Exercise Frequency and Subjective Well-Being among Chinese Adolescents: Competing Indirect Associations through Depressive Symptoms and Sleep Duration

Abstract Adolescent well-being is multidimensional, and a favorable behavioral indicator need not align with psychological and recovery-related indicators. Using the 2022 China Family Panel Studies, we identified 2,985 respondents aged 10–18 years; 2,864 had valid happiness and exercise measures, and the fully adjusted complete-case model included 2,849. Ordinary least squares models used HC1 standard errors, survey analyses applied the 2022 cross-sectional person weight with strata and PSU clustering, and coefficient products used 5,000 bootstrap resamples. Exercise frequency was positively associated with subjective well-being in the adjusted model (b = 0.045, SE = 0.016, p = .006) and in the design-adjusted model (b = 0.052, SE = 0.022, p = .016). The largest positive indirect association involved fewer depressive symptoms (ab = 0.0182, 95% CI [0.0087, 0.0281]); smaller negative associations involved sleep duration (ab = − 0.0024, 95% CI [− 0.0053, − 0.0003]) and the sequential sleep–depressive-symptoms path. Loneliness and self-rated-health products included zero. Tutoring-time moderation (p = .084), formal gender interactions (p = .395 and .050), and the matched PSM estimate (p = .138) did not meet p < .05. The cross-sectional findings describe differentiated indicator patterns rather than causal mechanisms.