Spinning Agent Orange Health Effects As A “Black Swan”
An Essay Review of Four Books: Dioxin, Agent Orange: The Facts (Michael Gough, 1986) Agent Orange: History, Science, and the Politics of Uncertainty (Edwin Martini, 2012) Toms River: A Story of Science and Salvation (Dan Fagin, 2013) Toxic War: The Story of Agent Orange (Peter Sills, 2014)
In an open democratic society, we citizens hope for scientific knowledge that is as close to the truth as possible. We reviled the Soviet Union partly because of the way they buried any facts that ran counter to State ideology. It is ironic that the U.S. Administration that took credit for finally dismantling the Soviet Union used many of its worst propaganda tactics here at home.
To a far greater degree than I ever realized until reading these books, the threat of compensation claims from victims of Agent Orange was the lightening rod which sparked the Reagan administration to completely refashion scientific thinking underlying American public health. As these books make plain in very different ways, that administration fostered a climate of opinion about the causes of disease, by controlling how health data is interpreted or whether they are produced at all.
That climate of scientific opinion was designed mainly to free the marketplace of constraints. Michael Gough tells us as much on page 205 of his book:
As part of a study that I directed at the Office of Techology Assessment (OTA), I contracted with Sir Richard Doll and Mr. Richard Peto of Oxford University, a highly esteemed epidemiologist and statistician, respectively, to compile and analyze estimates of what causes cancer. . . . .Doll and Peto estimated that 30% of all cancers were associated with smoking about 35% with diet, 4% with industrial exposures, and 2% with exposures through air, water, and soil. . . .
Two pages later he continues:
If one accepts their estimates, as many have, then one might conclude that eliminating environmental and occupational exposures will have a relatively small effect on preventing cancer. Their estimates, to some extent, took the wind out of regulating exposures in the environment and on the job. Moreover, the Reagan administration's stated position of taking the government off the people's back through less regulation reduces the need for testing. If regulation is hardly even contemplated, little importance could be attached to gathering information to support it. (Emphasis mine.)
Just how deeply Gough's maneuvering tore the fabric of the separation of powers, rendering the Doll/Peto report official doctrine of the whole U.S. Government, is apparent from two of the books under review. On page 171 of the Martini's Agent Orange, he explains that Gough had been a member of the Agent Orange Working Group (AOWG) while running the “White House Office of Technology Assessment.” However, Martini got it right and wrong: right in the sense that under Gough the OTA was an unofficial branch of the Reagan White House, but wrong, because the OTA was supposed to render unbiased assessments of technological issues for Congress (and was officially the Congressional OTA).
In addition, the success of Gough's campaign to bring the British epidemiological viewpoint to America is apparent in Dan Fagin's cogent work on pollution and childhood cancer in Toms River, New Jersey. Fagin's vivid account of the series of serendipitous events that led to identification of a cancer cluster in children and the scientific detective work leading to clear linkage between industrial chemicals in air/water and childhood cancers is blighted only by its decidedly anglophilic—one might say Royal--perspective on epidemiology.
Although Fagin is rightly dissatisfied with the ascendancy of lifestyle epidemiology, he does overstate the extent to which Doll's work legitimately disparaged the carcinogen-centered focus of leaders of environmental health science such as Wilhelm C. Hueper. On page 204, Fagin writes that:
The series of studies that Doll and Hill undertook [on cigarettes and lung cancer] would become the most celebrated pieces of research in the history of epidemiology. . . . They would also marginalize the workplace-centered research of rivals like Hueper and cast serious doubt on the usefullness of investigating cancer clusters like the one in Toms River. . . .
In belittling Hueper's careful human case studies in the workplace and the importance of cancer cluster detection as an early warning system, instead of putting them in the appropriate perspective of the whole of American epidemiology prior to 1981, Fagin accepts at face value a series of unwarranted assumptions including: 1) the first American case-control studies on cigarettes and the epidemiologists involved were not as important as the British ones, and 2) Hueper was mainly debunking the cigarettes-lung cancer link per se, and not actually seeing deeper problems with “risk factor” studies that look for effects of complex mixtures like cigarette tars on the average person.1,2,3
In fact, the framework of the controversy over Agent Orange health effects would have been quite different had Hueper's legacy, rather than Doll's4, become the main organizing principle. The assumption that hazardous chemicals should affect ALL individuals exposed would never have arisen in a Hueperian framework, because Hueper and his successors were always looking for the “susceptible subgroup”--those individuals who are most likely to sustain health damage5.
