Predictors of coronary artery calcification in middle-aged and older athletes: the MARC-2 study

Background Exercise improves cardiovascular health, but recent studies have found high-volume high-intensity exercise is associated with increased coronary artery calcification (CAC). While traditional risk factors are predictors of CAC, these are less common in athletes. Therefore, CAC in athletes may be driven by a different pathophysiology, related to factors such as exercise characteristics, diet, calcium homeostasis and inflammation. Since increased CAC scores are also prevalent in athletes without traditional risk factors, the standard risk prediction algorithms may not be appropriate for athletes. Purpose We aimed to identify predictors of CAC, by assessing the association of traditional cardiovascular and non-traditional risk factors with CAC in middle-aged and older athletes. Methods Participants of the Measuring Athlete’s Risk of Cardiovascular events 2 (MARC-2) study were included. CAC was assessed by computed tomography. Questionnaires including food frequency and stress questionnaires, blood biomarkers (cardiac biomarkers, mineral concentrations, inflammation and lipids) and blood pressure were collected. We assessed the association of risk factors with CAC using linear and logistic regression. Results We included 289 male athletes (60.0 [Q1–3 56.3–66.0] years old) with a lifelong weekly training load of 26 [17–35] MET-hours, BMI of 24.5 [22.9–26.6] kg/m2, systolic blood pressure of 139±18 mmHg, and reported 0.0 [0.0–8.0] smoking pack years. Twenty-nine % of athletes had a CAC score of 0, 40% 1–100, 18% 101–400, and 13% >400. Among traditional cardiovascular risk factors, higher age, systolic blood pressure, smoking pack years, and family history of coronary artery disease independently predicted greater CAC scores. Family history was the most important traditional risk factor. Among non-traditional risk factors, higher training loads, serum calcium and phosphate, and lower adjusted energy intake and fat percentage of energy intake independently predicted greater CAC scores. Lower proportion of fat in diet was the strongest predictor of greater CAC scores. The full model had higher accuracy in predicting CAC>100 (ROC-AUC 0.75, 95%CI [0.69–0.81]) and CAC>400 (0.83 [0.76–0.91]) than both traditional cardiovascular risk factors alone (0.71 [0.64–0.78], p=0.014, and 0.78 [0.69–0.87], p=0.018, respectively) and the cardiovascular risk algorithm SCORE2 (0.65 [0.58–0.72], p<0.001, and 0.72 [0.63–0.80], p=0.002, respectively). Conclusions Traditional cardiovascular risk factors and risk algorithms have low accuracy in predicting CAC in aging athletes. Non-traditional risk factors training load, energy intake, proportion of fat intake, serum calcium and serum phosphate were associated with CAC scores. The addition of these variables increased the accuracy of the prediction model, and the full model outperformed both traditional risk factors and SCORE2 in the prediction of CAC in athletes.