Physical inactivity trails only hypertension as a contributor to stroke risk, having an estimated population-attributable risk of 28.5%. Whether precise amounts or type of activity influences stroke risk has been unclear, the authors noted in their introduction.
Meta-analyses have suggested that regular physical activity reduces stroke risk by 25% to 30% as compared with little or no activity. Recent evidence has pointed toward differences in the impact of physical activity on stroke risk by sex, the authors added.
In an effort to clarify the association between physical activity and stroke, McDonnell and colleagues analyzed data from the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study, a national, multiracial prospective cohort study.
The analysis comprised 30,239 REGARDS participants ≥45, including oversampling in the Stroke Belt region of the southeastern U.S.
Baseline data collection included self-reported physical activity, defined as the weekly frequency of activity sufficiently intense to cause sweating. Physical activity could include leisure, commuting, and occupational activities.
On the basis of the activity frequency, investigators separated the participants into three groups: 1=no activity, 2=one to three times per week, and 3=four or more times weekly.
During a mean follow-up of 5.7 years, investigators documented 918 incidents of stroke and transient ischemic attack. After adjustment for age, sex, race, and interaction between age and race, a significant association between physical activity and stroke emerges.
Comparison of the lowest and highest frequencies of physical activity produced a hazard ratio of 1.20 (95% CI 1.01-1.42). Self-reported frequency of one to three times weekly was associated with a hazard ratio of 1.14, which did not achieve statistical significance (95% CI 0.96-1.35).
After adjustment for other stroke risk factors (diabetes, hypertension, body mass index, alcohol consumption, and smoking), the hazard declined by 30% to 1.14 and was no longer significant (95% CI 0.95-1.37, P=0.17).
Investigators also compared the highest frequency of physical activity with a weekly frequency of zero to three. The analysis produced an adjusted hazard ratio of 1.18 for stroke for the lower frequency (95% CI 1.01-1.36). Adjustment for region, urban/rural residence, and socioeconomic status attenuated the association and made it no longer significant (HR 1.17, 95% CI 0.99-1.36).
Analysis by stroke type (ischemic versus hemorrhagic) did not appreciably alter the results.
Separate analyses of men and women showed a significant association between frequency of physical activity and stroke risk among men (HR 1.26 to HR 1.30) but not women.
The study was supported by the National Institute of Neurological Disorders and Stroke.
The authors had no relevant disclosures.