Machine learning based model for predicting cardiovascular disease using dynamic triglyceride-glucose index: a...

Area of Science:

• Endocrinology
• Cardiology
• Public Health

Background:

• Cardiovascular disease (CVD) poses a significant global health burden, especially in aging populations.
• Insulin resistance, indicated by the Triglyceride-glucose (TyG) index, is a key factor in CVD development.
• The China Health and Retirement Longitudinal Study (CHARLS) provides valuable data for investigating CVD risk factors in older Chinese adults.

Purpose of the Study:

• To examine the association between dynamic changes in the Triglyceride-glucose (TyG) index and the risk of cardiovascular disease (CVD) in Chinese adults aged 45 and older.
• To compare the predictive power of dynamic TyG index changes versus static abnormal glucose metabolism for CVD events.
• To identify demographic subgroups with differential risks associated with TyG index dynamics.

Main Methods:

• Utilized five waves of the China Health and Retirement Longitudinal Study (CHARLS) data, including 5,625 participants with complete TyG index and CVD data.
• Categorized participants into four groups based on TyG index changes from 2011 to 2015: low-low, low-high, high-low, and high-high.
• Employed Cox proportional hazards models and machine learning algorithms (random forest, XGBoost, etc.) to analyze CVD risk and predictive performance.

Main Results:

• A stable high TyG index was significantly associated with increased risk of incident CVD (HR=1.18) and stroke (HR=1.45) compared to a stable low TyG index.
• Dynamic TyG index changes demonstrated a greater predictive value for CVD than abnormal glucose metabolism alone, particularly for stroke.
• Subgroup analyses revealed consistently elevated risks in the stable high TyG group, especially among individuals under 65, females, those with higher education, lower BMI, and higher depression scores.

Conclusions:

• Dynamic changes in the Triglyceride-glucose (TyG) index are significantly correlated with cardiovascular disease (CVD) risks in middle-aged and older adults.
• Monitoring TyG index trends offers a valuable approach for predicting and managing cardiovascular health.
• Implementing targeted interventions based on TyG index dynamics is essential for reducing CVD incidence in this population.