Age- and sex-specific muscle health cut-offs improve prognostic accuracy in older patients with heart failure: a secondary analysis of a nationwide, multicenter cohort registry

Abstract Aims Universal muscle health cut-offs for older patients with heart failure often overlook age- and sex-specific biological heterogeneity, potentially compromising prognostic accuracy. We aimed to develop and validate age- and sex-specific cut-off values for muscle health metrics to enhance prognostic performance. Methods and results From a nationwide, multicenter cohort registry, 9693 patients aged ≥65 years hospitalised for heart failure were analysed. We assessed five muscle health metrics (arm and calf circumference, handgrip strength, gait speed and Short Physical Performance Battery) and derived age- and sex-specific cut-off values for predicting 1-year all-cause mortality across three age groups (65–74, 75–84, 85+ years). Age- and sex-specific thresholds varied substantially across demographics (e. g. handgrip strength: 24. 5–19. 6 kg in men and 16. 6–11. 6 kg in women from the youngest to the oldest group) and showed robust prognostic associations with mortality (adjusted hazard ratio: 1. 44–3. 25, all P. 001). Importantly, age- and sex-specific cut-offs consistently outperformed fixed criteria by Asian Working Group for Sarcopenia in 2019, evidenced by substantial improvements in Akaike information criterion that ranged from 8. 1 to 119. 4, continuous net reclassification improvement (0. 050 to 0. 342), and integrated discrimination improvement (0. 002 to 0. 023) across four metrics. Conclusion Age- and sex-specific muscle health thresholds offer superior prognostic accuracy over universal fixed values in older patients with heart failure, evidenced by substantial improvements in model fit, reclassification and discrimination. This individualised approach enables more precise risk stratification and supports the transition toward personalised assessment in geriatric heart failure care. Data source registration University Hospital Medical Information Network Clinical Trials Registry, URL: https: //center6. umin. ac. jp/cgi-open-bin/ctr/ctrᵥiew. cgi? recptno=R000054579, Registration number: UMIN000047893.