Abstract
Electrocardiographic left ventricular hypertrophy (ECG‐LVH) is associated with both cardiovascular and all‐cause mortality. Obesity attenuates the sensitivity of several ECG‐LVH criteria, so body mass index (BMI) adjusted criteria have been developed. However, the prognostic significance of BMI‐adjusted ECG‐LVH criteria is not known. This analysis included 7812 participants (59.8 ± 13.4 years, 53% women, 50% non‐Hispanic‐whites) from the Third National Health and Nutrition Examination Survey. The Cornell criteria (R in aVL + S in V3 ≥ 2800 µV in men or ≥2200 µV in women) and Sokolow‐Lyon criteria (S in V1 + R in V5 or R in V6 ≥ 3500 µV) criteria were used for LVH. To account for the effects of obesity, the BMI‐adjusted Cornell criteria (product of R in aVL + S in V3 and BMI > 60 400 µV kg m−2) and the BMI‐adjusted Sokolow‐Lyon criteria (add 400 µV if overweight, add 800 µV if obese) were used. Compared to traditional ECG‐LVH criteria, more participants met criteria for ECG‐LVH with BMI‐adjusted Cornell voltage (9.9% vs 2.9%) and BMI‐adjusted Sokolow‐Lyon (13.1% vs 6.4%) criteria. In multivariable‐adjusted Cox proportional hazards models, the BMI‐adjusted Sokolow‐Lyon criteria performed no better than traditional criteria (HR 1.18, 95% CI 1.06‐1.32 for all‐cause, HR 1.38, 95% CI 1.17‐1.62 for cardiovascular mortality) and the BMI‐adjusted Cornell voltage criteria attenuated the association with all‐cause (HR 1.16, 95% CI 1.03‐1.32) and cardiovascular mortality (HR 1.34, 95% CI 1.13‐1.60). Despite potential improvements in the detection of LVH using BMI‐adjusted ECG‐LVH criteria, adjusting for BMI may result in the loss of prognostic information.
Keywords: BMI‐adjusted criteria, electrocardiography, left ventricular hypertrophy, obesity
1. INTRODUCTION
Left ventricular hypertrophy (LVH) is associated with increased risk of cardiovascular disease (CVD) morbidity and mortality, as well as all‐cause mortality. 1 , 2 , 3 Electrocardiography (ECG) is the recommended initial screening modality for LVH per both European and North American guidelines. 4 , 5 , 6 Obesity increases hemodynamic stress on cardiac muscle, which can lead to LVH, and is also independently associated with both all‐cause mortality and CVD mortality. 7 However, the most commonly used criteria for ECG‐LVH 8 are less sensitive in patients with obesity, 9 , 10 , 11 as body fat impedes the conduction of the electrical energy to the precordial leads. 11 , 12 , 13 In response to this, novel criteria for ECG‐LVH that account for obesity have been developed and improve the sensitivity in obese individuals, nearing the sensitivity of traditional ECG‐LVH criteria in non‐obese individuals. 10 , 14 , 15
While traditional ECG‐LVH criteria have been validated prognostically, 16 , 17 , 18 the utility of the newly developed body mass index (BMI)‐adjusted ECG‐LVH criteria has not been adequately explored. Therefore, we compared the prognostic significance of traditional ECG‐LVH criteria and BMI‐adjusted ECG‐LVH criteria among participants in the Third National Health and Nutrition Examination Survey (NHANES‐III) 19 with all‐cause and CVD mortality as outcomes of interest. In a secondary analysis, we explored the consistency of these relationships among sex, race, and obesity subgroups.
2. METHODS
2.1. Study design and population
NHANES‐III includes a nationally representative population sample to extrapolate disease prevalence and the general health status of the United States population. Baseline information was collected from 1988‐1994 through home interviews including factors such as demographics (age, sex, race/ethnicity), medications prescribed, past medical history, and behavioral data. All NHANES‐III participants with baseline ECG of adequate quality for interpretation, baseline height and weight measurements, and some follow‐up were included, yielding 7812 eligible participants.
2.2. Left ventricular hypertrophy
A 12‐lead ECG was obtained by a trained technician during a visit to a mobile examination center. Resting ECG was recorded on a Marquette MAC 12 system (Marquette Medical Systems, Milwaukee, Wisconsin), and data were automatically processed at a central core laboratory. Four ECG‐LVH criteria were used in the present analysis:
The original Cornell criteria 20 —R in aVL + S in V3 ≥ 2800 µV in men or ≥2200 µV in women
The original Sokolow‐Lyon criteria 21 —S in V1 + R in V5 or R in V6 ≥ 3500 µV
The BMI‐adjusted Cornell criteria 14 —product of R in aVL + S in V3 and BMI> 60 400 µV kg m−2
The BMI‐adjusted Sokolow‐Lyon criteria 15 —S in V1 + R in V5 or R in V6 ≥ 3500 µV (add 400 µV if overweight, add 800 µV if obese)
2.3. Other variables
Blood pressure values were determined by taking the mean of three home measurements and three mobile center measurements, using mercury sphygmomanometers. Hypertension was defined as systolic blood pressure (SBP) ≥140 mm Hg or diastolic blood pressure (DBP) ≥90 mm Hg. Diabetes was defined as fasting plasma glucose ≥126 mg/dL (>7 mmol/L), hemoglobin A1c values ≥ 6.5%, or previous use of antidiabetic medications. Dyslipidemia was defined as any of the following: total cholesterol >240 mg/dL (>6.21 mmol/L); low‐density lipoprotein (LDL) cholesterol >160 mg/dL (>4.14 mmol/L); high‐density lipoprotein (HDL) cholesterol ≤ 40 mg/dL (<1.03 mmol/L); or by the use of cholesterol‐lowering medications. Participant BMI was calculated as the weight in kilograms divided by the height in meters squared. Obesity was defined as a BMI ≥ 30 kg/m2.
2.4. Outcome variables
Through December 31, 2006, the mortality data for NHANES‐III participants became publicly available. A matching algorithm based on 12 identifiers was employed to correlate participants with corresponding mortality data that was reported in the National Death Index. Such identifying features included social security number, gender, and date of birth. Moreover, follow‐up time was determined based on the interval beginning with the start of the NHANES‐III examination and ending on one of the following three options: date of death, date of censoring, or December 31, 2006. The end date were determined based on which ever date preceded the others. The end points of all‐cause and cardiovascular mortality were investigated and analyzed using data from the NHANES‐III Linked Mortality File. Unmatched participants with a death record were considered to be alive throughout the duration of the follow‐up period.
2.5. Statistical methods
Baseline characteristics of the study population stratified by those who did and did not meet each of the LVH criteria were calculated. Results are reported as mean ± standard deviation for continuous variables and percentage for categorical variables. Cox proportional hazards modeling was used to compare the risk of all‐cause and CVD mortality in those who did and did not meet each criteria for LVH, generating hazard ratios (HRs) and 95% confidence intervals (95% CI). Models were adjusted for covariates previously demonstrated to be associated with ECG‐LVH or risk of death, including age, sex, race/ethnicity, diabetes, hypertension, dyslipidemia, coronary artery disease, congestive heart failure, current smoking, serum creatinine, and major ECG abnormalities. We then examined the consistency of the observed relationships between LVH and all‐cause and CVD mortality when participants were stratified by race/ethnicity, sex, and obesity using formal tests of interaction by including interaction terms in our model. Two‐sided P‐values below .05 were considered to be statistically significant. All statistical analyses were conducted at Wake Forest University School of Medicine using SAS version 9.4.
3. RESULTS
A total of 7812 participants (59.8 ± 13.4 years, 52.6% women, 49.5% non‐Hispanic whites, 23.4% non‐Hispanic black, and 22.8% Mexican‐American) were included in the study analysis. Characteristics of the study participants are provided in Table 1, stratified by those who did or did not meet criteria for ECG‐LVH by each of the four criteria used. More participants met criteria for ECG‐LVH using the BMI‐adjusted Cornell voltage criteria than the traditional Cornell voltage criteria (9.9% vs 2.9%). Similarly, more participants met criteria for ECG‐LVH using the BMI‐adjusted Sokolow‐Lyon criteria than the traditional Sokolow‐Lyon criteria (13.1% vs 6.4%). Of the 7812 participants, 7179 (91.9%) were classified identically by both Cornell criteria, with 6993 (89.5%) having no LVH and 186 (2.4%) having LVH. Discrepant classifications were seen in 633 (8.1%), with the use of the BMI‐adjusted Cornell voltage criteria reclassifying 591 (7.6%) as having LVH and 42 (0.5%) as not having LVH. Comparing the traditional and BMI‐adjusted Sokolow‐Lyon criteria, 7285 (93.3%) were classified identically, while 527 (6.7%) were reclassified as having LVH. Importantly, obese participants were more likely than non‐obese participants to be reclassified from no LVH to LVH with the use of the BMI‐adjusted criteria.
TABLE 1.
