for the National Human Adipose Tissue Survey. The geometric mean (± standard deviation) dioxin levels in adipose tissue for Vietnam veterans, non-Vietnam veterans and civilian controls were 11.7 (± 1.7), 10.9 (± 1.7), and 12.4 (± 1.9) parts per trillion on a lipid weight basis, respectively. The mean levels for these groups were not significantly different from each other with or without adjustment for age of individuals,

body mass index, and specimen collection year. In addition, none of the surrogate measures of Agent Orange exposure such as military branch, service within specific geographic region, military occupation, and troop location in relation to recorded Agent Orange spraying was associated with the dioxin levels in adipose tissue of Vietnam veterans. Our results suggest that heavy exposure to Agent Orange or dioxin for most U.S. troops was unlikely.

From 1965 to 1970, the U.S. Air Force sprayed more than 40 million liters of Agent Orange in South Vietnam. ¹ Approximately 2 million American soldiers served in Vietnam during this period. Agent Orange was the name used for a phenoxy herbicide consisting of a mixture of 2,4-dichloro-phenoxyacetic acid (2,4-D) and 2,4,5-trichlorophenoxyacetic acid (2,4,5-T). The 2,4,5-T contained 1-50 parts per million of the contaminant 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), also known as dioxin. Dioxin is extremely toxic in laboratory animals, promotes liver tumors in rats, and is teratogenic in mice. ¹ , ² The two parent herbicides, 2,4-D and 2,4,5-T are very short-lived in the human body and do not persist in the environment. ³ Because of these characteristics and their relatively low toxicity in humans, attention has been focused on the highly persistent and toxic chemical, dioxin.

Many Vietnam veterans believe that they were heavily exposed to Agent Orange and that the exposure is responsible for health problems such as skin rashes, rare types of cancer, and birth defects of their children. Since 1979, two hundred thousand Vietnam veterans have come to Department of Veterans Affairs (VA) hospitals for an Agent Orange Registry medical examination because of concerns about exposure to Agent Orange.

Two questions are paramount in dealing with the veterans' concerns. The first is whether dioxin causes birth defects, immune deficiencies, cancers, or other chronic health problems in humans. The second is whether Vietnam veterans, in general, received substantial exposure to dioxin while they served in Vietnam. This study addresses the second question.

Dioxin accumulates preferentially in the body fat of animals and man. The dioxin half-life in humans is estimated at 5 to 11 years. ⁴ , ⁵ , ⁶ , ⁷ Among the many 2,3,7,8-substituted polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) found in the environment and in samples of human adipose tissue, only 2,3,7,8-TCDD or dioxin was present in Agent Orange. ⁷ Dioxin levels in adipose tissue could, therefore, serve as a biological marker of exposure to Agent Orange.

In fact, several studies have reported that even after approximately 20 years, dioxin levels were elevated markedly in the adipose tissue or blood serum of Vietnam veterans who handled Agent Orange. ⁶ , ⁷ , ⁸ More recently, however, the Centers for Disease Control (CDC) reported that current serum dioxin levels of Army Vietnam combat troops did not differ significantly from those of non-Vietnam veterans and that dioxin levels in Vietnam veterans did not increase with increased exposure levels estimated from military records. ⁹ This study was criticized on the basis that blood specimens taken almost 20 years after exposure could not represent what happened in Vietnam and that veterans selected for the study were limited to Army enlisted men who served in only one geographic region of Vietnam. ¹⁰

The purpose of our study was twofold: (1) to determine if a group of individuals with military service in Vietnam have significantly higher levels of dioxin in adipose tissue than either a similar group of non-Vietnam veterans or civilian peers, and (2) to determine if dioxin levels in adipose tissue were associated with specific demographic and military service characteristics.

