ew

r2

0h

Association of Epicardial Fat, Hypertension, Subclinical CoronaryArtery Disease, and Metabolic Syndrome With Left Ventricular

Diastolic Dysfunction

João L. Cavalcante, MDa, Balaji K. Tamarappoo, MD, PhDa, Rory Hachamovitch, MD, MSca,Deborah H. Kwon, MDa, M. Chadi Alraies, MDa, Sandra Halliburton, PhDa,Paul Schoenhagen, MDa, Damini Dey, PhDb, Daniel S. Berman, MDb, and

Thomas H. Marwick, MD, PhD, MPHa,*

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 todiastolic function is unknown. We sought to (1) understand the relation of epicardial fatvolume (EFV) to diastolic function and (2) understand the role of EFV in relation topotential risk factors (hypertension, subclinical CAD, and metabolic syndrome) of diastolicdysfunction in apparently healthy subjects with preserved systolic function and no historyof CAD. We studied 110 consecutive subjects (65% men, 55 � 13 years old, mean bodymass index 28 � 5 kg/m2) who underwent cardiac computed tomography and transthoracicechocardiography within 6 months as part of a self-referred health screening program.Exclusion criteria included history of CAD, significant valvular disease, systolic dysfunc-tion (left ventricular ejection fraction <50%). Diastolic function was defined according toAmerican Society of Echocardiography guidelines. EFV was measured using validatedcardiac computed tomographic software by 2 independent cardiologists blinded to clinicaland echocardiographic data. Hypertension and metabolic syndrome were present in 60% and45%, respectively. Subclinical CAD was identified in 20% of the cohort. Diastolic dysfunctionwas present in 45 patients. EFV was an independent predictor of diastolic dysfunction, meanpeak early diastolic mitral annular velocity, and ratio of early diastolic filling to peak earlydiastolic mitral annular velocity (p � 0.01, <0.0001, and 0.001, respectively) with incrementalcontribution to other clinical factors. In conclusion, EFV is an independent predictor ofimpaired diastolic function in apparently healthy overweight patients even after accounting forassociated co-morbidities such as metabolic syndrome, hypertension, and subclinical

CAD. © 2012 Elsevier Inc. All rights reserved. (Am J Cardiol 2012;110:1793–1798)

eeat

The association of obesity with myocardial dysfunctionis probably mediated through its strong links with hyper-tension, dyslipidemia, and atherosclerotic coronary arterydisease (CAD). However, obesity may be associated withstructural changes in the myocardium independent of itseffects on risk factors or CAD. Subclinical changes of leftventricular (LV) structure and function including abnormalrelaxation and strain have been reported in overweight sub-jects even after adjustment for mean arterial pressure, age,gender, and LV mass.1 Obese patients may present withheart failure with normal ejection fraction even in the ab-sence of CAD; each 1-kg/m2 increase in body mass index(BMI) has been shown to increase the risk of heart failureby 5% for men and by 7% for women.2 In this context, it isxtremely important to understand the mechanisms byhich obesity may cause LV dysfunction with a preserved

aDepartment of Cardiovascular Medicine, Cleveland Clinic, Cleveland,Ohio; bCedars-Sinai Medical Center, Los Angeles, California. Manuscripteceived July 23, 2012; revised manuscript received and accepted July 30,012.

*Corresponding author: Tel: �61 (3) 6226-7703; fax: �61 (3) 6226-7704.

E-mail address: tom.marwick@utas.edu.au (T.H. Marwick).

002-9149/12/$ – see front matter © 2012 Elsevier Inc. All rights reserved.ttp://dx.doi.org/10.1016/j.amjcard.2012.07.045

jection fraction. Epidemiologic studies have shown thatpicardial fat, a metabolically active fat depot, is stronglyssociated with obesity, metabolic syndrome, and diabe-es.3,4 The proximity of epicardial fat to the coronary arter-

ies has been used to explain the association of epicardial fatvolume (EFV) with increased coronary artery calcium, ath-erosclerotic plaque, and myocardial ischemia.3,5–7 Althoughthese factors may be responsible for diastolic dysfunction,epicardial fat may also have a direct paracrine effect on themyocardium and thus alter the structural properties of the leftventricle. We sought to (1) understand the relation of EFV todiastolic dysfunction and (2) understand the role of EFV inrelation to potential risk factors (hypertension, subclinicalCAD, and metabolic syndrome) of diastolic dysfunction inhealthy subjects with preserved systolic function.

