Background Few studies examined socio-ecological factors and leisure time physical activities (LTPA) and rarely focused on self-regulation and social capital, which might play a significant role in impacting people’s physical activity behavior. This study aimed to examine the direct and indirect effects of individual level (perceived benefits, perceived barriers, and self-efficacy), interpersonal level (self-regulation), social level (social capital), and environmental level factors (perceived physical environment) on LTPA among older adults. Methods A cross-sectional study was conducted in 737 older adults from Sichuan, China. Structural equation modeling (SEM) analysis was used to examine the associations of individual, interpersonal, social, and environmental level factors with LTPA. Results The mean age of all participants was 71.22 (range, 60–97), and 56.1% of them were women. The SEM results showed that individual level variables (β = 0.32, ρ < 0.001), self-regulation (β = 0.18, ρ < 0.001) and social capital (β = 0.14, ρ < 0.001) could all directly affect LTPA while there was no significant association of perceived physical environment with LTPA. Self-regulation served as a bridge linking social capital and LTPA. Individual level variables contributed the largest total effect (0.32) on LTPA. Self-regulation and social capital had the same total effect (0.18) on LTPA. Conclusions Factors on three levels were all significantly associated with LTPA. Interventions that incorporate individual, interpersonal, social factors may be considered to promote LTPA in older adults. Self-regulation should receive more attention in future interventions.

INTRODUCTIONCurrent randomized trial evidence on the effect of physical activity (PA) intervention on cognitive function is scarce and mainly found in well‐educated populations, limiting its generalizability. Furthermore, the modification effect of education levels remains understudied. We conducted a cluster randomized trial to evaluate the effects of PA intervention on cognitive function in a general older population, and whether education modifies such effects.METHODSEight villages were randomized to the intervention (four villages, n = 240) or the control (four villages, n = 271). The intervention group received an 8‐week multilevel PA intervention based on a socio‐ecological model, while the control did not. The intervention has been condensed to three levels with activities occurring at the individual (telephone counseling, printed material, and training sessions), interpersonal (peer group), and community levels (group sharing and coaching). The primary outcome was changes in global cognition (overall cognitive function) measured by the Telephone Interview for Cognitive Status (TICS‐10) at 12‐month follow‐up. Repeated measurements were modeled using the linear mixed model, which assumed that the missing values were missing at random.RESULTSThe mean age was 70.94 years (standard deviation 5.71) and 55.6% were women. Compared to the control, the orientation improved from baseline at 12‐month follow‐up (0.24 points [95% confidence interval (CI), 0.03 to 0.46 points; P = 0.03]). Stratified analysis showed that the global cognition and orientation improved at 12‐month follow‐up among illiterate participants in the intervention (for global cognition, mean difference, 0.57 points [95% CI, 0.03 to 1.10], P = 0.04; for orientation, mean difference, 0.31 points [95% CI, 0.04 to 0.58], P = 0.03).DISCUSSIONThese results confirm the improvements in orientation in older adults, as well as in global cognition and orientation in the illiterate at 12‐month follow‐up.