Reassessing the Association of Sedentary Behavior and Physical Activity with Ischemic Stroke: A Mendelian Randomization Study

Findings from previous Mendelian randomization (MR) studies disagreed with the current scientific consensus regarding the role of physical activity (PA) and sedentary behavior in ischemic stroke (IS). We reassessed these associations with a focus on etiological subtypes of IS and the potential mediating roles of cardiometabolic traits and brain imaging-derived phenotypes (IDPs). We performed MR analyses using summary statistics from genome-wide association studies of sedentary behavior and PA (n = 88,411~608,595), cardiometabolic traits (n = 393,193~694,649), brain IDPs (n = 33,224) and the latest IS data (62,100 cases and 1,234,808 controls). Inverse-variance weighted regression was used as the primary method, complemented by several sensitivity analyses. A two-step MR approach was employed to assess the mediating effects of cardiometabolic traits and brain IDPs. Genetic liability to leisure-time moderate-to-vigorous PA (LTMVPA) and higher overall PA (OPA) were associated with reduced risks of IS and small vessel stroke (Benjamini-Hochberg adjusted P < 0.05). Suggestive associations were observed between longer leisure-screen time and higher IS risk and between higher OPA and lower cardioembolic stroke risk (P < 0.05). The isotropic volume fraction in the anterior limb of the left internal capsule (ALIC), as well as some cardiometabolic metrics, partially mediated these associations. There was no evidence for causal effects of overall MVPA, overall light-intensity PA or overall sedentary duration on IS. Longer LST, less OPA and not engaging in moderate-to-vigorous PA during leisure time were associated with higher risk of ischemic stroke. The associations between PA and ischemic stroke depended on different subtypes and were mediated by changes in ALIC and cardiometabolic biomarkers.