aClinica Medica and Department of Health Science, University of Milano-Bicocca, Monza
bIstituto Auxologico Italiano IRCCS
cDepartment of Clinical Sciences and Community Health, University of Milano and Fondazione Policlinico di Milano
dIstituto di Ricerca a Carattere Scientifico IRCCS Multimedica, Sesto San Giovanni, Milan, Italy
Correspondence to Professor Cesare Cuspidi, MD, Istituto Auxologico Italiano, Clinical Research Unit, Viale della Resistenza 23, 20036 Meda, Italy. Tel: +39 0362/772433; fax: +39 0362/772416; e-mail: email@example.com
In this issue of the Journal, Maunganidze et al. report the results of a study aimed at assessing the diagnostic accuracy of electrocardiographic (ECG-graphic) criteria for the screening of left ventricular hypertrophy (LVH) in a community sample of African ancestry with a high prevalence of obesity, echocardiographic left ventricular (LV) mass being the reference standard. Before addressing the details of the study, available evidence on this topic should be considered.
Detection of LVH has long been a major task in clinical electrocardiography, particularly in recent decades due to the increasing evidence that LVH is a powerful marker of cardiovascular morbidity/mortality in the general population as well as in different clinical settings and that regression of ECG-graphic LVH induced by treatment results in improved cardiovascular prognosis [2,3]. A recent analysis in patients with isolated systolic hypertension recruited in the Losartan Intervention For Event reduction in hypertension study showed that regression of time-varying Cornell product was associated with a significant reduction in the risk of new-onset heart failure (21%) and in the combined endpoint heart failure or death (17%) .
LVH is a complex cardiac phenotype resulting from the response of myocyte and nonmyocyte components to mechanical and hormonal stimuli. Although the mechanisms underlying this process remain incompletely understood, a solid body of evidence from experimental and clinical studies indicates that hemodynamic factors including pressure and volume overload play a key role in activating LV myocardial growth. Nonhemodynamic variables, including genetic, ethnic, demographic, environmental, neuro-hormonal and inflammatory factors contribute to modulate the hypertrophic response . In obese individuals a number of the above-mentioned factors contribute to development of LVH.
The prevalence of LVH not only depends on the demographic/clinical characteristics of the population study, but also on the methods and diagnostic criteria applied to define this marker of subclinical organ damage.
Body size accounts for up to 50% of adult LV dimensions; BMI has been shown to be directly related to LVH likelihood . The hemodynamic changes associated with obesity, namely increased stroke volume and cardiac output, tend to modify cardiac structure and function and cause LVH, systolic/diastolic dysfunction and left atrial enlargement. Studies in the general population and in hypertensive cohorts show that LVH prevalence is directly related to BMI and waist circumference. Soteriades et al. conducting a prospective analysis on the association between BMI and incidence of ECG-graphic abnormalities in 276 firefighters during a mean follow-up period of 4 years found that obese patients were about two times more likely to develop ECG-LVH than their normal-weight counterparts . In a study by Masaidi et al. evaluating the relationship between obesity and cardiac hypertrophy in 330 patients with systemic hypertension free from overt cardiovascular and pulmonary diseases, the prevalence of biventricular hypertrophy (i.e. combined LVH and right ventricular hypertrophy as assessed by echocardiography) in obese individuals was 1.5-fold and 4.5-fold higher compared with their overweight and lean counterparts, respectively. Despite a stepwise increase in prevalence rates of isolated LVH and biventricular hypertrophy from lean to overweight and obese hypertensives at the echocardiographic examination, the Sokolow–Lyon voltage and the prevalence of ECG-LVH showed the opposite trend. In 4468 essential hypertensives included in the Evaluation of Target Organ Damage in Hypertension study, LVH defined by LV mass indexed either to BSA (≥116 g/m2 in men, ≥96 g/m2 in women) and height (≥49 g/m2.7 in men, ≥45 g/m2.7 in women) increased from lean patients (BMI <20 kg/m2) to the obese ones (BMI ≥30 kg/m2) by approximately 1.6-fold when defined by LV mass/BSA, by 3.2-fold when defined by LV mass/height2.7 and by 3.3-fold when defined by both sets of criteria .
