Physical Activity, Cognitive Function, and Mortality in a US National Cohort

PURPOSE: Increasing physical activity is postulated to slow cognitive decline associated with aging. Low levels of both physical activity and cognitive function are associated with increased risk of mortality. We test the hypothesis that the relative protective effect of high physical activity level as related to mortality is greater in persons with impaired cognitive function than in others. METHODS: Data were analyzed from a longitudinal mortality follow-up study of 5903 American men and women aged 60 years and older examined in 1988 to 1994 who were followed an average of 8.5 years. Measurements at baseline included self-reported leisure-time physical activity (LTPA), a short index of cognitive function (SICF), sociodemographic data, health status, and physical and biochemical measurements. RESULTS: Death during follow-up occurred in 2431 persons. In bivariate cross-sectional analyses, more frequent LTPA was associated with greater cognitive function. In proportional hazards regression analysis, no significant interaction of LTPA with cognitive function was found; however, there was a significant age-LTPA interaction. After adjusting for confounding by baseline sociodemographic data and health status at ages 60 to 74, the hazards ratio (95% confidence intervals) was for LTPA more than 8 times weekly compared with none (0.51; 0.38-0.76, p < .001) and for low SICF score compared with high 1.43 (1.36; 1.00-1.84, p < .05). After controlling for health behaviors, blood pressure, and body mass, C-reactive protein, and high-density lipoprotein cholesterol, the LTPA hazards ratio was 0.52 (0.35-0.78; p = .002), but cognitive function was no longer significant. At ages 75 and older, results were similar for LTPA, but cognitive function remained significant after adjustment. CONCLUSIONS: In a nationwide cohort of older Americans, analyses demonstrated a lower risk of death independent of confounders among those with frequent LTPA. Much of the effect of low cognitive function could be explained by other risk factors at ages 60 to 74 but not 75 years and older. Ann Epidemiol 2010;20:251-257. (C) 2010 Elsevier Inc. All rights reserved.