A CVD Classification Model for Individuals With Obesity: Multi-Ethnic Validation Based on Multiple Metabolic Indicators.

Abstract

The global rise in obesity, a major risk factor for cardiovascular disease (CVD), contributes substantially to the disease burden. We developed and validated a model optimized for classifying prevalent CVD in individuals with obesity. In the training cohort, risk factors were screened by univariate and LASSO regression and then incorporated into a multivariate logistic model to establish the classifier. Based on this model, a nomogram was constructed. The area under the curve (AUC), calibration curve, and decision curve analysis (DCA) were used to evaluate model performance. Sensitivity was analyzed using a unified BMI standard in the Korea cohort. A web-based dynamic nomogram was also implemented (https://min115.shinyapps.io/CVDrisk/). The final nomogram incorporated age, hypertension, diabetes, metabolic score for visceral Fat (METS-VF), serum creatinine, and blood urea nitrogen. In internal validation, the model achieved an AUC of 0.799, with a Hosmer-Lemeshow p value of 0.823. In external validation, the AUC was 0.821, with a Hosmer-Lemeshow p value of 0.083. In both the internal and external validation sets, the model demonstrated good clinical utility. This study developed and validated a nomogram for the classification of prevalent CVD in individuals with obesity, providing a validated model optimized for the population with obesity.