Table 2.

EAT and Diastolic Function

First Author	Diastolic Function Reference	Subgroup Characteristics		Diastolic Parameter Correlations			Multivariable Regression Comments
		DD	Normal Function	E/A	e′	E/e′
Cavalcante9	ASE29	Grade 1 (n=29, 26%) Grade 2 (n=11, 10%)	n=70, 64%		Averaged 0.44a	0.34a	Multivariate model outcomes of grade 1 or higher DD, mean e′, and mean E/e′: EAT was an independent predictor (model included 10‐y Framingham Risk Score, metabolic syndrome, subclinical CAD, and LV mass index), β range, −0.02 to 0.04 (all P<0.05). Indexed EAT was found to increase clinical model for prediction of DD (adjusted R 2=0.16 vs 0.24; P=0.004) and mean e′ (adjusted R 2=0.17 vs 0.27; P=0.001) (ie, indexed EAT represents 8%–10% of the variation of predictors for DD
Ede13	Lang et al32	Grade 1 (n=39, 37%) Grade 2 (n=10, 9%) Grade 3 (n=2, 2%)	n=55, 52%	−0.404
Faustino14, b	Not specified	46 Patients with DD and EAT >44.1 mL	32 Patients with no DD and EAT <44.1 mL				EAT not significant on multivariable regression (results and covariates not reported). Relationship of EAT with DD by ROC AUC of 0.66 (P=0.02)
Fernando15, b	Not specified	EAT=164±118 mL (E/E′ >15)	EAT=114±54 mL (E/E′ <15)		−0.48a	0.22	On multivariable regression adjusted for age, BMI, LA volume, hypertension, and CAD, EAT associated with abnormal myocardial relaxation (OR, not specified; P=0.04)
Fontes‐Carvalho16	ASE29	EAT=116.7±67.9 cm3 Grade 1 (n=57, 28%) Grade 2 (n=58, 28%) Grade 3 (n=10, 5%)	EAT=93.0±52.3 cm3 n=80 (39%)		e′ Septal, −0.26a e′ lateral, −0.28a	0.25a	On multivariable regression adjusted for hypertension, age, sex, and other markers of adiposity (SAT, VAT, waist/height ratio, and fat mass %), EAT remained significantly predictive of E/e′ (β, 0.19 [0.06–0.32]; P<0.01), as did e′ septal and e′ lateral
Hachiya18	ASE29			−0.05	−0.31a	0.24a	Definition of diastolic dysfunction not specified. On different multivariate models, e′ inversely correlated with EAT (standardized β range, −0.30 to −0.36; all P<0.05) but not E/e′ (standardized β, 0.23; P=0.06), except when adjusted for age, sex, and BMI (model 1) and medication use (model 2) (standardized β range, 0.25–0.31; all P<0.05)
Konishi20	Defined as E/e′ >10	EAT=184±61 cm3 n=141 (62%)	EAT=154±58 cm3 n=88 (38%)			0.21a	On multivariable regression with age, hypertension, male sex, diabetes mellitus, and abdominal obesity, there was an independent effect of EAT on DD: OR, 2.09 (1.15–3.79; P=0.02) for EAT per 100 cm3
Lai21	Lang et al32	EAT=86.79±31.77 n=100	EAT=67.32±31.95 n=218		−0.38a	0.284a	On multivariable regression adjusted for age, sex, BMI, systolic blood pressure, LV mass index, hypertension, diabetes mellitus, hyperlipidemia, and smoking, EAT was significantly associated with E/A (β, −0.002)a and diastolic dyssynchrony (β, 0.197)a
	Gottdiener et al31			Men, −0.12)a women, −0.12a			On multivariable linear regression adjusted for age, height, smoking, alcohol, blood pressure, eGFR, hemoglobin, total physical activity score, medications, VAT, and weight, E/A no longer became significant (regression co‐efficient, −0.01±0.02 [P=0.41] in women and −0.0±0.02 [P=0.64] in men) (described as pericardial fat volume)
Longenecker23, b	Not specified	Grade 1 (n=29 [HIV+, n=19; HIV−, n=10]) Grade 2 (n=2 [HIV+, n=1; HIV−, n=2])	n=38 (HIV+) n=26 and n=12 (HIV−)	−0.392a			On multivariable regression adjusted for age, BMI, and sex, EAT remained independently associated with diastolic dysfunction (OR, 1.35; 95% CI, 1.02–1.79) per 10‐mL increase (described as pericardial fat volume)
Ng24	Not specified				e′ Septal, −0.263)a; e′ lateral, −0.285a
Vural26	Alnabhan et al30	EAT=164.4±54 cm3 Grade 1 (n=24, 38%) Grade 2 (n=4, 6%) Grade 3 (n=1, 1.5%)	EAT=114.1±46.6 cm3 n=34 (56%)		−0.437a		On multivariable regression adjusted for age, blood pressure, BMI, waist circumference, and cholesterol, EAT was an independent predictor of DD (OR, 1.03 [1.01–1.06]; P=0.006). ROC‐derived optimal cutoff for DD, 129.6 cm3 (ROC curve, 0.758)

Correlations represent the correlation co‐efficient.

Values are mean±SD or mean (range). ASE indicates American Society of Echocardiography; AUC, area under the curve; BMI, body mass index; CAD, coronary artery disease; CI, confidence interval; DD, diastolic dysfunction; e′, average mitral annular tissue Doppler velocity; E/e′, early inflow / annular velocity ratio; E/A, ratio of peak early (E) and late (A) transmitral inflow velocities; EAT, epicardial adipose tissue; eGFR, estimated glomerular filtration rate; LA, left atrial; LV, left ventricular; OR, odds ratio; ROC, receiver operating characteristic; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue.

^a P value for univariate correlation is significant at <0.05.

^bStudy is a conference abstract.