Background: This study aimed to investigate the association between epicardial adipose tissue (EAT) and coronary artery disease (CAD) as well as cardiovascular risk factors.

Methods: Complete medical records of subjects were reviewed and cardiovascular risk factors were recorded. Epicardial fat volume (EFV) and epicardial fat thickness (EFT) were measured using digital volumetry of acquired images using a 256-slice CT-scanner. Calcium score was measured using Agatston method in non-contrast images. After contrast administration, bolus-tract images were obtained. Coronary arteries were assessed using reconstructed images in arterial phase of contrast-enhanced images. EFV and EFT measurements were compared to computed tomography angiography (CTA) findings of coronary arteries.

Results: A total of 269 patients (Mean age: 55.5 ± 12.1, 44% female) were included. Higher means of EFT and EFV were associated with coronary artery stenosis. However, the correlation coefficients of the arterial stenosis with EFT and EFV were weak. EFV and EFT had a significant association with age (P < 0.001, P < 0.001 respectively), body mass index (BMI) (P < 0.001, P < 0.001 respectively) and hypertension (P < 0.016, P < 0.003 respectively). Diabetes mellitus (DM) and hyperlipidemia were not significantly associated with EFV (P = 0.069 and 0.639 respectively) and EFT (P = 0.103 and 0.366 respectively). EFV and EFT showed a weak correlation coefficient with calcium scoring (Spearman correlation coefficients: 0.26 and 0.22 respectively, both P < 0.001). In multivariate logistic regression models considering coronary stenosis as dependent variable and EFV, EFT and other CAD risk factors as independent variables, EFV and EFT did not show significant P values and were omitted from the model by other CAD risk factors.

Conclusion: Increased EFV and EFT are associated with CAD, age, BMI and hypertension. However, no remarkable association was found between them and calcium score, hyperlipidemia or DM. These variables could weakly predict CAD in univariate models but they are not independent predictive factors for CAD in multivariate models consisting of other CAD risk factors. Hence, EFT and EFV are not independent predictors for CADs when they are considered simultaneously with other CAD risk factors