The electrocardiogram (ECG) is widely used for detection of left ventricular hypertrophy (LVH). However, whether changes in ECG LVH during antihypertensive therapy predict changes in LV mass remains unclear. Baseline and year-1 ECGs and echocardiograms were assessed in 584 hypertensive patients with ECG LVH by Sokolow–Lyon or Cornell voltage–duration product criteria at entry into the Losartan Intervention For Endpoint reduction in hypertension (LIFE) echocardiographic substudy. A ⩾25% decrease in Cornell product defined regression of ECG LVH; a <25% decrease defined no significant regression; and an increase defined progression of ECG LVH. Regression of echocardiographic LVH was defined by a ⩾20% reduction in LV mass. After 1 year of therapy, 155 patients (27%) had regression of ECG LVH, 286 (49%) had no significant change, and 143 (25%) had progression of ECG LVH. Compared with patients with progression of ECG LVH, patients with no significant decrease and patients with regression of ECG LVH had stepwise greater absolute decreases in LV mass (−16±33 vs −29±37 vs −32±41 g, P<0.001), greater percent reductions in LV mass (−5.7±14.6 vs −11.3±13.6 vs −12.3±15.6%, P<0.001), and were more likely to decrease LV mass by ⩾20% (11.2 vs 24.8 vs 36.1%, P<0.001), even after adjusting for possible effects of baseline and change in systolic and diastolic pressures. Compared with progression of ECG LVH, regression of the Cornell product ECG LVH is associated with greater reduction in LV mass and a greater likelihood of regression of anatomic LVH.
Left ventricular hypertrophy (LVH) detected by the electrocardiogram (ECG)1,2,3 and echocardiogram4,5,6,7,8 are common manifestations of preclinical cardiovascular disease that strongly predict cardiovascular morbidity and mortality. Antihypertensive therapy can produce regression of LVH,3,4,9,10,11,12,13 and available data suggest that regression of LVH and prevention of progression to LVH can decrease cardiovascular morbidity.2,3,4,12,13 However, utility of ECG criteria for detection of LVH and for serial evaluation of changes in LV mass has been limited by the low sensitivity of standard voltage criteria for identification of anatomic LVH.14,15,16,17,18,19,20
Cornell voltage criteria modestly improve ECG detection of LVH,15,16 and the product of Cornell voltage and QRS duration (Cornell voltage–duration product),17,18 as an approximation of the true area under the QRS complex,19 further enhances sensitivity of the ECG for LVH while maintaining high specificity. As a consequence, Cornell voltage–duration product criteria were used in combination with standard Sokolow–Lyon voltage criteria14 to identify hypertensive patients at increased risk of cardiovascular morbidity and mortality in the Losartan Intervention For Endpoint reduction in hypertension (LIFE) study.21,22,23 These ECG criteria identified hypertensive patients with a >70% likelihood of having echocardiographic LVH and with high-normal indexed LV mass in those patients not fulfilling strict cutoff criteria for echocardiographic LVH.24 However, few data exist on the value of serial changes in ECG voltage for predicting changes in LV mass,25,26 and the ability of changes in more sensitive voltage–duration product criteria to predict changes in LV mass during antihypertensive therapy has not been examined. Therefore, the present study examined the ability of changes in the Cornell voltage–duration product criteria over the first year of antihypertensive therapy in LIFE to predict regression and progression of echocardiographic LVH.
The double-blind LIFE trial21,22,23 enrolled hypertensive patients with ECG LVH by the Cornell voltage–duration product17,18 and/or Sokolow–Lyon voltage criteria14 on a screening ECG to determine whether greater reduction in mortality and morbidity is associated with the use of losartan- as opposed to atenolol-based therapy.21 The study was approved by all ethics committees concerned. Study inclusion and exclusion criteria have been previously published.21,22,23 A previously described24 representative sample of the whole LIFE study, totaling 960 patients, underwent baseline echocardiograms, of which 584 had measurable echocardiographic LV mass and complete ECG measurements at baseline and year 1 of follow-up. There were 235 women and 349 men whose mean age was 66±7 years.
