Metabolic, Atherogenic, and Anthropometric Markers for Discrimination of Premature Coronary Artery Disease in a Predominantly Statin-Treated Cohort.

Purpose: Premature coronary artery disease (PCAD) is increasingly recognized in younger adults and requires improved risk stratification. Conventional lipid markers may have limited discriminative value, particularly when lipid levels reflect ongoing statin therapy. This study evaluated the associations and discriminatory performance of metabolic, atherogenic, and anthropometric indices for PCAD and developed internally validated dynamic nomograms for PCAD case discrimination. Patient and Methods: This multicenter retrospective case-control study used data from the Iran Premature Coronary Artery Disease registry. A total of 3063 adults undergoing coronary angiography were included, comprising 1929 patients with PCAD and 1134 controls. Twelve metabolic, lipid-based, and anthropometric indices were calculated from fasting glucose, lipid profiles, and body measurements. Participants were randomly divided into training and validation cohorts at a 7:3 ratio. Logistic regression, receiver operating characteristic analysis, calibration assessment, integrated discrimination improvement, net reclassification improvement, and decision-curve analysis were used to evaluate model performance. Results: Atherogenic index of plasma (AIP), triglyceride-glucose index (TyG), and a body shape index (ABSI) showed consistent independent associations with PCAD. Individually, these indices showed modest discrimination. The conventional clinical model achieved an AUC of 0.677. Adding AIP and binary ABSI increased the AUC to 0.721, while the TyG plus binary ABSI model showed the highest performance, with an AUC of 0.724, representing modest but statistically significant improvements over the conventional model. Both final models showed favorable internal calibration, and improvement in reclassification metrics, and decision-curve net benefit. Two dynamic web-based nomograms were developed for individualized PCAD case discrimination. Conclusion: TyG, AIP, and ABSI were independently associated with PCAD and provided modest incremental discrimination beyond conventional risk factors in this predominantly statin-treated case-control cohort. The resulting dynamic nomograms provide an accessible framework for PCAD case discrimination; however, their clinical applicability and estimation of absolute risk require external validation in independent, preferably prospective, populations.