Joint association of physical function and sleep duration with incident depressive symptoms among middle-aged and older Chinese

To explore the independent, joint, interactive, and mediating effects of physical function (PF) and sleep duration (SD) on incident depressive symptoms (IDS) among Chinese adults aged ≥ 45 years, and to recommend appropriate PF-SD combination strategy to lower IDS risk. Data were obtained from the China Health and Retirement Longitudinal Study (CHARLS), with a nine-year follow-up from 2011 to 2020. Physical function was assessed with a 9-item questionnaire, SD was self-reported, and IDS was defined by the Center for Epidemiological Studies Depression Scale ( CES-D). Cox regression, restricted cubic spline (RCS) model, Subgroup and sensitivity analysis, interaction, and mediation analyses were applied. Among 6,189 community-dwelling adults (mean age 56.6 ± 7.9 years; 52.4% male), 1,815 incident IDS events (29.3%) were documented over 9 years of follow-up. Both physical dysfunction (PD) and short sleep duration (SSD) were independently associated with IDS (PD: HR = 1.37; SSD: HR = 1.15–1.34; all P < 0.05). The SSD–IDS association remained robust across PD and non-physical dysfunction (NPD) subgroups (HR = 1.13–1.38; all P < 0.05 except for midday nap duration [MND] < 30 min in NPD). Relative to the PD + SSD referent, all other combinations exhibited significantly lower hazard ratios (HR = 0.53–0.85; all P < 0.05). Restricted cubic spline analyses revealed non-linear relationships of total sleep duration (TSD) and nighttime sleep duration (NSD) with IDS (P for non-linearity = 0.002), with no further risk reduction beyond TSD > 8 h or NSD > 7.5 h (both P > 0.05). Mediation analysis indicated that TSD and NSD accounted for 3.55% and 4.0%, respectively, of the PD–IDS effect. PD and SSD are independent risk factors for IDS among middle-aged and older Chinese, with PD + SSD combination conferring the highest risk. Recommended risk-mitigation strategies are: for NPD individuals, NSD ≥ 6 h; for PD individuals, NSD ≥ 6 h or MND ≥ 30 min. TSD and NSD partially mediate the PD-IDS relationship. Long sleep duration did not increase the risk of IDS.