Characteristics of NHANES study participants
| Cornell Voltage | BMI‐Adjusted Cornell Voltage | Sokolow‐Lyon | BMI‐Adjusted Sokolow‐Lyon | |||||
|---|---|---|---|---|---|---|---|---|
|
LVH n = 228 (2.9%) |
No LVH n = 7584 (97.1%) |
LVH n = 777 (9.9%) |
No LVH n = 7035 (90.1%) |
LVH n = 497 (6.4%) |
No LVH n = 7315 (93.6%) |
LVH n = 1024 (13.1%) |
No LVH n = 6788 (86.9%) |
|
| Age (y) | 59.6 ± 13.4 | 66.6 ± 12.6 | 60.3 ± 13.0 | 59.8 ± 13.5 | 61.3 ± 13.9 | 59.7 ± 13.4 | 60.5 ± 13.4 | 59.7 ± 13.4 |
| Women | 71.5% | 52.1% | 43.8% | 53.7% | 43.3% | 53.3% | 49.7% | 53.1% |
| Race | ||||||||
| Non‐Hispanic White | 37.3% | 49.9% | 35.4% | 51.1% | 32.6% | 50.7% | 31.3% | 52.3% |
| Non‐Hispanic Black | 32.9% | 23.2% | 34.0% | 22.3% | 53.1% | 21.4% | 50.7% | 19.3% |
| Mexican‐American | 25.9% | 22.8% | 28.2% | 22.3% | 10.7% | 23.7% | 15.0% | 24.1% |
| Others | 4.4% | 2.8% | 2.7% | 2.8% | 2.0% | 2.9% | 1.5% | 3.0% |
| Smoking (current) | 13.2% | 22.8% | 15.7% | 23.3% | 29.0% | 22.1% | 23.9% | 22.3% |
| Diabetes Mellitus | 22.8% | 11.3% | 22.7% | 10.4% | 10.9% | 11.7% | 13.7% | 11.3% |
| Hypertension | 63.2% | 35.3% | 59.8% | 33.5% | 53.9% | 34.9% | 53.9% | 33.5% |
| Coronary artery disease | 14.0% | 6.2% | 10.9% | 5.9% | 8.4% | 6.3% | 8.4% | 6.1% |
| Heart failure | 12.7% | 4.4% | 8.8% | 4.2% | 6.2% | 4.6% | 6.0% | 4.5% |
| Dyslipidemia | 20.6% | 24.8% | 24.5% | 24.7% | 19.1% | 25.0% | 21.3% | 25.1% |
| Body Mass Index | 29.2 ± 6.5 | 27.6 ± 5.5 | 33.6 ± 6.4 | 27.0 ± 5.0 | 26.0 ± 5.0 | 27.7 ± 5.5 | 28.8 ± 5.6 | 27.5 ± 5.5 |
| Serum Creatinine | 1.21 ± 0.84 | 1.12 ± 0.43 | 1.18 ± 0.63 | 1.11 ± 0.43 | 1.22 ± 0.71 | 1.11 ± 0.43 | 1.21 ± 0.77 | 1.11 ± 0.38 |
| Obesity | 25.9% | 18.7% | 34.6% | 17.2% | 12.1% | 19.4% | 26.3% | 17.8% |
| Major ECG Abnormalities | 44.3% | 11.1% | 28.4% | 10.3% | 31.0% | 10.8% | 26.0% | 10.0% |
Continuous variables described as mean ± standard deviation. Categorical variables described as frequency (percentage). Diabetes was defined as fasting plasma glucose ≥ 126 mg/dL, hemoglobin A1c values ≥ 6.5%, or history of glucose‐lowering medications. Hypertension was defined as systolic blood pressure ≥ 140 mm Hg, diastolic blood pressure ≥ 90 mm Hg, or use of blood pressure‐lowering medications. Obesity was defined as body mass index ≥ 30 kg m−2. Dyslipidemia was defined as any of the following: total cholesterol >240 mg/dL, low‐density lipoprotein cholesterol >160 mg/dL, high‐density lipoprotein cholesterol ≤40 mg/dL, or the use of cholesterol‐lowering medications.
The relationship between meeting each ECG‐LVH criteria and the risk of all‐cause and CVD mortality is provided in Table 2. Participants meeting each ECG‐LVH criteria had increased risks of all‐cause and CVD mortality. However, for Cornell voltage, the use of the BMI‐adjusted criteria attenuated the relationship between ECG‐LVH and both all‐cause and CVD mortality. For Sokolow‐Lyon, the use of the BMI‐adjusted criteria had minimal effect on the association between ECG‐LVH and outcomes.
TABLE 2.
Prognostic significance of BMI‐adjusted ECG‐LVH criteria in comparison with traditional ECG‐LVH criteria
| N = 7812 | ECG‐LVH | All‐cause mortality HR (95% CI) | Cardiovascular mortality HR (95% CI) |
|---|---|---|---|
| Cornell voltage | Original | 1.41 (1.17‐1.69) | 1.57 (1.21‐2.03) |
| BMI‐adjusted | 1.16 (1.03‐1.32) | 1.34 (1.13‐1.60) | |
| Sokolow‐Lyon | Original | 1.22 (1.06‐1.41) | 1.34 (1.09‐1.65) |
| BMI‐Adjusted | 1.18 (1.06‐1.32) | 1.38 (1.17‐1.62) |
Model is adjusted for age, sex, race, diabetes, hypertension, dyslipidemia, coronary artery disease, congestive heart failure, current smoking, serum creatinine, and major ECG abnormalities. The reference group is those participants not meeting ECG‐LVH criteria.
Abbreviations: CI, confidence interval; ECG‐LVH, electrocardiographic left ventricular hypertrophy; HR, hazard ratio.
