An analysis of reproductive outcomes for conceptions of participants of the Air Force Health Study

Abstract: Analyses were conducted of reproductive outcomes for conceptions of participants of the Air Force Heath Study. Participants were male Air Force veterans of the Vietnam War. Conceptions were categorized into conceived before and after the start of the participant's Vietnam War service. Analyses accounted for correlation between outcomes for multiple conceptions for each participant. For each of three non-sparsely occurring outcomes, including not live born, miscarriage, and preterm, the probability of occurrence increased substantially when conceived after compared to before the start of Vietnam War service. These results support the conclusion of an adverse effect due to Vietnam War service on these reproductive outcomes. Data for conceptions after the start of Vietnam War service for participants with measured dioxin values were used to estimate dose-response curves for the effect of dioxin exposure on the occurrence of each of the three non-sparsely occurring outcomes. These curves were assumed to be constant up to a threshold and then monotonic after that threshold. For each of the three non-sparsely occurring outcomes, the estimated dose-response curves increased nonlinearly after associated thresholds. These results support the conclusion that the adverse effect due to conception after the start of Vietnam War service is attributable to high enough exposures to dioxin, a toxic contaminant of Agent Orange used for herbicide spraying in the Vietnam War. Sensitivity analyses supported the conclusion that dioxin results were not substantially affected by the assumption of monotonicity, decay due to elapsed time from exposure to measurement, and available covariates.