Update 1998—and relevant papers published since the deliberations of the Update 1998 committee were completed.

Strength of Evidence in Epidemiologic Studies

Based on the scientific evidence reviewed in this report as well as the cumulative findings of research reviewed in the previous Veterans and Agent Orange reports, the committee finds that there is limited/suggestive evidence of an association between exposure to the herbicides used in Vietnam or the contaminant dioxin and Type 2 diabetes. This is a change in classification from previous Veterans and Agent Orange reports, which found inadequate/insufficient evidence to determine whether an association existed. 3

No one paper or study was determinative in reaching this decision. Instead, the committee found that the information accumulated over years of research now meets the definition established for limited/suggestive evidence—that is, evidence is suggestive of an association between herbicides and the outcome, but limited because chance, bias, and confounding could not be ruled out with confidence. In reaching this decision, the committee observed the following:

  • Positive associations are reported in many mortality studies, which may underestimate the incidence of diabetes. Morbidity (the rate of incidence of a disease) is thought to be a more informative end point than mortality (the rate of death) when conducting epidemiologic studies of Type 2 diabetes because the disease is not typically fatal, its known complications may be more likely to be implicated as the underlying cause of death, and reporting of contributory causes of death on death certificates may be spotty. These reasons also lead epidemiologists to suspect that mortality studies may underestimate the incidence of diabetes. Four mortality studies were reviewed in this report. Individuals living near the site of a 1976 industrial accident involving dioxin were found to have a higher risk of diabetes death than a reference population in all exposure zones where diabetes deaths were recorded. Two studies of a TCDD-exposed cohort of workers at 12 U.S. plants found positive but non-statistically significant associations between measures of exposure and notations of diabetes on death certificates. The fourth study, which examined workers in 12 countries who produced or sprayed phenoxy herbicides and chlorophenols, reported an elevated relative risk of mortality from diabetes in exposed workers versus non-exposed referents. Studies reviewed in previous Veterans and Agent Orange reports show an inconsistent but weakly positive association between exposure measures and Type 2 diabetes mortality.


The categories of association mentioned here were established in the original (1994) Veterans and Agent Orange report and have been used in all subsequent reports. A complete list of categories is contained in the “Organization and Framework” section of this report.

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement