Epicardial fat is a metabolically active fat depot that is strongly associated with obesity, metabolic syndrome, and coronary artery disease (CAD). The relation of epicardial fat to diastolic function is unknown. We sought to (1) understand the relation of epicardial fat volume (EFV) to diastolic function and (2) understand the role of EFV in relation to potential risk factors (hypertension, subclinical CAD, and metabolic syndrome) of diastolic dysfunction in apparently healthy subjects with preserved systolic function and no history of CAD. We studied 110 consecutive subjects (65% men, 55 ± 13 years old, mean body mass index 28 ± 5 kg/m 2 ) who underwent cardiac computed tomography and transthoracic echocardiography within 6 months as part of a self-referred health screening program. Exclusion criteria included history of CAD, significant valvular disease, systolic dysfunction (left ventricular ejection fraction <50%). Diastolic function was defined according to American Society of Echocardiography guidelines. EFV was measured using validated cardiac computed tomographic software by 2 independent cardiologists blinded to clinical and echocardiographic data. Hypertension and metabolic syndrome were present in 60% and 45%, respectively. Subclinical CAD was identified in 20% of the cohort. Diastolic dysfunction was present in 45 patients. EFV was an independent predictor of diastolic dysfunction, mean peak early diastolic mitral annular velocity, and ratio of early diastolic filling to peak early diastolic mitral annular velocity (p = 0.01, <0.0001, and 0.001, respectively) with incremental contribution to other clinical factors. In conclusion, EFV is an independent predictor of impaired diastolic function in apparently healthy overweight patients even after accounting for associated co-morbidities such as metabolic syndrome, hypertension, and subclinical CAD.
The association of obesity with myocardial dysfunction is probably mediated through its strong links with hypertension, dyslipidemia, and atherosclerotic coronary artery disease (CAD). However, obesity may be associated with structural changes in the myocardium independent of its effects on risk factors or CAD. Subclinical changes of left ventricular (LV) structure and function including abnormal relaxation and strain have been reported in overweight subjects even after adjustment for mean arterial pressure, age, gender, and LV mass. Obese patients may present with heart failure with normal ejection fraction even in the absence of CAD; each 1-kg/m 2 increase in body mass index (BMI) has been shown to increase the risk of heart failure by 5% for men and by 7% for women. In this context, it is extremely important to understand the mechanisms by which obesity may cause LV dysfunction with a preserved ejection fraction. Epidemiologic studies have shown that epicardial fat, a metabolically active fat depot, is strongly associated with obesity, metabolic syndrome, and diabetes. The proximity of epicardial fat to the coronary arteries has been used to explain the association of epicardial fat volume (EFV) with increased coronary artery calcium, atherosclerotic plaque, and myocardial ischemia. Although these factors may be responsible for diastolic dysfunction, epicardial fat may also have a direct paracrine effect on the myocardium and thus alter the structural properties of the left ventricle. We sought to (1) understand the relation of EFV to diastolic dysfunction and (2) understand the role of EFV in relation to potential risk factors (hypertension, subclinical CAD, and metabolic syndrome) of diastolic dysfunction in healthy subjects with preserved systolic function.
Methods
We studied consecutive patients who underwent cardiac computed tomography and transthoracic echocardiography at the Cleveland Clinic (Cleveland, Ohio) within 6 months as part of a self-referred health screening program. Exclusion criteria were (1) history of CAD (myocardial infarction and/or percutaneous or surgical revascularization), (2) moderate (≥2+) valvular regurgitation or any valvular stenosis, (3) systolic dysfunction (ejection fraction <50%), (4) incomplete echocardiographic data, and (5) an intercurrent event between the 2 imaging studies. Subjects with pathologic Q waves on electrocardiogram, wall motion abnormalities on echocardiogram, or myocardial wall thinning seen on echocardiogram and/or cardiac computed tomogram suggestive of previous myocardial infarction were also excluded. This study was approved by the institutional review board.
Hypertension was defined as systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥85 mm Hg at the time of the visit (mean of 2 readings) or history of hypertension but not on treatment. Diabetes mellitus was defined as fasting blood glucose ≥110 mg/dl or use of diabetes medications. Dyslipidemia was defined as total serum cholesterol ≥240 mg/dl or use of lipid-lowering treatment. BMI was calculated as weight (kilograms) divided by height (meters) squared. According to a standard definition, overweight patients had a BMI from 25.0 to 29.9 kg/m 2 , whereas obese patients had BMI ≥30 kg/m 2 . Metabolic syndrome was defined by the criteria proposed by the National Cholesterol Education Program Adult Treatment Panel III. At least 3 of the following components were needed to meet metabolic syndrome criteria: (1) fasting blood glucose ≥100 mg/dl or patient’s self-reported history of diabetes or use of diabetes medications; (2) blood pressure ≥130/85 mm Hg or patient’s self-reported history of hypertension or use of antihypertensive medications; (3) triglycerides ≥150 mg/dl; (4) high-density lipoprotein <40 mg/dl; and (5) BMI >30 kg/m 2 (where central obesity was assumed according to International Diabetes Federation guidelines ). Ten-year Framingham Risk Score was calculated according to guidelines and included the following risk factors: age, gender, systolic blood pressure, total cholesterol, high-density lipoprotein cholesterol and smoking history.
LV linear dimensions were measured from a parasternal long-axis view according to recommendations of the American Society of Echocardiography. LV mass index was calculated by the corrected American Society of Echocardiography simplified cubed equation and indexed for body surface area. LV ejection fraction and LV volume index were calculated by the biplane modified Simpson rule. LV diastolic function assessment was obtained according to American Society of Echocardiography guidelines and included peak velocities of the early phase (E) and late phase (A) of the transmitral inflow, with derivation of the E/A ratio. LV myocardial velocities were evaluated by tissue Doppler imaging with pulsed sample volume placed at the level of the lateral and septal mitral valve annuli. Peak early diastolic mitral annular velocities (e′) were measured and averaged over 3 cardiac cycles. Diastolic dysfunction was categorized as (1) E/A <0.8 and deceleration time >200 ms (stage I, impaired relaxation); or (2) E/A ≥0.8 and <1.5, deceleration time of 160 to 200 ms, and mean e′ <8 cm/s (stage II, pseudonormal); or (3) E/A ≥1.5, deceleration time <160 ms, and mean e′ <8 cm/s (stage III, restrictive). All echocardiograms were reviewed by 2 board-certified cardiologists blinded to clinical and cardiac computed tomographic information.
Noncontrast computed tomogram was acquired using a 64-slice computed tomographic scanner (Sensation 64, Siemens Medical Solutions, Erlangen, Germany) in the axial mode during a single breath-hold with prospective electrocardiographic triggering and 120-kVp tube voltage. Images were reconstructed using a medium sharp kernel (B35f) with a slice thickness of 3 mm. Coronary calcium was quantified on nonenhanced cardiac computed tomographic scans. Coronary calcium was defined as a plaque of ≥ 3 contiguous pixels (area 1.0 mm 2 ) with a density of >130 HU. A total coronary calcium score was determined by summing individual Agatston scores from each of 4 anatomic sites (left main, left anterior descending, left circumflex, and right coronary arteries). Contrast-enhanced coronary computed tomographic angiography was performed in the same scanner. Nonoverlapping images were reconstructed using a medium sharp kernel (B26f) with a thickness of 3 mm for quantification of epicardial fat. Scans were analyzed independently by 2 experienced investigators who were blinded to clinical and echocardiographic information.
Epicardial fat quantification was performed by QFAT (Cedars-Sinai Medical Center, Los Angeles, California) software as previously described. EFV was defined as adipose tissue enclosed by the visceral pericardium including fat directly surrounding the coronary arteries. Definition of epicardial contours was based in an upper slice limit (bifurcation of the pulmonary trunk) and lower slice limit (slice just below the posterior descending artery). Contiguous 3-dimensional voxels between the Hounsfield unit limits of −190 to −30 were defined as fat voxels by default. EFV index was calculated dividing EFV by body surface area.
Exclusion criteria for the present study removed any subjects with known CAD or any previous evaluation for CAD. Subclinical CAD was defined using cardiac computed tomographic data as a total Agatston score ≥400 and/or presence of any noncalcified or mixed plaque encompassing >25% of the luminal diameter of a coronary artery on contrast coronary computed angiogram. An Agatston score of 400 is currently accepted in the guidelines as sufficient to justify a functional study in an asymptomatic patient who is at increased risk for CAD and adverse clinical events. The upper 90th percentile of the total Agatston score was obtained from the Multi-Ethnic Study of Atherosclerosis (MESA) cohort population and used as a reference value.