ECGs were interpreted at the ECG Core Laboratory at Sahlgrenska University Hospital/Östra in Göteborg, Sweden, by experienced investigators blinded to clinical information. The product of QRS duration times the Cornell voltage combination (RaVL+SV3, with 8 mm added in women17,18) was used with a threshold value of 2440 mm × ms to identify LVH. After the design of the LIFE study, studies were published suggesting a smaller gender adjustment,27,28 and feedback from LIFE investigators showed that otherwise eligible patients had ECG LVH by highly specific but insensitive Sokolow–Lyon voltage criteria,14 but not by the Cornell product. Accordingly, changes were made in ECG entry criteria: the gender adjustment of Cornell voltage was reduced from 8 to 6 mm, and Sokolow–Lyon voltage (SV1+RV5/6)>38 mm was accepted for ECG eligibility.22
For the present study, additional ECG measurements were performed at Helsinki University Central Hospital, as previously described.29 QRS duration was measured to the nearest ms, and QRS amplitudes to the nearest 0.1 mm, by a trained technician unaware of clinical data who used a digitizer connected to a personal computer. Based on previous measures of repeat variability of Cornell voltage and QRS duration measurements30,31,32 and on the reproducibility of interval measurements in Helsinki,29 a ⩾25% decrease in the Cornell product was used to define regression of ECG LVH, a <25% decrease to represent no significant regression, and an increase in the Cornell product of any magnitude to define progression of ECG LVH. Regression of ECG LVH was not defined by a change in ECG status from LVH to no LVH by either criterion in order to avoid misclassifying patients with only minimal changes in ECG LVH or erroneously considering regression to the mean33,34 as true regression of LVH.
Studies were performed with commercially available phased-array echocardiographs using standardized recording procedures as previously reported in detail.24,29,34 LV mass was calculated35 and indexed for body surface area, and LVH was considered present if LV mass index (LVMI) was >104 g/m2 in women and >116 g/m2 in men.1,7 Regression of LVH was defined as a ⩾20% decrease in LV mass between the baseline and year-1 echocardiogram. This provides a >95% likelihood of representing a true decrease in LV mass based on assessment of interstudy variability and regression to the mean on changes in LV mass on serial echocardiograms,34 and approximates the upper quartile of change in LV mass in the study population.
Data management and analysis were performed with the SPSS version 10. Data are presented as mean±s.d. for continuous variables, and as proportions for categorical variables. Differences in prevalences between groups were compared using χ2-analyses. Mean values of continuous variables were compared using two-way ANOVA to adjust for the possible influence of gender, with post hoc testing of intergroup differences by Scheffe's method. Changes in LV mass between baseline and year 1 were further compared between groups using ANCOVA to adjust for differences between groups in age, gender, baseline diastolic blood pressure (DBP), and change in systolic and diastolic pressures from baseline to year 1. The independent relation of regression of echocardiographic LVH to change in Cornell product was determined using stepwise logistic regression analyses. For all tests, a two-tailed P-value <0.05 was required for statistical significance.
After 1 year of antihypertensive therapy in the LIFE study, regression of ECG LVH by the Cornell product was present in 155 patients (26.5%); there was no significant decrease in the Cornell product in 286 patients (49.0); and progression of ECG LVH was present in 143 patients (24.5%). The mean reduction in the Cornell product after 1 year was 371±715 mm × ms. There were only minor differences in baseline demographic characteristics among groups defined by the change in the Cornell product between baseline and year 1 (Table 1).
