Unveiling Shared and Disease-Specific Metabolic Disruptions in Three Chronic Cardiovascular Diseases

Abstract

Multimorbidity, defined as the coexistence of multiple chronic conditions, is a growing global health challenge. Understanding the metabolic relationships among these conditions, as well as identifying specific metabolites that can differentiate between them, will offer valuable biological insights into their co-occurrence. We employed untargeted plasma metabolomics and machine learning to analyze metabolic changes and associations linked to three chronic cardiovascular diseases—deep venous thrombosis (DVT), hypertension (HPT) and coronary heart disease (CHD)—in a cohort of 196 subjects. Our analysis revealed that 134 metabolites that were common to at least two of these diseases, representing 54.3% of the total 409 significant metabolite-disease associations. The results shows that lipids, amino acids and peptides, hypoxanthine, carnitines and Glucose exhibited interconnected roles across multiple chronic cardiovascular diseases. Additionally, numerous carnitines were found to be specifically linked to DVT and CHD, while the DVT displayed pronounced disruptions in amino acid metabolism. By employing logistic regression models, we identified differential metabolites associated with the three chronic cardiovascular diseases, which demonstrated strong diagnostic performance in both the discovery and validation cohorts. Altogether, our study uncovered extensive and interconnected metabolic dysregulation across the three chronic cardiovascular diseases. The identified differential metabolites hold potential for diagnosing these conditions and offer valuable insights for future clinical interventions and management strategies grounded in metabolomics approaches.