A practical estimation equation for appendicular skeletal muscle mass in stroke rehabilitation: validating diagnostic accuracy and predicting functional outcomes.

Background: The gold standard methods for assessing muscle mass in sarcopenia diagnosis are often impractical due to cost and accessibility, necessitating simpler tools. This study evaluates the validity of diagnosing sarcopenia using skeletal muscle mass estimated by a prediction equation in post-stroke rehabilitation patients.

Methods: This retrospective cohort study analyzed hospitalized post-stroke patients. Skeletal muscle mass was assessed using bioelectrical impedance analysis (BIA) and a validated prediction equation (ASM = 0.485 × 0.998^age × 0.814^[female] × 1.006^height × weight^0.680). Sarcopenia was diagnosed following the Asian Working Group for Sarcopenia 2019 criteria. The accuracy of the prediction equation was assessed by correlation with BIA-derived skeletal muscle mass and diagnostic metrics (κ and AUC). Functional outcomes, including motor and cognitive scores from the Functional Independence Measure (FIM), were analyzed using multivariate regression to adjust for confounders.

Results: A total of 748 participants were analyzed. The prediction equation demonstrated a strong correlation with BIA-derived skeletal muscle mass (R² = 0.84, RMSE = 2.04). Diagnostic accuracy for sarcopenia was moderate (men κ = 0.47, AUC = 0.74; women κ = 0.55, AUC = 0.75), with high sensitivity (men 83%, women 96%) and moderate specificity (men 55%, women 65%). Sarcopenia diagnosed using the prediction equation was independently associated with lower FIM motor (men: 87 vs. 78, p  < 0.001; women: 85 vs. 74, p < 0.001) and cognitive scores (men: 32 vs. 28, p  < 0.001; women: 33 vs. 27, p < 0.001) at discharge.

Conclusions: The prediction equation offers a practical and accessible tool for estimating skeletal muscle mass and diagnosing sarcopenia, demonstrating strong correlations with established methods and associations with functional outcomes.