PERSONALIZED CARDIOVASCULAR RISK STRATIFICATION IN YOUNG ADULTS IN ROMANIA USING ARTIFICIAL INTELLIGENCE

Objective: While cardiovascular disease (CVD) has traditionally been linked to older populations, its prevalence among young adults is on the rise. Conventional tools for cardiovascular risk (CVR) assessment often underestimate risk levels in this age group. This study evaluates the role of artificial intelligence (AI) in improving CVR stratification for young adults. Design and method: A multicenter observational study was conducted involving 311 young adults aged 18 to 35 (47.57% male). In addition to traditional CVR markers—such as hypertension, diabetes, dyslipidemia, smoking, and obesity, our analysis incorporated novel biomarkers, including inflammatory indicators (C-reactive protein, interleukins 1 and 6, tumor necrosis factor-alpha), LDL subfractions, triglycerides, and tissue parameters (e.g., pulse wave velocity). AI-based models were applied to redefine CVR categories. Results: AI analysis generated four distinct risk groups: low (2.89%, n=9), moderate (4.82%, n=15), high (32.47%, n=101), and very high (59.8%, n=186). Comparisons with the SCORE2 model (designed for individuals over 45 and based on traditional risk markers) revealed substantial discrepancies. AI demonstrated superior precision, reclassifying 85.53% of individuals previously identified as low-to-moderate risk by SCORE2 into higher risk categories. Conclusions: By integrating novel biomarkers alongside conventional factors, AI offers a more precise tool for CVR assessment in young adults. This improved accuracy enables timely interventions and enhances strategies for primary cardiovascular disease prevention. Further research on larger cohorts is necessary to validate these findings.