C© 2008, the AuthorsJournal compilation C© 2008, Wiley Periodicals, Inc.DOI: 10.1111/j.1540-8175.2008.00764.x

Lack of Association between Obesity and LeftVentricular Systolic DysfunctionMohammad Reza Movahed, M.D., Ph.D.∗ and Yuji Saito, M.D., Ph.D.†∗Division of Cardiology, Department of Medicine, Sarver Heart Center and The Southern Arizona

VA Health Care System, University of Arizona School of Medicine, Tucson, Arizona; and †Divisionof Cardiology, Department of Medicine, University of California, Irvine Medical Center, Orange,California

Background: Previous studies have demonstrated that obesity is one of the risk factors for congestiveheart failure (CHF). By analyzing a large database, we investigated any association between bodymass index (BMI) and left ventricular (LV) systolic dysfunction. Methods: We retrospectively analyzed24,265 echocardiograms performed between 1984 and 1998. Fractional shortening (FS) and BMI wereavailable for 13,382 subjects in this cohort which were used for data analysis. FS was stratified intofour groups: (1) FS > 25%, (2) FS 17.5–25%, (3) FS 10–17.5%, and (4) FS < 10%. Furthermore, wealso used final diagnosis that was coded by the reading cardiologist as mild, moderate, and severe LVdysfunction separately for data analysis. BMI was divided into four groups: BMI < 18.5 kg/m2 (un-derweight), 18.5–24.9 kg/m2 (normal), 25–30 kg/m2 (overweight), and >30 kg/m2 (obese). Results:There was no association between different BMI categories and LV systolic function. The preva-lence of mild, moderate, or severely decreased LV function (based on FS or subjective interpretationof reading cardiologists) was equally distributed between the groups. Obese patients (BMI > 30%)had normal FS of >25 in 16.9%, mildly decreased FS in 18%, moderately decreased FS in 18.4%,and severely decreased FS in 20.1% P = ns. Conclusion: Our study is consistent with previous trialssuggesting that obesity is not related to systolic LV dysfunction. The underlying mechanism for theoccurrence of congestive heart failure in obese patients needs further investigation. (ECHOCARDIO-GRAPHY, Volume 26, February 2009)

obesity, body mass index (BMI), fractional shortening (FS), left ventricular function, EF, congestiveheart failure, diastolic function, CHF

Obesity is associated with increased car-diovascular morbidity and mortality. Extremeobesity is recognized to be a risk factor for con-gestive heart failure (CHF). Large community-based studies have demonstrated that obesityis associated with a significantly increased riskfor heart failure.1,2 The cause of this associa-tion is not known. Several studies have eval-uated body mass index (BMI) as a risk factorfor left ventricular (LV) remodeling, such as LV

Address for correspondence and reprint requests:Mohammad Reza Movahed, M.D., Ph.D., F.A.C.P., F.A.C.C.,F.C.C.P., Director of Coronary Care Unit, Universityof Arizona Sarver Heart Center, 1501 North CampbellAvenue, Tucson, AZ 85724. Fax: +1-520-626-5181; E-mail:rmovahed@email.arizona.edu, rmova@aol.com

hypertrophy and dilatation, which could lead toheart failure.3–6 However, conflicting data existin the literature with regard to the associationbetween BMI and LV function.7–9 Furthermore,a direct effect of isolated obesity on cardiac sys-tolic function is not well established.

In the present study, we used a largedatabase which was generated from echocar-diograms performed at the University of Cali-fornia, Irvine. The objective of this study wasto investigate any association between obesityand LV systolic dysfunction.

