Abstract 11582: Association Between Cardiorespiratory Fitness and Stroke: Investigating Possible Mediators of Disease

Abstract

BACKGROUND: Cardiorespiratory fitness (fitness) is inversely associated with stroke. The extent to which this association is explained by the development of stroke risk factors such as diabetes (DM) and hypertension (HTN) is unknown. We hypothesized that the association between fitness and stroke would be attenuated after adjustment for incident DM & HTN.

METHODS: Linking participant data from the Cooper Center Longitudinal Study with Medicare claims files, we studied 20,838 individuals (20.7% women, mean age 49) without prior stroke, who underwent baseline examination from 1970 to 2009 and survived to receive Medicare coverage from 1999-2009. Fitness in metabolic equivalents was estimated at baseline by Balke protocol treadmill time and categorized into age and sex specific quintiles (Q1=low fitness). Stroke hospitalization, incident DM & HTN were determined using Medicare data. Associations between fitness and stroke hospitalization after age 65 were assessed by applying a proportional hazards model to the multivariate failure time data with DM and HTN as time dependent covariates.

RESULTS: After 135,066 person years of Medicare follow up, we observed 808 stroke hospitalizations. For men and women, compared to Q1, higher midlife fitness (Q4-5) was associated with a lower rate of stroke hospitalizations [9.0 vs 4.9 (men); 8.9 vs 3.3 (women) per 1,000 person years]. There was a strong association between midlife fitness and future stroke risk after adjustment for baseline risk factors. As expected, interval development of DM and HTN during Medicare follow up were strongly associated with a higher risk of stroke. However, the association between higher midlife fitness and stroke risk was not substantially attenuated after adjustment for the development of later life DM and HTN (Table 1).

CONCLUSIONS: There is a strong, inverse association between midlife fitness and stroke risk in later life. This association is independent of the interval development of DM or HTN.