Physical activity and depressive symptoms among Chinese adults a cross sectional analysis
TL;DR
Low and moderate activity were associated with lower continuous depressive-symptom scores than inactivity, and the possible nonlinear pattern depended on modeling choices and did not establish an optimal dose or a stable J-shaped relationship.
Abstract
To examine the association between total physical activity and depressive symptoms among Chinese adults, including possible departures from a linear dose–response relationship. We analyzed the 2024 Chinese National Sports and Society Survey, which obtained 11,385 adult responses across 31 provincial-level regions of mainland China. The complete-case sample comprised 10,663 adults. Activity duration was confirmed against the questionnaire as a weekly total and combined using intensity-specific MET coefficients. The primary outcome was the five-category ten-item depressive-symptom score (10–50). We fitted sampling-weighted estimating equations with an observation-level sandwich variance, categorical models and a two-part restricted cubic spline model. A study-specific recoded screening outcome was exploratory. Models adjusted for nonlinear age, sex, education, household registration, marital status, income and chronic disease. Relative to inactivity, adjusted symptom-score differences were −3.11 (95% CI −5.94 to −0.27) for low, −3.41 (−6.13 to −0.69) for moderate and −2.20 (−4.96 to 0.57) for high activity. Low and moderate activity did not differ clearly (moderate minus low: −0.30, 95% CI −1.38 to 0.78). The primary nonlinear-term Wald test gave P = 0.034, but evidence weakened with three knots ( P = 0.309) and exclusion above the 95th activity percentile ( P = 0.063). All primary exploratory screening prevalence-ratio intervals included 1. Low and moderate activity were associated with lower continuous depressive-symptom scores than inactivity. The possible nonlinear pattern depended on modeling choices and did not establish an optimal dose or a stable J-shaped relationship. High activity was not shown to be more harmful than inactivity. Activity volume is distinct from intensity, and inactivity is not a measure of prolonged sitting. The cross-sectional design does not establish temporal order, causality or an exercise prescription.