Long-Term Risk of Incident Arrhythmias Associated With Early Repolarization Pattern　― The Atherosclerosis Risk in Communities (ARIC) Study ―

Abstract

Background: The early repolarization pattern (ERP) is associated with cardiovascular death, but its connection with arrhythmias remains unknown. This study evaluated relationships between ERP and incident arrhythmias, including ventricular arrhythmias, bradyarrhythmias, and atrial fibrillation (AF)/flutter (Afl).

Methods and Results: We analyzed 14,679 middle-aged (45–64 years) participants from the Atherosclerosis Risk in Communities cohort, a prospective population-based study in the US. Participants were monitored for ERP status at baseline and at 3 subsequent follow-up visits. We examined associations between incident arrhythmias and baseline ERP, time-varying ERP, time-updated ERP, and changes in ERP over time using Cox models to estimate hazard ratios (HRs) adjusted for potential confounders. Over a 20-year follow-up, there were 1,252 ventricular arrhythmias, 890 bradyarrhythmias, and 2,202 cases of AF. Time-updated ERP was associated with increased HRs for ventricular arrhythmias (1.55; 95% confidence interval [CI] 1.35–1.77), bradyarrhythmias (1.76; 95% CI 1.48–2.08), and AF (1.25; 95% CI 1.10–1.43). Time-varying ERP also showed associations with these outcomes. Compared with individuals with consistently normal electrocardiogram results, those with new-onset or persistent ERP had increased risks of incident arrhythmias. In subjects with time-updated ERP, anterior leads and J wave amplitudes ≥0.2 mV were associated with a higher incidence of arrhythmias.

Conclusions: Several types of ERP, including time-varying, time-updated, new-onset, and consistent, are associated with the incidence of arrhythmias in the middle-aged biracial (Black and White) population.

The early repolarization pattern (ERP) is a common variant in electrocardiograms (ECGs) and has been traditionally considered a benign phenomenon. The ERP is characterized by an elevation of the QRS–ST junction (J point) across multiple leads, accompanied by a notch or slurring at the end of the QRS complex, and occurs without any chest pain.^1,2 However, recent studies suggest that ERP may be linked to a higher risk of idiopathic ventricular fibrillation and isolated atrial fibrillation (AF).^3–7 In addition, individuals with ERP often experience bradycardia.^8,9 The connection between ERP and bradyarrhythmias like sick sinus syndrome and atrioventricular block remains unconfirmed. Although ERP is a variable factor that can change over time,^10,11 existing population studies have not assessed the relationship between changes in ERP over time and the risk of arrhythmia-related events. Therefore, further investigation is needed into the relationship between changes in ERP over time and the occurrence of various types of arrhythmia events.

The aim of this study was to prospectively explore the association between long-term ERP, variations in ERP dynamics, and the risk of developing incident arrhythmias, covering ventricular arrhythmias, bradyarrhythmias, and AF among participants of the Atherosclerosis Risk in Communities (ARIC) study. This research encompasses a vast, multicenter, biracial (Black and White), community-based cohort, with outcomes tracked for more than 2 decades.

Methods

Study Design and Population

This study is based on the ARIC study, a prospective cohort initiated in 1987 to investigate cardiovascular disease risk factors. From 1987 to 1989, 15,792 participants aged 45–64 years were recruited from 4 US communities: Jackson, Mississippi; Washington County, Maryland; Minneapolis suburbs, Minnesota; and Forsyth County, North Carolina. The cohort was predominantly Black and White,^12,13 with follow-up visits in 1990–1992, 1993–1995, and 1996–1998. Non-Black and non-White racial groups were excluded to ensure data validity. The median follow-up interval was 3.0 years (interquartile range 2.9–4.0). Examinations included 12-lead ECG, from which ERP data were derived. In addition, to rule out Brugada syndrome, we excluded leads V1 and V2 when selecting anterior leads, which specifically refer to leads V3 and V4. The lateral leads are defined as Leads I, aVL, V5, and V6, whereas the inferior leads are defined as Leads II, III, and aVF.

The ARIC study received approval from the ethics review boards of all participating institutions, and all participants provided written informed consent at enrollment. The present study was approved by the Ethics Review Committee of Guangdong Provincial People’s Hospital (KY2023-183-03).

