Plasma metabolomics differentiating and predicting prognosis of coronary artery disease patients with distinct nutritional status.

This study investigated the metabolic mechanisms underlying the association between malnutrition and poor prognosis in coronary artery disease (CAD). We hypothesized that specific metabolites associated with nutritional status impact all-cause mortality and Major Adverse Cardiovascular Events in CAD patients. To test this hypothesis, we evaluated the nutritional status of 5182 CAD patients from multiple centers using three nutritional risk screening tools and analyzed the impact on CAD outcomes with restricted cubic splines and Cox regression. Poor nutritional status was found to be linked to increased adverse outcomes. Further analysis using multiple linear regression and mediation analysis identified elevated concentrations of β-pseudouridine and dulcitol, and decreased concentrations of l-tryptophan and LPC (18:2/0:0), among other metabolites, as mediators of this association. Employing Least Absolute Shrinkage and Selection Operator for variable selection, we integrated these metabolites with clinical variables, which significantly improved the predictive accuracy for adverse outcomes. Our results highlight significant metabolic disparities in CAD patients based on nutritional status and provide novel insights into the role of nutrition-associated metabolites in CAD prognosis. These findings suggest that customized nutritional interventions targeting these metabolites could positively influence the progression of CAD.