Proteomics assisted improvement prediction of risk for adverse cardiovascular events in individuals with metabolic syndrome

Traditional clinical models, designed for the general population, are insufficient for capturing the complex molecular pathology that drives major adverse cardiovascular events (MACEs) in individuals with metabolic syndrome (MetS). This study aimed to leverage large-scale proteomics to improve risk stratification in MetS and to investigate the potential causal effects of identified proteins using Mendelian Randomization (MR). We analyzed 2,911 plasma proteins in 12,308 MetS participants from the UK Biobank. A protein prediction model was constructed using the least absolute shrinkage and selection operator method to select MACE-associated proteins in the training set. The model’s predictive performance was evaluated in a validation set and compared to the clinical PREVENT model. To investigate whether these identified proteins have a causal relationship with MACE, we utilized cis -pQTLs as genetic instruments to perform a two-sample MR analysis. Over a median 14-year follow-up, 1,899 incident MACEs occurred. In the validation set, the protein model significantly outperformed the clinical PREVENT model (C-statistic: 0.730 vs. 0.673, P = 0.003). Crucially, among individuals classified as intermediate-risk by the PREVENT model, the protein score achieved a total net reclassification index of 0.44, driven by the correct reclassification of 64% of individuals who developed MACE. Time-dependent analysis revealed exceptional predictive accuracy for short-term heart failure risk (3-year AUC = 0.90), which was superior to long-term predictions. MR analysis against individual MACE components identified four proteins with causal evidence: genetically higher AGER (soluble RAGE) was associated with lower heart failure risk, and higher MMP-12 with lower stroke risk, whereas higher VAMP5 and NCAN were associated with increased coronary heart disease risk. A proteomics-derived model deconstructs the molecular heterogeneity of MACEs risk in individuals with MetS, significantly improving conventional models. Furthermore, integrating MR analysis distinguishes causal drivers from biomarkers, providing novel molecular insights to guide personalized prevention.