High-Frequency Functional Trajectories Predict Depressive Worsening in Singapore’s Community-Dwelling Older Adults: A K-means Longitudinal Clustering Analysis

Background/Objectives: Functional decline and depression often coexist in older adults, yet local Singapore-based research often lacks detailed temporal resolution due to heterogeneity in ageing. This study employs non-parametric, data-driven longitudinal clustering to analyse functional trajectories and their association with depression, using high-frequency data to pinpoint key intervention periods. Methods: Data were drawn from 4,273 older adults from Singapore Life Panel® (2020–2024). Participants completed quarterly self-reported assessments of ADL, IADL and depressive symptoms (8-item CES-D). We employed k-means longitudinal clustering (kml) to identify functional trajectory groups and Cox regression to evaluate the hazard of worsening depression (≥5-point increase in CES-D). Results: Three trajectories emerged: Stable, Medium (gradual increase in functional difficulty), and High (rapid increase in functional difficulty). The High cluster, comprising older and socioeconomically disadvantaged individuals, exhibited worse baseline health and psychosocial factor scores. Depression scores escalated in the Medium and High groups. Kaplan-Meier analysis revealed a faster rate of symptom worsening in these groups than in the Stable group. The High ADL trajectory predicted a 1.65-fold increased hazard of depression worsening after adjusting for sociodemographic and psychosocial confounders. Conclusions: Rapid functional decline acts as a precursor to worsening depressive symptoms. Routine monitoring of functional trajectories offers a strategic window for proactive mental health interventions in at-risk older adults.