Continuous variables were described as mean ± SD. Differences between groups were assessed by 1-way analysis of variance and post hoc multiple comparisons were performed using the Bonferroni correction when appropriate. Spearman correlation was used for nonparametric distributions. Multivariate binary logistic regression models were constructed to identify the relation between diastolic dysfunction and EFV, 10-year Framingham Risk Score, subclinical CAD, and metabolic syndrome. Multiple linear regression models were constructed to assess the independent association of these parameters to other continuous diastolic function parameters (mean e′ and mean E/e′ ratio). Nonstandardized (beta) coefficient estimates with relative SEs were reported. Statistical analyses were performed using SPSS 19.0 (SPSS, Inc., Chicago, Illinois) and a 2-tailed p value <0.05 was considered statistically significant.
Results
Clinical and demographic characteristics of 110 subjects who met the inclusion criteria are listed in Table 1 . Calculated mean 10-year Framingham Risk Scores were 10% for men and 4% for women (p <0.0001), which were considered low-risk thresholds for major adverse cardiovascular events. There was a correlation between metabolic syndrome and diabetes and between metabolic syndrome and dyslipidemia (r = 0.30, p = 0.002 for the 2 comparisons).
| Age (years) | 55 ± 13 |
| Men | 72 (65%) |
| Body mass index (kg/m 2 ) | 28 ± 5 |
| Systolic blood pressure (mm Hg) | 124 ± 19 |
| Diastolic blood pressure (mm Hg) | 76 ± 10 |
| Diabetes mellitus | 13 (12%) |
| Dyslipidemia | 56 (50%) |
| Hypertension | 65 (59%) |
| Metabolic syndrome | 44 (40%) |
| 10-year Framingham Risk Score (%) ⁎ | 6 (0.5–34) |
| Women | 4 ± 3 |
| Men | 10 ± 7 |
Most subjects had normal echocardiographic findings including left atrial volume index, LV mass index, LV ejection fraction, and diastolic function parameters ( Table 2 ). According to American Society of Echocardiography guidelines, there were 40 subjects with abnormal diastolic function further classified as stage I (29 subjects, 72%) and stage II (11 subjects, 28%). Median duration between the 2 imaging studies was 1 day. Most patients (73%) underwent noncontrast cardiac computed tomography for coronary calcium scoring for screening of CAD. Subclinical CAD, as defined before, was present in 20% of the cohort (20 men vs 2 women, p = 0.005). EFV followed a non-normal distribution and was associated with metabolic disturbance, LV mass index, and diastolic dysfunction ( Table 3 ). There was no correlation between calcium score and EFV.
| Variable | Overall (n = 110) | Normal Values | Subjects (%) With | |
|---|---|---|---|---|
| <2 SD | >2 SD | |||
| Left atrial volume index (cm 3 /m 2 ) | 26 ± 9 | 22 ± 6 | 0% | 15% |
| Left ventricular mass index (g/m 2 ) | ||||
| Women | 74 ± 14 | 44–88 | 0 | 24% |
| Men | 91 ± 11 | 50–102 | 0 | 16% |
| Left ventricular ejection fraction (%) | 58 ± 4 | ≥55% | — | — |
| Diastolic mitral inflow ratio (early/late diastolic filling at time of atrial systole) | 1.23 ± 0.5 | 1.28 ± 0.25 | 13% | 10% |
| Deceleration time (ms) | 226 ± 57 | 181 ± 19 | 2% | 45% |
| Average tissue Doppler velocity from septal and lateral annuli (cm/s) | 10 ± 3 | ≥8 | 21% | — |
| Average left ventricular filling pressures (early diastolic filling/tissue Doppler velocity averaged from septal and lateral annuli) | 8 ± 3 | ≤8 | — | 36% |
| Epicardial fat volume (cm 3 ) | ||||
| Women | 67 ± 40 | 110 ± 41 | 8% | 0% |
| Men | 101 ± 51 | 137 ± 53 | 4% | 3% |
| Epicardial fat volume index (cm 3 /m 2 ) | 14 (13%) ⁎ | |||
| Women | 32.8 (18.7–53.7) | 31.8 (24.2–41.3) | — | |
| Men | 46.6 (30.0–59.4) | 34.2 (24.8–45.5) | — | |
| Agatston score (n = 65) | ||||
| Women | 40 ± 110 | 126 † | ||
| Men | 556 ± 1,109 | 324 † | — | — |
⁎ Epicardial fat volume index >68 cm 3 /m 2 .
