Preprints.org · Nov 10, 2026
Comparative Validity of Skeletal Muscle Index Normalization Methods in Predicting Cardiopulmonary Efficiency in Liver Transplant Recipients
By Lo Russo E, Corsi M, Gitto S et al.
Objectives: To perform a head-to-head comparative evaluation of three Skeletal Muscle Index (SMI) normalization methods (SMIH², SMI100, and SMIBMI) in predicting cardiopulmonary efficiency (oxygen pulse, VO2/HR) and local muscle architecture in male liver transplant (LT) recipients, and to establish population-specific ROC-derived screening cut-off values for detecting low cardiopulmonary reserve. Method: In this cross-sectional study, 32 stable male LT recipients (mean age 61.03 ± 8.53 years) underwent whole-body bioelectrical impedance vector analysis (BIVA) to derive SMI variants, B-mode muscle ultrasound of the rectus femoris cross-sectional area (RF CSA), and cardiopulmonary exercise testing (CPET) on a cycle ergometer. Hierarchical multiple linear regression models evaluated the incremental predictive contribution (ΔR²) of each SMI metric over RF CSA for VO2/HR. Discriminative performance and high-sensitivity screening cut-offs were determined using ROC curve analysis and Youden's index. Result: Hierarchical regression revealed that adding SMIBMI (ΔR² = 0.099, p = 0.033) and SMIH² (ΔR² = 0.084, p = 0.046) significantly increased the explained variance of VO2/HR over the baseline model containing RF CSA alone (R² = 0.195). Conversely, SMI100 did not significantly improve model fit (ΔR² = 0.038, p = 0.131). Both SMIH² (AUC = 0.791, p = 0.009) and SMIBMI (AUC = 0.755, p = 0.022) exhibited good discriminative capacity for impaired oxygen pulse (VO2/HR < 13.5 mL/beat). Optimal high-sensitivity screening cut-off thresholds were identified at SMIBMI ≤ 0.802 (100.0% sensitivity, 45.0% specificity) and SMIH² ≤ 6.89 kg/m² (100.0% sensitivity, 45.0% specificity). Conclusion: SMI normalization methods are not clinically interchangeable when assessing cardiopulmonary efficiency in male LT recipients. SMIBMI and SMIH² demonstrate superior predictive validity for oxygen pulse compared to SMI100, with SMIBMI offering the highest incremental explanatory power over local muscle ultrasound metrics. The population-specific cut-offs (SMIBMI ≤ 0.802 and SMIH² ≤ 6.89 kg/m²) provide reliable clinical decision limits to identify recipients at risk of exercise intolerance and guide referral to targeted rehabilitation.
doi.org/10.20944/preprints202611.1080.v1Open accessBiomarkers