Cumulative intrinsic capacity impairment and the risk of cognitive and depressive symptoms among rural community-dwelling older adults in Taiwan: a cross-sectional and three-year longitudinal analysis using the WHO ICOPE framework

The WHO Integrated Care for Older People (ICOPE) framework operationalises intrinsic capacity (IC) for community screening, but the combined and longitudinal effects of cumulative IC impairment in rural Asian populations remain insufficiently characterised. To examine cross-sectional and three-year longitudinal associations of single and cumulative ICOPE impairments with cognitive impairment and depressive symptoms in rural Taiwan. Dual-design analysis of routine ICOPE Step 1 data (2023–2025). Mutually adjusted and Firth penalized logistic regression ( N = 1,658) estimated adjusted odds ratios (aOR). A pre-specified “Double-Hit” indicator (concurrent sensory and physical/nutritional impairment) was evaluated by ROC analysis (DeLong tests). Cumulative impairment was modelled continuously. Generalised estimating equations (GEE) modelled depressive trajectories (454 participants; 1,018 observations), with inverse-probability-of-retention weighting. Depressive symptoms were independently associated with hearing (aOR = 2.57, 95% CI 1.58–4.16), mobility (aOR = 2.38, 1.46–3.88) and nutritional impairment (aOR = 4.97, 2.46–10.07); cognitive impairment with age (aOR = 1.03 per year, 1.01–1.04) and mobility (aOR = 1.52, 1.15–2.02). The Double-Hit indicator was associated with depressive symptoms (aOR = 3.48, 2.13–5.57) and improved discrimination over a sociodemographic model (AUC 0.55→0.62; p = 0.025) but not over an individual-domain model (AUC 0.71). Cumulative impairment showed a linear dose-response (aOR = 1.79 per domain). The Time × Mobility GEE interaction was non-significant (β = 0.56, p = 0.15). Associations were robust to penalized estimation and alternative Double-Hit specifications; causal inference is precluded, and the longitudinal subsample was underpowered and selectively retained. Findings support multi-domain ICOPE monitoring in rural primary care; adequately powered longitudinal confirmation is needed.