Epicardial adipose tissue predicts incident cardiovascular disease and mortality in patients with type 2 diabetes

Table 3 Performance of adipose tissue in predicting the composite end-point

All patients	C-statistics		Net reclassification index
	Model 1 (95% CI)	Model 3 (95% CI)	Model without vs. with adipose tissue
	Model 1 (95% CI)	Model 3 (95% CI)	Percent (95% CI)	p value
Model without adipose tissue		0.67 (0.63–0.72)
Model including
EAT	0.53 (0.49–0.58)	0.68 (0.63–0.72)	19.6% (1.4–37.8)	0.035
High EAT	0.54 (0.50–0.58)	0.67 (0.63–0.72)	20.8% (2.5–39.1)	0.026
Men
Model without adipose tissue		0.65 (0.60–0.70)
Model including
EAT	0.54 (0.49–0.59)	0.66 (0.61–0.70)	17.7% (− 3.1 to 38.5)	0.09
High EAT	0.55 (0.50–0.59)	0.65 (0.60–0.70)	19.6% (1.1–40.0)	0.06
Women
Model without adipose tissue		0.66 (0.55–0.76)
Model including
EAT	0.50 (0.39–0.60)	0.66 (0.56–0.77)	− 4.3% (− 42.8 to 34.1)	0.83
High EAT	0.52 (0.43–0.60)	0.66 (0.55–0.77)	20.3% (20.6–61.2)	0.33

Adipose tissue variables are presented as continuous or as binary variables: high (> median) vs. low (< median) levels of adipose tissue
Model 1 is unadjusted
Model 3 is adjusted for age, sex, LDL, diabetes duration, HbA_1C, systolic blood pressure, smoking and BMI
CI confidence interval, EAT epicardial adipose tissue, HR hazard ratio, PAT pericardial adipose tissue, CAT total cardiac adipose tissue

ISSN: 1475-2840