Methods

We studied consecutive patients who underwent cardiaccomputed tomography and transthoracic echocardiographyat the Cleveland Clinic (Cleveland, Ohio) within 6 monthsas part of a self-referred health screening program. Exclu-sion criteria were (1) history of CAD (myocardial infarction

and/or percutaneous or surgical revascularization), (2) mod-

www.ajconline.org

B(w

wl

tFl

c

c

oisnaw

(wtrdtpbdd

sCpacassp

1794 The American Journal of Cardiology (www.ajconline.org)

erate (�2�) valvular regurgitation or any valvular stenosis,(3) systolic dysfunction (ejection fraction �50%), (4) incom-plete echocardiographic data, and (5) an intercurrent eventbetween the 2 imaging studies. Subjects with pathologic Qwaves on electrocardiogram, wall motion abnormalities onechocardiogram, or myocardial wall thinning seen on echocar-diogram and/or cardiac computed tomogram suggestive ofprevious myocardial infarction were also excluded. This studywas approved by the institutional review board.

Hypertension was defined as systolic blood pressure�130 mm Hg or diastolic blood pressure �85 mm Hg at thetime of the visit (mean of 2 readings) or history of hyper-tension but not on treatment. Diabetes mellitus was definedas fasting blood glucose �110 mg/dl or use of diabetesmedications. Dyslipidemia was defined as total serum cho-lesterol �240 mg/dl or use of lipid-lowering treatment.

MI was calculated as weight (kilograms) divided by heightmeters) squared. According to a standard definition, over-eight patients had a BMI from 25.0 to 29.9 kg/m2, whereas

obese patients had BMI �30 kg/m2. Metabolic syndromeas defined by the criteria proposed by the National Cho-

esterol Education Program Adult Treatment Panel III.8 Atleast 3 of the following components were needed to meetmetabolic syndrome criteria: (1) fasting blood glucose�100 mg/dl or patient’s self-reported history of diabetes oruse of diabetes medications; (2) blood pressure �130/85mm Hg or patient’s self-reported history of hypertension oruse of antihypertensive medications; (3) triglycerides �150mg/dl; (4) high-density lipoprotein �40 mg/dl; and (5) BMI�30 kg/m2 (where central obesity was assumed accordingo International Diabetes Federation guidelines9). Ten-yearramingham Risk Score was calculated according to guide-

ines8 and included the following risk factors: age, gender,systolic blood pressure, total cholesterol, high-density lipo-protein cholesterol and smoking history.

LV linear dimensions were measured from a parasternallong-axis view according to recommendations of the Amer-ican Society of Echocardiography.10 LV mass index wascalculated by the corrected American Society of Echocar-diography simplified cubed equation10 and indexed for bodysurface area. LV ejection fraction and LV volume indexwere calculated by the biplane modified Simpson rule. LVdiastolic function assessment was obtained according toAmerican Society of Echocardiography guidelines and in-cluded peak velocities of the early phase (E) and late phase(A) of the transmitral inflow, with derivation of the E/Aratio. LV myocardial velocities were evaluated by tissueDoppler imaging with pulsed sample volume placed at thelevel of the lateral and septal mitral valve annuli. Peak earlydiastolic mitral annular velocities (e=) were measured andaveraged over 3 cardiac cycles.11 Diastolic dysfunction wasategorized 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). Allechocardiograms were reviewed by 2 board-certified cardi-ologists blinded to clinical and cardiac computed tomo-graphic information.