Our study used adipose tissue specimens that had been collected from the general population by the U.S. Environmental Protection Agency (EPA). The EPA has conducted the National Human Adipose Tissue Survey (NHATS) since 1970 to monitor the human body burden of pesticides and other selected chemicals. Up to 1,000 adipose tissue specimens have been collected annually from pathologists and medical examiners across the country and analyzed by the EPA for the selected chemicals. After analysis, the unused tissue specimens were sent to a central facility to be stored at 0ºC to -20ºC.

The NHATS utilized a probability sample of the Standard Metropolitan Statistical Areas (SMSA) that is designated to represent a sample of the U.S. population in terms of age, sex, and race. ¹¹ The target population for the NHATS program was all non-institutionalized persons in the conterminous United States. However, due to the invasive nature of collecting adipose tissue samples, the sampling population was limited to individuals who died from external causes (90%) and surgical patients (10%). Within each SMSA, hospitals or medical examiners were identified and asked to contribute tissue specimens according to the design specifications of age (0-14 years, 15-44 years, 45+ years), sex, and race (white, non-white). Since the vast majority of Vietnam veterans were men

born between 1936 and 1954, this study was restricted to specimens from men born in that period.

An Inventory File was created for the 8,000 specimens that were recorded to have an adequate amount of tissue. It was found that 528 of 8,000 specimens were from males born between 1936 and 1954. The hospitals or medical examiners who originally collected the 528 specimens were recontacted to obtain enough identifying information on the donors to determine their military service status. The collection effort yielded information for 494, or 94%, of the 528 specimens.

The military service status for these 494 men, including any Vietnam service, was determined by reviewing records archived at the National Personnel Records Center (NPRC) in St. Louis and military records maintained at other locations. From this effort, 134 men (or 27% of 494) were initially found to have served in the military, 40 of whom served in Vietnam. According to the 1980 Census, one would expect about 34% of men in this age group to be veterans. All 40 Vietnam veterans were selected for the study. From the 94 remaining veterans, 80 were selected randomly for the non-Vietnam comparison group. Two civilians were closely matched to each Vietnam veteran by birth year (±2 years) and sample collection year (±2 years). Age and sample collection year were considered important matching variables because of probable accumulation of dioxin in the body with each year of exposure and the possibility of degradation of fat or dioxin while in storage. All adipose tissue specimens were analyzed during 1987. Demographic data were taken from the NHATS file and the official death certificates. Body mass index (BMI) was calculated from weight and height as follows: BMI = weight in kg/height in m².

A precise estimate of the exposure of each Vietnam veteran to Agent Orange is not considered feasible based on either military records or self-reported data. In this study the probable opportunity for exposure was determined from the following: branch of service, military occupational specialty code (MOSC), and location of the individual 's unit in Vietnam in relation to recorded Agent Orange spray. Ground troops in Vietnam might have had a higher probability of contact with Agent Orange than other Vietnam veterans due to the nature of their military operations. Ground troops engaged in combat were assumed more likely to be in herbicide-sprayed areas. As another surrogate for exposure, a veteran's military unit was assigned to one of the four broad military regions in Vietnam. According to the records of the U.S. Air Force Ranch Hand Operation, 20 million liters of Agent Orange were sprayed in Military

Region III from 1965 to 1970. During the same period, Military Regions II, I, and IV received 9.5, 8.3, and 4.5 million liters, respectively. Finally, troop locations (company size units) were determined on a 100-meter grid map of Vietnam at intervals of 90 days or less. Computer matching of troop location with respect to time and distance from recorded herbicide spray tracts was carried out using the HERBS tape and services HERB tape databases. The HERBS tape contained information on most of the herbicide missions flown by fixed-wing aircraft from 1965 to 1971, and on crop destruction missions flown by helicopter between 1968 and 1971. The tape contained information on the type of herbicide, gallons, dates, and where spray runs started and ended. Services HERBS Tape prepared by U.S. Army and Joint Services Environmental Support Group identified and documented an additional 6 million liters of herbicide sprayed mainly by Army personnel around the perimeter of base camps, fire bases, airbases and other fixed military installations. The opportunity for Agent Orange exposure was determined in two ways: an individual's company was ever located either within 2 kilometers (km) of a recorded Agent Orange spray tract within 3 days of application or within 8 km of a spray tract within 90 days application.