The relations of baseline BP and LV mass, and the change in BP and in body weight over the 1 year of therapy, to regression vs progression of ECG LVH are shown in Table 2. Baseline diastolic pressure was significantly lower in patients with progression of ECG LVH, and there was a trend towards lower baseline systolic pressure in this group as well. BP decreased in all groups, but patients with regression of ECG LVH had the greatest decreases in both systolic and diastolic pressure; patients with no significant decrease in the Cornell product had intermediate decreases in BP; and patients with progression of ECG LVH had the smallest decreases in systolic and diastolic pressure after 1 year of antihypertensive therapy. There were no differences between groups in the change in body weight over 1 year. The mean LV mass and indexed LV mass were elevated in all three groups at study baseline, without significant differences between groups.
Regression of ECG LVH and change in LV mass
The relations of absolute and percent changes in LV mass to the change in the Cornell product are examined in Table 3. After adjusting for gender, both the absolute and percent change in LV mass were strongly associated with the presence of regression of ECG LVH by the Cornell product criteria. Compared to patients with no significant decrease in the Cornell product, patients with progression of ECG LVH had smaller, and those with regression of ECG LVH had greater, decreases in LV mass in absolute terms or percent change. The differences persisted after adjusting for the possible effects of age, sex, baseline diastolic pressure, and change in systolic and diastolic pressures (Table 3).
The ability of regression of ECG LVH by the Cornell product to identify patients most likely to experience regression of echocardiographic LVH as determined by a ⩾20% decrease in LV mass is examined in Figure 1 and Table 4. After 1 year of antihypertensive therapy, 36.1% of patients with regression of ECG LVH by the Cornell product had ⩾20% reduction in LV mass, compared with 24.8% of patients who had no significant decrease in the Cornell product, and only 11.2% of patients who had progression of ECG LVH (P<0.001, Figure 1). Of note, the absolute degree of association between regression of ECG and echocardiographic LVH was moderate (gamma=0.41, P<0.001). Univariate logistic regression analysis demonstrated that, compared with patients who had no significant decrease in ECG LVH, patients with regression of ECG LVH had a nearly 180% greater likelihood of regression of echocardiographic LVH, whereas those with progression of ECG LVH had a >60% lower likelihood of experiencing a ⩾20% reduction in LV mass (Table 4). After adjusting for effects of sex, baseline LV mass, and baseline and change in systolic and diastolic BP on regression of hypertrophy, regression of ECG LVH by the Cornell product criteria remained strongly associated with regression of echocardiographic LVH, with no attenuation of odds ratios (Table 4). Regression of ECG LVH by the Cornell product criteria remained a significant predictor of regression of echocardiographic LVH when change in the Cornell product was considered as a continuous variable (χ2=20.48, P<0.001) and when examined separately in men and women. Of note, regression of ECG LVH by simpler Cornell voltage criteria was a univariate (P=0.011) but not multivariate (P=0.208) predictor of regression of echocardiographic LVH. Regression of ECG LVH by Sokolow–Lyon voltage criteria was not a significant predictor of regression of anatomic LVH (P=0.068).
This study demonstrates that regression of ECG LVH by the Cornell product criteria is an independent predictor of the presence and degree of regression of echocardiographic LVH after 1 year of antihypertensive therapy in the LIFE study. Compared with patients with no significant decrease in ECG LVH, patients with regression of ECG by the Cornell product criteria had greater absolute and percent decreases in LV mass, and were nearly twice as likely to decrease LV mass by ⩾20%, even after adjusting for the greater reductions in systolic blood pressure (SBP) and DBP in this group. These findings further support the use of voltage–duration product criteria for ECG assessment of the presence of LVH, and for identifying significant serial changes in LV mass over time.
