Sleep duration, depressive symptoms, and memory aging: longitudinal patterns and temporal network evidence from four-wave CHARLS study (2011–2018)

Sleep is a modifiable behavioral factor closely linked to cognitive aging and depressive symptoms. However, evidence remains limited on whether nocturnal sleep and midday nap duration are nonlinearly associated with heterogeneous longitudinal patterns of subjective memory complaints (SMCs), objective episodic memory (OEM), and depressive symptoms, as well as their temporal interrelationships. We analyzed four-wave data (2011–2018) from the China Health and Retirement Longitudinal Study (CHARLS), including 7,198 participants aged ≥ 45 years (mean age: 56.12 ± 7.93 years; 51.8% female).

Sleep duration was self-reported. SMCs, OEM (Word Recall Test, reverse-coded), and depressive symptoms (CESD-10) were assessed. K-means clustering was applied to identify longitudinal patterns of SMCs, OEM, and depressive symptoms.

Restricted cubic spline models were applied to examine potential nonlinear dose–response associations between sleep duration and longitudinal patterns. Cross-lagged panel network (CLPN) analysis was used to examine the temporal and directional associations among these variables across waves. Distinct longitudinal patterns were identified for SMCs, OEM, and depressive symptoms.

Sleep duration showed nonlinear associations with trajectories reflecting poorer memory outcomes and greater depressive symptoms, with estimated nadirs of approximately 7.6, 6.9, and 7.7 h for SMCs, OEM, and depressive symptoms, respectively. CLPN analysis demonstrated stable network structures with good centrality stability. SMCs showed the highest out-expected influence, while OEM demonstrated the highest in-expected influence, suggesting a consistent cross-wave predictive association from SMCs to subsequent OEM scores.

Deviations from the nocturnal sleep duration at the estimated nadir are associated with adverse memory and depressive symptoms. Longitudinal network analysis suggested a directional pattern from SMCs to OEM, suggesting that SMCs may potentially represent an early indicator of memory decline. These findings highlight the potential value of integrating subjective memory assessment with sleep monitoring for early prevention of cognitive decline in aging populations.

China Health and Retirement Longitudinal Study Center for Epidemiologic Studies Depression Scale The authors would like to express our sincere gratitude to the China Health and Retirement Longitudinal Study (CHARLS) for providing access to the 2011–2018 survey data. This study was supported by National Key R&D Program of China (No. 2022YFC2503901). These authors contributed equally: Ying He and Ying Ou.