National Meeting 2009

1036 — Physical Activity in a National Sample of Veterans

Littman AJ (Seattle Epidemiologic Research and Information Center) , Forsberg CW (Seattle Epidemiologic Research and Information Center), Koepsell TD (Seattle Epidemiologic Research and Information Center)

Veterans are unique in part because they have a history of high levels of physical activity (PA) and physical fitness due to their military service. The aim of this study was to describe and compare the prevalence of PA in veterans and non-veterans, and among VA users and non-users.

Data were obtained from the 2003 Behavioral Risk Factor Surveillance System surveys of US adults. Veteran status, VA health care use, and PA were determined in 245,564 adults. Individuals were classified as inactive, insufficiently active, or meeting recommendations for moderate or strenuous PA. We used model-based direct adjustment to the distribution of all veterans, as estimated from the original weighted sample and chi-square tests corrected for the survey design.

After adjusting for age and gender, the prevalence of inactivity was significantly lower (16.2% vs. 20.5%), and meeting PA recommendations was significantly greater (46.0% vs. 42.0%) in veterans than in non-veterans (p < 0.0001). VA users were more likely to be inactive than VA non-users (20.8% vs. 14.7%) and less likely to be insufficiently active or meet PA recommendations (p < 0.0001). Differences in PA levels between veterans and non-veterans did not change substantially after additional adjustment for education, race/ethnicity, and comorbidities, while differences in inactivity between VA users and VA non-users largely disappeared (16.8% vs. 15.2%) and the difference in meeting PA recommendations reversed, with a higher prevalence in VA users (49.5%) than in non-users (46.5%) (p = 0.009).

Despite the high level of PA required of active duty military personnel, only a minority of veterans met PA guidelines, and the prevalence of inactivity was particularly high in VA users. Lack of longitudinal information precludes inferring whether lower PA in VA users was a consequence or precursor of comorbidities.

Findings from this study suggest a large potential to increase PA and improve health in VA users.