Unraveling the role of resistin, retinol-binding protein 4 and adiponectin produced by epicardial adipose tissue in cardiac structure and function: evidence of a paracrine effect

**Background:** Adipokines produced by adipose tissue are implicated in metabolic and cardiovascular diseases. Epicardial adipose tissue lies in direct proximity to the myocardium and coronary arteries, suggesting a potential paracrine effect. While serum levels of resistin, retinol-binding protein 4 (RBP4), and adiponectin have been linked to coronary artery disease (CAD) and heart failure, the role of these adipokines specifically produced by epicardial adipose tissue—and whether regional expression corresponds to pathology in the adjacent myocardium or coronary artery—remains unclear.

**Methods:** Forty-one non-diabetic males (median age 66 years, range 38–83; mean BMI 28.4±3.5 kg/m²) scheduled for cardiothoracic surgery were prospectively recruited. Exclusion criteria included diabetes, severe kidney/liver disease, malignancy, systemic inflammation, and use of antidiabetic/antiobesity drugs or glucocorticoids within 3 weeks. Preoperatively, all participants underwent anthropometric measurements, echocardiography, coronary angiography, and blood sampling. During surgery, three adipose tissue biopsies were collected: one from thoracic subcutaneous adipose tissue, and two from epicardial adipose tissue—one adjacent to the left anterior descending artery (LAD) and one adjacent to the right coronary artery (RCA). mRNA expression of resistin, RBP4, and adiponectin was measured by quantitative real-time PCR normalized to β-actin. Serum levels were measured by ELISA. CAD was defined as ≥50% stenosis on angiography. Statistical analysis used Spearman correlations and backward stepwise regression.

**Key Results:** Thirty participants had CAD (1 one-vessel, 14 two-vessel, 15 three-vessel disease); 29 had CAD in the LAD and 21 in the RCA. LV ejection fraction (LVEF) correlated negatively with adiponectin-LAD (rho=−0.390, p=0.025). The ratio of early to late diastolic transmitral flow velocity (MVE/A), an index of LV diastolic function, correlated negatively with resistin-LAD (rho=−0.529, p=0.024) and RBP4-LAD (rho=−0.458, p=0.049). RBP4-LAD also correlated positively with an estimate of LV stiffness (MVE/Ea(s-l)/LVEDV: rho=0.762, p=0.028). There was no difference in epicardial adipose tissue mRNA expression of resistin, RBP4, or adiponectin between individuals with and without CAD, nor between those with and without CAD in the corresponding coronary artery territory. Serum adiponectin was significantly lower in CAD patients (2.50 mg/L vs. 6.55 mg/L, p<0.001). In backward stepwise regression, logadiponectin-LAD was the only independent predictor of LVEF (standardized β=−0.382, p=0.050). For MVE/A, logresistin-LAD (standardized β=−0.455, p=0.031) and logRBP4-LAD (standardized β=−0.500, p=0.014) were the only independent determinants in separate models.

**Clinical Implications:** This study provides evidence that adipokines produced locally by epicardial adipose tissue may influence adjacent myocardial function through a paracrine mechanism. The finding that adiponectin-LAD was inversely associated with LVEF, while resistin-LAD and RBP4-LAD were inversely associated with diastolic function, suggests region-specific effects. Notably, the lack of association between regional epicardial adipokine expression and CAD in the corresponding artery challenges the hypothesis that these adipokines directly promote coronary atherosclerosis via paracrine signaling. Instead, previously reported associations between epicardial adipokine expression and CAD may be confounded by underlying LV dysfunction. These findings highlight epicardial adipose tissue-derived resistin, RBP4, and adiponectin as potential therapeutic targets for heart failure, though further studies are needed to establish causality. Limitations include the cross-sectional design, relatively small sample size, recent myocardial infarction in some participants, wide ranges of mRNA expression, and statin use in the majority of CAD patients.