Prognostic utility of the GLIS model for sarcopenia in older patients with heart failure: a post-hoc analysis of the FRAGILE-HF study

Background Sarcopenia is associated with poor prognosis in heart failure (HF) patients. While the Asian Working Group for Sarcopenia (AWGS) 2019 criteria are widely used, the emergence of new diagnostic frameworks has led to inconsistencies. The Global Leadership Initiative on Sarcopenia (GLIS) proposes a consensus definition incorporating muscle mass, muscle strength, and muscle-specific strength while redefining physical performance as an outcome rather than a diagnostic criterion. However, the GLIS model lacks standardized measurement methods and cut-off values, and its prognostic relevance remains unclear. Purpose This study aimed to evaluate the prognostic utility of the GLIS model in older patients with HF and compare its predictive performance against the AWGS 2019 criteria. Methods This post-hoc analysis of the FRAGILE-HF study included 891 older patients hospitalized for HF (median age: 81 years, 41.9% female). Patients were classified as sarcopenia, possible sarcopenia, or non-sarcopenia using the GLIS model, where sarcopenia was defined as reduced muscle mass with either low muscle strength or low muscle-specific strength. Possible sarcopenia was defined as a reduction in at least one of these components without meeting full criteria. Sarcopenia was also assessed using the AWGS 2019 criteria for comparison. The primary outcome was 2-year all-cause mortality, and the secondary outcome was impaired physical performance (gait speed, five-time chair stand test, Short Physical Performance Battery [SPPB], and six-minute walk distance [6MWD]). Prognostic performance was compared between AWGS 2019- and GLIS-based models using net reclassification improvement (NRI). Results According to the GLIS model, sarcopenia, possible sarcopenia, and non-sarcopenia were identified in 186 (20.9%), 539 (60.5%), and 166 (18.6%) patients, respectively. During the 2-year follow-up (98.7% completion rate), 159 (17.8%) patients died, with mortality rates of 8.4%, 16.9%, and 29.0% in the non-sarcopenia, possible sarcopenia, and sarcopenia groups, respectively (log-rank test, P < 0.001). Cox proportional hazards analysis, adjusted for log-transformed BNP and the MAGGIC risk score, showed increased mortality risk in possible sarcopenia (HR 1.79, 95% CI 0.95–3.38, P = 0.072) and sarcopenia (HR 3.38, 95% CI 1.74–6.56, P < 0.001). The GLIS-based model demonstrated superior prognostic discrimination over the AWGS 2019-based model (NRI 0.269, 95% CI 0.141–0.397, P < 0.001). Additionally, both possible sarcopenia and sarcopenia were significantly associated with impaired physical performance across all measures (P for trend < 0.001). Conclusions Sarcopenia diagnosed using the GLIS model was significantly associated with worse prognosis and impaired physical performance in older patients with HF. Compared to the AWGS 2019 model, the GLIS model provided superior prognostic discrimination, supporting its clinical utility for risk stratification in older patients with HF.