Author + information
- Received December 23, 2010
- Revision received March 4, 2011
- Accepted April 12, 2011
- Published online May 31, 2011.
- Peter A. Noseworthy, MD⁎,†,
- Jani T. Tikkanen, MB‡,
- Kimmo Porthan, MD§,
- Lasse Oikarinen, MD, PhD§,
- Arto Pietilä, MSc¶,
- Kennet Harald, MSc¶,
- Gina M. Peloso, MA#,⁎⁎,
- Faisal M. Merchant, MD⁎,
- Antti Jula, MD, PhD¶,
- Heikki Väänänen, Lic Sc§∥,
- Shih-Jen Hwang, PhD#,††,
- Christopher J. O'Donnell, MD, MPH#,††,
- Veikko Salomaa, MD, PhD¶,⁎ (, )
- Christopher Newton-Cheh, MD, MPH⁎,†,#,⁎ ( and )
- Heikki V. Huikuri, MD‡,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Heikki V. Huikuri, Department of Internal Medicine, P.O. Box 5000, FIN-90014, University of Oulu, Oulu, Finland
- ↵⁎Dr. Christopher Newton-Cheh, Cardiovascular Research Center and Center for Human Genetic Research, Massachusetts General Hospital, 185 Cambridge Street, CPZN 5.242, Boston, Massachusetts 02114
- ↵⁎Dr. Veikko Salomaa, National Institute for Health and Welfare, Pl 30, FI-00271 Helsinki, Finland
Objectives This study sought to describe the clinical correlates and heritability of the early repolarization pattern (ERP) in 2 large, population-based cohorts.
Background There is growing recognition that ERP is associated with adverse outcomes.
Methods Participants of the Framingham Heart Study (FHS) (N = 3,995) and the Health 2000 Survey (H2K) (N = 5,489) were included. ERP was defined as a J-point elevation ≥0.1 mV in ≥2 leads in either the inferior (II, III, aVF) or lateral (I, aVL, V4–6) territory or both. We tested the association between clinical characteristics and ERP, and estimated sibling recurrence risk.
Results ERP was present in 243 of 3,955 (6.1%) of FHS and 180 of 5,489 (3.3%) of H2K subjects. Male sex, younger age, lower systolic blood pressure, higher Sokolow-Lyon index, and lower Cornell voltage were independently associated with the presence of ERP. In the FHS sample, siblings of individuals with ERP had an ERP prevalence of 11.6% (recurrence risk ratio of 1.89). Siblings of individuals with ERP had an increased unadjusted odds of ERP (odds ratio: 2.22, 95% confidence interval: 1.01 to 4.85, p = 0.047).
Conclusions ERP has strong association with clinical factors and has evidence for a heritable basis in the general population. Further assessment of the genetic determinants of ERP is warranted.
The electrocardiographic early repolarization pattern (ERP) is found in approximately 5% of the general population (1) and has long been thought to be a “normal variant”(2). However, emerging data (3–5) demonstrate an association between ERP in the inferior leads and sudden cardiac death (SCD).
It has been proposed that the early repolarization syndromes may exist on a continuous spectrum of disease (6). At one end of this spectrum, the early repolarization syndrome presents with SCD, often in the context of a structurally normal heart. Several genetic variants have been identified and point to an underlying electrophysiological pathology (3,7). At the other end of the spectrum, the early repolarization pattern is frequently observed in young, athletic, healthy males.
We examined 2 large, community-based cohorts: 1) to define the prevalence of ERP in the general population; 2) to identity clinical correlates of ERP; and 3) to assess the heritability of ERP.
The Framingham Heart Study (FHS) is a prospective, community-based study that spans 3 generations in the United States (8). Third generation subjects (n = 4,095) with available electrocardiograms (ECGs) were included. The Health 2000 Survey (H2K) is a cross-sectional, population-based cohort in Finland (9). Subjects were excluded for poor ECG quality, missing data, QRS duration ≥120 ms, atrial fibrillation/flutter on the ECG, Wolff-Parkinson-White syndrome, paced rhythm, age >80 years, age <18 years, or self-reported non-Caucasian race. The final samples included 3,955 subjects from FHS and 5,489 from H2K. Both studies were approved by local ethics boards.
ECG measurements and ERP definition
The ECGs were recorded as previously reported (10). In FHS, paper ECGs were read manually. In H2K, the J-point measurements were performed from digital ECGs using custom software. An ECG was considered ERP positive if there was J-point elevation of ≥0.1 mV in ≥2 leads in the inferior (II, III, aVF) or lateral (I, aVL, V4–6) territory, or both. Interobserver reliability was assessed within FHS reviewers (kappa = 0.65, 95% confidence interval [CI]: 0.48 to 0.82) using a set of 100 ECGs selected to be one-third ERP positive, and within H2K reviewers (kappa = 0.63, 95% CI: 0.58 to 0.69) using the entire H2K ECG cohort. Interobserver reliability was assessed between centers (kappa = 0.86, 95% CI: 0.72 to 0.99).
In FHS, bivariate logistic regression using generalized estimating equations to account for family structure was used to estimate the variable relationships to ERP in a model adjusting for age and sex. In H2K, logistic regression was used to adjust for age and sex. In FHS and H2K separately, multivariable logistic regression models were then constructed, forcing in age and sex, using forward selection (entry p < 0.05, retention p < 0.05). SAS proc logistic (version 9.2, SAS Institute, Cary, North Carolina) was used for model selection in both cohorts. In H2K, the sex-specific ERP prevalence by decade was estimated and a test for age by sex interaction performed.
