Associations of exercise frequency, exercise session duration, and sleep duration with depressive symptoms among...

Exercise and sleep are modifiable behaviors related to adolescent mental health, but the relevance of specific exercise-frequency categories, participant session duration, and nonlinear sleep duration remains uncertain.

We analyzed 2,861 adolescents aged 10-19 years from the 2022 China Family Panel Studies. Survey-weighted regression incorporated individual weights, strata, and primary sampling units. Exercise frequency, any exercise participation, participant session duration, and spline-modeled sleep duration were examined in separate models with continuous eight-item Center for Epidemiologic Studies Depression Scale (CES-D8) scores as the primary outcome.

CES-D8 scores differed across exercise-frequency categories (p = 0.002), but the pattern was not monotonic. The largest adjusted difference was observed for once-daily exercise vs. never exercising (B = -1.08; 95% CI, -1.62 to -0.53). Any exercise participation was associated with a modestly lower score (B = -0.57; 95% CI, -1.01 to -0.12), whereas session duration among 2,167 exercise participants was not associated with CES-D8 scores (B per 30 min = 0.004; 95% CI, -0.16-0.17). Sleep duration was nonlinearly associated with CES-D8 scores (overall p < 0.001; nonlinearity p = 0.004); relative to 8 h, scores were higher at 6 h (B = 1.03; 95% CI, 0.55-1.51) and 7 h (B = 0.53; 95% CI, 0.30-0.76). Sensitivity analyses were broadly consistent, and multiple-imputation estimates for exercise participation, exercise-frequency contrasts, and sleep-duration contrasts were similar to the complete-case estimates.

Exercise frequency and sleep duration were modestly associated with depressive-symptom scores, whereas participant session duration showed no clear association. The cross-sectional findings do not establish causality or support a specific exercise prescription.