Physical activity domains and patterns with risk of depressive symptoms: A cross-sectional study in China

The health benefits of domain-specific physical activity (PA) on depressive symptoms were inconclusive. Few studies explored PA patterns and depressive symptoms. This study aimed to investigate the associations of PA domains and patterns with depressive symptoms. We conducted a cross-sectional study in China with 5047 adults. Latent class analysis was applied to identify the PA patterns and logistic regression analysis was used to calculate odds ratios (ORs) and 95 % confidence intervals (CIs). The ORs (95 % CIs) for the active versus inactive groups were 0.79 (0.69–0.91) for leisure-time PA, 0.57 (0.49–0.65) for transport PA, 0.95 (0.82–1.09) for household PA, and 1.38 (1.18–1.62) for occupational PA. We found non-linear associations between leisure-time PA, transport PA and depressive symptoms, with the lowest risk at 11 METs-h/week of leisure-time PA (equal to 147 min/week moderate PA or 88 min/week vigorous PA) and 23 METs-h/week of transport PA. There was a marginal inverse association with household PA for men while not for women. We identified four PA patterns and found a lower risk of depressive symptoms associated with “low occupational PA pattern” versus “moderate PA level pattern” (0.45 (0.38–0.52)). Given the cross-sectional design, causality cannot be inferred. Our study supported an inverse association of leisure-time PA and transport PA with depressive symptoms and a positive association of occupational PA. The observed inconsistent association of household PA among men and women, and the finding that “low occupational PA pattern” was associated with a lower risk of depressive symptoms warrant further investigation.