Pulation, so we decided that the cut-off point for obesity was 27 kg/m2 instead of 30 kg/ m2.BFigure 3 Distribution of the thickness of epicardial adipose tissue according to body mass index and metabolic score. (A) Non-high BMI group. (B) High BMI group.Figure 5 Distribution of the thickness of epicardial adipose tissue according to body mass index and coronary artery disease.Park et al. Cardiovascular Diabetology 2010, 9:29 http://www.cardiab.com/content/9/1/Page 6 ofFigure 6 ROC curve analysis to demonstrate the discriminatory power of the thickness of epicardial adipose tissue in the diagnosis of coronary artery disease according to body mass index. AUC, the area under the curveSeveral biomolecular studies in humans have shown that EAT is metabolically active and an important source of both pro-inflammatory adipokines, such as tumor necrosis factor-, interleukin 1, interleukin 6 and nerve growth factor, and anti-inflammatory adipokines, such as adiponectin and adrenomedullin [13-15]. The regulationand metabolism of pro- and anti-inflammatory mediators secreted by EAT have been linked to insulin sensitivity [23]. EAT has been shown to express a pathogenic mRNA profile of pro-inflammatory adipokines in patients with CAD [13]. One study showed that expression of adiponectin, an anti-inflammatory adipokine, in EAT was significantly lower in patients with CAD compared to those without CAD [14]. EAT thickness has also been shown to be related to markers of insulin resistance and inflammation [2]. For example, one study demonstrated that obesity leads to adipocyte hypertrophy, which increases the secretion of pro-inflammatory adipokines and decreases the secretion of anti-inflammatory adipokines by EAT [23]. PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/27725455 The change of EAT thickness by obesity might have obscured the difference between patients with and without CAD, or with and without MS, in the high BMI group of our study. The second possible explanation for the weak correlation between the PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/1724526 EAT thickness and presence of MS and CAD in the high BMI group, might be the different proportion of the EAT to total amount of VAT according to BMI. Patients with high BMI generally have more VAT, which includes EAT. In the high BMI group, the EAT might make up a smaller proportion of the total VATTable 2: Multiple logistic analysis of coronary artery disease risk factors(A) Non-high BMI group (n = 478) Risk factors Age Male 45 years Female 55 years Smoking Hypertension Diabetes mellitus EAT thickness (3 mm) (B) High BMI group (n = 165) Risk factors Age Male 45 BzATP triethylammonium salt years Female 55 years Smoking Hypertension Diabetes mellitus EAT thickness (3 mm) 1.775 (0.763-4.129) 1.276 (0.6-2.712) 2.687 (1.222-5.911) 2.341 (1.136-4.827) 0.183 0.527 0.014 0.021 Odds ratio (95 CI) 6.531 (2.089-20.420) p < 0.001 1.576 (1.014-2.448) 2.115 (1.385-3.230) 2.139 (1.282-3.569) 3.504 (2.297-5.346) 0.043 < 0.001 0.004 < 0.001 Odds ratio (95 CI) 2.847 (1.603-5.057) p < 0.EAT, epicardial adipose tissue; CI, confidence intervalPark et al. Cardiovascular Diabetology 2010, 9:29 http://www.cardiab.com/content/9/1/Page 7 ofcompared to the non-high BMI group. In the high BMI group, the EAT thickness by echocardiography might not be representative of the total VAT owing to its smaller proportion of the total VAT. That reason might attenuate the predictive value of EAT thickness by echocardiography for MS and CAD in the high BMI group. Further study is necessary to clarify the different roles of EAT according to BMI. Considering r.