Work

Frailty accumulation predicts depression across cultures: evidence from four aging cohorts

Year: 2026
Type: article
Abstract: In the context of global population aging, frailty is closely associated with geriatric depression, yet large-scale longitudinal evidence examining the dose-response relationship across diverse cultural settings remains scarce. The consistency and magnitude of a potential indirect association through cognitive function require evaluation across cohorts. Using representative cohorts from four countries, this study examined the prospective association between cumulative frailty exposure and incident depression, examined cognitive function as a candidate mediator, and secondarily explored the role of sleep quality. We integrated data from CHARLS, HRS, ELSA, and SHARE, comprising 12,042 adults aged ≥ 45 years. In separate samples, 10,303 participants remained after excluding those with depression at wave 1, and 10,325 remained after excluding those with depression at wave 3. A unified 29-item frailty index (FI; range 0–1) was constructed across all cohorts. Cumulative frailty exposure was operationalized as the area under the curve of FI over the first three waves (CFI). Cognitive function was assessed at wave 4 using a harmonized battery (episodic memory, orientation, and working memory). Sleep measures were cohort-specific and examined only in exploratory analyses. Depressive symptoms were classified using cohort-specific validated scales; the mediation outcome was the binary classification at the fifth measurement. Logistic regression, Cox proportional hazards models, mediation analysis, restricted cubic splines (RCS), and random intercept cross-lagged panel models (RI-CLPM) were applied. Greater cumulative frailty exposure was robustly associated with incident depression across all cohorts in fully adjusted logistic models (ORs 3.46–6.09; all p < 0.001). In the primary complete-case models, the estimated indirect component through cognitive function accounted for 2.05%–5.97% of the total model-based association; support for this component was not consistent across sensitivity analyses, particularly in HRS and ELSA. Exploratory sleep mediation analyses showed significant indirect effects within individual cohorts, yet measurement heterogeneity precluded cross-cultural interpretation. In Cox models, each 1-unit increase in CFI was associated with a higher hazard of depression across the four cohorts (HRs, 2.46–3.62). RCS analyses revealed significant nonlinear risk acceleration at higher exposure levels in ELSA, SHARE, and the pooled sample, whereas HRS showed a more linear pattern. RI-CLPM results indicated that stable between-person traits accounted for substantial covariation, and within-person cross-lagged paths were inconsistent in early waves, becoming positive only in later waves. Cumulative frailty exposure is consistently associated with incident depression across cultures. The primary models suggested a small indirect component through cognitive function, with limited consistency across sensitivity analyses, while sleep-related pathways require harmonized measurement for future evaluation. These observational findings support frailty assessment as a potential risk stratification tool for late-life depression, pending validation through randomized intervention trials.
Language: English
Field: Medicine
Open Access status: gold
Estimated APC paid: $3,290
OpenAlex ID: W7226012828
FWCI:
Cites:
Cited by:
Related to: 10