Next, we compared the relationships between meeting each ECG‐LVH criteria and the risk of all‐cause and CVD mortality when participants are stratified by race/ethnicity, sex, and prevalent obesity. Comparing white to non‐white individuals, the associations between ECG‐LVH and the risk of adverse outcomes are generally greater in both magnitude and statistical significance in whites (Table 3), though P‐values for interaction did not meet criteria for statistical significance (.41‐.96). However, the trend of the BMI‐adjusted criteria providing decreased prognostic significance is generally conserved across races (interaction P‐values .15‐.74). Similarly, utilization of the BMI‐adjusted criteria generally weakens the association between ECG‐LVH and the risk of mortality in both men and women (Table 4; interaction P‐values .23‐.97). When comparing those with and without obesity, the association between ECG‐LVH and the risk of mortality by the traditional criteria is generally stronger in the non‐obese (Table 5). Utilization of the BMI‐adjusted criteria has more effect on weakening the association between ECG‐LVH and the risk of mortality among the obese than the non‐obese.
TABLE 3.
Prognostic significance of BMI‐adjusted ECG‐LVH criteria in comparison with traditional ECG‐LVH criteria by race
| ECG‐LVH | All‐cause mortality | Cardiovascular mortality | |||
|---|---|---|---|---|---|
| HR (95% CI) | Interaction P‐value | HR (95% CI) | Interaction P‐value | ||
| Whites (n = 5699) | |||||
| Cornell voltage | Original | 1.40 (1.22‐1.75) | .70 | 1.63 (1.21‐2.20) | .56 |
| BMI‐adjusted | 1.18 (1.01‐1.36) | .96 | 1.39 (1.13‐1.71) | .41 | |
| Sokolow‐Lyon | Original | 1.26 (1.05‐1.53) | .47 | 1.30 (0.99‐1.70) | .69 |
| BMI‐adjusted | 1.23 (1.07‐1.42) | .95 | 1.45 (1.19‐1.76) | .54 | |
| Non‐Whites (n = 2113) | |||||
| Cornell voltage | Original | 1.41 (1.00‐2.00) | .70 | 1.37 (0.81‐2.29) | .56 |
| BMI‐adjusted | 1.15 (0.92‐1.44) | .96 | 1.21 (0.86‐1.69) | .41 | |
| Sokolow‐Lyon | Original | 1.12 (0.90‐1.40) | .47 | 1.28 (0.91‐1.79) | .69 |
| BMI‐adjusted | 1.06 (0.88‐1.28) | .95 | 1.13 (0.85‐1.50) | .54 | |
Model is adjusted for age, sex, diabetes, hypertension, dyslipidemia, coronary artery disease, congestive heart failure, current smoking, serum creatinine, and major ECG abnormalities. The reference group is those participants not meeting ECG‐LVH criteria.
Abbreviations: CI, confidence interval; ECG‐LVH, electrocardiographic left ventricular hypertrophy; HR, hazard ratio.
TABLE 4.
Prognostic significance of BMI‐adjusted ECG‐LVH criteria in comparison with traditional ECG‐LVH criteria by sex
| ECG‐LVH | All‐Cause Mortality | Cardiovascular mortality | |||
|---|---|---|---|---|---|
| HR (95% CI) | Interaction P‐value | HR (95% CI) | Interaction P‐value | ||
| Men (n = 3696) | |||||
| Cornell Voltage | Original | 1.48 (1.07‐2.05) | .59 | 1.75 (1.15‐2.68) | .23 |
| BMI‐adjusted | 1.16 (0.99‐1.36) | .74 | 1.40 (1.12‐1.76) | .15 | |
| Sokolow‐Lyon | Original | 1.12 (0.93‐1.36) | .32 | 1.22 (0.92‐1.62) | .56 |
| BMI‐adjusted | 1.08 (0.92‐1.26) | .49 | 1.28 (1.02‐1.61) | .62 | |
| Women (n = 4116) | |||||
| Cornell Voltage | Original | 1.34 (1.07‐1.69) | .59 | 1.47 (1.06‐2.04) | .23 |
| BMI‐adjusted | 1.21 (0.99‐1.46) | .74 | 1.31 (0.99‐1.74) | .15 | |
| Sokolow‐Lyon | Original | 1.39 (1.12‐1.73) | .32 | 1.53 (1.12‐2.09) | .56 |
| BMI‐adjusted | 1.30 (1.10‐1.53) | .49 | 1.46 (1.16‐1.85) | .62 | |
Model is adjusted for age, race, diabetes, hypertension, dyslipidemia, coronary artery disease, congestive heart failure, current smoking, serum creatinine, and major ECG abnormalities. The reference group is those participants not meeting ECG‐LVH criteria.
Abbreviations: CI, confidence interval; ECG‐LVH, electrocardiographic left ventricular hypertrophy; HR, hazard ratio.
TABLE 5.
