Population‐Based Prediction Equations for Appendicular Skeletal Muscle Mass: A Three‐Phase Validation Study in 13 582 Asian Adults
- 1Graduate Institute of Clinical Pharmacy, College of Medicine National Taiwan University Taipei Taiwan
- 2National Center for Geriatrics and Gerontology Obu Aichi Japan
- 3Center for Healthy Longevity and Aging Sciences National Yang Ming Chiao Tung University Taipei Taiwan
Abstract
Background Accurate estimation of appendicular skeletal muscle mass (ASM) is crucial for sarcopenia diagnosis, but device‐dependent approaches (dual‐energy X‐ray absorptiometry [DXA] or bioelectrical impedance analysis [BIA]) limit widespread application in community settings. Developing prediction equations using readily available anthropometric measurements and common laboratory data (such as serum creatinine) offers a practical alternative for large‐scale screening. However, no population‐based prediction equations using these accessible parameters have been developed specifically for Asian populations, where body composition differs significantly from Western cohorts. Methods We conducted a three‐phase investigation: Phase 1 validated existing NHANES‐derived equations across three cohorts (n = 13 582) from Taiwan (ILAS, n = 2780; NAHSIT, n = 6853) and Japan (NILS‐LSA, n = 3949); Phase 2 developed new population‐based equations using the NAHSIT cohort; and Phase 3 performed external validation of newly developed NAHSIT‐derived equations in an independent Japanese cohort (NILS‐LSA). Model performance was assessed using R‐squared, root mean square error (RMSE) and area under the receiver operating characteristic curve (AUC) for detecting low muscle mass based on Asian Working Group for Sarcopenia (AWGS) 2019 criteria. Results In Phase 1 external validation, existing equations showed acceptable to strong predictive accuracy across populations (R2 = 0.52–0.89). Equation 4 (incorporating sex, age, height and weight) demonstrated robust performance across ILAS (R2 = 0.80, RMSE = 1.88), NILS‐LSA (R2 = 0.86, RMSE = 1.51) and NAHSIT (R2 = 0.89, RMSE = 1.52) cohorts. For detecting low muscle mass using AWGS 2019 criteria, AUC values ranged from 0.79 to 0.91 in males and 0.76 to 0.89 in females across cohorts. In Phase 2, our newly developed population‐based Equation 4 achieved superior performance (R2 = 0.90, RMSE = 1.45) with minimal prediction bias (median error: 0.06, IQR: 1.81) and excellent diagnostic capability (AUC = 0.90 in both sexes; sensitivity/specificity: 0.85/0.80 in males and 0.87/0.80 in females). Phase 3 external validation in the independent Japanese cohort (NILS‐LSA) demonstrated robust cross‐population generalizability, with Equation 4 achieving the highest AUC values (0.90 in males and 0.85 in females). Conclusions Our three‐phase investigation has established and validated the first population‐based prediction equations for ASM estimation in Asian populations. These clinically accessible equations overcome device‐dependency limitations and enable practical, large‐scale sarcopenia screening in community settings, representing a significant advance for early identification of sarcopenia and promotion of healthy muscle aging across Asian populations.
