Does physical exercise inhibit smoking behavior? Evidence from the China General Social Survey

Examining how physical exercise influences smoking behavior and the mediating mechanisms involved helps clarify the broader social benefits of exercise and informs efforts to improve population health and build a more harmonious society.

This cross-sectional observational study analyzed the nationally representative 2021 China General Social Survey (CGSS). After sample screening, we retained 2,373 valid observations. We estimated OLS models to examine the association between physical exercise and residents' smoking behavior. For robustness, we re-estimated ordered probit and ordered logit specifications and used propensity score matching (PSM) as an endogeneity check against the baseline estimates. We further assessed mediation by mental health and social support and tested the moderating role of individual income.

The results showed a significant positive association between physical exercise and reduced smoking (β = 0.0276), which remained robust after sensitivity and endogeneity checks. Subgroup analyses indicated heterogeneity: the association was significant at the 1% level for middle-aged adults and at the 10% level for younger adults, but not for older adults. For both men and women, the association was statistically significant at the 5% level, with a larger coefficient for men than for women (0.0359 and 0.0167). Mediation tests suggested that improvements in mental health and increases in social support were key pathways through which exercise was associated with lower smoking. Moderation analyses further showed that income positively moderated the exercise-smoking association, with higher income strengthening this association.

Physical exercise is associated with reduced smoking, and the mechanisms of mental health and social support help explain this association. These findings deepened understanding of how exercise relates to smoking behavior and provided practical implications for public health policy and tobacco-control interventions in China. The study also extended research on the social and psychological effects of exercise and offered a theoretical basis for future work. However, because the analysis is cross-sectional and relies on self-reported measures from a single country and a single-year dataset, the findings are correlational and context-specific.