Prognostic significance of BMI‐adjusted ECG‐LVH criteria in comparison with traditional ECG‐LVH criteria by BMI
| ECG‐LVH | All‐cause mortality | Cardiovascular mortality | |||
|---|---|---|---|---|---|
| HR (95% CI) | Interaction P‐value | HR (95% CI) | Interaction P‐value | ||
| Obese (n = 1477) | |||||
| Cornell voltage | Original | 1.32 (0.88‐1.98) | .70 | 1.15 (0.64‐2.07) | .97 |
| BMI‐adjusted | 1.23 (0.97‐1.56) | .23 | 1.25 (0.89‐1.76) | .67 | |
| Sokolow‐Lyon | Original | 1.09 (0.68‐1.74) | .63 | 1.42 (0.67‐2.30) | .59 |
| BMI‐adjusted | 1.03 (0.80‐1.33) | .67 | 1.00 (0.69‐1.46) | .54 | |
| Non‐obese (n = 6335) | |||||
| Cornell voltage | Original | 1.40 (1.13‐1.73) | .70 | 1.61 (1.21‐2.15) | .97 |
| BMI‐adjusted | 1.13 (0.98‐1.32) | .23 | 1.33 (1.08‐1.64) | .67 | |
| Sokolow‐Lyon | Original | 1.24 (1.06‐1.44) | .63 | 1.36 (1.09‐1.70) | .59 |
| BMI‐adjusted | 1.22 (1.07‐1.38) | .67 | 1.45 (1.21‐1.74) | .54 | |
Model is adjusted for age, sex, race, diabetes, hypertension, dyslipidemia, coronary artery disease, congestive heart failure, current smoking, serum creatinine, and major ECG abnormalities. The reference group is those participants not meeting ECG‐LVH criteria.
Abbreviations: CI, confidence interval; ECG‐LVH, electrocardiographic left ventricular hypertrophy; HR, hazard ratio.
4. DISCUSSION
In this analysis from NHANES‐III, we found that the BMI‐adjusted criteria enhance the detection of ECG‐LVH compared with traditional ECG‐LVH criteria, but do not improve prediction of all‐cause and CVD mortality. In fact, for the Cornell voltage criteria, utilizing the BMI‐adjusted formula actually attenuates the association with both all‐cause and cardiovascular mortality. These findings were consistent in subgroups stratified by sex, race, and prevalent obesity.
4.1. Results in the context of the literature
Prior studies have demonstrated that ECG‐LVH is a pathological process that is highly prevalent in obese patients, usually of the eccentric subtype, and strongly associated with subsequent clinical cardiovascular events. 22 ECG‐LVH is an independent risk factor for CVD and all‐cause mortality and is also associated with a higher risk of sudden cardiac death. 23 , 24 In obese patients, adjusting Sokolow‐Lyon voltage criteria by 4 mm and 8 mm in overweight and obese patients, respectively, has been shown to increase the sensitivity of the ECG‐LVH voltage criteria to values similar to that of the normal weight population without a significant decrease in specificity. 15 Adjusting Cornell voltage criteria for BMI showed similar results; 14 hence, both adjustment methods for BMI were utilized for the present study.
Compared to the detection of LVH by cardiac magnetic resonance (CMR), which is the gold standard for diagnosing anatomic LVH, 12 Sokolow‐Lyon and Cornell ECG‐LVH voltage criteria have limited sensitivity but high specificity. 15 Adjusting for BMI can improve the sensitivity of ECG criteria to detect LVH and optimize the use of widely available ECG in the primary care setting. In addition, echocardiography has higher sensitivity than ECG for detecting anatomic LVH. 25
In our study, the use of BMI‐adjusted ECG‐LVH criteria resulted in increased prevalence of LVH, compared to the traditional ECG‐LVH criteria. Identifying more patients with LVH is important, as successful lowering of blood pressure can lead to the regression of LVH and ultimately decrease CVD mortality and morbidity. 26 , 27 From this perspective, using BMI‐adjusted ECG‐LVH criteria in obese people would be useful since it will enhance detection of ECG‐LVH. However, focusing on the diagnostic utility of various criteria for ECG‐LVH in detecting anatomic LVH (as assessed by echocardiography or MRI) may not be the most important question, as there is mounting evidence that ECG‐LVH and anatomic LVH are distinct processes, both of which are clinically important. 28 Several studies have demonstrated that ECG‐LVH is independently predictive of the risk of death, even after accounting for the risk conferred by anatomic LVH as assessed by echocardiography. 29 , 30 In light of this, assessing the prognostic utility of ECG‐LVH criteria may be more important than their diagnostic utility.
ECG‐LVH is unique from anatomic LVH, 31 as detected by CMR and echocardiography, and abnormal ECG findings can precede underlying anatomical cardiac pathologies. ECG‐LVH is not only a reflection of increase in LV mass but also may be attributed to secondary increases in myocardial strain, intraventricular electrical conduction disturbance, or neurohumoral and/or biochemical changes. 28 , 32 This may explain the disconnect observed between the diagnostic and prognostic abilities of ECG‐LVH.
Previous literature, with regard to the prognostic utility of BMI‐adjusted ECG‐LVH criteria in a European population, reported an increased risk of cardiovascular events and all‐cause mortality with BMI‐adjusted Cornell criteria, but not with BMI‐adjusted Sokolow‐Lyon criteria. 33 Our study examined the prognostic utility of BMI‐adjusted ECG‐LVH in a larger, more diverse, and nationally representative study population. We showed that BMI‐adjusted Cornell and Sokolow‐Lyon voltage criteria are similar to traditional criteria for risk of all‐cause and CVD mortality and that adjusting for BMI using the ECG‐LVH criteria does not improve the prognostic significance for all‐cause and CVD mortality.
