Machine Learning Analysis of Sex Differences in Cardiovascular-Kidney-Metabolic Risk Factors and Prognosis Among Patients With Moderate-to-Severe Coronary Artery Calcification: Prospective Cohort Study

Background Patients with coronary artery calcification exhibit notable sex differences in clinical presentation, particularly concerning the role of cardiovascular-kidney-metabolic (CKM) risk factors and their impact on prognoses. However, the precise nature of these sex-specific differences remains incompletely understood. Objective This study aimed to investigate sex disparities in CKM risk factors among patients with moderate-to-severe coronary artery calcification (MSCAC) and elucidate their association with adverse clinical outcomes. Methods A total of 2418 patients with MSCAC undergoing their first percutaneous coronary intervention were included. Hazard ratios (HRs) were computed to evaluate sex differences in the prognostic significance of various CKM risk factors, including chronic kidney disease (CKD), diabetes mellitus (DM), obesity, hypertension, and hypertriglyceridemia. Four machine learning models—logistic regression, extreme gradient boosting (XGBoost), random forest, and support vector machine—were constructed to predict adverse events. The best-performing model was interpreted using Shapley additive explanations (SHAP) values to identify the relative importance of CKM risk factors and clarify potential sex disparities. Major adverse cardiovascular events (MACEs) were defined as all-cause mortality, nonfatal myocardial infarction, and unplanned repeat revascularization. Results Among the participants, 86.9% (2101/2418) had ≥1 CKM risk factor, while 3.1% (74/2418) had ≥4 risk factors. CKD was independently associated with the occurrence of MACEs in both female patients (HR 2.65, 95% CI 1.50-4.69) and male patients (HR 1.53, 95% CI 1.02-2.60). Notably, the association was stronger in female patients, with a female-to-male multivariate-adjusted HR ratio for CKD of 1.68 (95% CI 1.04-2.97). DM was also associated with MACEs in both sexes, with adjusted HRs of 1.20 (95% CI 1.02-1.92) in female patients and 1.59 (95% CI 1.19-2.12) in male patients. Among the models evaluated, XGBoost demonstrated the highest predictive performance in the test set (area under the curve 0.92; average precision 0.92; F1-score=0.86). XGBoost maintained good predictive performance in the external validation cohort (area under the curve 0.86; average precision 0.71; F1-score=0.68). Baseline DM was identified as the CKM risk factor with the highest feature importance for MACEs, with SHAP values of 0.17 in female patients and 0.23 in male patients. Conversely, CKD emerged as the most important CKM risk factor in female patients (SHAP value 0.18), while DM ranked highest in male participants. Conclusions This study confirmed that CKM risk factors and their influence on prognosis in patients with MSCAC exhibit significant sex differences. The application of machine learning, particularly XGBoost, facilitates a deeper understanding of these disparities and provides a basis for personalized, sex-specific risk assessment and targeted interventions for patients with MSCAC.