Noncontrast computed tomogram was acquired using a

64-slice computed tomographic scanner (Sensation 64, Sie-

mens Medical Solutions, Erlangen, Germany) in the axialmode during a single breath-hold with prospective electro-cardiographic triggering and 120-kVp tube voltage. Imageswere reconstructed using a medium sharp kernel (B35f)with a slice thickness of 3 mm. Coronary calcium wasquantified on nonenhanced cardiac computed tomographicscans. Coronary calcium was defined as a plaque of � 3ontiguous pixels (area 1.0 mm2) with a density of �130

HU. A total coronary calcium score was determined bysumming individual Agatston scores from each of 4 ana-tomic sites (left main, left anterior descending, left circum-flex, and right coronary arteries).12 Contrast-enhanced cor-nary computed tomographic angiography was performedn the same scanner.13 Nonoverlapping images were recon-tructed using a medium sharp kernel (B26f) with a thick-ess of 3 mm for quantification of epicardial fat. Scans werenalyzed independently by 2 experienced investigators whoere blinded to clinical and echocardiographic information.Epicardial fat quantification was performed by QFAT

Cedars-Sinai Medical Center, Los Angeles, California) soft-are as previously described.14 EFV was defined as adipose

issue enclosed by the visceral pericardium including fat di-ectly surrounding the coronary arteries. Definition of epicar-ial contours was based in an upper slice limit (bifurcation ofhe pulmonary trunk) and lower slice limit (slice just below theosterior descending artery). Contiguous 3-dimensional voxelsetween the Hounsfield unit limits of �190 to �30 wereefined as fat voxels by default. EFV index was calculatedividing EFV by body surface area.

Exclusion criteria for the present study removed anyubjects with known CAD or any previous evaluation forAD. Subclinical CAD was defined using cardiac com-uted tomographic data as a total Agatston score �400nd/or presence of any noncalcified or mixed plaque en-ompassing �25% of the luminal diameter of a coronaryrtery on contrast coronary computed angiogram. An Agat-ton score of 400 is currently accepted in the guidelines asufficient to justify a functional study in an asymptomaticatient15 who is at increased risk for CAD and adverse

clinical events. The upper 90th percentile of the total Ag-atston score was obtained from the Multi-Ethnic Study ofAtherosclerosis (MESA) cohort population and used as a

Table 1Clinical baseline characteristics of subjects (n � 110)

Age (years) 55 � 13Men 72 (65%)Body mass index (kg/m2) 28 � 5Systolic blood pressure (mm Hg) 124 � 19Diastolic blood pressure (mm Hg) 76 � 10Diabetes mellitus 13 (12%)Dyslipidemia 56 (50%)Hypertension 65 (59%)Metabolic syndrome 44 (40%)10-year Framingham Risk Score (%)* 6 (0.5–34)

Women 4 � 3Men 10 � 7

Values are expressed as mean � SD, number of subjects (percentage), ormedian (range).

* Likelihood of coronary heart disease in 10 years.

reference value.16

Dypatwdsrdd

rd7

b(pSta

LL

LDDAA

E

E

MBSW1MSHLE

E

A

t

1795Preventive Cardiology/Epicardial Fat and Diastolic Dysfunction

Continuous variables were described as mean � SD.ifferences between groups were assessed by 1-way anal-sis of variance and post hoc multiple comparisons wereerformed using the Bonferroni correction when appropri-te. Spearman correlation was used for nonparametric dis-ributions. Multivariate binary logistic regression modelsere constructed to identify the relation between diastolicysfunction and EFV, 10-year Framingham Risk Score,ubclinical CAD, and metabolic syndrome. Multiple linearegression models were constructed to assess the indepen-ent association of these parameters to other continuousiastolic function parameters (mean e= and mean E/e= ratio).

Nonstandardized (beta) coefficient estimates with relativeSEs were reported. Statistical analyses were performed us-ing SPSS 19.0 (SPSS, Inc., Chicago, Illinois) and a 2-tailedp value �0.05 was considered statistically significant.

Results

Clinical and demographic characteristics of 110 subjectswho met the inclusion criteria are listed in Table 1. Calcu-lated mean 10-year Framingham Risk Scores were 10% formen and 4% for women (p �0.0001), which were consid-ered low-risk thresholds for major adverse cardiovascularevents. There was a correlation between metabolic syn-drome and diabetes and between metabolic syndrome anddyslipidemia (r � 0.30, p � 0.002 for the 2 comparisons).

