Deciphering the role of epicardial adipose tissue as a predictor of vascular and cardiac outcomes

Abstract

Background and Aims: Increase in epicardial adipose tissue volume (EATV) is associated with cardiovascular disease risk. We sought to investigate the relationship between EATV and specific-cardiovascular outcomes using flexible data modelling approaches. Method(s): In 80,503 patients undergoing routine coronary CT angiography, enrolled in the Oxford Risk Factors and Non-invasive (ORFAN) study, EATV was quantified with an automated deep-learning model. Clinical and mortality data were linked through NHS England and ONS. Fractional polynomial Cox regressions models with competing risks were fitted to estimate hazard ratios (HR) and 95% confidence intervals (CI) for cardiac death, heart failure and fatal/non-fatal myocardial infarction, adjusted for age, gender and ethnicity. Result(s): Heart failure [4,024 (5.0%)], cardiac death [2,846 (3.5%)] and fatal/non-fatal myocardial infarction [4,464 (5.5%)] occurred over a median [IQR] of 5.6[4.1-8.1] years. The relationship between EATV and heart failure or cardiac mortality was U-shaped. Optimum EATV in cm3 associated with minimum risk was 65.47 (99%CI, 55.98-74.96) for heart failure [HR, 0.94, 95%CI (0.91-0.97)] and 85.46 (99%CI,73.55-97.37) for cardiac mortality [HR 0.99, 95%CI (0.98-1.00)]. Above the optimal EATV ranges, risk of heart failure: HR 1.10, 95%CI (1.07-1.12) and cardiac mortality: HR 1.13, 95%CI (1.09-1.16) were increased. Below the optimal EATV range, increase in risk was observed in cardiac mortality: HR 1.13, 95%CI (1.05-1.21) but not in heart failure: HR 1.05, 95%CI (0.98-1.13). The relationship between EATV and fatal/non-fatal myocardial infarction was linear (HR per standard deviation 1.05, 95%CI (1.02-1.08). (Figure) [Formula presented] Conclusion(s): There is a U-shape association between EATV and cardiac outcomes (i.e. heart failure and cardiac mortality), while the association takes a linear shape for the prediction of fatal and non-fatal myocardial infarction. These findings suggest that the relationship between epicardial obesity and cardiovascular outcomes should be interpreted based on different models for different outcomes. Copyright © 2025