Vetting Agent Orange

by Howard Fienberg
March 11, 2002

The Institute of Medicine (IOM) reported last year that exposure to Agent Orange, an herbicide used in the Vietnam War, might have caused leukemia in the children of Vietnam veterans. But on February 27, they revised their analysis and changed their position. “The evidence is too weak to draw any conclusions or even make tentative ones,” the head of the IOM panel told the Associated Press.


A large portion of evidence dissipated following the release of the Agent Orange-leukemia study last year. It turned out that an Australian study the IOM panel evaluated, which found an association between Australian veterans of Vietnam and acute myelogenous leukemia (AML) in their children, contained a critical miscalculation. Once corrected, the research showed veterans children had a rate of AML within the expected norm for their community. That revelation, coupled with two new studies, led the panel to its revised conclusions.


Unfortunately, the assertion that no association can be found does not differ all that much, scientifically speaking, from their original conclusion. Last year, panel chairman Irva Hertz-Picciotto announced to the media that while there was no firm evidence, new research suggested “some kind of connection.” But the IOM report actually states on page 424 that “Evidence is suggestive of an association between herbicide and the outcome, but limited because chance, bias, and confounding could not be ruled out with confidence.”


It appears that in studies of Agent Orange, “connections” can be pretty tenuous.


In 2000, the IOM announced a connection between adult-onset diabetes and Agent Orange exposure. The Baltimore Sun headlined, “Agent Orange is linked to diabetes in veterans”; the Associated Press, “Study finds strong link between Vietnam War herbicide and diabetes.” It sounded like a strong connection, but the IOM only called it “limited and suggestive” and a closer reading of the study shows why. The diabetes study found that many of the risk estimates in the studies examined were “not statistically significant” and had methodological problems. Moreover, one of the biggest problems in assessing the connection was obesity. Dioxin, a major component of Agent Orange, is stored in fat, so greater obesity tends to lead to greater dioxin exposure. That makes separating the two factors difficult.


Although only a few reporters caught on, the IOM report actually said that “the increased risk [of diabetes], if any... appears to be small. The known predictors for diabetes risk family history, physical activity, and obesity continue to greatly outweigh any suggested increased risk from wartime exposure to herbicides.”


What was the reaction of veterans groups when they were told the latest new from the IOM? Were they relieved to know that their constituents’ children were not subject to an increased risk of a terrible disease?


It did not seem that way. Len Selfon, director of benefits programs for Vietnam Veterans of America, told the Associated Press, “Obviously, we're disappointed.”


Why? Because these connections, while statistically insignificant, are very significant politically. For purposes of compensation, the U.S. military’s Department of Veterans Affairs (VA) now assumes that every person who served in Vietnam was exposed to Agent Orange. A veteran claiming disability from an officially-linked ailment doesn’t have to prove exposure. The VA's list of presumably Agent Orange-linked disabilities includes:

  • chloracne
  • Hodgkin's disease
  • multiple myeloma
  • non-Hodgkin's lymphoma
  • acute and subacute peripheral neuropathy
  • porphyria cutanea tarda
  • prostate cancer
  • respiratory cancers (lung, bronchus, larynx and trachea)
  • soft-tissue sarcoma
  • And adult onset diabetes mellitus recently joined the list.


Why such a lengthy list? The VA (and therefore the IOM) evaluates Agent Orange based on “statistical association” rather than causality. The slightest scientific hints can eventually lead to public policy facts.


The VA’s rearguard action against health claims from former soldiers, giving up ground a little at a time to the scientific equivalent of supposition and innuendo, sets up the U.S. for a much greater long-term problem.


As of last year, only 8,600 of the 2.3 million surviving Vietnam veterans had sought disability compensation for Agent Orange-linked disease. According to Washington Post science writer David Brown, the average age of Vietnam veterans is 54, and the group is just entering the age when lung cancer, prostate cancer and diabetes become more prevalent. Demographic reality means that veterans’ health will continue to deteriorate and the costs to the VA will skyrocket.


In addition, Vietnam intends to sue the U.S. government for losses from Agent Orange. The communist government estimates that at least 1 million people suffer diseases linked to Agent Orange, including about 150,000 malformed babies. While the Vietnamese refuse to allow independent verification of their research, they will have a strong case as long as the U.S. continues with its current policies of veteran compensation.


As we have seen with the supposed “Gulf War Syndrome,” the symptoms and diseases befalling Vietnam veterans differ little from those suffered by veterans of all major conflicts. The urge to link veterans suffering to some particular aspect of their service is as natural as it is scientifically foolhardy.


Soon, we will face a difficult decision as a nation. If we provided health benefits for all veterans, regardless of “presumptive exposure” to something, we could save the integrity of our scientific institutions, prevent future lawsuits, shore up trust in the military and boost the morale of current and future soldiers.   

See the original:

return to Howard Fienberg's page