Our findings were overall similar in subgroups, with no evidence of effect measure modification in subgroups (interaction P‐value >.05) when stratified by sex, race, or prevalent obesity. We observed a non‐significant trend toward decreased prognostic utility of ECG‐LVH criteria in participants with prevalent obesity, in comparison with those without obesity. Both the magnitude and statistical significance of the associations between ECG‐LVH criteria and the risk of cardiovascular and all‐cause mortality were more pronounced in the non‐obese, though the lack of statistical significance may be explained by the relatively small sample size of participants with prevalent obesity. However, it should be noted that use of the BMI‐adjusted ECG‐LVH criteria did not improve prognostic utility in either those with or those without obesity, which emphasizes that the use of the BMI‐adjusted criteria is not helpful in refining prognosis in any subgroup.
It is worth noting that the use of weight‐based indexing may mask the prognostic significance of ECG‐LVH in obese patients, as we and others have observed. 34 The European Society of Cardiology hypertension guidelines recommend indexing echocardiographic left ventricular mass to height, rather than body surface area, in obese individuals. Similar indexing of ECG‐LVH to height should be considered for application in BMI ≥ 30 kg/m2 and studied for changes in the prognostic significance of BMI‐adjusted ECG‐LVH criteria.
Although ECG specificity for LVH is high, it is less specific in certain ethnic groups such as the black population. 35 , 36 In particular, ECG‐LVH criteria is less specific in black patients with hypertension who experience more complications than their white counterparts. Independent from vascular hemodynamic effect or vascular stiffness, left ventricular mass, and relative wall thickness are greater in black than white individuals. 37 In the current study, adjusting for BMI in the non‐white population yielded no significance in the prognostic value of the traditional and BMI‐adjusted Cornell and Sokolow‐Lyon criteria, likely secondary to their small sample size in our study. In contrast, our study showed a significant prognostic value for the BMI‐adjusted criteria for both all‐cause and CVD mortality in the white population.
We found no difference in women between the traditional and BMI‐adjusted Sokolow‐Lyon voltage criteria for risk of both all‐cause and CVD mortality. No significant prognostic value was detected when adjusting for BMI using the Cornell voltage criteria in women. In men, adjusting for BMI for both Sokolow‐Lyon and Cornell voltage criteria showed prognostic significance for CVD mortality but not all‐cause mortality. Previous studies showed that ECG‐LVH voltage criteria has higher sensitivity in men than in women. 12 Moreover, compared to women, men tend to exhibit more concentric LVH, 9 which is strongly associated with higher CVD mortality. 38 Current literature reports discrepancies in the significance of LVH reduction between men and women. 39 Gender differences in body size, obesity, and LVM cannot fully explain the gender differences in QRS duration and voltage readings. 40
4.2. Limitations and strengths
Our study should be interpreted in the context of its limitations. While we adjusted for covariates believed to be associated with the risk of death, the possibility of residual confounding remains. Our classification of participants by ECG‐LVH criteria was based on their ECG at enrollment and does not account for variability over time. The relatively small sample size of obese and non‐white participants may limit our ability to find a difference in risk should it exist. Despite these limitations, the use of a central ECG reading center maximizes internal validity and the use of a multiracial population increases the generalizability of our findings.
5. CONCLUSIONS
Compared with traditional Cornell and Sokolow‐Lyon criteria, the use of BMI‐adjusted Cornell and Sokolow‐Lyon criteria for LVH does not improve their prognostic utility, even in subgroups stratified by sex, race, and prevalent obesity. Though adjusting for BMI may improve the performance of ECG‐LVH criteria in diagnosing anatomic LVH, this comes at the expense of some loss of prognostic information. Further studies exploring alternative ECG‐LVH criteria that are indexed to height or ideal body surface area are needed.
AUTHOR CONTRIBUTION
Dr. Matthew Singleton is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. MS and ES conceived and designed the analysis. MS performed the analysis. MS and ES interpreted the results. HA and HL drafted the manuscript. MS and ES revised for critical intellectual content. All authors approved of the final manuscript for submission.
Disclosures
None.
Afify H, Lim Lee H, Soliman EZ, Singleton MJ. Prognostic significance of body mass index‐adjusted criteria for left ventricular hypertrophy. J Clin Hypertens. 2020;22:1476–1483. 10.1111/jch.13973
Afify and Lim Lee contributed equally and are co‐first authors.
The authors have no relationships with industry to report.
