Exploring the Coronary Calcium Scores in Indiana Firefighters from Risk Factors to AI Predictions

Abstract

Firefighters, facing toxic exposure and mandatory personal protective equipment use, are at increased risk for cardiovascular strain. Coronary artery disease (CAD) can lead to sudden heart attacks, making early detection and risk assessment critical. Coronary calcium score (CCS), obtained via computed tomography, serve as precise indicators of pre-clinical CAD, and are linked to increased cardiovascular events and mortality. However, specific risk factors affecting CCS in firefighters remain underexplored. In my study, I utilized existing health data from Indiana firefighters aged 35-68, gathered during their health screenings at Ascension Public Safety Medical. Focusing on those with complete evaluations—including physicals, lab tests, fitness assessments, and CT scans for CCS — I first examined the clinical risk factors influencing CCS. Then, I explored the association between maximal oxygen uptake MaxVO2, an essential measure of aerobic capacity and cardiovascular fitness, and CCS among different age groups (< 45, and >= 45 years). Subsequently, I developed machine learning models using these risk factors to predict CCS. I observed significant positive associations between age, monocyte percentage, and CCS. Also, I identified significant positive associations between alkaline phosphatase and CCS. Higher MaxVO2 levels were associated with lower CCS, especially in firefighters over 45. Finally, utilizing these findings, I developed machine learning models to predict CCS and selected the most precise one. In conclusion, this research provides a novel perspective on the cardiovascular risks faced by firefighters. The associations and predictive models I've established not only contribute to the understanding of CCS in this population but pave the way for targeted interventions. These findings emphasize the importance of age, monocyte percentage, alkaline phosphatase, and regular cardiovascular fitness assessments and may influence future guidelines for firefighter health monitoring, ultimately aiming to reduce CAD incidence and enhance occupational safety.