| Variables | R | p Value |
|---|---|---|
| Age | 0.34 | <0.0001 |
| Male gender | −0.35 | <0.0001 |
| Body mass index | 0.43 | <0.0001 |
| Systolic blood pressure (mm Hg) | 0.22 | 0.02 |
| Waist circumference (n = 35) | 0.43 | 0.009 |
| 10-year Framingham Risk Score ⁎ | 0.41 | <0.0001 |
| Metabolic syndrome | 0.13 | 0.15 |
| Subclinical coronary artery disease | 0.21 | 0.03 |
| Hyperlipidemia | 0.25 | 0.007 |
| Left ventricular mass index | 0.41 | <0.0001 |
| Early and late diastolic fillings at time of atrial systole, early/late diastolic filling at time of atrial systole, deceleration time | −0.05 to 0.08 | NS † |
| Tissue Doppler velocity averaged from septal and lateral annuli | 0.44 | <0.0001 |
| Early diastolic filling/tissue Doppler velocity averaged from septal and lateral annuli | 0.34 | <0.0001 |
| Agatston score (n = 65) | 0.08 | 0.49 |
⁎ Framingham Risk Score included age, gender, systolic blood pressure, total cholesterol, high-density lipoprotein, and smoking status.
† The p value is nonsignificant for each of the 4 diastolic parameters.
To answer how these 3 parameters related to diastolic dysfunction, a multivariate binary logistic regression model was constructed. EFV was the only independent predictor (hazard ratio 1.09, 95% confidence interval 1.03 to 1.15, p = 0.003), whereas BMI (p = 0.36) and waist circumference (p = 0.09) were not. When we considered BMI >30 kg/m 2 as a surrogate for visceral adiposity and equivalent risk factor for metabolic syndrome development, EFV was the only predictor of diastolic dysfunction (hazard ratio 1.02, 95% confidence interval 1.007 to 1.03, p = 0.001).
Age, systolic blood pressure, Framingham Risk Score, LV mass index, EFV, and EFV index were independently associated with diastolic dysfunction. The independent and incremental value of EFV in addition to these traditional cardiovascular risk factors was sought in a multivariable binary logistic regression model; only Framingham Risk Score (p = 0.048) and EFV (p = 0.02) were independent predictors of diastolic dysfunction (stage ≥I). Furthermore, EFV had incremental prognostic value (R 2 change from 0.16 to 0.21, p = 0.02) when added to the clinical model that included Framingham Risk Score, metabolic syndrome, LV mass index, and subclinical CAD. A better model fit was seen when EFV index was used instead of EFV (model R 2 increased from 0.16 to 0.24, p = 0.004). EFV index remained the sole independent predictor of diastolic dysfunction (hazard ratio 1.03, p = 0.01) after adjusting for Framingham Risk Score, metabolic syndrome, LV mass index, and subclinical CAD ( Table 4 and Figure 1 ). The association of with mean e′ with Framingham Risk Score, metabolic syndrome, subclinical CAD, LV mass index, and EFV index was sought in a separate multiple linear regression model ( Table 4 ). EFV index was not only independently associated with mean e′ but also incremental to the association with clinical and echocardiographic data (model R 2 increased from 0.17 to 0.27, p = 0.001; Figure 2 ). In contrast, only EFV index was independently associated with mean E/e′ ( Table 4 ). EFV index was not only an independent predictor of mean E/e′ (p = 0.001) but also incrementally added to the model (R 2 of the model increased from 0.04 to 0.14, p = 0.001). The relation between EFV and age in the determination of LV filling pressures (mean E/e′ ratio) is shown in Figure 3 . Note that the age-associated increase in LV filling pressures (higher mean E/e′) is amplified by the increase in EFV.