Most subjects had normal echocardiographic findingsincluding left atrial volume index, LV mass index, LVejection fraction, and diastolic function parameters (Table2).4,12,13,17 According to American Society of Echocardiog-aphy guidelines, there were 40 subjects with abnormaliastolic function further classified as stage I (29 subjects,

Table 2Imaging characteristics of subjects

Variable

eft atrial volume index (cm3/m2)eft ventricular mass index (g/m2)WomenMeneft ventricular ejection fraction (%)iastolic mitral inflow ratio (early/late diastolic filling at time of atrial syeceleration time (ms)verage tissue Doppler velocity from septal and lateral annuli (cm/s)verage left ventricular filling pressures (early diastolic filling/tissue Dop

velocity averaged from septal and lateral annuli)picardial fat volume (cm3)WomenMenpicardial fat volume index (cm3/m2)WomenMen

Agatston score (n � 65)WomenMen

Values are expressed as mean � SD.* Epicardial fat volume index �68 cm3/m2.† Upper 90th percentile.

2%) and stage II (11 subjects, 28%). Median duration b

etween the 2 imaging studies was 1 day. Most patients73%) underwent noncontrast cardiac computed tomogra-hy for coronary calcium scoring for screening of CAD.ubclinical CAD, as defined before, was present in 20% of

he cohort (20 men vs 2 women, p � 0.005). EFV followednon-normal distribution and was associated with meta-

Overall(n � 110)

NormalValues4,12,13,17

Subjects (%) With

�2 SD �2 SD

26 � 9 22 � 6 0% 15%

74 � 14 44–88 0 24%91 � 11 50–102 0 16%58 � 4 �55% — —

1.23 � 0.5 1.28 � 0.25 13% 10%226 � 57 181 � 19 2% 45%10 � 3 �8 21% —8 � 3 �8 — 36%

67 � 40 110 � 41 8% 0%101 � 51 137 � 53 4% 3%

14 (13%)*32.8 (18.7–53.7) 31.8 (24.2–41.3) —46.6 (30.0–59.4) 34.2 (24.8–45.5) —

40 � 110 126†

556 � 1,109 324† — —

Table 3Univariate correlations of epicardial fat volume

Variables R p Value

Age 0.34 �0.0001ale gender �0.35 �0.0001ody mass index 0.43 �0.0001ystolic blood pressure (mm Hg) 0.22 0.02aist circumference (n � 35) 0.43 0.009

0-year Framingham Risk Score* 0.41 �0.0001etabolic syndrome 0.13 0.15

ubclinical coronary artery disease 0.21 0.03yperlipidemia 0.25 0.007eft ventricular mass index 0.41 �0.0001arly and late diastolic fillings at time of

atrial systole, early/late diastolicfilling at time of atrial systole,deceleration time

�0.05 to 0.08 NS†

Tissue Doppler velocity averaged fromseptal and lateral annuli

0.44 �0.0001

arly diastolic filling/tissue Dopplervelocity averaged from septal andlateral annuli

0.34 �0.0001

gatston score (n � 65) 0.08 0.49

* Framingham Risk Score included age, gender, systolic blood pressure,otal cholesterol, high-density lipoprotein, and smoking status.

† The p value is nonsignificant for each of the 4 diastolic parameters.

stole)

pler

olic disturbance, LV mass index, and diastolic dysfunction

ekrt1

LaicbSpE

r

fFi

irip

D

dFr

Ftmd

, total

1796 The American Journal of Cardiology (www.ajconline.org)

(Table 3). There was no correlation between calcium scoreand EFV.

To answer how these 3 parameters related to diastolicdysfunction, a multivariate binary logistic regression modelwas 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 circumfer-nce (p � 0.09) were not. When we considered BMI �30g/m2 as a surrogate for visceral adiposity and equivalentisk factor for metabolic syndrome development, EFV washe only predictor of diastolic dysfunction (hazard ratio.02, 95% confidence interval 1.007 to 1.03, p � 0.001).

Age, systolic blood pressure, Framingham Risk Score,V mass index, EFV, and EFV index were independentlyssociated with diastolic dysfunction. The independent andncremental value of EFV in addition to these traditionalardiovascular risk factors was sought in a multivariableinary logistic regression model; only Framingham Riskcore (p � 0.048) and EFV (p � 0.02) were independentredictors of diastolic dysfunction (stage �I). Furthermore,FV had incremental prognostic value (R2 change from

0.16 to 0.21, p � 0.02) when added to the clinical modelthat included Framingham Risk Score, metabolic syndrome,LV mass index, and subclinical CAD. A better model fitwas seen when EFV index was used instead of EFV (modelR2 increased from 0.16 to 0.24, p � 0.004). EFV index

Figure 1. Incremental value of epicardial fat volume index for prediction ofdiastolic function class (stage �I). *The clinical model included 10-year

ramingham Risk Score (age, gender, systolic blood pressure, total choles-erol, high-density lipoprotein, and smoking history), metabolic syndrome, LVass index and subclinical coronary artery disease subclinical coronary artery

isease.