Specimens from the three groups were assigned randomly to one of 20 batches. Each batch typically consisted of 10 study specimens and 4 quality control (QC) samples. The QC samples provided data on method accuracy and precision. In addition, the external quality control audit samples were prepared by another laboratory and incorporated as blind samples into the various batches. Study specimens and QC samples were coded with a unique laboratory number and submitted to the analysts as blind samples. Following sample cleanup and preparation, instrumental analyses were achieved using a high-resolution gas chromatograph coupled to a high-resolution double-focusing mass spectrometer. Overall method accuracy and precision for 2,3,7,8-TCDD analyses of spiked lipid samples were 113% recovery and 8.8% coefficient of variation.

The analytical protocol provided for the detection and quantitative determination of 17 congeners of 2,3,7,8-substituted polychlorinated dibenzo-p-dioxins (PCDD) and dibenzofurans (PCDF). The minimum measurable concentration ranged from 1 picogram per gram (pg/g) for 2,3,7,8-TCDD and 2,3,7,8-TCDF, to 5 pg/g for OCDD and OCDF based on a 10-gram aliquot of human adipose tissue. This protocol was evaluated for method performance prior to being used in this study. ¹²

Multiple comparisons and testing for differences were done by using the F test in one way analysis of variance (ANOVA) and analysis of covariance with adjustments for demographic variables such as age, collection year, and body mass index. ¹³ A paired t-test was conducted to compare the means of Vietnam veterans with their matched civilian controls. ¹⁴ In all analyses, the dioxin values were transformed to a natural logarithmic scale because the dioxin values were found to have approximately log-normal distributions in this study and another study. ¹⁵

A stepwise linear regression model was also used to determine whether dioxin levels were associated with demographic and military service characteristics. ¹⁶ Factors considered a priori as covariates were age, adipose tissue sample collection year, race, and body mass index. A regression model specific to Vietnam veterans included such covariates as military occupation, calendar year of tour in Vietnam, geographic region in Vietnam, time of and distance from recorded Agent Orange spray, and adipose tissue sample collection year. All statistical tests were conducted at the 05 level of significance.

Table 1 presents the arithmetic and geometric means, and various percentile values for dioxin for the three study groups. One-way analysis of variance did not demonstrate a statistically significant difference in the mean dioxin levels among groups (p = 0.35). Analysis of covariance, testing the effect of Vietnam service on dioxin levels after adjusting for age, sample collection year, or body mass index, did not indicate a statistically significant association between service in Vietnam and dioxin levels. A paired t-test between Vietnam veterans and their matched civilian pairs did not show a significant difference in mean dioxin levels (p = 0.52; 95% confidence interval for the difference between two means = -1.32, 1.16).

For Vietnam veterans, dioxin levels were also evaluated by four factors related to the likelihood of Agent Orange exposure ( Table 2 ). None of the surrogate measures of Agent Orange exposure was associated with the dioxin levels in adipose tissue of Vietnam veterans. Furthermore, the mean dioxin level of 7 Vietnam veterans whose specimens were taken within 4 years (less than one dioxin half-life estimated for humans) since their last service in Vietnam were compared with 19 non-Vietnam veterans whose tissue specimens were collected on or before 1974 (the last sample collection year for the 7 Vietnam veterans), and also with their matched civilian pairs. The geometric mean dioxin levels (± standard deviation) for

the Vietnam group (n = 7), non-Vietnam veteran group (n = 19), and civilian controls (n = 14) were at the levels of 16.6 (± 1.6), 15.5 (± 1.5), and 18.4 (± 1.6) parts per trillion (ppt) respectively. The difference among the means was not statistically significant (p = 0.56). The 95% confidence limits for a mean dioxin difference between Vietnam veterans and civilian controls were -1.59 and 1.31. Stepwise linear regression analysis for 36 Vietnam veterans indicated that Vietnam service characteristics could account for only 14% of the variation in dioxin levels (p = 0.3), whereas collection year alone could account for 21% of variance (p = 0.005). Five other 2,3,7,8-substituted dioxins and 10 other dibenzofurans were measured, and their mean levels were calculated from specimens with levels above the detection limit ( Table 3 ). There were no group differences in the mean level of any of the dioxin congeners. The differences in TCDD levels in the three groups were also evaluated while adjusting for levels of other 2,3,7,8-substituted dioxin congeners not found in Agent Orange by a stepwise multiple regression technique. TCDD levels were taken as the dependent variable and Vietnam service status and other PCDD congener levels as independent variables. There was a significant association between the TCDD levels and the levels of PCDD congeners (R² = 0.54, p = 0.0001), but not with the Vietnam service status (R² = 0.0025, p = 0.3).

The levels of dioxins increased with an increase in the number of chlorine except for 1,2,3,7,8,9-HxCDD. Levels of dibenzofurans were always lower than their dioxin counterparts. In each study group, the levels of dioxin tended to be inversely related to the specimen collection year, i.e., the earlier the collection year, the higher the dioxin levels (p = 0.0001, n = 195).

In this study, military service in Vietnam was not associated with elevated dioxin levels in adipose tissue with or without adjustment for demographic variables. In addition, no Vietnam service characteristic measured singly or in combination was a good predictor of dioxin levels in adipose tissue. There were no consistent trends in the dioxin levels according to the surrogate measures of Agent Orange exposure. The mean levels of dioxin did rise slightly with combat MOSC and having been within 3 days/2 km of sprayed areas. But the other two surrogates, branch of service and service location in Vietnam, were not associated with the dioxin levels. The small magnitude of the mean difference and the variation within each group suggest that the small difference in mean values could have been easily due to the sampling and measurement variation.

Analyses of adipose tissue from the general populations of industrialized countries have indicated the presence of a number of 2,3,7,8-substituted dioxins and dibenzofurans at ppt level. ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ These dioxins and dibenzofurans could have originated from a number of sources: incineration of municipal waste and wood products; manufacturing, use and disposal of pesticides, herbicides, and wood preservatives; and PCBs from electric transformers and capacitors. Because 2,3,7,8-TCDD was the only congener found in Agent Orange as a contaminant, ¹ , ⁷ knowing the levels of other dioxins and dibenzofurans would help determine whether TCDD levels in adipose tissue of Vietnam veterans were the result of Agent Orange exposure in Vietnam or exposure to other sources. For example, if most dioxins and dibenzofurans as well as 2,3,7,8-TCDD levels were found to be elevated among Vietnam veterans, contribution from sources other than Agent Orange should not be ruled out. However, if only the 2,3,7,8-TCDD level remained elevated and other PCDD levels were comparable to the comparison groups, Agent Orange would be considered as the contributor. There were no group differences in the mean levels of any PCDD congeners. The lack of group differences suggested that sources other than Agent Orange may have contributed to the Vietnam veterans' levels of 2,3,7,8-TCDD. In a study by Kahn et al., ⁷ of 10 Vietnam veterans with heavy potential for exposure to Agent Orange (e.g. Ranch Hand personnel, Army Chemical Corps specialists), the levels of 12 other 2,3,7,8-substituted dioxins and dibenzofurans were similar to the levels found among 10 Vietnam veteran controls and 7 non-Vietnam veteran controls. Only 2,3,7,8-TCDD levels were elevated approximately 10-fold among Vietnam veterans with a heavy exposure potential.

This study may have failed to detect a small difference in mean dioxin levels among groups because of the relatively small sample size. The study had an adequate statistical power (90%) to detect a mean difference of 5 ppt or more between groups. Elimination of dioxin from the body after Vietnam service is an unlikely explanation because dioxin levels of seven Vietnam veterans whose specimens were taken within four years from their return from Vietnam (which was considerably less than the estimated half-life of dioxin in humans) were not significantly different from their appropriate comparison groups.

