Cross-Sectional Associations of Cognitive Function, Frailty, Depressive Symptoms, and Self-Care Dependence in Residents of a Long-Term Care Institution

Background: Cognitive and functional difficulties commonly coexist in long-term care. Frailty and depressive symptoms may also co-occur with cognitive function and self-care dependence, but cross-sectional data cannot determine temporal order. We characterized the association structure among these measures. Methods: We analyzed complete standardized assessments from 182 residents of one long-term care institution. Cognitive function (Mini-Mental State Examination), physical frailty (FRAIL scale), depressive symptoms (15-item Geriatric Depression Scale), and self-care dependence (20-item Chinese Activities of Daily Living scale; higher scores indicate greater dependence) were assessed. Prespecified regression equations were adjusted for age, sex, and education. Path-product quantities and participant-resampling bias-corrected intervals (BC, not BCa) were used as descriptive stability summaries, not cluster-robust confidence intervals. Residential heterogeneity was examined with intraclass correlations and leave-one-building-out estimates. Results: Lower cognitive scores co-occurred with greater frailty, more depressive symptoms, and greater dependence. In the adjusted model, the total cognition–self-care association was c = −1.501 (BC interval −1.716 to −1.274), and the conditional association coefficient was c′ = −1.082 (−1.357 to −0.779). The frailty, depression, and ordered path-product quantities were −0.221, −0.166, and −0.033, respectively. Self-care dependence had a cohort intraclass correlation of 0.42. Across six building-omission analyses, the ordered quantity ranged from −0.050 to −0.018 and the frailty quantity from −0.272 to −0.164. Conclusions: These findings describe interrelated measures within one institution. They do not establish causal mediation, temporal order, or an intervention sequence.