OBJECTIVES: Although studies have indicated that physical activity (PA) is related to cardiovascular disease, the specific association between PA and incident cerebrovascular disease (CBVD) remains uncertain. The current study aimed to investigate the associations between PA levels and the CBVD incidence or all-cause mortality. DESIGN: Prospective cohort study. SETTINGS AND PARTICIPANTS: Older participants (aged >60 years) from the UK Biobank. METHODS: The baseline PA was classified as total, light, moderate, and vigorous PA based on the metabolic equivalent-minutes per week (MET-min/wk) and considered as exposures, whereas CBVD incidence and all-cause mortality were considered as the outcomes. Cox proportional hazards were used to calculate the hazard ratios (HRs) and 95% CIs for the influence of the association between PA and CBVD incidence and all-cause mortality. RESULTS: A total of 146,742 participants aged 60 years and older were included. During a median follow-up period of 13.5 years (interquartile range of 12.8-14.2), 9338 older individuals developed CBVD and 3033 death were recorded (including 767 CBVD-related deaths). High volumes of PA were consistently associated with lower risks of CBVD and all-cause mortality. The lowest risk of CBVD incidence was observed at 2001-2500 MET-min/wk of total PA (HR 0.61, 95% CI 0.53-0.70), and the lowest risk of all-cause mortality was observed at 2501-5000 MET-min/wk (HR 0.52, 95% CI 0.43-0.63) in older adults. Total PA at 2001-2500 MET-min/wk significantly reduced the CBVD incidence in older women (HR 0.57, 95% CI 0.46-0.71), which was more pronounced than that in older men (HR for 2001-2500 MET-min/wk: 0.64, 95% CI 0.50-0.77). CONCLUSIONS AND IMPLICATIONS: Total PA at 2001-2500 MET-min/wk significantly reduced the risk of incident CBVD and all-cause mortality in adults aged >60 years, although the extents of risk reduction vary in men and women.