Methods

We retrospectively analyzed 24,265 comput-erized echocardiogram reports from patients

128 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 26, No. 2, 2009

OBESITY AND SYSTOLIC LV FUNCTION

who underwent echocardiographic examinationfor clinical reasons at the University ofCalifornia, Irvine, between 1984 and 1998. Weselected adult patients between the age of16 to 90 years for this study. These patientswere referred by clinicians for in- or outpatientechocardiographic examination for various rea-sons. BMI and fractional shortening (FS) wereavailable in 13,382 patients which were usedfor statistical analysis. Weight and height of thepatients at the time of the echocardiographicexamination were utilized for BMI calculation.BMI was calculated using standard formula us-ing height and weight. The mean age of the pop-ulation was 51.6 ± 18 SD (standard deviation).FS was measured in the standard parasternallong axis view. We divided FS into four groupsaccording to the magnitude of LV systolic func-tion: (1) FS > 25% (normal), (2) FS 17.5–25%(mild LV systolic dysfunction), (3) FS 10–17.5%(moderate LV systolic dysfunction) and (4) FS <10% (severe LV systolic dysfunction). BMI wascategorized into four groups: BMI < 18.5 kg/m2

(underweight), 18.5–24.9 kg/m2 (normal), 25–29.9 kg/m2 (overweight) and >30 kg/m2 (obese).We did not have ejection fraction (EF) docu-mentation in our database. Therefore, in orderto overcome the limitation of FS in the accu-rate assessment of global LV systolic function,we also used the final diagnosis that was codedin the computerized report by the reading car-diologists as normal, mild, moderate or severelyreduced LV function and correlated it with thepresence of obesity. All reading cardiologistswere board certified or eligible and were trained

Figure 1. Distribution of body massindexes (BMI) and age stratified in5-year increments.

for interpretation of echocardiograms. We usedthe Pearson chi-square test for univariate anal-ysis. SPSS program version 1.3 (SPSS, Chicago,IL, USA) was utilized for this study. Thisstudy was approved by institutional reviewboard.

Results

There were 10,916 males (46.6%) and 12,431females (53.4%) in this study. The distributionbetween different BMI ranges and age can beseen in Figure 1 showing a peak prevalence ofobesity occurring between the age of 55 and60 years.

BMI of <18.5 kg/m2 (underweight) waspresent in 5.7%, BMI of 18.5–24.9 kg/m2 (nor-mal) in 49.5%, BMI of 25–29.9 kg/m2 (over-weight) in 27.8%, and BMI of >30 kg/m2 (obese)in 17.1% of the patients. There was no cor-relation between the different BMI categoriesin regard to normal or abnormal LV systolicfunction. Mild, moderate, or severely decreasedLV function based on FS or subjective inter-pretation of LV systolic function was equallydistributed between the groups (Table I andFig. 2). Obese patients (BMI > 30%) had aprevalence of normal FS (FS of >25) in 16.9%,mildly decreased FS in 18%, moderately de-creased FS in 18.4%, and severely decreasedFS in 20.1% P = ns (Fig. 2). Furthermore,when subjective criteria for decreased LV sys-tolic function were used (coded as final diagno-sis by reading cardiologists), the incidence ofabnormal LV systolic function or the severity of

Vol. 26, No. 2, 2009 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. 129

MOVAHED AND SAITO

TABLE I

The Number (N) and Percentage of Patients with Body Mass Index (BMI) and Fractional Shortening Divided into FourSubgroups Showing no Significant Differences between the Groups.

Fractional Shortening Total

Over 25 17.5–25 10–17.5 Less than 10BMI <18.5 kg/m2 Underweight N = 670 N = 53 N = 38 N = 7 N = 768

5.8% 4.9% 6.2% 4.7% 5.7%BMI 18.5–24.9 kg/m2 Normal N = 5,741 N = 504 N = 298 N = 75 N = 6,618

49.7% 46.8% 48.5% 50.3% 49.5%BMI 25–29.9 kg/m2 Overweight N = 3,182 N = 326 N = 166 N = 37 N = 3,711

27.6% 30.3% 27.0% 24.8% 27.7%BMI ≥30 kg/m2 Obese N = 1,948 N = 194 N = 113 N = 30 N = 2,285

16.9% 18.0% 18.4% 20.1% 17.1%Total N = 11,541 N = 1,077 N = 615 N = 149 N = 13,382

86.2% 8.0% 4.6% 1.1% 100.0%

systolic LV dysfunction did not defer betweendifferent BMI categories (Fig. 3).

