Predictive models of cardiovascular risk based on metabolomic data

Cardiovascular disease (CVD) remains the leading cause of morbidity and mortality worldwide, largely due to the lack of effective tools for early detection. This project proposes the development of an innovative system for early prediction of cardiovascular risk based on serum metabolomic analysis using nuclear magnetic resonance (NMR) spectroscopy combined with advanced artificial intelligence techniques. The approach is based on a metacohort of more than 50,000 individuals from ten international cohorts, providing a solid foundation for the discovery of robust biomarkers through the analysis of metabolites and lipoprotein profiles quantified by NMR. From these data, statistical and AI-based models will be developed to discriminate between healthy and at-risk populations, identify altered metabolic pathways, and establish early CVD markers with greater accuracy than current clinical tools. The project includes comprehensive validation of the models using the BioSilver cohort, composed of older individuals with varying degrees of frailty, enabling the recording of cardiovascular events within a reasonable timeframe and ensuring the clinical applicability of the model. In addition, the project will carry out the technological translation of the predictive models to benchtop NMR spectrometers, facilitating their future implementation in clinical and primary care settings. The final outcome will be a reproducible, accessible, and clinically interpretable metric for early cardiovascular risk assessment, contributing to improved prevention, patient stratification, and personalized intervention planning.