Association between friend gatherings and leisure-time exercise among Chinese adults.

INTRODUCTION: Broad social-capital measures may obscure the concrete relationships through which leisure-time exercise is organized. This study examined whether leisure-time exercise among Chinese adults is more strongly associated with friend gatherings than with neighborhood interaction, generalized trust, or perceived fairness. METHODS: We pooled nine anonymous repeated cross-sectional waves of the Chinese General Social Survey from 2010 to 2023. Friend gatherings were treated as a proxy for friendship-oriented leisure socializing, not as direct evidence that respondents exercised with friends. Standardized OLS models included socioeconomic and demographic covariates, survey-year fixed effects, and harmonized province fixed effects. RESULTS: In the primary social-contrast model, friend gatherings showed the largest social-relational association with exercise (standardized beta = 0.149), exceeding the coefficients for neighborhood interaction, generalized trust, and perceived fairness (all coefficient-comparison p < 0.001). The association remained positive after adjustment for other leisure activities (beta = 0.081, p < 0.001), and robustness analyses supported the same bounded interpretation. Boundary analyses did not show that the pattern was reducible to venue/gym exercise, and supplementary health-profile analyses did not identify exercise-friendship health synergy. DISCUSSION: Leisure-time exercise in China co-patterns with friendship-oriented leisure socializing more strongly than with broader social-capital attitudes or neighborhood contact. These findings are associative and do not establish causation or direct exercise companionship.