In parodying Hueper's “combative” personality as the source of his downfall, and comparing him to the patrician Doll, Fagin falls into another trap of popular histories of epidemiology: portraying iconoclasts of public health science as having abrasive personalities and/or extreme viewpoints. I think it is safe to say that since Ignaz Semmelweis first linked childbed fever to dirty hands in the 1840s that each generation has had at least one or two scientists whose discoveries upset the powers that be, and were subsequently dismissed as abrasive and uncompromising6.
Unfair as it may be, the epidemiologists Alice Stewart, Irwin Bross, Thomas Mancuso, and Rosalie Bertell who had uncovered low-level radiation hazards in the 1970s were often accused of being obnoxious and deserving the trashbin of anonymity. In that light, an account of the U.S. government's role in suppressing low-level radiation hazards, which preceded the Agent Orange controversy, would have greatly enriched the analysis in Edwin Martini's and Peter Sills' books (and Fagin's for that matter).
Most of the technical debates over AO effects regarding such issues as determining doses of pollutants to which individuals were exposed (“dosimetry”) and whether standards designed to protect the average person would protect the most vulnerable “susceptibles” had already been fought tooth and nail between Stewart, Bross, Mancuso, Bertell on one side, and government scientists on the other (and heralded in such books as Wasserman's Killing Our Own: The Disaster of Our Experience with Atomic Radiation).
In rehashing these debates over AO, as if they had never occurred before in another pollution-research context, authors of the AO books miss some potentially important insights and theses. This isn't surprising for Gough, who wasn't interested in past results as clues to future work, particularly on the issue of AO dosimetry, the ambiguities in which led him to help stop most of the government AO work, but is more so for the other authors. It is more than the dictum that “he who ignores history is doomed to repeat it,” but we know from modern probability theory that we don't know how seriously to take new observations, unless we have some gauge of whether something like them has happened before, and how many times.
Importantly, repeated “positive findings” of many of the same health observations in both radiation and AO studies makes it clear that at least some of the findings were not the unprecedented “black swans” government scientists made them out to be. Most strikingly, that certain birth defects are increased in offspring of fathers who were exposed to AO during the Vietnam war had been portrayed as “statistically significant but not biologically significant” by the authors of the first CDC study published in 19847. That this was but one of many semantic gimmicks used by government scientists and public relations people to downplay multgenerational effects of AO is apparent from several data sources (see second essay on this webpage), which had previously shown similar genetic effects.
In fact, the worst military pollution “experiments” on humans beings—the atomic bombing of Hiroshima/Nagasaki and the AO-ing of Vietnamese—have produced results which are mutually confirmatory: both show increased incidence of certain types of birth defects8. In an accompanying essay, I highlight the fact that the U.S. Government set up a monitoring study of birth defects, miscarriages and other “genetic damage” in Japanese victims of the atomic bombing that was never intended to reveal the actual damage9, but to serve a diplomatic function.
In several studies of low-level radiation effects on father's sperm cells conducted at U.S. Atomic weapons labs and British nuclear plants, the incidence of neural tube defects (NTDs) was markedly increased in their children10–10,13,14. The fact that children of AO-exposed fathers also suffered more NTDs than unexposed fathers in uncontrolled Vietnamese studies15 and in controlled epidemiological studies by the U.S. Centers for Disease Control (CDC) including the Vietnam Experience Study and Operation Ranch Hand study shows a consistent pattern16,17, and implies that increased rates of NTDs are very likely a marker of genetic damage which occurs in the sperm of a susceptible subgroup of chemical or radiological pollution-exposed men. It is remarkable that when sudden increases in incidence of NTDs recently occurred near the Hanford atomic weapons plant, scientific authorities still treated such events as unprecedented “black swans.” http://ac360.blogs.cnn.com/2014/02/26/birth-defect-surge-in-washington-state-stumps-medical-officials/?hpt=ac_mid
Instead of warning the world community that low level exposure to pollutants does damage the genetic material of human beings, at first only in certain susceptible people, but that the pool of susceptibles steadily increases with ongoing exposures8, the U.S. Government interpreted the results narrowly in terms of their responsibilities only to soldiers exposed to AO in Vietnam. Politics have always influenced AO reporting more than actual biological results and theory.
Hence, government reports on birth defects and other reproductive problems associated with AO are paternalistic: the reports are obviously worded to provide “reassurance” (and that word is itself often used), as if the main issue were the nervous condition of the exposed soldiers, and not an important biomedical problem. Virtually all studies emphasize that no general increase in birth defects in exposed Vietnam veterans was found, rather than zeroing in on the biological meaning of the increases in the types of birth defects that were consistently found.
That paternalistic spin is apparent in the first CDC report in the Journal of the American Medical Association in 1984, when Erickson et al write that:
This suggests that if there is any increased risk related to exposure to Agent Orange, either the risk must be small, must be limited to select groups of Vietnam veterans, or the increased risk must be limited to specific types of defects.” (Emphasis added.)
