Association of depressive symptoms with longitudinal changes in activities of daily living among middle-aged and older Chinese adults with hip fracture: an Andersen Behavioral Model perspective

BACKGROUND: Depression is common after hip fracture, but its independent association with early functional decline remains unclear. Using the Andersen Behavioral Model (ABM), we examined whether baseline predisposing, enabling, and need factors-particularly depressive symptoms-were associated with short-term activities of daily living (ADL) decline in middle-aged and older Chinese adults with incident hip fracture. METHODS: Using China Health and Retirement Longitudinal Study Waves 2-5, we conducted a longitudinal cohort analysis. Incident hip fracture was defined as the first report in Waves 2-4 with no prior report; this wave served as baseline for the 10-item Center for Epidemiologic Studies Depression Scale (CESD-10) and ADL. ADL decline was defined as a decrease in total ADL score from baseline to the immediately subsequent follow-up wave (ΔADL < 0). Multivariable logistic regression estimated odds of ADL decline. CESD-10 was modeled categorically and continuously using restricted cubic splines (RCS). Sensitivity analyses included continuous-exposure models, multiple imputation (20 imputations), and subgroup analyses. RESULTS: We analyzed 472 participants (n = 231/118/123 with no/mild/severe depressive symptoms). Crude cumulative incidence of functional decline was 9.5% (45/472). Older age was associated with higher odds of decline (per 1-year increase: OR, 1.08; 95% CI, 1.04-1.11; p < 0.001). Severe-but not mild-depressive symptoms were independently associated with higher odds of decline (severe versus none: OR, 2.51; 95% CI, 1.19-5.40; p = 0.017). Life satisfaction remained associated with decline in the parsimonious model (dissatisfied versus neutral: OR, 2.56; 95% CI, 1.15-5.70; p = 0.021), whereas enabling resources and lifestyle factors were not independently associated. RCS and continuous-exposure models were compatible with an approximately linear dose-response association between CESD-10 and decline risk (nonlinearity p > 0.050), with each 1-point increase associated with 6%-7% higher odds (OR, 1.06-1.07; p ≤ 0.016). CONCLUSIONS: Within the ABM framework, older age and severe depressive symptoms were baseline correlates of early post-fracture ADL decline, with an additional association for life satisfaction. These findings support consideration of depressive-symptom screening and mental-health assessment in post-fracture rehabilitation; prospective studies should evaluate clinical effectiveness.