Table 4Predictors of diastolic dysfunction

Stage �I

(Model R2 � 0.24)

Beta � SE p Value

10-year Framingham Risk Score* 0.07 � 0.04 0.09Metabolic syndrome �0.01 � 0.46 0.98Subclinical coronary artery disease �0.11 � 0.58 0.84Left ventricular mass index 0.009 � 0.01 0.41Epicardial fat volume index 0.03 � 0.01 0.01

Values are nonstandardized correlation coefficients (betas) � SEs and* Framingham Risk Score included age, gender, systolic blood pressure

emained the sole independent predictor of diastolic dys- o

unction (hazard ratio 1.03, p � 0.01) after adjusting forramingham Risk Score, metabolic syndrome, LV mass

ndex, and subclinical CAD (Table 4 and Figure 1). Theassociation of with mean e= with Framingham Risk Score,metabolic syndrome, subclinical CAD, LV mass index, andEFV index was sought in a separate multiple linear regres-sion model (Table 4). EFV index was not only indepen-dently associated with mean e= but also incremental to theassociation with clinical and echocardiographic data (modelR2 increased from 0.17 to 0.27, p � 0.001; Figure 2). Incontrast, only EFV index was independently associated withmean E/e= (Table 4). EFV index was not only an indepen-dent predictor of mean E/e= (p � 0.001) but also incremen-tally added to the model (R2 of the model increased from0.04 to 0.14, p � 0.001). The relation between EFV and agen the determination of LV filling pressures (mean E/e=atio) is shown in Figure 3. Note that the age-associatedncrease in LV filling pressures (higher mean E/e=) is am-lified by the increase in EFV.

iscussion

This study defines several important associations of car-iac computed tomogram-based quantification of EFV.irst, our results indicate that EFV is an independent cor-elate of impaired diastolic function in apparently healthy

Figure 2. Incremental value of epicardial fat volume index for prediction ofmean peak early diastolic mitral annular velocities. *The clinical model in-cluded 10-year Framingham Risk Score (age, gender, systolic blood pressure,total cholesterol, high-density lipoprotein, and smoking history), metabolicsyndrome, LV mass index and subclinical coronary artery disease subclinicalcoronary artery disease.

Mean e= Mean E/e=

(Model R2 � 0.27) (Model R2 � 0.14)

Beta � SE p Value Beta � SE p Value

0.11 � 0.04 0.01 0.02 � 0.04 0.590.07 � 0.54 0.90 0.59 � 0.54 0.280.36 � 0.69 0.60 �0.03 � 0.69 0.96

0.005 � 0.01 0.70 �0.006 � 0.01 0.630.02 � 0.006 �0.0001 0.04 � 0.01 0.001

s after adjusting for all covariates listed.cholesterol, high-density lipoprotein, and smoking status.

���

�

p value

verweight patients after accounting for associated co-mor-

pwmeicr

idm

pfbcdr(fscif

mbebstcrcg

adl

ftnp

1797Preventive Cardiology/Epicardial Fat and Diastolic Dysfunction

bidities such as CAD risk factors, hypertension, metabolicsyndrome, and subclinical CAD. Two other important ad-ditional findings are (1) measurement of EFV adds incre-mentally to the prediction of diastolic dysfunction, mean e=,and mean E/e=; and (2) EFV is a stronger correlate of meane= and mean E/e= than are BMI and metabolic syndrome.We interpret these findings to suggest that changes inmyocardial function may arise from the paracrine effectsof epicardial fat (maybe even more than the metaboliceffects of visceral adiposity) and that these effects arerelated to myocardial dysfunction as measured by early/sensitive markers of impaired diastolic function such ase= and E/e= ratio.