Relation of changes in ECG LVH criteria to changes in LV mass
Previous studies have demonstrated that antihypertensive therapy can produce significant decreases in LV mass and regression of LVH;3,4,9,10,11,12,13 however, there are only limited data examining the ability of serial changes in ECG LVH criteria to predict changes in LV mass.25,26 Ditchey et al25 examined short-term changes in ECG and echocardiographic LVH in 15 normal subjects and in 18 patients undergoing aortic or mitral valve replacement, and found that serial changes in R-wave amplitude in aVL, but not Sokolow–Lyon voltage, paralleled serial changes in LV mass. However, this study did not quantify the predictive value of changes in ECG voltages for changes in LV mass, and was limited by the small number of patients examined and by inclusion of postoperative studies in which other clinical factors could have affected ECG variables. Crow et al26 found poor correlations between average changes in LV mass and eight ECG LVH criteria, including Sokolow–Lyon voltage and the Cornell product, over 4 years in 834 participants in the Treatment of Mild Hypertension Study. However, participants had a low prevalence of echocardiographic LVH at baseline (15.3%), and no data were provided on the degree of change in LV mass or ECG criteria. The mild hypertension and expected small changes in LV mass over time could truncate the relationship between changes in LV mass and ECG LVH criteria, which could have been further reduced by the use of average changes in these measures compared with baseline. In contrast, the present study clearly demonstrates that changes in the Cornell product independently predict significant changes in LV mass.
Regression of ECG LVH and prognosis
A number of previous studies have demonstrated that regression of ECG LVH and prevention of LVH development are associated with a reduced risk of cardiovascular morbidity.2,3,12,13 In 524 participants in the Framingham Heart Study who had ECG LVH by various criteria at a qualifying examination, Levy et al2 found that a significant decline in Cornell voltage, defined by a change to a lower quartile, was associated with lower risk for cardiovascular disease; whereas a significant increase in Cornell voltage, defined by change to a higher quartile of voltage, identified subjects at increased risk. Importantly, these relationships persisted when change in Cornell voltage was examined as a continuous variable and when voltage change was defined by a >10% serial increase or decrease.2 Prineas et al13 demonstrated that incident LVH and increases in ECG LVH by the Cornell product and Novacode criteria were associated with increased risk of mortality in the usual care arm of MRFIT. In contrast, increases in Sokolow–Lyon voltage were associated with decreased risk in MRFIT.13 However, these investigators averaged changes in ECG LVH criteria over 6 years of follow-up, which could underestimate the predictive value of serial increases or decreases in ECG measures over this time period. Furthermore, their findings were limited to men. In contrast, recent findings from the HOPE trial using Sokolow–Lyon voltage criteria for LVH3 suggest that reduced development of ECG LVH and increased regression of ECG LVH by these criteria, in response to ramipril therapy are associated with reduced risk of death, myocardial infarction, stroke, and congestive heart failure. However, HOPE did not assess the degree of change in Sokolow–Lyon voltage, only whether voltage was less than or greater than the threshold value of 3.5 mV, and did not adjust for other risk factors.3 The present findings taken together with the greater sensitivity of Cornell product criteria for LVH,17,18 and the findings of Prineas et al,13 further suggest that serial changes in the Cornell product may be useful for stratifying risk in hypertensive patients. Further analysis of data from the LIFE study will allow this issue to be addressed.
The present study defined significant changes in ECG LVH and LV mass based on percent changes expected to exceed changes due to normal variability and regression to the mean of these measures.30,31,32,33,34 Regression of LVH was specifically not defined by either the Cornell product or LV mass measurements declining below threshold values used to define the presence of LVH by these methods. Use of this approach would overestimate true regression by including small changes in LV mass or the Cornell product in patients just above or below the threshold value as reflecting significant changes, whereas these changes are more likely to reflect regression to the mean or normal intrinsic variability in these measurements.30,31,32,33,34 At the same time, this approach would underestimate regression of LVH in patients with large reductions in LV mass that fail to fall below this artificial threshold for defining hypertrophy.