Discussion

Obesity has been found to be associated withheart failure. In an analysis from the Fram-ingham Heart Study, heart failure developedin 8.4% and the risk of heart failure was in-creased approximately twofold in obese individ-uals (BMI ≥30) when compared with nonobesesubjects.1 They concluded that approximately11% of heart failure cases in men and 14% inwomen could be attributed to obesity alone. Asimilar finding was reported by He et al.2 Theyfound that being overweight is an independentrisk for heart failure by studying a cohort of13,643 patients with the average follow-up of19 years.

However, these epidemiological studies donot evaluate the type or cause of heart fail-ure in the obese patients. In our study, we

Figure 2. No significant differences inthe prevalence of FS (fractional shorten-ings) and BMI (body mass indexes) cat-egories (n = number of cases, P = ns).

found no correlation between obesity and anydegree of LV systolic dysfunction measured us-ing both objective data such as FS or subjectiveinterpretation of LV function by board certifiedcardiologists. Our study is the largest studythat confirms previous finding suggesting lackof any association between obesity and LV sys-tolic dysfunction. Our result is in concordancewith a recently published study of 1,806 pa-tients. In this study, there was no associationbetween obesity and LV systolic dysfunctionsimilar to our observation.10 Our study eval-uated over 13,000 echocardiograms, the largeststudy available in the literature in this regard.Furthermore, in our study despite the use oftwo different definitions of LV systolic func-tion, we found identical results showing lackof any association between obesity and LV sys-tolic dysfunction.

Our finding in concordance to Dorbala et al.10

strongly suggests that the mechanism forobesity-associated heart failure is not due to

130 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 26, No. 2, 2009

OBESITY AND SYSTOLIC LV FUNCTION

Figure 3. No significant differencesin the prevalence of BMI (body massindex) categories and subjective as-sessment of left ventricular functionby reading cardiologist. Diagnosiswas coded as mild, moderate, or se-vere left ventricular systolic dysfunc-tion (P = ns).

LV systolic dysfunction. Obesity is associatedwith several changes in cardiovascular func-tion, some of which increase cardiac work pa-rameters like LV dimension, stroke volume, andcardiac output leading to eccentric LV hypertro-phy. These changes could cause diastolic dys-function as a possible cause of heart failure inthis population.5,6,9,11

It has recently been recognized that in-creased BMI is protective against mortality inpatients with heart failure (so-called reverseepidemiology).8,12,13 The cause of this paradoxis not understood. It is postulated that lowerbody weight is associated with the increasedcatabolic state which is associated with higherlevels of tumor necrosis factor and other cy-tokines.14,15 Our study could, in part, explainthe cause of reverse epidemiology in obese pa-tients with heart failure. Based on our re-sults and available literature, the increased oc-currence of congestive heart failure in obesepatients could be related to LV diastolic dys-function. Using our database, we found that ab-normal diastolic parameters such as left atrialenlargement, LV hypertrophy, LV mass, andE/A reversal are directly associated with in-creasing BMI categories.16 This finding sug-gests abnormal diastology as an important fac-tor in the pathogenesis of heart disease in obesepatients. Tumuklu and colleagues corroborateour observation by finding early diastolic ab-normalities in obese patients.17 The prevalenceof valvular abnormalities such as aortic regur-gitation 18 has been found to be higher in obesepatients which may contribute to a higher riskfor heart failure. Future studies are necessaryfor better understanding of diastolic function inthe obese population.

Limitations

This study was a retrospective cross-sectionalstudy, which limits our results. We did not havebaseline characteristics such as diabetes, hy-pertension, or other comorbid conditions limit-ing our results. Referring diagnoses were notavailable in the majority of the patients pre-venting us to adjust our data for underlyingdisease at the time when echocardiography wasperformed.

