Risk Stratification and Early Warning with Personalized Intervention for Hyperlipidemia in Middle-Aged and Older Adults by Integrating Traditional Chinese Medicine Constitution, Lipid Profiles, and Activity Capacity

This study develops a three-level risk warning and personalized intervention framework integrating lipid indicators, traditional Chinese medicine constitution scores, and activity capacity using 1,000 health management records. A dual-response strategy identifies phlegm-dampness constitution as informationally independent from lipid levels.

Logistic regression and XGBoost reveal total cholesterol and triglycerides as dominant risk drivers, with constitution variables serving as conditional modulators. The population is stratified into low-risk (20.7%), medium-risk (56.9%), and high-risk (22.4%) groups via Youden index and management-proportion thresholds. Three high-risk patterns are identified: phlegm-dampness with multiple lipid abnormalities, low activity capacity, and elevated triglycerides.