Obesity is frequently associated with comorbidities such as diabetes mellitus, metabolic syndrome, hypertension carrying a high risk of LVH. Detection of this marker of subclinical cardiac damage, however, is quite challenging in obese patients. Standard 12-lead ECG, the first-line method for diagnosing LVH , owing to its limited cost, large availability and good reproducibility, has a lower sensitivity in obese patients compared with lean and normal-weight counterparts, due to the increased amount of interposed adipose tissue that reduces QRS voltage in precordial leads. Reduced voltages also in peripheral limb leads have been reported in severe obese patients . More than 30 ECG-graphic indexes for LVH diagnosis have been proposed during the last six decades . These criteria have shown varying performance characteristics in identifying cardiac hypertrophy, particularly in relation to age, ethnicity, sex and body size. Some of the proposed criteria have a limited application, but others, mostly based on the measurement of QRS voltages, are commonly used for clinical and research purposes. In a recent review by our group aimed at investigating the prevalence of ECG-graphic LVH in a pooled population of 40 440 hypertensive patients the Sokolow–Lyon index was the prevalent LVH diagnostic criterion (20 studies), followed by Cornell voltage (13 studies) and Cornell voltage-duration product (six studies) . Multiple ECG criteria for LVH (including those previously mentioned) have been studied in the attempt to determine their accuracy in obese individuals. In the vast majority of studies, echocardiographic LV mass was the reference standard, MRI being less frequently used. In their pioneering report Nath et al. tested the sensitivity and specificity of 10 ECG-graphic criteria against echocardiographic parameters for LVH (i.e. LV internal dimension, wall thickness and LV mass) in 65 normotensive obese individuals. The sensitivity of ECG in detecting LV dilatation, LV wall hypertrophy and increased LV mass was poor, ranging from 0–13, 0–20 and 0–12%, respectively. ECG-graphic criteria only partially (i.e. Cornell voltage or product) or totally independent of precordial voltages (i.e. amplitude of QRS in aVL) have been proposed to perform better than those relying on precordial voltages (i.e. Sokolow–Lyon) in the diagnosis of LVH, although this view is not widely supported.
In a study assessing the performance of 11 ECG criteria in 95 severely obese patients (BMI ≥40 kg/m2) undergoing elective bariatric surgery, none of the voltage-based criteria were of value for LVH diagnosis; also the Cornell product criterion showed a poor sensitivity . Overall, these data indicate that ECG criteria based on precordial voltages have a poor accuracy in obese patients. Although alternative criteria such as Cornell voltage or Cornell voltage product are recommended for LVH diagnosis in this setting, recent evidence supports the view that the performance of such indexes is largely unsatisfactory. Vernojj et al. found that in the 137 patients with abdominal obesity enrolled in the Second Manifestation of Arterial Diseases cohort the Cornell voltage and Sokolow–Lyon criteria exhibited a similar sensitivity (18 versus 13%) in detecting LVH as assessed by MRI.
Maunganidze et al. in their study examined the performance of eight ECG criteria (i.e. five based on QRS voltage: Sokolow–Lyon, Cornell, RaVL, Gubner–Ungerleider, Lewis, and three based on QRS voltage-duration product: Sokolow–Lyon, Cornell and Gubner–Ungerleider product) in detecting LVH, as defined by the nongender-specific echocardiographic partition value of 51 g/h2.7, in a community sample of 661 patients of African ancestry (65% women, 43% obese, 43% hypertensives and 25% diabetics) characterized by a 20% prevalence of increased LV mass index. The principal findings of the study can be summarized as follows: neither Sokolow–Lyon voltage nor product criteria showed significant correlations with LV mass index. The amplitude of R wave in aVL lead showed the closest correlation with LV mass index, followed by the Lewis and Gubner–Ungerleider voltage; the strength of the correlation between RaVL and criteria relying on limb lead recordings, such as Cornell voltage and product, was markedly reduced in obese patients compared with the nonobese counterparts, resulting in a poor accuracy in the diagnosis of LVH, as shown by receiver-operating characteristic curves analysis. These results are in keeping with previous findings obtained by Dominiek-Karlowicz et al. in a small white cohort of obese patients (BMI 49 ± 9 kg/m2) with a similar sex distribution (85% women) and extend the observation to a population of African ancestry with a less severe degree of obesity (BMI 37 ± 5 kg/m2). Among the factors responsible for the poor diagnostic performance of the ECG criteria tested, in particular the poor specificity of limb lead voltage criteria, Maunganidze et al. included ethnic-related alterations in skin conductivity. This hypothesis should be taken with caution as a similar reduction in specificity has been reported in non-African ethnic groups .
A few aspects and limitations of this interesting work should be mentioned. First, the results of the study should be limited to the female sex, as men represented less than 15% of the obese sample analyzed. It has been shown that sex may affect the ability of ECG criteria to detect LVH . Adult women have a slightly lower upper limit of QRS voltage than men. This is due to sex-related differences in body size and in LV compensatory responses to hemodynamic stimuli as well as in the distance between the precordial electrodes and the cardiac mass. Second, ECG-graphic threshold criteria for LVH in the present analysis were derived from the upper 95% confidence intervals in a population sample of 140 normotensive, nonobese individuals; ECG-graphic cut-offs derived from health obese individuals without echocardiographic evidence of LVH might have produced different results. Third, echocardiographic LVH prevalence among the 284 obese patients examined in the present study was approximately 70%. Unfortunately, no data are provided about LVH degrees (i.e. mild, moderate or severe hypertrophy) and, more importantly, no separate ROC curves have been performed according to LVH severity in order to examine the sensitivity and specificity of ECG criteria by the level of increased LV mass. Fourth, this study clearly indicates that echocardiography performs better than ECG in detecting obesity-related LVH. In obese patients, however, the ultrasonographic view of LV wall interfaces is hampered by the increased chest wall impendence; this major limitation is testified by the exclusion of as many as 151 patients (92% women, mean BMI 37 ± 8 kg/m2) from the final analysis of the present study because of poor quality echocardiograms. Finally, for proper interpretation of the present findings it should be stressed that ECG, at difference from the ‘gold standard’ reference tools such as echocardiography, cardiac tomography or MRI, does not provide a direct measurement of cardiac wall dimensions but only displays the alterations of cardio-electric properties associated with the hypertrophic process.
