Clinical Report: Machine Learning Model for Sarcopenia Risk in CVD

Overview

This study identifies TyG-BMI as a key composite biomarker for sarcopenia risk in cardiovascular disease (CVD) patients and develops a CatBoost machine learning model for risk stratification. The model demonstrates strong discriminative ability across multiple cohorts.

Background

Sarcopenia is prevalent among patients with cardiovascular disease and is linked to adverse outcomes such as functional decline and increased mortality. Current diagnostic methods for sarcopenia are often impractical in routine cardiovascular settings, highlighting the need for accessible risk stratification tools. The integration of composite metabolic biomarkers into machine learning models presents a potential approach for early identification of sarcopenia risk in this population.

Data Highlights

Key Findings

• TyG-BMI was identified as the most informative composite biomarker for sarcopenia risk.
• The CatBoost model achieved an AUC of 0.938 for TyG-BMI in discriminating sarcopenia risk.
• The model demonstrated strong performance across various cohorts, with AUCs ranging from 0.891 to 0.982.
• The Cardiovascular Disease–Sarcopenia Risk Score (CVD-SRS) provided consistent risk stratification across validation cohorts.

Clinical Implications

Incorporating composite biomarkers into routine assessments could enhance risk stratification in cardiovascular practice.

Conclusion

The integration of TyG-BMI into a machine learning model offers a potential approach for effective sarcopenia risk stratification in patients with cardiovascular disease.

Related Resources & Content

1. npj Digital Medicine, 2026 -- Thigh Volume Measurement via Neck-to-Knee Dixon MRI as an Enhanced Biomarker for Sarcopenia: Insights from the UK Biobank Study
2. Frontiers in Cardiovascular Medicine, 2026 -- Evaluation of coronary heart disease risk prediction based on simple physical examination parameters by machine learning model: a retrospective cohort model development and validation study
3. Frontiers in Endocrinology, 2026 -- Artificial intelligence-driven assessment of sarcopenia in orthopedic geriatrics: technical progress and clinical implications
4. Health outcomes of sarcopenia: a consensus report by the outcome working group of the Global Leadership Initiative in Sarcopenia (GLIS), 2025
5. Frontiers in Medicine — Prediction models for sarcopenia in older adults in China: a scoping review
6. A focus shift from sarcopenia to muscle health in the Asian Working Group for Sarcopenia 2025 Consensus Update
7. 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure
8. Health outcomes of sarcopenia: a consensus report by the outcome working group of the Global Leadership Initiative in Sarcopenia (GLIS)
9. JCS/JHFS 2025 Guideline on Diagnosis and Treatment of Heart Failure
10. Overview | Chronic heart failure in adults: diagnosis and management | Guidance | NICE
11. Frontiers | Sarcopenia defined by multidimensional factors and its prognostic role in heart failure: a systematic review and meta-analysis
12. Sarcopenic obesity and risk of cardio-cerebrovascular disease and mortality: a systematic review and meta-analysis | International Journal of Obesity
13. Association Between Sarcopenia Index and Incident Cardiovascular Events in a Chinese Aging Population: Prospective Analysis of the CHARLS Study - PubMed
14. From Renal Markers to Muscle Mass: A Sarcopenia Index Predicts Outcomes in Coronary Artery Disease Patients with Aortic Valve Sclerosis - PubMed
15. Incident cardiovascular diseases and changes in the ratio of serum creatinine to cystatin C: the sarcopenia index - ScienceDirect
16. Frontiers | Association between the triglyceride-glucose-waist-to-height ratio and cardiovascular disease in Chinese adults with sarcopenia or probable sarcopenia
17. JMIR Aging - Association of Insulin Resistance, Sarcopenia, and Risk of Cardiovascular Disease: Findings From the China Health and Retirement Longitudinal Study