Although the NHATS sampling scheme was designed to collect a representative sample of the SMSA in terms of age, sex, and race, subjects selected for the study may not have represented their respective groups for several reasons. First, over 90% of the NHATS sample were collected from deceased persons whose deaths in most instances were due to traumatic injury. Second, tissue samples for this study were selected from the archived NHATS specimens rather than original NHATS samples. Third, 6% of the men who were eligible for the study had to be excluded

because of missing personal identifiers. Despite these problems, demographic and military characteristics of the Vietnam veterans selected for the study did not differ substantially from the overall Vietnam veteran population. They were predominantly white (75%), draft eligible during the Vietnam war (age 18 to 25), enlisted men (89%), who served in the Army and Marine Corps (72%) with military occupational specialties related to combat support roles (67%). None of the Vietnam veterans in the study had a record of routinely handling or spraying Agent Orange.

The mean background levels of dioxin reported here were generally higher than the values reported by others. ⁶ , ⁷ , ⁸ , ⁹ It is unlikely that this difference is due solely to an interlaboratory variation. Our laboratory (the Midwest Research Institute) participated in an interlaboratory validation study for dioxin measurement and produced a satisfactory result. Furthermore, external quality control audit samples prepared by the Battelle Columbus Division were incorporated as blind samples into the various batches and analyzed for dioxin and dibenzofuran congeners. The analytical results were found to be acceptable. We believe that the difference in sampling years between this study and several other studies could account for the higher values reported in our study. We found that the dioxin levels in adipose tissue were significantly associated with the sample collection year (p = 0.0001); the earlier the collection year, the higher the levels of dioxin irrespective of veteran status. In fact, this general trend was observed for other dioxins. The median sample collection year in our study was 1978, whereas in other studies the specimens were mostly collected in the mid-1980s or later. The observed dioxin decline from 1971 to 1982 is consistent with the general trend for chlorinated hydrocarbon chemical compounds in human adipose tissue. The U.S. Environmental Protection Agency's National Human Adipose Tissue Survey Program indicates that the median levels of BHC, HCB, and PCB had been steadily decreasing over time between 1970 and 1983. ¹¹ In Sweden, the levels of dioxins and dibenzofurans in human milk decreased significantly from 1972 to 1985. ²² The Swedish authors attributed the decline to the reduction in use of certain organochlorine compounds such as PCBs, PCP and 2,4,5-T. A study involving a large sample of specimens representative of the U.S. population will be needed to confirm this observation.

We concluded that the results of our study did not support the hypothesis that most U.S. troops were heavily exposed to dioxin in Vietnam. Furthermore, none of the surrogate measures of Agent Orange exposure based on military service characteristics was associated with the dioxin levels in adipose tissue of Vietnam veterans. These results are consistent with those of CDC ⁹ and not inconsistent with Kahn et al. ⁷ and Schecter et al. ⁸

1. Young AL Kang HK Shepard BS : Chlorinated dioxins as herbicide contaminants Environ SCi Technol 1983 ; 17 : 530A-540A

2. Neal RA Olson JR Gasiewicz TA et al : The toxicokinetics of 2,3,7,8-tetrachlorodibenzo-p-dioxin in mammalian systems Drug metabolism Review 1982 ; 13 : 355-385

3. World Health Organization : IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Man V. 15 Lyon, France 1977

4. Poiger H Schlatter C : Pharmacokinetics of 2,3,7,8-TCDD in man Chemosphere 1986 ; 15 : 1489-1494

5. Pirkle JL Wolff WH Patterson DG Jr et al : Estimates of the half-life of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Vietnam veterans of Operation Ranch Hand J Toxicol Environ Health 1989 ; 27 : 165-171