Related
- Association of glycemic coefficient of variation with reduced appendicular skeletal muscle mass index among patients with type 2 diabetes mellitus: a cross‑sectional study
- Focusing on VaIN2+: Development and validation of a prediction nomogram based on a large colposcopy population
- Association between statin use and longitudinal changes in skeletal muscle mass: A population-based cohort study
- Photograph-Based Skeletal Malocclusion Screening with Geometric Machine Learning: External Validation Across Turkish and Non-European Cohorts
- GLP-1 Receptor Agonist Initiation Is Associated With Excess Skeletal Muscle Loss in Adults With Diabetes: A Prospective Population-Based MRI Study
BibTeX
@article{lai2026populationbased,
title = {{Population‐Based Prediction Equations for Appendicular Skeletal Muscle Mass: A Three‐Phase Validation Study in 13 582 Asian Adults}},
author = {Hsi‐Yu Lai and Shu Zhang and I‐Tzu Chen and Rei Otsuka and Wei‐Ju Lee and Chih‐Kuang Liang and Ko‐Han Yen and Hidenori Arai and Fei‐Yuan Hsiao and Liang‐Kung Chen},
journal = {Journal of Cachexia, Sarcopenia and Muscle},
year = {2026},
month = oct,
doi = {10.1002/jcsm.70360},
url = {https://doi.org/10.1002/jcsm.70360},
note = {CC BY},
}
RIS
TY - JOUR
TI - Population‐Based Prediction Equations for Appendicular Skeletal Muscle Mass: A Three‐Phase Validation Study in 13 582 Asian Adults
AU - Hsi‐Yu Lai
AU - Shu Zhang
AU - I‐Tzu Chen
AU - Rei Otsuka
AU - Wei‐Ju Lee
AU - Chih‐Kuang Liang
AU - Ko‐Han Yen
AU - Hidenori Arai
AU - Fei‐Yuan Hsiao
AU - Liang‐Kung Chen
PY - 2026
DA - 2026/10/26
JO - Journal of Cachexia, Sarcopenia and Muscle
DO - 10.1002/jcsm.70360
UR - https://doi.org/10.1002/jcsm.70360
AB - Background Accurate estimation of appendicular skeletal muscle mass (ASM) is crucial for sarcopenia diagnosis, but device‐dependent approaches (dual‐energy X‐ray absorptiometry [DXA] or bioelectrical impedance analysis [BIA]) limit widespread application in community settings. Developing prediction equations using readily available anthropometric measurements and common laboratory data (such as serum creatinine) offers a practical alternative for large‐scale screening. However, no population‐based prediction equations using these accessible parameters have been developed specifically for Asian populations, where body composition differs significantly from Western cohorts. Methods We conducted a three‐phase investigation: Phase 1 validated existing NHANES‐derived equations across three cohorts (n = 13 582) from Taiwan (ILAS, n = 2780; NAHSIT, n = 6853) and Japan (NILS‐LSA, n = 3949); Phase 2 developed new population‐based equations using the NAHSIT cohort; and Phase 3 performed external validation of newly developed NAHSIT‐derived equations in an independent Japanese cohort (NILS‐LSA). Model performance was assessed using R‐squared, root mean square error (RMSE) and area under the receiver operating characteristic curve (AUC) for detecting low muscle mass based on Asian Working Group for Sarcopenia (AWGS) 2019 criteria. Results In Phase 1 external validation, existing equations showed acceptable to strong predictive accuracy across populations (R2 = 0.52–0.89). Equation 4 (incorporating sex, age, height and weight) demonstrated robust performance across ILAS (R2 = 0.80, RMSE = 1.88), NILS‐LSA (R2 = 0.86, RMSE = 1.51) and NAHSIT (R2 = 0.89, RMSE = 1.52) cohorts. For detecting low muscle mass using AWGS 2019 criteria, AUC values ranged from 0.79 to 0.91 in males and 0.76 to 0.89 in females across cohorts. In Phase 2, our newly developed population‐based Equation 4 achieved superior performance (R2 = 0.90, RMSE = 1.45) with minimal prediction bias (median error: 0.06, IQR: 1.81) and excellent diagnostic capability (AUC = 0.90 in both sexes; sensitivity/specificity: 0.85/0.80 in males and 0.87/0.80 in females). Phase 3 external validation in the independent Japanese cohort (NILS‐LSA) demonstrated robust cross‐population generalizability, with Equation 4 achieving the highest AUC values (0.90 in males and 0.85 in females). Conclusions Our three‐phase investigation has established and validated the first population‐based prediction equations for ASM estimation in Asian populations. These clinically accessible equations overcome device‐dependency limitations and enable practical, large‐scale sarcopenia screening in community settings, representing a significant advance for early identification of sarcopenia and promotion of healthy muscle aging across Asian populations.
SN - 2190-6009
SN - 2190-5991
ER -