Abstract
BACKGROUND: Cardiovascular multimorbidity (CMM), defined as the coexistence of multiple cardiometabolic diseases, has posed an escalating global health burden associated with premature mortality. Systemic inflammation has been increasingly recognized as a central mechanism linking cardiometabolic diseases, yet the prognostic implications of routine inflammatory and nutritional biomarkers in patients with CMM remained unclear. METHODS: This cohort study analyzed 1,928 CMM patients from the National Health and Nutrition Examination Survey (NHANES) and 364 patients from a Chinese cohort (Gaoyou). Ten inflammatory and nutritional parameters were evaluated. Associations with all-cause and cardiovascular mortality were assessed using multivariable Cox regression and restricted cubic splines. Feature selection (SHAP, Boruta, and Lasso) was employed to identify optimal predictors, followed by construction and validation of nomogram and machine learning (ML) models. RESULTS: The systemic inflammation response index (SIRI) emerged as the strongest independent predictor of mortality. Patients in the highest SIRI quartile exhibited significantly increased risks of all-cause mortality (HR = 2.34, 95% CI: 1.88-2.90) and cardiovascular mortality (HR = 2.09, 95% CI: 1.47-2.98), with consistent performance across various subgroups. Nomograms incorporating SIRI demonstrated excellent discrimination (AUCs > 0.7) and clinical utility. Among the ML models, XGBoost achieved the highest predictive efficiency at 60, 120, and 150 months. CONCLUSION: SIRI, reflecting the combined influence of inflammatory responses and nutritional status, provided an available and independent biomarker for mortality risk stratification in CMM patients. The validated nomograms and web-based prediction tool offered clinicians a practical approach for individualized prognosis and informed future strategies targeting systemic inflammation and nutrition in multimorbidity management.