Frailty across the lifespan in heart failure: age-modified drivers and prognostic implications

Background Frailty is a multidimensional syndrome of reduced physiologic reserve associated with adverse outcomes in heart failure (HF). Whether its underlying drivers and prognostic significance differ by age remains poorly understood. Purpose To assess age-specific correlations of frailty, predictors of incident frailty, and the prognostic implications of frailty across the adult lifespan in patients with HF. Methods In this retrospective cohort study, we analyzed 23,556 adults with HF from a tertiary care registry (2009-2025). Frailty was defined using a multimodal phenotype incorporating functional, nutritional, and biochemical domains. Patients were stratified into age tertiles (≤70, 70-82, and ≥82 years [reference group]). Among those non-frail at baseline (≥30 days post-HF diagnosis), incident frailty was ascertained during follow-up. Multivariable logistic regression identified determinants of prevalent frailty at index admission. Multivariable Cox proportional consisting of frailty, sex, left ventricular ejection fraction, and key comorbidities, evaluated predictors of incident frailty and all-cause mortality, with tests for age interaction. Results The cohort had a median age of 74 years, with 45% frail at baseline (27% ≤70; 35% 70–82; 50% ≥82). Among those non-frail, 1,536 (10%) developed incident frailty over a median 8.8 years (IQR 3.4-14.3), increasing stepwise by age (7.5%, 11%, 16%; figure). Over 3.5 years (IQR 1.2-7.3), 15,631 patients (57%) died, with mortality rising across tertiles (40%, 63%, 81%). Frailty independently predicted all-cause mortality across all age groups, with graded attenuation with advancing age (HR 2.35 [2.16-2.56] in ≤70 years; 1.97 [1.83-2.12] in 70-82 years; 1.78 [1.64-1.94] in ≥82 years). In multivariable models, the association between chronic kidney disease (CKD) and prevalent frailty attenuated with advancing age (OR 2.31 [95% CI 2.04-2.62] in ≤70 years; 1.65 [1.47-1.85] in 70-82 years; 1.20 [1.03-1.40] in ≥82 years; p-interaction <0.001), with a similar pattern observed for peripheral artery disease (p-interaction=0.046). In contrast, diabetes demonstrated a progressively stronger inverse association with frailty in older age groups (p-interaction=0.002). For incident frailty, CKD increased risk in patients ≤70 years (HR 1.40, 95% CI 1.16-1.69) though was not significant in ≥82 years (HR 0.85, 95% CI 0.68-1.04; p-interaction=0.001). Similarly, female sex was associated with higher risk in ≤70 years (HR 1.27, 95% CI 1.05-1.52) but lower risk in ≥82 years (HR 0.79, 95% CI 0.64-0.98; p-interaction=0.001). Conclusions In HF, frailty is a dynamic, age-modified phenotype with distinct clinical drivers across the lifespan. Comorbidity-frailty coupling is strongest in younger patients and attenuates with advancing age. Nevertheless, frailty confers substantial and independent mortality risk at every age, underscoring its central role in HF risk stratification and management.