These findings have important clinical implications. The strong relationship between changes in the Cornell voltage–duration product and significant changes in LV mass, taken together with previous studies relating changes in ECG voltage to prognosis2,3 and the greater accuracy of the Cornell product criteria for detecting LVH,17,18 suggest that serial changes in Cornell product may be able to be used to determine whether patients are having an appropriate reduction in LV mass in response to antihypertensive therapy. As the LIFE study was restricted to patients with ECG LVH on a screening ECG,21,22,23 additional studies in large groups of hypertensive patients without ECG LVH will be required to validate these findings in a less-selected population. Further study will be necessary to elucidate the value of changes in the Cornell product criteria for predicting cardiovascular morbidity and mortality in the LIFE study.21,22,23
Verdecchia P et al. Prognostic value of a new electrocardiographic method for diagnosis of left ventricular hypertrophy in essential hypertension. J Am Coll Cardiol 1998; 31: 383–390.
Levy D et al. Prognostic implications of baseline electrocardiographic features and their serial changes in subjects with left ventricular hypertrophy. Circulation 1994; 90: 1786–1793.
Mathew J 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.
Verdecchia P et al. Prognostic significance of serial changes in left ventricular mass in essential hypertension. Circulation 1998; 97: 48–54.
Koren MJ et al. Relation of left ventricular mass and geometry to morbidity and mortality in uncomplicated essential hypertension. Ann Intern Med 1991; 114: 345–352.
Levy D et al. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990; 322: 1561–1566.
Liao Y et al. The relative effects of left ventricular hypertrophy, coronary artery disease, and ventricular dysfunction on survival among black adults. JAMA 1995; 273: 1592–1597.
Schillaci G et al. Continuous relation between left ventricular mass and cardiovascular risk in essential hypertension. Hypertension 2000; 35: 580–586.
Dahlöf B, Pennert K, Hansson L . Reversal of left ventricular hypertrophy in hypertensive patients. A metaanalysis of 109 treatment studies. Am J Hypertens 1992; 5: 95–110.
Schlaich MP, Schmieder RE . Left ventricular hypertrophy and its regression: pathophysiology and therapeutic approach: focus on treatment by antihypertensive agents. Am J Hypertens 1998; 11: 1394–1404.
Neaton JD et al. Treatment of mild hypertension study. Final results. Treatment of Mild Hypertension Study Research Group. JAMA 1993; 270: 713–724.
Five-year findings of the Hypertension Detection and Follow-up Program. Prevention and reversal of left ventricular hypertrophy with antihypertensive drug therapy. Hypertension Detection and Follow-up Program Cooperative Group. Hypertension 1985; 7: 105–112.
Prineas RJ et al. Independent risk for cardiovascular disease predicted by modified continuous score electrocardiographic criteria for 6-year incidence and regression of left ventricular hypertrophy among clinically disease free men: 16-year follow-up for the multiple risk factor intervention trial. J Electrocardiol 2001; 34: 91–101.
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.
Casale PN et al. Electrocardiographic detection of left ventricular hypertrophy: development and prospective validation of improved criteria. J Am Coll Cardiol 1985; 6: 572–580.
Casale PN et al. Improved sex-specific criteria of left ventricular hypertrophy for clinical and computer interpretation of electrocardiograms: validation with autopsy findings. Circulation 1987; 75: 565–572.
Molloy TJ, Okin PM, Devereux RB, Kligfield P . Electrocardiographic detection of left ventricular hypertrophy by the simple QRS voltage–duration product. J Am Coll Cardiol 1992; 20: 1180–1186.
Okin PM, Roman MJ, Devereux RB, Kligfield P . Electrocardiographic identification of increased left ventricular mass by simple voltage–duration products. J Am Coll Cardiol 1995; 25: 417–423.
Okin PM et al. Time–voltage QRS area of the 12-lead electrocardiogram: detection of left ventricular hypertrophy. Hypertension 1998; 31: 937–942.
Levy D et al. Determinants of sensitivity and specificity of electrocardiographic criteria for left ventricular hypertrophy. Circulation 1990; 81: 815–820.
Dahlöf B et al. The Losartan Intervention For Endpoint reduction (LIFE) in Hypertension study: rationale, design, and methods. The LIFE Study Group. Am J Hypertens 1997; 10: 705–713.
Dahlöf B et al. Characteristics of 9194 patients with left ventricular hypertrophy: the LIFE study. Losartan Intervention For Endpoint Reduction in Hypertension. Hypertension 1998; 32: 989–997.
Kjeldsen SE et al. Lowering of blood pressure and predictors of response in patients with left ventricular hypertrophy: the LIFE Study. Losartan Intervention For Endpoint. Am J Hypertens 2000; 13: 899–906.
Wachtell K et al. Impact of different partition values on prevalences of left ventricular hypertrophy and concentric geometry in a large hypertensive population: the LIFE study. Hypertension 2000; 35: 6–12.
Ditchey RV, Schuler G, Peterson KL . Reliability of echocardiographic and electrocardiographic parameters in assessing serial changes in left ventricular mass. Am J Med 1981; 70: 1042–1050.
Crow RS, Hannan P, Grandits G, Liebson P . Is the echocardiogram an appropriate ECG validity standard for the detection and change in left ventricular size? J Electrocardiol 1996; 29(Suppl): 248–255.
Schillaci G et al. Improved electrocardiographic diagnosis of left ventricular hypertrophy. Am J Cardiol 1994; 74: 714–719.
Norman Jr. JE, Levy D . Improved electrocardiographic detection of echocardiographic left ventricular hypertrophy: results of a correlated data base approach. J Am Coll Cardiol 1995; 26: 1022–1029.
Oikarinen L et al. Relation of QT interval and QT dispersion to echocardiographic left ventricular hypertrophy and geometric pattern in hypertensive patients. The LIFE study. The Losartan Intervention For Endpoint Reduction. J Hypertens 2001; 19: 1883–1891.
Farb A, Devereux RB, Kligfield P . Day-to-day variability of voltage measurements used in electrocardiographic criteria for left ventricular hypertrophy. J Am Coll Cardiol 1990; 15: 618–623.
Willems JL, Poblete PF, Pipberger HV . Day-to-day variation of the normal orthogonal electrocardiogram and vectorcardiogram. Circulation 1972; 45: 1057–1064.
Zhou SH et al. Improved ECG models for estimation of left ventricular hypertrophy progression and regression incidence by redefinition of the criteria for a significant change in left ventricular hypertrophy status. The MRFIT Research Group. Multiple Risk Factor Intervention Trial. J Electrocardiol 1993; 26(Suppl): 108–113.
Davis CE . The effect of regression to the mean in epidemiologic and clinical studies. Am J Epidemiol 1976; 104: 493–498.
Palmieri V et al. Reliability of echocardiographic assessment of left ventricular structure and function: the PRESERVE study. Prospective Randomized Study Evaluating Regression of Ventricular Enlargement. J Am Coll Cardiol 1999; 34: 1625–1632.
Devereux RB et al. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol 1986; 57: 450–458.
We thank Paulette A Lyle for assistance with preparation of the manuscript. This work is supported in part by Grant COZ-368 from Merck & Co., Inc., West Point, PA, USA.
About this article
Cite this article
Okin, P., Devereux, R., Liu, J. et al. Regression of electrocardiographic left ventricular hypertrophy predicts regression of echocardiographic left ventricular mass: the LIFE study. J Hum Hypertens 18, 403–409 (2004) doi:10.1038/sj.jhh.1001707
- blood pressure
Relationships between the arterial velocity pulse index as a novel marker of atherosclerosis and biomarkers of cardiac or renal condition in patients with type 2 diabetes mellitus
Diabetology International (2017)
Clinical Hypertension (2017)
Combining ECG Criteria for Left Ventricular Hypertrophy Improves Risk Prediction in Patients With Hypertension
Journal of the American Heart Association (2017)
Electrocardiographic Left Ventricular Hypertrophy Predicts Cardiovascular Morbidity and Mortality in Hypertensive Patients: The ALLHAT Study
American Journal of Hypertension (2017)
British Journal of Cardiac Nursing (2016)