Characterizing  High-risk Population with Metabolic Syndrome based on Extreme Logistic Regression

Metabolic syndrome (MetS), as a potential factor leading to diabetes and cardiovascular diseases, has attracted increasing global attention, especially on its identification and intervention.This study aims to characterize high-risk individuals with MetS based on an extreme logistic regression (ELR) model. Five key factors of waist circumference (WC), fasting plasma glucose (FPG), blood pressure (including systolic blood pressure (SBD) and diastolic blood pressure (DBP), triglycerides (TG), and high-density lipoprotein cholesterol (HDL-C) are incorporated in an extreme function to predict the severity of MetS. The individuals with high-risk MetS are uniformly identified with the estimated cut-offs of predictors and prediction rules of the MetS population simultaneously. Typically, the identified individuals with higher-risk MetS under extreme diagnostic criteria are more likely to experience a greater burden of relevant diseases. We conclude that the innovative diagnosis criteria for MetS have significant implications for clinical warning, effective intervention strategies, and better resource allocation in clinical settings.