REFERENCES
- 1. Bayes‐Genis A, Guindo J, Vinolas X, et al. Cardiac arrhythmias and left ventricular hypertrophy in systemic hypertension and their influences on prognosis. Am J Cardiol. 1995;76:54D‐59D. [DOI] [PubMed] [Google Scholar]
- 2. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med. 1990;322:1561‐1566. [DOI] [PubMed] [Google Scholar]
- 3. Yildiz M, Oktay AA, Stewart MH, Milani RV, Ventura HO, Lavie CJ. Left ventricular hypertrophy and hypertension. Prog Cardiovasc Dis. 2020;63:10‐21. [DOI] [PubMed] [Google Scholar]
- 4. Williams B, Mancia G, Spiering W, et al. ESC/ESH Guidelines for the management of arterial hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension. J Hypertens. 2018;2018(36):1953‐2041. [DOI] [PubMed] [Google Scholar]
- 5. Goff DC Jr, Lloyd‐Jones DM, Bennett G, et al. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2014;63:2935‐2959. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Hancock EW, Deal BJ, Mirvis DM, et al. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part V: electrocardiogram changes associated with cardiac 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. Endorsed by the International Society for Computerized Electrocardiology. J Am Coll Cardiol. 2009;53:992‐1002. [DOI] [PubMed] [Google Scholar]
- 7. Csige I, Ujvarosy D, Szabo Z, et al. The impact of obesity on the cardiovascular system. J Diabetes Res. 2018;2018:3407306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Hancock EW, Deal BJ, Mirvis DM, et al. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part V: electrocardiogram changes associated with cardiac 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: endorsed by the International Society for Computerized Electrocardiology. Circulation. 2009;119:e251‐e261. [DOI] [PubMed] [Google Scholar]
- 9. Nomsawadi V, Krittayaphong R. Diagnostic performance of electrocardiographic criteria for left ventricular hypertrophy among various body mass index groups compared to diagnosis by cardiac magnetic resonance imaging. Ann Noninvasive Electrocardiol. 2019;24:e12635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Snelder SM, van de Poll SWE, de Groot‐de Laat LE, Kardys I, Zijlstra F, van Dalen BM. Optimized electrocardiographic criteria for the detection of left ventricular hypertrophy in obesity patients. Clin Cardiol. 2020;43(5):483‐490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Okin PM, Jern S, Devereux RB, Kjeldsen SE, Dahlof B, Group LS . Effect of obesity on electrocardiographic left ventricular hypertrophy in hypertensive patients: the losartan intervention for endpoint (LIFE) reduction in hypertension study. Hypertension. 2000;35:13‐18. [DOI] [PubMed] [Google Scholar]
- 12. Krittayaphong R, Nomsawadi V, Muenkaew M, Miniphan M, Yindeengam A, Udompunturak S. Accuracy of ECG criteria for the diagnosis of left ventricular hypertrophy: a comparison with magnetic resonance imaging. J Med Assoc Thai. 2013;96(Suppl 2):S124‐S132. [PubMed] [Google Scholar]
- 13. Levy D, Labib SB, Anderson KM, Christiansen JC, Kannel WB, Castelli WP. Determinants of sensitivity and specificity of electrocardiographic criteria for left ventricular hypertrophy. Circulation. 1990;81:815‐820. [DOI] [PubMed] [Google Scholar]
- 14. Angeli F, Verdecchia P, Iacobellis G, Reboldi G. Usefulness of QRS voltage correction by body mass index to improve electrocardiographic detection of left ventricular hypertrophy in patients with systemic hypertension. Am J Cardiol. 2014;114:427‐432. [DOI] [PubMed] [Google Scholar]
- 15. Rider OJ, Ntusi N, Bull SC, et al. Improvements in ECG accuracy for diagnosis of left ventricular hypertrophy in obesity. Heart. 2016;102:1566‐1572. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Wachtell K, Okin PM, Olsen MH, et al. Regression of electrocardiographic left ventricular hypertrophy during antihypertensive therapy and reduction in sudden cardiac death: the LIFE Study. Circulation. 2007;116:700‐705. [DOI] [PubMed] [Google Scholar]
- 17. Porthan K, Kentta T, Niiranen TJ, et al. ECG left ventricular hypertrophy as a risk predictor of sudden cardiac death. Int J Cardiol. 2019;276:125‐129. [DOI] [PubMed] [Google Scholar]
- 18. Iribarren C, Round AD, Lu M, Okin PM, McNulty EJ. Cohort Study of ECG left ventricular hypertrophy trajectories: ethnic disparities, associations with cardiovascular outcomes, and clinical utility. J Am Heart Assoc. 2017;6:4954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Plan and operation of the Third National Health and Nutrition Examination Survey, 1988–94. Series 1: programs and collection procedures. Vital Health Stat. 1994;1:1‐407. [PubMed] [Google Scholar]
- 20. Casale PN, Devereux RB, Alonso DR, Campo E, Kligfield P. Improved sex‐specific criteria of left ventricular hypertrophy for clinical and computer interpretation of electrocardiograms: validation with autopsy findings. Circulation. 1987;75:565‐572. [DOI] [PubMed] [Google Scholar]