Diastolic dysfunction is widely prevalent in the generalpopulation and its severity is prognostically important. Alarge longitudinal data from a study of Olmsted Countyshowed that patients with advanced diastolic dysfunctionhad a 10-fold higher risk of all-cause death compared tosubjects with normal diastolic function after adjustment forage, gender, and ejection fraction.18 Abhayaratna et al19 alsoshowed that diastolic dysfunction severity is associated withcardiovascular risk factors including hypertension, diabetes,obesity, and metabolic syndrome. These findings were con-firmed in a recent study by Russo et al17 who found arogressively worsening of individual diastolic parametersith worsening BMI. Diastolic dysfunction seen in diabetesellitus and metabolic syndrome has been a matter of

xtensive investigation with several mechanisms proposedncluding but not limited to metabolic disturbances, myo-ardial fibrosis, microvascular dysfunction, autonomic neu-opathy, and insulin resistance.20–22

Epicardial fat is a metabolically active fat depot that is inanatomic proximity to the myocardium, shares the samemicrocirculation, and may have important paracrine ef-fects.23 Epicardial fat correlates with BMI and is increasedn obese subjects.3 The association of increased EFV withiastolic dysfunction may be confounded by factors such as

Figure 3. Relation among age (years), epicardial fat volume (cubic centimetto peak early diastolic mitral annular velocity).

etabolic syndrome4,24 and hypertension,25 which are often u

resent in obese patients. Furthermore, increased epicardialat appears to be related to increased atherosclerotic plaqueurden, concomitant CAD, myocardial ischemia, and sub-linical microvascular dysfunction that is commonly seen iniabetics and in patients with metabolic syndrome.3,5,20 Ouresults indicate that EFV has an independent, albeit modest8% to 10% of variance), role in affecting myocardial dys-unction after adjusting for age, gender, systolic blood pres-ure, lipids, metabolic syndrome, LV mass index, and sub-linical CAD. Furthermore, measurement of EFV addsncrementally to the prediction of impaired LV diastolicunction.

Epicardial fat thickness can be readily visualized andeasured by echocardiography.26 Several thresholds have

een proposed to indicate increased cardiovascular risk.4 How-ver, the reproducibility of this method is imperfect probablyecause these measurements represent only 1 planar dimen-ion, which is subject to probe angulation, adequate visualiza-ion of the cardiac structure, and the given phase of the cardiacycle chosen to be measured. A recent study showed pooreproducibility of echocardiographic measurements and lowoncordance compared to cardiac computed tomography, aeometry-independent technique.27

There are several limitations. Waist circumference wasavailable only in a minority of subjects (�10%) but BMI�30 kg/m2 has been endorsed as a surrogate for visceraldiposity and equivalent risk factor for metabolic syndromeevelopment by International Diabetes Federation guide-ines.9 Definition of diabetes chosen (fasting blood sugar

�110 mg/dl) might have included subjects with impairedfasting glucose, although an important proportion of theseprediabetic patients (12% to 37%) carries a risk for futurecardiovascular disease.28 Our evaluation of LV diastolicunction by Doppler flow analysis did not include parame-ers such as isovolumic relaxation time or pulmonary ve-ous flow. However, those parameters have high load de-endence, and use of tissue Doppler parameters allowed

d mean left ventricular filling pressures (mean ratio of early diastolic filling

ers), an

s to detect diastolic abnormalities even when a pseudo-

eTmo

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

1798 The American Journal of Cardiology (www.ajconline.org)

normalized flow pattern was present. The retrospectivenature of the study and small sample might have contrib-uted to the smaller estimations of the predictive modelsbut are certainly consistent in conveying the message thatEFV might play an important role in the impairment ofdiastolic function.

EFV is an independent correlate of impaired LV diastolicfunction in apparently healthy overweight patients evenafter accounting for associated co-morbidities such as met-abolic syndrome, hypertension, and subclinical CAD. Twoimportant additional findings are (1) cardiac computed to-mogram-based measurement of EFV adds incrementally tothe prediction of diastolic functional class (stage �I), meane=, and E/e= ratio and (2) EFV is a stronger correlate of mean= and E/e= ratio than are BMI and metabolic syndrome.hese findings suggest that EFV may exert changes inyocardial function by a paracrine mechanism independent

f subclinical CAD, metabolic syndrome, or hypertension.