In the FHS sample, the sibling recurrence risk ratio (λS) was estimated by comparing the ERP prevalence in siblings of ERP-positive individuals to the ERP prevalence in the general population. Since FHS includes varying sibship sizes, a set of independent participants was sampled from all siblings to obtain the population prevalence of ERP. One sibling was then chosen at random for each of the index subjects. Bootstrap confidence intervals were estimated for λS by resampling with 1,001 replicates. Additionally, odds ratios were obtained predicting the ERP status of siblings of index individuals using logistic regression, with and without adjustment for age and sex. In secondary analyses, sibships were stratified based on the age of the index participant using ≤40 and >40 years (≈median age).
ERP was present in 243 of 3,955 (6.1%) of FHS subjects, 180 of 5,489 (3.3%) of H2K subjects, and 423 of 9,444 (4.5%) overall. Clinical characteristics by ERP status are shown in Table 1. Several characteristics were significant predictors of ERP status in age- and sex-adjusted regression models (Table 2). The cohort-specific independent associations with ERP status are shown in Table 3. In H2K, there was a higher ERP prevalence in men, decreasing prevalence with increasing age (Fig. 1,Table 4), and evidence for an age-by-sex interaction (p = 0.048).
In FHS, siblings of individuals with ERP had an ERP prevalence of 11.6% (sibling recurrence risk ratio = 11.6%/6.1% = 1.89). Siblings of ERP-positive individuals had increased odds of ERP that was attenuated after adjustment for age and sex (Table 5).
Consistent with published estimates ranging from 1% to 13% (1,4), we found an overall prevalence of ERP of 4.5%. Variation in estimates may be due to differences in trait definition, adjudication technique, and study sample demographics. For instance, individuals in FHS were approximately 10 years younger than those in the H2K cohort and the ERP prevalence was nearly twice as high.
Our study demonstrates independent associations of ERP with male sex, lower systolic blood pressure, higher Sokolow-Lyon voltage, and lower Cornell voltage. Prior reports have shown a positive association between Sokolow-Lyon index and ERP (1,5), but the contrasting effects of Sokolow-Lyon and Cornell voltage have not previously been described. The Sokolow-Lyon and Cornell voltages are poorly correlated to cardiac magnetic resonance imaging–based estimates of left ventricular hypertrophy (11) and have different clinical associations. Higher Cornell voltage has been associated with female sex and increased blood pressure, whereas higher Sokolow-Lyon index has been associated with male sex and lower body mass index (12). It is particularly high in young men with a thin chest wall resulting in high precordial voltage. Further work is needed to better understand the relationship of ERP to various estimates of left ventricular mass.
ERP appears to have a heritable basis based on the results of our study and a recent report (13). If ERP in the general population represents a stable, heritable phenomenon rather than a consequence of physical training or a transient characteristic of the “juvenile” ECG, it raises the question of whether there are similarities in the genetic basis of ERP and other J-wave syndromes. Of course, the observation that the early repolarization pattern is heritable suggests that ERP has genetic determinants, but these factors may not confer risk of arrhythmia. This concept has parallels to the relationship of the QT interval with SCD. The long QT syndrome is associated with an increased risk of SCD, and prolongation of the QT interval is a risk factor for SCD at the population level (14). The common genetic underpinnings of normal QT variation are increasingly understood (10), but these variants have modest effects on SCD risk (15). Similarly, although the relative risk of SCD is higher in patients with ERP, the likelihood of SCD remains low in absolute terms. Furthermore, since siblings of ERP-positive individuals are more likely to have ERP, the value of screening ECGs in siblings of SCD victims is uncertain since even “benign” ERP may cluster in families.
The strengths of this study are its large, community-based cohorts with application of a uniform adjudication process. The major limitation is the lack of outcome data, which precludes survival analysis. Additionally, the findings may not translate to individuals of non-European ancestry, especially those of African descent in whom ERP is particularly prevalent.
We demonstrate that ERP shows strong association with clinical factors and has a heritable basis. We believe that these findings motivate further assessment of the genetic determinants of the pattern.
This work was supported by the Max Schaldach Fellowship in Cardiac Pacing and Electrophysiology, and a Burroughs Wellcome Fund travel grant (to Dr. Noseworthy); the National Institutes of Health/National Heart, Lung, and Blood Institute (HL080025; HL098283), the Doris Duke Charitable Foundation, and the Burroughs Wellcome Fund (to Dr. Newton-Cheh); the Academy of Finland, grant number 129494 (to Dr. Salomaa); the Finnish Foundation for Cardiovascular Research (to Drs. Porthan, Tikkanen, Oikarinen, and Huikuri); the Finnish Cultural Foundation (to Dr. Tikkanen); and the Sigrid Juselius Foundation (to Dr. Huikuri). The Framingham Heart Study was supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health and Boston University School of Medicine (Contract No. N01-HC-25195). Dr. Newton-Cheh is on a Merck scientific advisory board. All other authors have reported that they have no relationships to disclose. Drs. Noseworthy, Tikkanen, and Porthan contributed equally to this work. Drs. Salomaa, Newton-Cheh, and Huikuri are joint senior authors of this work.
- Abbreviations and Acronyms
- confidence interval
- early repolarization pattern
- Framingham Heart Study
- sudden cardiac death
- Received December 23, 2010.
- Revision received March 4, 2011.
- Accepted April 12, 2011.
- American College of Cardiology Foundation
- Heistaro S.
- Okin P.M.,
- Devereux R.B.,
- Jern S.,
- Kjeldsen S.E.,
- Julius S.,
- Dahlof B.the Life Study Investigators.
- Reinhard W.,
- Kaess B.M.,
- Debiec R.,
- et al.
- Chugh S.S.,
- Reinier K.,
- Singh T.,
- et al.
- Kao W.H.,
- Arking D.E.,
- Post W.,
- et al.