Physical exercise frequency and subjective well-being among older adults in China: a cross-sectional analysis

Physical exercise is associated with subjective well-being in later life, but estimates from cross-sectional surveys require measurement-aligned models and cautious interpretation. We analyzed 1,872 adults aged 60 years or older from the 2021 Chinese General Social Survey. A five-category global happiness item was modeled with ordered probit regression, and standardized predicted probabilities quantified the magnitude of association. An exploratory bootstrap decomposition estimated cross-sectional indirect associations through self-rated health and lower depressive mood. Higher exercise frequency was associated with greater well-being (b = 0.0795, 95% CI 0.0482–0.1109, P = 6.55 × 10^-7). Setting all respondents to daily rather than no exercise corresponded to a 6.8-percentage-point higher average predicted probability of well-being of 4 or 5 (87.9% vs. 81.1%). Indirect associations through self-rated health and lower depressive mood accounted for 7.33% and 7.20% of the total association, respectively; together they accounted for 14.53%. After false-discovery-rate correction, age, gender, and spouse co-residence did not modify the association, whereas the slope weakened as household economic status increased (interaction b = − 0.0487, adjusted P = 0.025). These findings identify a modest, robust cross-sectional association but do not establish causal or temporal pathways.