In this analysis we excluded participants who: had missing or incomplete ECG data; exhibited significant ventricular conduction issues, such as complete left or right bundle branch block, Brugada syndrome, Wolff-Parkinson-White syndrome, QRS duration >120 ms, or pacemaker implantation; had arrhythmia events, such as ventricular arrhythmias, bradyarrhythmias, or AF, either previously or currently; experienced acute chest pain with ST-segment elevation on ECG; and were of non-Black or non-White races.

ECG Measurement

Our analysis used 12-lead ECGs collected during the baseline period from 1987 to 1989, as well as during subsequent periods from 1993 to 1995 and from 1996 to 1998, using the General Electric (GE) 12SL algorithm for automated measurements. For J-wave detection, we defined the J-wave as a notch or slur on the descending slope of the terminal portion of the QRS complex and used an automated J-wave detection method developed by Wang et al.,¹⁴ which has a sensitivity of 100% and a specificity of 94%. Other ECG variables included in this analysis were heart rate, left ventricular hypertrophy (LVH), and J-point elevation. LVH was defined according to the Cornell voltage criteria,¹⁵ whereas J-point elevation was defined as a J-point amplitude ≥0.1 mV in any lead.

ERP can only be confirmed when all specified standards are met,¹ namely an end-QRS notch or slur on the last 50% of the downslope of a prominent R wave, J wave amplitudes ≥0.1 mV in 2 or more contiguous ECG leads, and a QRS duration <120 ms.

Measurement of Temporal Variations in ERP

To assess temporal changes in ERP, we created a time-updated covariate representing the latest ERP for each participant throughout the study. This covariate is termed the time-updated ERP. Although we analyzed data from the second to the fourth visits, which involved updating the ERP status 3 times after the initial visit, not every subject had ECG data at each visit. For participants missing ERP data from the second to fourth visits, the last recorded ERP value was used as the most recent continuous ERP value. Consequently, different subjects may have between 1 and 4 ECG data points.

During the 4 visits, we defined the “time-updated ERP” as the ERP manifestation from its last occurrence on the ECG to the first occurrence of incident arrhythmias, death, loss to follow-up, or the end of the follow-up period. “Time-varying ERP” was defined as the ERP manifestation from its first occurrence on the ECG to the first occurrence of incident arrhythmias, death, loss to follow-up, or the end of the follow-up period. Detailed definitions are provided in the Supplementary Table.

Temporal changes in ERP status were assessed by comparing the presence and variations in ERP ECG findings between the latest visit and the preceding visit. We categorized these changes into 4 binary groups: (1) no ERP at both assessments (normal–normal, or consistently normal); (2) initially no ERP but later exhibiting ERP (normal–ERP, or new-onset ERP); (3) ERP present initially, but not in subsequent assessments (ERP–normal, or transient ERP); and (4) ERP at both assessments (ERP–ERP, or consistent ERP). For example, if a subject’s ERP status was ERP in the first examination, reverted to normal in the second and third examinations, and then presented ERP in the fourth examination, this subject would be classified as having new-onset ERP.

Ascertainment of Outcomes

In this study we primarily examined the incidence rates of ventricular arrhythmias, bradyarrhythmias, and AF. We identified new cases of arrhythmia using 3 methods: ECGs during research visits, hospital discharge codes, and death certificate analyses. Ventricular arrhythmias encompassed ventricular tachycardia (International Classification of Disease, Ninth Revision, Clinical Modification [ICD-9-CM] code 427.1), ventricular flutter/fibrillation (ICD-9-CM codes 427.4, 427.41, 427.42), cardiac arrest (ICD-9-CM code 427.5), and sudden cardiac death, defined as death originating from the heart, occurring without any apparent circumstances unrelated to arrhythmias.^16,17 Bradyarrhythmias were detected through diagnoses of atrioventricular block (ICD-9-CM codes 426, 426.0, 426.1, 426.10, 426.12, 426.13, 426.8, 426.89, 426.9) and sick sinus syndrome (ICD-9-CM code 427.81), along with pacemaker implantation procedures (ICD-9-CM codes V45.0, V45.01, V53.3, V53.31, 37.78, 37.8, 89.45, 89.46, 89.47, 89.48).^18,19 AF/flutter included AF (ICD-9-CM codes 427.3, 427.31; International Classification of Diseases, Tenth Revision [ICD-10] code I48) and atrial flutter (ICD-9-CM code 427.32).^16,20,21

