Relationship between the Echocardiographic Epicardial Adipose Tissue Thickness and Serum Adiponectin in Patients with Angina

Recently, the interest in epicardial adipose tissue (EAT) is rapidly growing. Several clinical and experimental evidences suggest that epicardial fat may play a role in coronary artery disease (CAD) and atherosclerosis.1-7) Because of its close anatomical relationship to the heart, EAT may locally modulate the coronary arteries and myocardium through paracrine secretion of anti- and pro-inflammatory adipokines. Some studies showed that adipokines and cytokines were locally expressed in EAT and their levels were significantly lower in patients with CAD.8-10) Additionally, it is well-known that EAT is associated with blood markers such as serum lipid profile, inflammatory marker, and insulin resistance.2), 5) Accordingly, we presumed that EAT might be related also to serum adiponectin.

When EAT was observed by echocardiography, we can also see mediastinal adipose tissue (MAT) in the outside of parietal pericardium. Pericardial adipose tissue (PAT) is defined as EAT plus MAT. Several studies on both EAT and MAT revealed that cardiac fat is associated with coronary risk factors and insulin resistance.11), 12) However, few clinical studies examined EAT and MAT separately with regard to their relation with CAD. So, we measured EAT and MAT using echocardiography and examined their correlations with CAD and serum adiponectin.

The mean EAT of the patients was 2.88±1.94 mm (range 0.20-9.60 mm), the mean MAT 3.22±2.65 mm (range 0.00-12.70 mm), and the mean PAT 6.10±3.38 mm (range 0.50-18.20 mm). Table 1 shows the baseline clinical characteristics according to the presence of obstructive CAD. Compared to those without significant stenosis, patients with CAD were older (58.1±10.0 vs. 62.5±10.1 years, p=0.010), showed higher prevalence of diabetes (14.7 vs. 35.3%, p=0.004), and showed a higher LDL cholesterol level (108.6±33.9 vs. 125.4±36.8 mg/dL, p=0.004) and lower HDL cholesterol level (49.9±11.0 vs. 46.1±11.0 mg/dL, p=0.032). EAT was thicker in patients with CAD (1.8±1.4 vs. 3.8±1.9 mm, p<0.001), but MAT was not different according to the presence of CAD (2.9±2.8 vs. 3.5±2.5 mm, p=0.121). The serum adiponectin level was significantly lower in patients with CAD (8.2±7.1 vs. 5.9±4.7 µg/mL, p=0.019).

Table 1
Baseline characteristics

• Click for larger image

EAT showed a significant positive correlation with age (r=0.225. p=0.005), homocystein (r=0.289, p=0.001), fasting glucose (r=0.167, p=0.042), and fibrinogen (r=0.218, p=0.009), and a significant negative correlation with serum adiponectin (r=-0.194, p=0.016) (Table 2). Particularly in male patients, the correlation between MAT and serum adiponectin level was more significant (Fig. 2). MAT showed a correlation with BMI (r=0.185, p=0.023), but not with other biochemical markers.

Fig. 2
Correlation between serum adiponectin level and epicaridal adipose tissue according to gender. EAT: epicardial adipose tissue.

• Click for larger image

Table 2
Correlation between the cardiac adipose tissue thickness and clinical variables

• Click for larger image

EAT increased significantly with increase in the severity of CAD (Fig. 3). EAT was 1.76±1.36 mm in patients without significant stenosis (n=68), 3.39±1.64 mm in single vessel disease patients (n=41), 4.12±2.03 mm in multi-vessel disease patients (n=44), showing significant differences among the groups (p<0.001 by ANOVA). However, MAT was not proportional to the severity of CAD. The serum adiponectin level showed the tendency of decreasing with increase in the severity of CAD but the tendency was not statistically significant (p=0.065 by ANOVA).

Fig. 3
Cardiac adipose tissue and serum adiponectin level according to the severity of coronary artery disease. ^*p<0.05.

