Risk factors for incident diabetes in US military participants of the Millennium Cohort Study

Little prospective data exist on the risk of diabetes among persons in active military service. We examined risk of newly-reported diabetes (DM) in members of the Millennium Cohort Study over a 3-yr follow-up period that included the initiation of military conflicts in Afghanistan and Iraq. Study subjects included a random sample of all persons on active service in the U.S. military on Oct 1 2000. Of the 77,047 subjects who agreed to participate by completing a baseline self-administered paper or online survey in 2001, 55,021 (median age 36 yrs) returned a follow-up survey in 2004. Survey questions elicited information on demographics, height and weight, military service, clinician-diagnosed diabetes, and other physical and mental health conditions and lifestyle characteristics. Occurrence and timing of deployment was obtained from electronic DoD data. Odds of DM in relation to exposures of interest were estimated using logistic regression analysis. Of the 51,605 subjects without self-reported diabetes at baseline, 0.9% reported DM at follow-up, a rate of 3/1000 person years. Greater age (odds ratio, 95% confidence limits 1.06, 1.05-1.07), body mass index (1.20, 1.17-1.22), and non-Caucasian ethnicity (1.49, 1.24-1.80) were related to increased odds of DM. Those deployed with and without self-reported combat exposures had a statistically significant lower odds of DM compared with nondeployees. Ongoing analyses are designed to help clarify the associations between exposures due to military service and physical and mental health disorders and incident diabetes in a large military and veteran population while controlling for known diabetes risk factors.