Statistical Analysis

The baseline characteristics of the study population are presented as frequencies (percentages) for categorical variables and as the mean±SD for continuous variables. To assess the relationship between baseline ERP, time-updated ERP, time-varying ERP, and temporal changes in ERP status with various previously identified outcomes, we used the Cox proportional hazards model to calculate multivariable-adjusted hazard ratios (HR) for all study outcomes. The multivariable analysis was adjusted for known outcome predictors, including sex, race, and time-updated variables such as age, hypertension, smoking status, diabetes, body mass index (BMI), blood pressure, total cholesterol, fasting blood glucose, heart rate, QTc duration, the presence of LVH on the ECG, and the use of cardiac medications. Missing values in the covariates were imputed, and models were run for each imputed result, with the final estimates averaged using Rubin›s rules. In addition, we plotted Kaplan-Meier survival curves for different outcomes.

ERP status was classified based on the amplitude of the J wave (0.1–0.19 or ≥0.2 mV) and lead territories (inferior, lateral, or anterior). In the study, subgroup analyses were performed on various prespecified clinical factors potentially linked to the risk of arrhythmias, and P values for interactions were calculated for each subgroup. These risk factors encompassed age, race, sex, hypertension, diabetes, dyslipidemia, smoking status, alcohol consumption, and BMI. Two-sided P<0.05 was considered statistically significant. However, for subgroup analyses, Bonferroni correction was used for multiple comparisons, with P<0.005 (0.05/9) indicating statistically significant differences. Analyses were performed using Stata version 15.0.

Results

Baseline Characteristics

The study included 6,700 (45.6%) men and 10,932 (74.5%) White participants, with a mean age of 54.3±5.8 years. At baseline, 5,136 (35.0%) participants had hypertension, 1,778 (12.1%) had diabetes, and 2,781 (18.9%) were former drinkers. In terms of cardiac medications, 533 (3.6%) participants were prescribed calcium antagonists, 244 (1.7%) were taking digitalis, and 799 (5.4%) were on β-blockers. Detailed baseline characteristics according to ERP status are provided in Table 1.

Table 1.

Characteristics of Participants According to Baseline ERP Status

	All subjects (n=14,679)	No ERP (n=12,757)	ERP (n=1,922)	P value
Age (years)	54.3±5.8	54.3±5.8	54.2±5.8	0.41
Male sex	6,700 (45.6)	5,235 (41.0)	1,465 (76.2)	<0.001
White	10,932 (74.5)	9,974 (78.2)	958 (49.8)	<0.001
Education				<0.001
<High school	3,500 (23.8)	2,880 (22.6)	620 (32.3)
High school/vocational school	5,995 (40.8)	5,355 (42.0)	640 (33.3)
College, graduate, or professional school	5,184 (35.3)	4,522 (35.4)	662 (34.4)
Income (US$)				<0.001
<16,000	3,167 (21.6)	2,613 (20.5)	554 (28.8)
16,000 to <25,000	2,068 (14.1)	1,780 (14.0)	288 (15.0)
25,000 to <35,000	3,299 (22.5)	2,897 (22.7)	402 (20.9)
35,000 to <50,000	2,679 (18.3)	2,378 (18.6)	301 (15.7)
≥50,000	3,466 (23.6)	3,089 (24.2)	377 (19.6)
Smoking status				<0.001
Never	6,027 (41.1)	5,399 (42.3)	628 (32.7)
Former	4,768 (32.5)	4,124 (32.3)	644 (33.5)
Current	3,884 (26.5)	3,234 (25.4)	650 (33.8)
Alcohol drinking				<0.001
Never	3,637 (24.8)	3,209 (25.2)	428 (22.3)
Former	2,781 (18.9)	2,331 (18.3)	450 (23.4)
Current	8,261 (56.3)	7,217 (56.6)	1,044 (54.3)
Physical activity (MET min/week)	3.3±3.0	3.3±3.0	3.3±3.2	0.17
Hypertension	5,136 (35.0)	4,346 (34.1)	790 (41.1)	<0.001
Diabetes	1,778 (12.1)	1,498 (11.7)	280 (14.6)	<0.001
Blood pressure (mmHg)
Systolic blood pressure	121.4±19.0	120.8±18.4	125.3±22.1	<0.001
Diastolic blood pressure	73.7±11.3	73.2±10.9	76.7±13.0	<0.001
Body mass index (kg/m2)	27.7±5.4	27.8±5.5	26.6±4.7	<0.001
Laboratory values, mean
Total cholesterol (mmol/L)	5.6±1.1	5.6±1.1	5.5±1.1	<0.001
LDL-C (mmol/L)	3.6±1.0	3.6±1.0	3.6±1.0	0.43
HDL-C (mmol/L)	1.3±0.4	1.3±0.4	1.3±0.5	0.33
Triglycerides (mmol/L)	1.5±1.0	1.5±1.0	1.4±0.9	<0.001
Fasting blood glucose (mmol/L)	6.1±2.3	6.0±2.2	6.2±2.6	0.002
Electrocardiographic findings
Heart rate (beats/min)	66.3±10.3	66.6±10.2	64.5±10.7	<0.001
QTc duration (ms)	416.4±19.8	416.7±19.1	415.0±24.1	<0.001
LVH on ECG	332 (2.3)	167 (1.3)	165 (8.6)	<0.001
Medication
Calcium antagonist	533 (3.6)	445 (3.5)	88 (4.6)	0.02
Digitalis	244 (1.7)	191 (1.5)	53 (2.8)	<0.001
β-blocker	799 (5.4)	706 (5.5)	93 (4.8)	0.21

