Dynamic association between sleep duration and depressive symptoms in middle-aged and older Chinese population: A secondary analysis of CHARLS data using panel vector autoregression

BACKGROUND: The temporal directional relationship between sleep duration and depressive symptoms remains unclear. This study explored their dynamic links among middle-aged and older Chinese adults using a panel vector autoregressive framework. METHODS: We used nationally representative longitudinal data from the China Health and Retirement Longitudinal Study (CHARLS, 2011-2020). Analyses included participants with ≥3 repeated measurements of sleep duration and depressive symptoms to compute inter-wave change scores. Sleep duration was self-report. Depressive symptoms were measured by the 10-item CES-D scale. A panel vector autoregressive (PVAR) model was applied to examine the dynamic bidirectional association between sleep duration and depressive symptoms across multiple survey waves. RESULTS: 12,747 participants were included in this study. An initial change in sleep duration was negatively associated with subsequent changes in sleep duration (β= -0.340, 95% CI: -0.358, -0.322). Similarly, an initial change in depressive symptoms was associated with subsequent reductions in depressive symptoms (β = -0.329, 95% CI: -0.346, -0.311). No significant association was observed between initial changes in depressive symptoms and subsequent changes in sleep duration (β = -0.001, 95% CI: -0.007, 0.005). In contrast, longer sleep duration significantly showed a directional lagged association with subsequent reductions in depressive symptoms (β = -0.064, 95% CI: -0.110, -0.018). CONCLUSIONS: Sleep duration was associated with lower subsequent depressive symptoms, while depressive symptoms did not predict sleep duration. These results indicate a directional temporal association and suggest that sleep duration may be a potential intervention target for improving mental health in middle-aged and older adults.