Scientific Reports · Nov 18, 2026
Development and validation of a clinical nomogram for predicting sarcopenia risk in elderly patients with chronic heart failure
By Han F, Zhai Y, Wu Y et al.
To identify risk factors for sarcopenia in elderly chronic heart failure (CHF) patients and develop a predictive nomogram. Cross-sectional study with random split-sample (7:3) derivation and internal validation. Tertiary hospital in Shanxi, China. 317 elderly CHF patients (July 2023–June 2024); 221 in derivation cohort, 96 in validation. The primary outcome was sarcopenia. Primary evaluation metrics included the area under the ROC curve (AUC), calibration plots, and decision curve analysis (DCA). Secondary measures included sensitivity, specificity, and the Kappa coefficient. Sensitivity analyses were performed to assess model robustness, including exclusion of grip strength from the model and reclassification of sarcopenia according to the AWGS 2025 criteria. Five independent risk factors for sarcopenia were identified: fatigue, physical activity level, calf circumference, body mass index (BMI), and grip strength (all p < 0.05). The model showed good performance, with a C-index of 0.951. The AUC was 0.955 (95% CI: 0.929–0.980) in the derivation cohort and 0.951 (95% CI: 0.907–0.995) in the validation cohort. Calibration curves showed good agreement between predicted and observed probabilities, and DCA confirmed the model’s clinical utility. Sensitivity analysis excluding grip strength yielded an AUC of 0.949 (95% CI: 0.925–0.973). Reclassification using AWGS 2025 criteria resulted in an AUC of 0.964 (95% CI: 0.946–0.982), confirming model robustness across different diagnostic frameworks. Fatigue, low physical activity, reduced calf circumference, low BMI, and low grip strength are independent risk factors for sarcopenia in elderly CHF patients. The validated nomogram has strong predictive performance and may help clinicians identify high-risk individuals. Sensitivity analyses confirmed the model’s robustness across different variable compositions and diagnostic frameworks.