The very fact that they felt the need to report what would have been predicted by anyone who had studied biological hazards from any chemical or radiological hazard-- that the effects would be limited to select groups of Vietnam veterans, or the increased risk must be limited to specific types of defects—highlights the extent to which the Doll/Peto framework had already become the foundation for analysis of environmental hazards. It provides the ideal platform for a government attempting to belittle its own positive findings on genetic hazards from low-level pollution.
In public relations flyers sent out by the U.S. Government agencies in the 1980s, the same unscientific but paternalistic tone is apparent in reporting the results of the CDC birth defects studies. One such flyer takes the downplaying of positive findings on birth defects one step further, by changing the actual results from “specific types” to “rare types” of defects. The flyer states that:
The conclusion that Vietnam veterans in general do not appear to have been at increased risk suggests that if effects have been caused by exposure, those effects are small and are limited to select groups of veterans, and/or are limited to rare types of defects.” (Emphasis added.)
Of all of the books under review, only Fagin's Toms River pays close attention to actual data and reporting of the results. He makes a strong case that we have no organized system in the United States for monitoring, detecting, and coping with low-level pollution hazards. None of the authors of AO books provide direct analysis of data, or interpretation of results in the way that Fagin does. Most of the authors are interested principally in the legalistic notion of “cause and effect” as it plays out in court trials.
On the Loose Cannons Inc. website, we hope to further contrast legal notions of causality with biomedical ones, which eventually require deeper mechanistic scientific understanding. We will begin to explain more about how AO and other chemicals cause genetic damage and specific types of birth defects.
1. Sellers, C. Discovering environmental cancer: Wilhelm Hueper, post-World War II epidemiology, and the vanishing clinician’s eye. Am. J. Public Health87, 1824–35 (1997).
2. Schneiderman, M. A., Davis, D. L. & Wagener, D. K. Lung cancer that is not attributable to smoking. JAMA261, 2635–6 (1989).
3. Stellman, S. D. et al. Smoking and lung cancer risk in American and Japanese men: an international case-control study. Cancer Epidemiol. Biomarkers Prev.10, 1193–9 (2001).
4. Doll, R. & Peto, R. The causes of cancer: quantitative estimates of avoidable risks of cancer in the United States today. J. Natl. Cancer Inst.66, 1191–308 (1981).
5. Bross, I. D. & Natarajan, N. Genetic damage from diagnostic radiation. JAMA237, 2399–401 (1977).
6. Peller, S. Quantitative Research in Human Biology and Medicine. (Bristol: John Wright and Sons, Ltd., 1967). at <https://books.google.com/books? id=KDngBAAAQBAJ&pgis=1>
7. Erickson, J. D. Vietnam Veterans’ Risks for Fathering Babies With Birth Defects. JAMA J. Am. Med. Assoc.252, 903 (1984).
8. Bross, I. D. J. Scientific Strategies to Save Your Life: A Statistical Approach to Primary Prevention. (Marcel Dekker Incorporated, 1981). at <https://books.google.com/books/about/Scientific_Strategies_to_Save_Your_Life.html?id=eekgAQAAIAAJ&pgis=1>
9. Bross, I. D. Regarding ‘Genetic effects of the Atomic Bombs: a reappraisal’ by Schull, et al. Health Phys.44, 283–5 (1983).
10. Parker, L. Fetal death and radiation exposure. Lancet357, 556–7 (2001).
11. Parker, L., Pearce, M. S., Dickinson, H. O., Aitkin, M. & Craft, A. W. Stillbirths among offspring of male radiation workers at Sellafield nuclear reprocessing plant. Lancet354, 1407–14 (1999).
12. Sever, L. E. Looking for Causes of Neural Tube Defects: Where Does the Environment Fit in? Environ. Health Perspect.103, 165 (1995).
13. Wertelecki, W. Malformations in a chornobyl-impacted region. Pediatrics125, e836–43 (2010).
14. Sever, L. E., Gilbert, E. S., Hessol, N. A. & McIntyre, J. M. A case-control study of congenital malformations and occupational exposure to low-level ionizing radiation. Am. J. Epidemiol.127, 226–42 (1988).
15. Ngo, A. D., Taylor, R. & Roberts, C. L. Paternal exposure to Agent Orange and spina bifida: a meta-analysis. Eur. J. Epidemiol.25, 37–44 (2009).
16. Wolfe, W. H. et al. Paternal serum dioxin and reproductive outcomes among veterans of Operation Ranch Hand. Epidemiology6, 17–22 (1995).
17. Wolfe, W. H. et al. Paternal Serum Dioxin and Reproductive Outcomes among Veterans of Operation Ranch Hand. 6, 17–22 (2015).