considered secondary. Response rates were 53% for male Gulf War veterans, 72% for female Gulf War veterans, 42% for male nondeployed veterans and 60% for female nondeployed veterans. The authors observed no effect of Gulf War service on miscarriages reported by female veterans. The number of stillbirths among females was too small for meaningful assessment. There was a 40% increase in the odds of miscarriage among pregnancies reported by male Gulf War veterans compared with their nondeployed counterparts (OR 1.4, 95% CI 1.3-1.5), and the effect was stronger for early miscarriages (OR 1.5, 95% CI 1.3-1.6). However, in the Nuclear Industry Family Study, Maconochie et al. (1999) found evidence of underreporting of miscarriages among the nonexposed workers. The potential selection bias could explain the associations observed among the Gulf War veterans, if there was selective participation related to pregnancy outcome.
In the Kang et al. (2001) study described above, adverse pregnancy outcomes, including spontaneous abortions, stillbirths and preterm births, were compared between Gulf War deployed and nondeployed veteran mothers and fathers. There was an excess prevalence of self-reported spontaneous abortions (OR 1.62, 95% CI 1.32-1.99) among pregnancies conceived by Gulf War deployed fathers. Stillbirths were also reported more frequently among deployed fathers (OR 1.65, 95% CI 0.91-2.98), but this difference was not significant. Among veteran mothers, the odds of spontaneous abortion were modestly increased for those deployed to the Gulf War, but the 95% confidence interval did not exclude the null value (OR 1.35, 95% CI 0.97-1.89). No differences for preterm birth or infant death were observed among males or females. The limitations of this study include self-reported outcomes and differential participation rates for deployed (75% response) and nondeployed (65% response) veterans.
Five additional secondary studies evaluating the effect of deployment on adverse pregnancy outcomes were identified (Ishoy et al., 2001a; Kang et al., 2009; Kelsall et al., 2007; Verret et al., 2008; Wells et al., 2006). All of these studies were based on self-reported data.
Interview data from 661 male Danish peacekeepers who served in the gulf in 1990-1997 and 215 male Danish military personnel who were not deployed to the gulf revealed no differences in prevalence of spontaneous abortions, live births, or infant deaths (Ishoy et al., 2001a). In addition to self-reported outcomes, the study did not control for the influence of important confounders.
One of the largest studies of reproductive outcomes in female Gulf War veterans was conducted by Wells et al. (2006). In this 1996-1997 survey of 8742 married male and female US Gulf War era veterans, no associations between deployment status and number of pregnancies reported between 1991 and 1995 were observed for males or females. Among the 2159 men and 2233 women reporting one or more pregnancies during this 4-year period, Gulf War deployment status was not associated with an increased odds of miscarriage, stillbirth, ectopic pregnancy, low birth weight, or macrosomic (> 4000 g) births among females. Among males, a weak, marginal association was observed between Gulf War deployment and miscarriage (OR 1.24, 95% CI 0.96-1.61), but no associations with other adverse pregnancy outcomes were observed. Limitations of this study include self-reported outcomes, poor response rate (51%), and limited information on maternal risk factors. Thus, the results may be susceptible to recall bias, selection bias, and confounding.
In the Australian veteran cohort described above, Kelsall et al. (2007) assessed the self-reported outcomes of pregnancies occurring in 1991 or later among 1424 Gulf War veterans