Epicardial adipose tissue density is a better predictor of cardiometabolic risk in HFpEF patients: a prospective cohort study

BACKGROUND: Epicardial adipose tissue (EAT) accumulation is associated with multiple cardiometabolic risk factors and prognosis of heart failure with preserved ejection fraction (HFpEF). The correlation between EAT density and cardiometabolic risk and the effect of EAT density on clinical outcome in HFpEF remain unclear. We evaluated the relationship between EAT density and cardiometabolic risk factors, also the prognostic value of EAT density in patients with HFpEF. METHODS: We included 154 HFpEF patients who underwent noncontrast cardiac computed tomography (CT) and all patients received follow-up. EAT density and volume were quantified semi-automatically. The associations of EAT density and volume with cardiometabolic risk factors, metabolic syndrome and the prognostic impact of EAT density were analyzed. RESULTS: Lower EAT density was associated with adverse changes in cardiometabolic risk factors. Each 1 HU increase in fat density, BMI was 0.14 kg/m(2) lower (95% CI 0.08–0.21), waist circumference was 0.34 cm lower (95% CI 0.12–0.55), non-HDL-cholesterol was 0.02 mmol/L lower (95% CI 0–0.04), triglyceride was 0.03 mmol/L lower (95% CI 0.01–0.04), fasting plasma glucose was 0.05 mmol/L lower (95% CI 0.02–0.08), TyG index was 0.03 lower (95% CI 0.02–0.04), Log(2)(TG/HDL-C) was 0.03 lower (95% CI 0.02–0.05), METS-IR was 0.36 lower (95% CI 0.23–0.49), MetS Z-score was 0.04 lower (95% CI 0.02–0.06), and Log(2)(CACS + 1) was 0.09 lower (95% CI 0.02–0.15). After adjusting for BMI and EAT volume, the associations of non-HDL-cholesterol, triglyceride, fasting plasma glucose, insulin resistance indexes, MetS Z-score, and CACS with fat density remained significant. The area under the curve (AUC) for the presence and severity of metabolic syndrome was greater in EAT density than volume (AUC: 0.731 vs 0.694, 0.735 vs 0.662, respectively). Over a median follow-up of 16 months, the cumulative incidence of heart failure readmission and composite endpoint increased with lower level of EAT density (both p < 0.05). CONCLUSIONS: EAT density was an independent impact factor of cardiometabolic risk in HFpEF. EAT density might have better predictive value than EAT volume for metabolic syndrome and it might have prognostic value in patients with HFpEF.