Sex-specific adipose-muscle patterns from CT reveal metabolic syndrome risk in the elderly with machine learning
Authors' conclusion
Theses findings identified distinct sex-specific adipose-muscle patterns in MS, providing new insights for personalized prevention and treatment strategies. The robust predictive performance of our models supported the clinical utility of these biomarkers for MS risk assessment.
Study context
- Population
- Elderly participants with and without metabolic syndrome; CT quantification of subcutaneous adipose tissue (SAT), skeletal muscle (SM) and intramuscular adipose tissue (IMAT)180 metabolic syndrome (MS) patients
- Participants
- 316 participants
- Exposure
- MS patients vs controls, by sex
- MSn = 180
- Outcome
- Prediction of MS by sex-specific XGBoost models
Methodological detailsCT-based body composition quantification, XGBoost machine learning with SHAP interpretation, IRB approved
- CT-based body composition quantification
- XGBoost machine learning with SHAP interpretation
- IRB approved
Main finding
As reported in the abstract
- AUC in females
- 0.875 (95 %CI:0.854-0.897)
- AUC in males
- 0.902 (95 %CI:0.885-0.918)
- Female MS patients had significantly more IMAT vs controls
- P = 0.043, 0.030, 0.029
- Male MS subjects had higher SMI in different parts of lower extremities
- P = 0.045, 0.043, 0.005
Observational data: associations, not effects.
Further findings
Waist circumference and BMI
- Greater waist circumference in MS group
- P < 0.001
- Higher BMI in MS group
- P = 0.003, 0.038
SHAP dependence analysis
- U-shaped SMI-MS risk relationship in males; females showed greater risk at low skeletal mass index
Original publication
Sex-specific adipose-muscle patterns from CT reveal metabolic syndrome risk in the elderly with machine learning. Diabetes Res Clin Pract. Published 7 November 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →