Author + information
- David A. McNamara, MD, MPH∗ (, )
- Ari J. Bennett, MD,
- Colby Ayers, MS,
- James A. de Lemos, MD,
- Jarett D. Berry, MD, MS and
- Mark S. Link, MD
- ↵∗University of Texas Southwestern Medical Center, Division of Cardiology, 5323 Harry Hines Boulevard, Dallas, Texas 75390-8830
Early repolarization pattern (ERP) on the 12-lead electrocardiogram (ECG) is associated with sudden cardiac death (SCD) and increased mortality in certain populations. The mechanism of this association has been hypothesized to relate to regional heterogeneity of epicardial and endocardial repolarization. However, whether differences in subclinical left ventricular wall thickness may affect this association is unknown (1). Individuals with ERP are more likely to meet Sokolow-Lyon ECG criteria for left ventricular hypertrophy (LVH-SL) but are paradoxically less likely to meet Cornell voltage criteria (LVH-C) (2,3). Studies to date are limited by use of ECG measurements to define LVH and are thus not sufficient to determine whether there is a relationship between left ventricular mass (LVM) and ERP. We hypothesized that participants with ERP have a higher LVM and a greater prevalence of cardiac magnetic resonance imaging–determined LVH (LVH-CMR) compared with those without ERP.
The population-based DHS (Dallas Heart Study) is a single-site, multiethnic, population-based cohort of Dallas County residents oversampled for black participants (4). Of 3,067 participants with a high-quality ECG, those with a QRS duration >120 ms (n = 46) and missing contemporaneous CMR (n = 268) were excluded, resulting in 2,753 study participants. ECG measurements, including LVH-C and LVH-SL, were electronically measured by the Marquette 12SL analysis system (General Electric, Boston, Massachusetts). J-point amplitude was defined by the amplitude at the QRS offset in relation to the QRS onset. ERP was defined as J-point amplitude ≥1 mm in any 2 contiguous leads in 1 of 3 regions: “anterior” (V1 to V4), “inferior” (II, III, aVF), and/or “lateral” (I, aVL, V4 to V6). Lean-body mass and percentage of body fat were measured by dual-energy x-ray absorptiometry (Delphi W scanner, Hologic Inc., Bedford, Massachusetts; Discovery software, version 12.2, Hologic Inc.).
LVM was calculated from the short-axis sequence of CMR using 2 1.5-T systems (Philips Medical Systems, Best, the Netherlands), as previously described (4). All measurements were completed using QMass software (version 6.2.3, Medis Medical Imaging Systems, Leiden, the Netherlands). LVM and LVH-CMR were compared in participants with and without ERP. Logistic regression models assessed the association between ERP and LVM indexed to body surface area in unadjusted and fully adjusted models (age, sex, black race, hypertension, diabetes, hyperlipidemia, QRS width, PR interval, corrected QT interval, percentage of body fat, and lean body mass) with LVM as the dependent variable.
ERP prevalence was 9.9% (273 of 2,753). Of those with ERP, 23% demonstrated ERP in the inferior and/or lateral leads and 92% anteriorly. Participants with ERP were more likely to be male, black, have slower heart rates, longer QRS duration, shorter corrected QT interval, have greater lean-body mass, and less total body fat. No differences were seen in age or prevalence of established cardiovascular risk factors (Table 1). Participants with ERP were more likely to have LVH-SL, but not LVH-C. Individuals with ERP had greater indexed LVM and higher rates of LVH-CMR compared with those without ERP, a difference that was consistent when stratified by lead location (Table 2) and sex (Table 3). Unadjusted regression analyses demonstrated a 2-fold increase in the prevalence of ERP per standard deviation increase in indexed LVM (g/m2) (odds ratio: 2.16; 95% confidence interval: 1.91 to 2.42), which remained significant in fully adjusted models (odds ratio: 1.58; confidence interval: 1.35 to 1.85).
Our main finding is that greater imaging-derived LVM is independently associated with an ERP pattern in healthy, ambulatory men and women. The mechanism underlying the association between ERP and SCD has remained elusive. Greater LVM increases the susceptibility to fatal arrhythmias and SCD. Several potential mechanisms have been proposed to explain this association of LVM and SCD, including increased interstitial myocardial fibrosis, early after-depolarizations, dispersion of repolarization, and longer action potential duration (5). Differences in LVM may thus account for some of the increased risk of SCD in individuals with ERP, a finding with potentially important implications. For example, it is plausible that LVH regression may have favorable effects on outcomes among individuals with concomitant ERP and LVH.
Our study’s strengths include utilization of a highly accurate and reproducible method of LVM assessment (CMR), which has been previously validated (4), and a fully automated ascertainment of J-point elevation. Several limitations of our study merit discussion. First, our study investigates ERP with ST-segment elevation. We included anterior, lateral, and inferior leads to provide a detailed description of ERP with J-point elevation, consistent with updated American Heart Association Scientific Statements (1). Although measures of fitness were not obtained in study participants, professional athletes, and by extension true “athletic hearts” were not included in our population. As our goal was to utilize an automated identification of the J-point to facilitate examination in a clinical setting, we did not assess for J-point notching/slurring or the slope of the ST-segment in this study.
In summary, our study demonstrates an independent association between greater LVM and ERP in a multiethnic, population-based study. Further studies are needed to explore whether increased LVM may in part mediate the association between ERP and SCD.
Please note: †Drs. McNamara and Bennett contributed equally to this work. Dr. McNamara is supported by the National Heart, Lung, and Blood Institute (grant T32-HL125247). The Dallas Heart Study was supported by a grant from the Reynolds Foundation and grant UL1TR001105 from the National Center for Advancing Translational Sciences of the National Institutes of Health. Dr. Berry has received honoraria from AstraZeneca, American Heart Association, and Abbott Diagnostics. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2019 American College of Cardiology Foundation
- Patton K.K.,
- Ellinor P.T.,
- Ezekowitz M.,
- et al.
- Walsh J.A. 3rd.,
- Ilkhanoff L.,
- Soliman E.Z.,
- et al.
- Noseworthy P.A.,
- Tikkanen J.T.,
- Porthan K.,
- et al.