Abstract
BACKGROUND: Frailty, a geriatric syndrome associated with adverse outcomes, lacks a universal definition. No consensus exists on the most effective frailty scale for predicting mortality. METHODS: This prospective observational study followed community-dwelling volunteers for 6 years. Frailty was measured with the Frailty Index (FI) and the Frailty Phenotype (FP). Concordance was assessed using Cohen's Kappa coefficients. Age-and sex-adjusted Cox regression analyses were conducted to evaluate the association with mortality. RESULTS: Out of 1,114 participants (median age 72 years, IQR 69-77), 186 were classified as frail by the FI, 13 by the FP and 48 by both definitions. The concordance between the two measures was fair (κ = 0.26). Thirty-nine individuals died during the follow-up period. The FI showed a stronger association with mortality (HR 75.29, 95% CI 8.12-697.68, p < 0.001) compared to the FP (HR 3.3, 95% CI 1.45-7.51, p = 0.004). Individuals classified as frail by both definitions had the highest mortality risk and the highest FI scores (median 0.36). CONCLUSION: Definitions of frailty identify different individuals as frail. The FI was more closely related to mortality than the FP. Individuals classified as frail according to both definitions displayed the highest complexity (corresponding also ho higher FI scores) and the greatest mortality. The FI demonstrated a more accurate ability to predict mortality due to its comprehensive nature.