Acknowledgment: We would like to thank Ms. Good Galeand Dr. Mehrnoosh Hashemzadeh for their support in theediting of this manuscript.

References

1. Kenchaiah S, Evans JC, Levy D, et al: Obesity and therisk of heart failure. N Engl J Med 2002;347:305–313.

2. He J, Ogden LG, Bazzano LA, et al: Risk factorsfor congestive heart failure in US men and women:NHANES I epidemiologic follow-up study. Arch In-tern Med 2001;161:996–1002.

3. Messerli FH, Sundgaard-Riise K, Reisin ED, et al:Dimorphic cardiac adaptation to obesity and arterialhypertension. Ann Intern Med 1983;99:757–761.

4. Hammond IW, Devereux RB, Alderman MH, et al: Re-lation of blood pressure and body build to left ventricu-lar mass in normotensive and hypertensive employedadults. J Am Coll Cardiol 1988;12:996–1004.

5. Lauer MS, Anderson KM, Kannel WB, et al: The im-pact of obesity on left ventricular mass and geometry.The Framingham Heart Study. JAMA 1991;266:231–236.

6. Alpert MA, Lambert CR, Terry BE, et al: Influenceof left ventricular mass on left ventricular diastolicfilling in normotensive morbid obesity. Am Heart J1995;130:1068–1073.

7. Krishnan R, Becker RJ, Beighley LM, et al: Impact ofbody mass index on markers of left ventricular thick-ness and mass calculation: Results of a pilot analysis.Echocardiography 2005;22:203–210.

Vol. 26, No. 2, 2009 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. 131

MOVAHED AND SAITO

8. Peterson LR, Waggoner AD, Schechtman KB, et al: Al-terations in left ventricular structure and function inyoung healthy obese women: Assessment by echocar-diography and tissue Doppler imaging. J Am Coll Car-diol 2004;43:1399–1404.

9. Pascual M, Pascual DA, Soria F, et al: Effects of iso-lated obesity on systolic and diastolic left ventricularfunction. Heart 2003;89:1152–1156.

10. Dorbala S, Crugnale S, Yang D, et al: Effect of bodymass index on left ventricular cavity size and ejectionfraction. Am J Cardiol 2006;97:725–729.

11. Alpert MA, Hashimi MW: Obesity and the heart. AmJ Med Sci 1993;306:117–123.

12. Davos CH, Doehner W, Rauchhaus M, et al: Bodymass and survival in patients with chronic heart fail-ure without cachexia: The importance of obesity. JCard Fail 2003;9:29–35.

13. Kalantar-Zadeh K, Block G, Horwich T, et al: Reverseepidemiology of conventional cardiovascular risk fac-tors in patients with chronic heart failure. J Am CollCardiol 2004;43:1439–1444.

14. Anker SD, Clark AL, Kemp M, et al: Tumor necrosisfactor and steroid metabolism in chronic heart failure:Possible relation to muscle wasting. J Am Coll Cardiol1997;30:997–1001.

15. Anker SD, Ponikowski P, Varney S, et al: Wasting asindependent risk factor for mortality in chronic heartfailure. Lancet 1997;349:1050–1053.

16. Movahed MR, Saito Y: Obesity is associated withleft atrial enlargement, E/A reversal and left ven-tricular hypertrophy. Exp Clin Cardiol 2008;13:89–91.

17. Tumuklu MM, Etikan I, Kisacik B, et al: Effect ofobesity on left ventricular structure and myocardialsystolic function: Assessment by tissue Doppler imag-ing and strain/strain rate imaging. Echocardiography2007;24:802–809.

18. Gardin JM, Constantine G, Davis K, et al: Aor-tic valvular regurgitation: Prevalence and clinicalcharacteristics in a predominantly obese adult pop-ulation not taking anorexigens. Echocardiography2006;23:569–576.

132 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 26, No. 2, 2009