1. Wong CY, O’Moore-Sullivan T, Leano R, Byrne N, Beller E, Mar-wick TH. Alterations of left ventricular myocardial characteristicsassociated with obesity. Circulation 2004;110:3081–3087.

2. Kenchaiah S, Evans JC, Levy D, Wilson PW, Benjamin EJ, LarsonMG, Kannel WB, Vasan RS. Obesity and the risk of heart failure.N Engl J Med 2002;347:305–313.

3. Rosito GA, Massaro JM, Hoffmann U, Ruberg FL, Mahabadi AA,Vasan RS, O’Donnell CJ, Fox CS. Pericardial fat, visceral abdominalfat, cardiovascular disease risk factors, and vascular calcification in acommunity-based sample: the Framingham Heart Study. Circulation2008;117:605–613.

4. Mookadam F, Goel R, Alharthi MS, Jiamsripong P, Cha S. Epicardialfat and its association with cardiovascular risk: a cross-sectional ob-servational study. Heart Views 2010;11:103–108.

5. Tamarappoo B, Dey D, Shmilovich H, Nakazato R, Gransar H, ChengVY, Friedman JD, Hayes SW, Thomson LE, Slomka PJ, Rozanski A,Berman DS. Increased pericardial fat volume measured from noncon-trast CT predicts myocardial ischemia by SPECT. JACC CardiovascImaging 2010;3:1104–1112.

6. Cheng VY, Dey D, Tamarappoo B, Nakazato R, Gransar H, Miranda-Peats R, Ramesh A, Wong ND, Shaw LJ, Slomka PJ, Berman DS.Pericardial fat burden on ECG-gated noncontrast CT in asymptomaticpatients who subsequently experience adverse cardiovascular events.JACC Cardiovasc Imaging 2010;3:352–360.

7. Mahabadi AA, Massaro JM, Rosito GA, Levy D, Murabito JM, WolfPA, O’Donnell CJ, Fox CS, Hoffmann U. Association of pericardialfat, intrathoracic fat, and visceral abdominal fat with cardiovasculardisease burden: the Framingham Heart Study. Eur Heart J 2009;30:850–856.

8. Expert Panel on Detection, Evaluation, and Treatment of High BloodCholesterol in Adults. Executive summary of the third report of theNational Cholesterol Education Program (NCEP) expert panel ondetection, evaluation, and treatment of high blood cholesterol in adults(Adult Treatment Panel III). JAMA 2001;285:2486–2497.

9. Alberti KG, Zimmet P, Shaw J. Metabolic syndrome—a new world-wide definition. A consensus statement from the International DiabetesFederation. Diabet Med J Br Diabet Assoc 2006;23:469–480.

0. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pel-likka PA, Picard MH, Roman MJ, Seward J, Shanewise JS, SolomonSD, Spencer KT, Sutton MS, Stewart WJ; Chamber QuantificationWriting Group, American Society of Echocardiography’s Guidelinesand Standards Committee, European Association of Echocardiogra-phy. Recommendations for chamber quantification: a report from theAmerican Society of Echocardiography’s Guidelines and StandardsCommittee and the Chamber Quantification Writing Group, developedin conjunction with the European Association of Echocardiography, abranch of the European Society of Cardiology. J Am Soc Echocardiogr2005;18:1440–1463.

1. Nagueh SF, Appleton CP, Gillebert TC, Marino PN, Oh JK, SmisethOA, Waggoner AD, Flachskampf FA, Pellikka PA, Evangelista A.

Recommendations for the evaluation of left ventricular diastolicfunction by echocardiography. J Am Soc Echocardiogr 2009;22:107–133.

2. Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M Jr,Detrano R. Quantification of coronary artery calcium using ultrafastcomputed tomography. J Am Coll Cardiol 1990;15:827–832.

3. Tamarappoo BK, Dey D, Nakazato R, Shmilovich H, Smith T, ChengVY, Thomson LE, Hayes SW, Friedman JD, Germano G, Slomka PJ,Berman DS. Comparison of the extent and severity of myocardialperfusion defects measured by CT coronary angiography and SPECTmyocardial perfusion imaging. JACC Cardiovasc Imaging 2010;3:1010–1019.