6. Centers for Disease Control : Serum 2,3,7,8-tetra-chlorodibenzo-p-dioxin levels in Air Force Health study participants- Preliminary report JAMA 1988 ; 259 : 3533-3535

7. Kahn PC Gochfeld M Nygren M et al : Dioxins and dibenzofurans in blood and adipose tissue of Agent Orange exposed Vietnam veterans and matched controls JAMA 1988 ; 259 : 1661-1667

8. Schecter A Constable JD Arghestani S et al : Elevated levels of 2,3,7,8-tetrachlorodibenzodioxin in adipose tissue of certain U.S. veterans of the Vietnam war Chemosphere 1987 ; 16 : 1997-2002

9. Centers for Disease Control : Serum 2,3,7,8-tetra-chlorodibenzo-p-dioxin levels in US Army Vietnam era veterans JAMA 1988 ; 260 : 1249-1254

10. : Unmasking Agent Orange Discover April 1988 ; 9 : 42-48

11. US Environmental Protection Agency : Baseline Estimates and time trends for beta-benzene hexachloride, hexachlorobenzene, and polychlorinated biphenyls in human adipose tissue 1970-1983 EPA No. 560/5-85-025 Office of Toxic Substances Washington, D.C. September 30, 1985

12. U.S. Environmental Protection Agency : Analysis for Polychlorinated Dibenzo-p-dioxins (PCDD) and Dibenzofurans (PCDF) in Human Adipose Tissue: Method Evaluation Study EPA-560/5-86-020 Washington, D.C. October 1986

13. SAS Institute Inc : The GLM Procedure In : SAS User's Guide: Statistics Version 5 Edition Cary, N.C. SAS Institute, Inc. 1985

14. SAS Institute Inc : The MEANS Procedure In : SAS User's Guide: Basics Version 5 Edition Cary, N.C. SAS Institute, Inc. 1985

15. Patterson DG Hoffman RE Needham LL et al : 2,3,7,8-tetrachlorodibenzo-p-dioxin levels in adipose tissue of exposed and control persons in Missouri JAMA 1986 ; 256 : 2683-2686

16. Dixon WJ Jennrick R : Stepwise regression In: BMDP Statistical Software University of California Press 1983

17. Patterson DG Holler JS Smith SJ et al : Human adipose data for 2,3,7,8-tetrachlorodibenzo-p-dioxin in certain U.S. samples Chemosphere 1986 ; 15 : 2055-2060

18. Graham M Hileman FD Orth RG et al : Chlorocarbons in adipose tissue from a Missouri population Chemosphere 1986 ; 15 : 1595-1600

19. Ryan JJ Lizotte R Lan BP : Chlorinated dibenzo-p-dioxins and chlorinated dibenzofurans in Canadian human adipose tissue Chemosphere 1985 ; 14 : 697-706

20. Rappe C Nygren M Lindstrom G et al : Dioxins and dibenzofurans in biological samples of European origin Chemosphere 1986 ; 15 : 1635-1639

21. Stanley JS Boggess KE Onstot J et al : PCDDs and PCDFs in human adipose from the EPA FY82 NHATS repository Chemosphere 1986 ; 15 : 1605-1612

22. Noren K : Changes in the levels of organochlorine pesticides, polychlorinated biphenyls, dibenzo-p-dioxin and dibenzofurans in human milk from Stockholm, 1972-1985 Chemosphere 1988 ; 17 : 39-49

We are grateful to Aaron Blair, Ph.D. (National Cancer Institute), Carl Keller, Ph.D. (formerly, National Institute of Environmental Health Sciences), Donald Patterson, Ph.D. (Centers for Disease Control), Linda Pickle, Ph.D. (Georgetown University), and John Ryan, Ph.D. (Canada) for their review of a study report from which the manuscript was prepared. The guidance provided by Alvin Young, Ph.D. (U.S. Department of Agriculture) and Joseph Carra and Mary Frankenberry (U.S. Environmental Protection Agency) in planning for the study is greatly appreciated.