Unless indicated otherwise, data are given as the mean±SD or n (%). ECG, electrocardiogram; ERP, early repolarization pattern; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; LVH, left ventricular hypertrophy; MET, metabolic equivalents.

Among subjects with ERP, 1,465 (76.2%) were male and 958 (49.8%) were White. These subjects exhibited lower total cholesterol and triglyceride levels, higher systolic and diastolic blood pressures, and increased fasting blood glucose levels. ECG findings showed that those with ERP had reduced heart rates and QTc durations, as well as signs of LVH. In addition, those with ERP were more likely to use digitalis, have less than a high school education, earn less than US$16,000 annually, smoke (33.8% vs. 25.4%; P<0.001), be former alcohol drinkers (23.4% vs. 18.3%; P<0.001), have diabetes (14.2% vs. 10.7%; P<0.001), and have hypertension (42.9% vs. 36%; P<0.001) than those without ERP.

Associations Between Baseline ERP and Incident Arrhythmias

Over a follow-up period of over 20 years, there were 1,252 cases of ventricular arrhythmias, 890 cases of bradyarrhythmias, and 2,202 cases of AF. Table 2 presents unadjusted and adjusted risk ratios for arrhythmic events based on baseline ERP. Among those with ERP, the incidence of ventricular arrhythmias, bradyarrhythmias, and AF was 8.1, 4.4, and 8.8 per 1,000 person-years, respectively. Compared with those without ERP, the rate changes were significant at 3.3 and 0.9 per 1,000 person-years for ventricular arrhythmias and bradyarrhythmias, respectively, with no significant difference in AF, as indicated in Table 2.

Table 2.

Unadjusted and Adjusted HRs of Incident Arrhythmias According to ERP Phenotypes at Baseline