- 21. Sokolow M, Lyon TP. The ventricular complex in left ventricular hypertrophy as obtained by unipolar precordial and limb leads. Am Heart J. 1949;37:161‐186. [DOI] [PubMed] [Google Scholar]
- 22. Cuspidi C, Rescaldani M, Sala C, Grassi G. Left‐ventricular hypertrophy and obesity: a systematic review and meta‐analysis of echocardiographic studies. J Hypertens. 2014;32:16‐25. [DOI] [PubMed] [Google Scholar]
- 23. Kannel WB, Gordon T, Castelli WP, Margolis JR. Electrocardiographic left ventricular hypertrophy and risk of coronary heart disease. The Framingham study. Ann Intern Med. 1970;72:813‐822. [DOI] [PubMed] [Google Scholar]
- 24. Kannel WB, Gordon T, Offutt D. Left ventricular hypertrophy by electrocardiogram. Prevalence, incidence, and mortality in the Framingham study. Ann Intern Med. 1969;71:89‐105. [DOI] [PubMed] [Google Scholar]
- 25. Woythaler JN, Singer SL, Kwan OL, et al. Accuracy of echocardiography versus electrocardiography in detecting left ventricular hypertrophy: comparison with postmortem mass measurements. J Am Coll Cardiol. 1983;2:305‐311. [DOI] [PubMed] [Google Scholar]
- 26. Lindholm LH, Ibsen H, Dahlof B, et al. Cardiovascular morbidity and mortality in patients with diabetes in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet. 2002;359:1004‐1010. [DOI] [PubMed] [Google Scholar]
- 27. Mathew J, Sleight P, Lonn E, et al. Reduction of cardiovascular risk by regression of electrocardiographic markers of left ventricular hypertrophy by the angiotensin‐converting enzyme inhibitor ramipril. Circulation. 2001;104:1615‐1621. [DOI] [PubMed] [Google Scholar]
- 28. Aro AL, Chugh SS. Clinical diagnosis of electrical versus anatomic left ventricular hypertrophy: prognostic and therapeutic implications. Circ Arrhythm Electrophysiol. 2016;9:e003629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Sundstrom J, Lind L, Arnlov J, Zethelius B, Andren B, Lithell HO. Echocardiographic and electrocardiographic diagnoses of left ventricular hypertrophy predict mortality independently of each other in a population of elderly men. Circulation. 2001;103:2346‐2351. [DOI] [PubMed] [Google Scholar]
- 30. Narayanan K, Reinier K, Teodorescu C, et al. Electrocardiographic versus echocardiographic left ventricular hypertrophy and sudden cardiac arrest in the community. Heart Rhythm. 2014;11:1040‐1046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Leigh JA, O'Neal WT, Soliman EZ. Electrocardiographic left ventricular hypertrophy as a predictor of cardiovascular disease independent of left ventricular anatomy in subjects aged >/=65 years. Am J Cardiol. 2016;117:1831‐1835. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Bacharova L, Schocken D, Estes EH, Strauss D. The role of ECG in the diagnosis of left ventricular hypertrophy. Curr Cardiol Rev. 2014;10:257‐261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Cuspidi C, Facchetti R, Bombelli M, et al. Does QRS voltage correction by body mass index improve the accuracy of electrocardiography in detecting left ventricular hypertrophy and predicting cardiovascular events in a general population? J Clin Hypertens. 2016;18:415‐421. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Singh M, Sethi A, Mishra AK, Subrayappa NK, Stapleton DD, Pellikka PA. Echocardiographic imaging challenges in obesity: guideline recommendations and limitations of adjusting to body size. J Am Heart Assoc. 2020;9:e014609. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Lee DK, Marantz PR, Devereux RB, Kligfield P, Alderman MH. Left ventricular hypertrophy in black and white hypertensives. Standard electrocardiographic criteria overestimate racial differences in prevalence. JAMA. 1992;267:3294‐3299. [PubMed] [Google Scholar]
- 36. Jingi AM, Noubiap JJ, Kamdem P, Kingue S. Determinants and improvement of electrocardiographic diagnosis of left ventricular hypertrophy in a black African population. PLoS One. 2014;9:e96783. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Kizer JR, Arnett DK, Bella JN, et al. Differences in left ventricular structure between black and white hypertensive adults: the Hypertension Genetic Epidemiology Network study. Hypertension. 2004;43:1182‐1188. [DOI] [PubMed] [Google Scholar]
- 38. Krumholz HM, Larson M, Levy D. Prognosis of left ventricular geometric patterns in the Framingham Heart Study. J Am Coll Cardiol. 1995;25:879‐884. [DOI] [PubMed] [Google Scholar]
- 39. Bella JN, Palmieri V, Wachtell K, et al. Sex‐related difference in regression of left ventricular hypertrophy with antihypertensive treatment: the LIFE study. J Hum Hypertens. 2004;18:411‐416. [DOI] [PubMed] [Google Scholar]
- 40. Okin PM, Roman MJ, Devereux RB, Kligfield P. Gender differences and the electrocardiogram in left ventricular hypertrophy. Hypertension. 1995;25:242‐249. [DOI] [PubMed] [Google Scholar]