14. Dey D, Suzuki Y, Suzuki S, Ohba M, Slomka PJ, Polk D, Shaw LJ,Berman DS. Automated quantitation of pericardiac fat from noncon-trast CT. Invest Radiol 2008;43:145–153.

5. Brindis RG, Douglas PS, Hendel RC, Peterson ED, Wolk MJ, AllenJM, Patel MR, Raskin IE, Hendel RC, Bateman TM, Cerqueira MD,Gibbons RJ, Gillam LD, Gillespie JA, Hendel RC, Iskandrian AE,Jerome SD, Krumholz HM. ACCF/ASNC appropriateness criteria forsingle-photon emission computed tomography myocardial perfusionimaging (SPECT MPI): a report of the American College of Cardiol-ogy Foundation quality strategic directions committee appropriatenesscriteria working group and the American Society of Nuclear Cardiol-ogy endorsed by the American Heart Association. J Am Coll Cardiol2005;46:1587–1605.

6. McClelland RL, Chung H, Detrano R, Post W, Kronmal RA. Distri-bution of coronary artery calcium by race, gender, and age: resultsfrom the Multi-Ethnic Study of Atherosclerosis (MESA). Circulation2006;113:30–37.

7. Russo C, Jin Z, Homma S, Rundek T, Elkind MS, Sacco RL, Di TullioMR. Effect of obesity and overweight on left ventricular diastolicfunction: a community-based study in an elderly cohort. J Am CollCardiol 2011;57:1368–1374.

8. Pritchett AM, Mahoney DW, Jacobsen SJ, Rodeheffer RJ, Karon BL,Redfield MM. Diastolic dysfunction and left atrial volume: a popula-tion-based study. J Am Coll Cardiol 2005;45:87–92.

9. Abhayaratna WP, Marwick TH, Smith WT, Becker NG. Characteris-tics of left ventricular diastolic dysfunction in the community: anechocardiographic survey. Heart 2006;92:1259–1264.

0. Campbell DJ, Somaratne JB, Jenkins AJ, Prior DL, Yii M, Kenny JF,Newcomb AE, Schalkwijk CG, Black MJ, Kelly DJ. Impact of type 2diabetes and the metabolic syndrome on myocardial structure andmicrovasculature of men with coronary artery disease. CardiovascDiabetol 2011;10:80.

1. Fang ZY, Prins JB, Marwick TH. Diabetic cardiomyopathy: evidence,mechanisms, and therapeutic implications. Endocr Rev 2004;25:543–567.

2. van Heerebeek L, Hamdani N, Handoko ML, Falcao-Pires I, MustersRJ, Kupreishvili K, Ijsselmuiden AJ, Schalkwijk CG, Bronzwaer JG,Diamant M, Borbély A, van der Velden J, Stienen GJ, Laarman GJ,Niessen HW, Paulus WJ. Diastolic stiffness of the failing diabeticheart: importance of fibrosis, advanced glycation end products, andmyocyte resting tension. Circulation 2008;117:43–51.

3. Rabkin SW. Epicardial fat: properties, function and relationship toobesity. Obes Rev 2007;8:253–261.

4. Iacobellis G, Willens HJ, Barbaro G, Sharma AM. Threshold values ofhigh-risk echocardiographic epicardial fat thickness. Obesity (SilverSpring) 2008;16:887–892.

5. Sironi AM, Pingitore A, Ghione S, De Marchi D, Scattini B, PositanoV, Muscelli E, Ciociaro D, Lombardi M, Ferrannini E, Gastaldelli A.Early hypertension is associated with reduced regional cardiac func-tion, insulin resistance, epicardial, and visceral fat. Hypertension 2008;51:282–288.

6. Iacobellis G, Willens HJ. Echocardiographic epicardial fat: a review ofresearch and clinical applications. J Am Soc Echocardiogr 2009;22:1311–1319.

7. Saura D, Oliva MJ, Rodríguez D, Pascual-Figal DA, Hurtado JA, PinarE, de la Morena G, Valdés M. Reproducibility of echocardiographicmeasurements of epicardial fat thickness. Int J Cardiol 2010;141:311–313.

8. Ford ES, Zhao G, Li C. Pre-diabetes and the risk for cardiovascular

disease: a systematic review of the evidence. J Am Coll Cardiol2010;55:1310–1317.