Xin-Yi Cai, Jia-Feng Zhang, Hao Huang, Wei Tang, Qian Lv, Yan Song, Xing-Xing Zhang, Quan-Cai Cai, Jian Zou, Yong-Quan Shi, Tuo Li
Cardiovascular diseases (CVD) remain a leading cause of death globally, yet assessing the combined impact of multiple metabolic disorders on CVD risk is challenging due to a lack of comprehensive tools. We aimed to develop and validate a novel CVD screening model based on metabolic clustering networks. Using data from 4066 adults enrolled in the STONE study (China) through stratified sampling, we constructed a holistic network map of metabolic health, incorporating obesity, dyslipidemia, glucose disorders, and hepatic, renal, thyroid, and bone conditions. Through network analysis, 12 central and clinically accessible indicators-including abdominal obesity, fatty liver, LDL-C, HbA1c, eGFR, TSH, and bone density-were selected to establish the CardioMet12 scoring system, with an AUC of 0.771 (95% CI: 0.733-0.808), sensitivity of 0.756, and specificity of 0.732. A score above 56 corresponded to an observed CVD prevalence exceeding 50% in the STONE cohort. External validation in an independent US population-based epidemiological sample demonstrated satisfactory performance. Higher scores were significantly associated with metabolic syndrome diagnosis and advanced cardiovascular-kidney-metabolic staging in both cohorts. In conclusion, CardioMet12 is a robust metric that captures the complex interplay among multiple metabolic disorders, offering a comprehensive and proactive tool for enhanced CVD screening and risk evaluation.