• Click for larger image

The results of multivariate logistic regression analysis for prediction of CAD were presented in Table 3. In addition to the well-known CAD risk factors, such as age, diabetes, LDL cholesterol and hsCRP, EAT thickness and serum adiponectin were independent predictors of obstructive CAD. However, MAT thickness was not associated with CAD.

Table 3
Multiple regression analysis for prediction of the presence of significant coronary artery disease

• Click for larger image

The present study demonstrated that EAT was an independent predictor of CAD and negatively correlated with serum adiponectin level. In contrast with EAT, MAT was not associated with CAD and adiponectin.

It has been reported that EAT plays a role in the pathogenesis of CAD. The studies using epicardial fat obtained during coronary artery bypass surgery revealed that significantly higher expression of interleukin-1, interleukin-6, and tumor necrosis factor-α mRNA was shown in epicardial fat than in leg subcutaneous adipose tissue.8) In addition, the expression of adiponectin, a protective cytokine, was significantly lower in the EAT of patients with CAD than in that of patients without CAD.9), 10) EAT thickness was also correlated with LDL cholesterol, HDL cholesterol, fasting glucose, hsCRP and blood pressure.2), 5), 14), 15) Accordingly, EAT is believed to be closely related with the incidence and development of CAD, and in our study as well, EAT was an independent predictor of obstructive CAD. Recently, it is reported that EAT is correlated not only with tissue expression but also with serum pro-inflammatory mediator interleukin-6, monocyte chemoattractant protein-1, visfatin, and plasminogen activator inhibitor-1.16), 17) In this study, the serum adiponectin level showed a negative correlation with EAT. These evidences suggested that EAT plays as a endocrine organ of the heart.

Several imaging modalities have been used for the assessment of epicardial fat. Since Iacobellis et al.13) introduced the measurement of EAT thickness using echocardiography for the first time in 2003, epicardial fat measuring on the right ventricular free wall has been generalized. It is common recently to measure the volume of epicardial fat using multislice computed tomography (MSCT) or magnetic resonance imaging (MRI).1), 4), 6) Fat volume also shows a high correlation with coronary calcium score or the risk of CAD. Because EAT is a three-dimensional structure, it may be more accurate to measure the fat volume, but as its high correlation with echocardiographic measurement has been proved, MSCT or MRI may not be necessary if it is only for the evaluation of cardiac adiposity.

When we examine cardiac fat, we can see that adipose tissue is divided by the parietal pericardium. The tissue between the epicardium and the parietal pericardium is called EAT, and the tissue in the outside is called MAT, and the whole of the tissues is called PAT.18) Research using CT reported that PAT shows a correlation with abdominal fat and is associated with coronary risk factors.12) In our study, MAT did not show a correlation with coronary risk factors, so the association of PAT, the sum of EAT and MAT, with such factors appeared weak. Furthermore, although PAT was related with the severity of CAD, it was not an independent factor of CAD. In fact, EAT is a different type of tissue from MAT. EAT originates from the splanchnopleuric mesoderm associated with gut. On the other hand, MAT originates from the primitive thoracic mesenchyme, which splits to form the parietal pericardium and the outer thoracic wall. EAT is supplied by branches of the coronary arteries, whereas MAT is supplied by the branches of the internal mammary arteries.18) Accordingly, it is believed that, compared to MAT, EAT is more closely associated with the incidence of CAD and the development of atherosclerosis.

Our analysis was limited by the studied population because it included only those patients pre-selected to undergo coronary angiography. A prospective cohort study might be necessary to elucidate the clinical significance of EAT and MAT in the general population. In addition, as epicardial adipose tissue has a three-dimensional distribution, two-dimensional echocardiography may not assess the total amount of epicardial adiposity completely. When we measure EAT on the free wall of the right ventricle, we may measure from the parasternal long axis view and from the short axis view and obtain the mean of the two values, but because the two measurements are highly correlated with each other, some studies including ours use only the value measured from the parasternal long axis view.5), 19)

In conclusion, compared to MAT, EAT showed higher association with the severity and risk factors of CAD, and a good negative correlation with the serum adiponectin level. Echocardiographic epicardial fat measurement might be used as an easy and reliable cardiovascular risk indicator.