Sex-specific association of epicardial adipose tissue with coronary artery disease in an Indian cohort: a cross-sectional study

Figures

• 

  Fig. 1. Representative coronary computed tomography angiography (CCAT) images of epicardial adipose tissue (EAT) of a control and a patient with coronary artery disease (CAD). (A–B) Thoracic CCAT images of a control (A) and a CAD patient (B). EAT, pericardium, and paracardial adipose tissue are as indicated in the images. (C–D) Isolated EAT fraction of the same control (C) and the same CAD patient (D).

• 

  Fig. 2. Association between epicardial adipose tissue (EAT) volume and coronary artery disease (CAD) or its risk factors. The generalized additive model (GAM) identified a nonlinear association of EAT volume with CAD in overall subjects (A) and in men (B), but not in women (C). GAM determined a linear correlation between EAT volume and BMI in men (D). Among the covariates (age, gender, body mass index, smoking, triglycerides, hypertension, diabetes mellitus, family history CAD, total cholesterol, low-density lipoprotein cholesterol, and high-density lipoprotein cholesterol), except the tested outcome, the rest covariates were adjusted correspondingly. The red solid lines show the estimated risk, and the purple dotted lines represent pointwise 95 % CI.

• 

  Fig. 3. The association EAT volume and CAD risk by multivariable logistic regression analysis. In the partially-adjusted model, age was adjusted. In fully-adjusted models, age, smoking, hypertension, diabetes mellitus, family history of CAD, TG, TC, LDL-C and HDL-C were adjusted. Horizontal error bars denote 95 % confidence intervals from the data using EAT as the continuous variable (Table 4).

• 

  Fig. 4. The association of EAT volume and BMI by multivariable logistic regression analysis. In the partially-adjusted model, age was adjusted. In fully-adjusted models, age, smoking, hypertension, diabetes mellitus, family history of CAD, TG, TC, LDL-C and HDL-C were adjusted. Horizontal error bars denote 95 % confidence intervals from the data using EAT as the continuous variable (Supplementary Table 4).

• 

  Fig. 5. The association of EAT volume and CAD risk according to subgroups in men. The effect of EAT volume on the risk of CAD was similar in various subgroups. All analyses were adjusted for age, smoking, BMI, hypertension, diabetes mellitus, family history of CAD, TG, TC, LDL-C and HDL-C. Risk estimates were for every 1-SD increase (31.8 ml) in EAT volume. The size of squares is proportional to CAD prevalence in the specified subgroup. Horizontal error bars indicate 95 % confidence intervals. N = subjects in each subgroup. OR, odds ratio.