In conclusion, identification of LVH by ECG in obese patients, particularly in morbid obese, may appear a ‘mission impossible’, this technique, however, remains a first-line diagnostic tool in this setting as it reveals a number of abnormalities of paramount clinical relevance frequently associated with obesity such as alterations of P wave, QRS and T wave, prolongation of QRS and QT duration, supraventricular and ventricular arrhythmias . Obesity is a growing epidemic condition strongly associated to LVH and future studies for improving ECG detection of this cardiac phenotype are greatly needed.
Conflicts of interest
The authors report no conflicts of interest.
The authors alone are responsible for the content and writing of the article.
1. Maunganidze F, Woodiwiss AJ, Libhader CD, Maseko MJ, Majane OHI, Norton GR. Obesity markedly attenuates the validity and performance of all electrocardiographic criteria for left ventricular hypertrophy detection in a group of black African ancestry. J Hypertens
2. Verdecchia P, Staessen JA, Angeli F, de Simone G, Achilli A, Ganau A, et al. Usual versus tight control of systolic blood pressure in nondiabetic patients with hypertension (Cardio-Sis): an open-label randomised trial. Lancet
3. Lonn E, Mathew J, Pogue J, Johnstone D, Danisa K, Bosh J, et al. Relationship between electrocardiographic left ventricular hypertrophy to mortality and cardiovascular morbidity in high risk patients. Eur J Cardiovasc Prev Rehabil
4. Larstop AC, Okin PM, Devereux RB, Olsen MH, Ibsen H, Dahlof B, et al. Regression of ECG-LVH is associated with lower risk of new-onset heart failure and mortality in patients with isolated systolic hypertension: the LIFE study. Am J Hypertens
5. Schmieder RE. The role of nonhaemodynamic factors in the genesis of LVH. Nephron Dial Transplant
6. Dewey FE, Rosenthal D, Murphy DJ, Froelicher VF, Ashley EA. Does size matter? Clinical applications of scaling cardiac size and function for body size. Circulation
7. Soteriades ES, Targino MC, Talias MA, Hauser R, Kawachi I, Christiani DC, Kales SN. Obesity and risk of LVH and ECG abnormalities in US firefighters. J Occup Environ Med
8. Masaidi M, Cuspidi C, Negri F, Giudici V, Sala C, Zanchetti A, Mancia G. Left and right ventricular structural changes in obese hypertensives. Blood Press
9. Cuspidi C, Giudici V, Lonati L, Sala C, Valerio C, Mancia G. Left ventricular hypertrophy and body size in essential hypertension. Blood Press
10. 2007 Guidelines for the Management of Arterial Hypertension. The Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). J Hypertens
11. Shirani J, Berezowoski K, Roberts WC. Quantitative measurement of normal and excessive (cor adiposum) subepicardial adipose tissue, its clinical significance, and its effect on electrocardiographic QRS voltages. Am J Cardiol
12. Hancock FW, Deal BJ, Mirvis DM, Okin P, Kligfield P, Gettes LS. AHA/ACCF/HRS Recommendations for the standardization and interpretation of the electrocardiogram. Part V: electrocardiogram changes associated with chamber hypertrophy a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology: the American College of Cardiology Foundation and the Heart Rhythm Society. J Am Coll Cardiol
13. Cuspidi C, Rescaldani M, Sala C, Negri F, Grassi G, Mancia G. Prevalence of electrocardiographic left ventricular hypertrophy in human hypertension: an updated review. J Hypertens
14. Nath A, Alpert MA, Terry BE, Kelly DL. Sensitivity and specificity of electrocardiographic criteria for left and right ventricular hypertrophy in morbid obesity. Am J Cardiol
15. Dominiek-Karlowicz J, Lichodziejewska B, Lisik W, Ciurzynski M, Bienias P, Chmura A, Pruszczyk P. Electrocardiographic criteria of left ventricular hypertrophy in patients with morbid obesity. Ann Noninvasive Electrocardiol
16. Fraley MA, Birchem JA, Senkottayan N, Alpert MA. Obesity and the electrocardiogram. Obes Rev
17. Vernojj JWP, Cramer MJM, Visseren FLJ, Korndewal MJ, Bots ML, Meijs MFL, et al. Relation between abdominal obesity, insulin resistance and left ventricular hypertrophy diagnosed by electrocardiogram and magnetic resonance imaging in hypertensive patients. Am J Cardiol
18. Alfakih K, Walters K, Jones T, Ridgway J, Hall AS, Sivanatham. New gender-specific partition values for ECG criteria of left ventricular hypertrophy. Recalibration against cardiac MRI. Hypertension