Epicardial Fat, Cardiac Geometry and Cardiac Function in Patients with Nonalcoholic Fatty Liver Disease: Association with the Severity of Liver Disease

Nonalcoholic fatty liver disease (NAFLD) has been associated with increased cardiovascular risk, including coronary artery disease and cardiac dysfunction. In addition, recent evidence highlighted the possible role of epicardial fat as a new cardiometabolic risk factor. We tested the correlation between epicardial fat, alterations in cardiac geometry and function and the severity of liver damage in patients with biopsy-proven NAFLD. Methods The anthropometric, biochemical and metabolic features were recorded in 147 consecutive biopsy-proven NAFLD cases (Kleiner score). Epicardial fat thickness was measured by echocardiography. Results Epicardial fat was higher in patients with severe vs. milder fibrosis (8.5±3.0 vs. 7.2±2.3 mm; p=0.006); this association was maintained at multivariate logistic regression analysis (OR 1.22, 95%C.I. 1.01-1.47; p=0.04) after correction for gender, age>50 years, visceral obesity, IFG/diabetes, non-alcoholic steatohepatitis and severe steatosis. Of note 37.1% of patients with epicardial fat >7mm (median value) had severe liver fibrosis, compared to 18.3% of cases with lower epicardial fat (p=0.01). As for echocardiographic indices, after adjusting for cardiometabolic confounders, diastolic posterior-wall thickness (p=0.01), left ventricular mass (p=0.03), relative wall thickness (p=0.02), and left atrial volume (0.04), as well as ejection fraction (p=0.004), lower lateral TDI e’ (p=0.009), E/A ratio (0.04) (cardiac geometry alterations and diastolic dysfunction) were linked to severe liver fibrosis. Conclusions In patients with NAFLD, a higher epicardial fat thickness is associated with the severity of liver fibrosis, in keeping with a possible pathogenic role of ectopic fat depots in whole body organ damage. In addition, morphological and functional cardiac alterations are more pronounced according to the severity of fibrosis. Further studies are needed to validate our results.