Abstract 4144064: AI-enabled bone mineral density (AutoBMD AI) measurement in coronary artery calcium (CAC) scans associated with high CAC score independently of conventional risk factors: Multi-Ethnic Study of Atherosclerosis (MESA)

Background: The association between low bone mineral density (BMD) and coronary artery calcium (CAC) has been reported before using manual measurement of thoracic BMD. Such manual measurements are time-consuming and subject to operator errors. We therefore used an AI-enabled BMD measurement tool to explore the relationship between CAC and BMD independent of traditional risk factors. Methods: The validation of AutoBMD AI (HeartLung.AI, Houston TX) has been reported previously. We applied AutoBMD to CAC scans of 6043 individuals (53.3% female, age 61.9±10.2 years) from the baseline examination (2000-2002) of MESA. Thoracic BMD was measured in T7-T9 vertebrae. Osteoporosis was defined as a T-Score below -2.5. Cox proportional hazards regression was used to calculate hazard ratios (HR) per unit change for 10-year coronary heart disease (CHD) prediction. Results: Average BMD in men and women were 164.4 ± 45.1 and 163.1 ± 50.0 g/cm 3 respectively (P=0.3467). Average CAC score for men and women were 223.4 ± 544.0 and 76.2 ± 241.3 cm 3 respectively (P<0.0001). A total of 1672 cases were classified as osteoporotic, which included 986 (58.0%) men and 714 (42.0%) women. After adjusting for age, gender and conventional CVD risk factors, average CAC score in osteoporotic cases versus normal BMD was 170.1 ± 531.1 vs. 130.4 ± 272.8 respectively (P = 0.004). Decreased BMD was associated with CHD independently of CAC score and CVD risk factors (HR=0.997, p=0.0143). Conclusion: Low BMD and high CAC are associated independently of age, gender, and traditional risk factors of cardiovascular diseases. Further studies are warranted to evaluate the potential added value of BMD to CAC for prediction of adverse events.