	Baseline ERP		J wave amplitude		J wave territory
	No ERP (n=12,757)	Any ERP (n=1,922)	0.1–0.19 mV (n=1,731)	≥0.2 mV (n=191)	Anterior leads (n=1,717)	Lateral leads (n=642)	Inferior leads (n=119)
Ventricular arrhythmias
No. events	1,252	298	259	39	273	97	13
Incidence rate (95% CI) per 1,000 person years	4.8 (4.5–5.1)	8.1 (7.2–9.1)	7.7 (6.8–8.7)	12.2 (8.7–16.7)	8.4 (7.4–9.4)	7.8 (6.3–9.5)	5.7 (3.1–9.8)
Unadjusted HR (95% CI)	1.00	1.73 (1.52–1.96)	1.64 (1.44–1.88)	2.66 (1.93–3.66)	1.78 (1.56–2.03)	1.51 (1.23–1.86)	1.05 (0.61–1.82)
Age-, sex-, and race-adjusted HRs (95% CI)	1.00	1.23 (1.08–1.42)	1.20 (1.04–1.39)	1.55 (1.12–2.14)	1.26 (1.09–1.45)	1.02 (0.82–1.26)	0.86 (0.50–1.48)
Multivariate adjustedA HR (95% CI)	1.00	1.31 (1.14–1.51)	1.28 (1.11–1.48)	1.69 (1.21–2.35)	1.32 (1.15–1.53)	1.18 (0.95–1.46)	1.07 (0.62–1.85)
Bradyarrhythmias
No. events	890	161	138	23	150	46	11
Incidence rate (95% CI) per 1,000 person years	3.4 (3.2–3.6)	4.4 (3.7–5.1)	4.1 (3.5–4.9)	7.2 (4.6–10.8)	4.6 (3.9–5.4)	3.7 (2.7–4.9)	4.8 (2.4–8.7)
Unadjusted HR (95% CI)	1.00	1.33 (1.12–1.57)	1.24 (1.04–1.49)	2.22 (1.47–3.37)	1.40 (1.18–1.67)	1.02 (0.76–1.38)	1.26 (0.69–2.27)
Age-, sex-, and race-adjusted HR (95% CI)	1.00	1.28 (1.07–1.53)	1.22 (1.01–1.47)	1.84 (1.21–2.81)	1.35 (1.12–1.62)	1.06 (0.78–1.44)	1.13 (0.62–2.05)
Multivariate adjustedA HR (95% CI)	1.00	1.44 (1.21–1.73)	1.37 (1.14–1.66)	2.18 (1.42–3.35)	1.51 (1.25–1.82)	1.24 (0.91–1.68)	1.38 (0.76–2.51)
Atrial fibrillation
No. events	2,202	316	279	37	290	101	16
Incidence rate (95% CI) per 1,000 person years	8.6 (8.2–8.9)	8.8 (7.9–9.9)	8.5 (7.5–9.5)	13.1 (9.2–18.1)	9.1 (8.1–10.3)	8.3 (6.7–10.0)	7.6 (4.3–12.3)
Unadjusted HR (95% CI)	1.00	1.04 (0.92–1.17)	1.00 (0.88–1.14)	1.51 (1.07–2.13)	1.09 (0.96–1.23)	0.94 (0.77–1.15)	0.68 (0.39–1.21)
Age-, sex-, and race-adjusted HR (95% CI)	1.00	1.01 (0.89–1.15)	0.98 (0.86–1.12)	1.37 (0.96–1.93)	1.06 (0.93–1.20)	1.02 (0.83–1.26)	0.70 (0.40–1.23)
Multivariate adjustedA HRs (95% CI)	1.00	1.11 (0.98–1.26)	1.07 (0.94–1.22)	1.58 (1.11–2.24)	1.15 (1.01–1.31)	1.19 (0.96–1.47)	0.79 (0.45–1.40)

^AVariables that were included in the multivariate analyses were age, race, sex, education, income, hypertension, smoking status, alcohol drinking, diabetes, physical activity, body mass index, systolic blood pressure, diastolic blood pressure, total cholesterol, LDL-C, HDL-C, triglycerides, fasting blood glucose, heart rate, QTc duration, presence of LVH on ECG, and the use of cardiac medications. CI, confidence interval; HR, hazard ratio. Other abbreviations as in Table 1.

Individuals with baseline ERP had higher risks of ventricular arrhythmias (unadjusted HR 1.73; 95% CI 1.52–1.96) and bradyarrhythmias (unadjusted HR 1.33; 95% CI 1.12–1.57) than those without ERP. The HR for AF was 1.04 (95% CI 0.92–1.17). After adjusting for age, sex, and race, the HRs were 1.23 (95% CI 1.08–1.42) for ventricular arrhythmias, 1.28 (95% CI 1.07–1.53) for bradyarrhythmias, and 1.01 (95% CI 0.89–1.15) for AF. Multivariate adjustments yielded HRs of 1.31 (95% CI 1.14–1.51) for ventricular arrhythmias, 1.44 (95% CI 1.21–1.73) for bradyarrhythmias, and 1.11 (95% CI 0.98–1.26) for AF.

Associations Between J-Wave Amplitude in ERP and Incident Arrhythmias

In our study cohort, ERP was present in 1,922 (13.1%) individuals. Among these individuals, 1,731 (90.1%) had a J wave amplitude between 0.1 and 0.19 mV, and 191 (9.9%) individuals had an amplitude of at least 0.2 mV.

Adjustments for multiple factors confirmed that individuals with J wave amplitudes of 0.1–0.19 mV had significantly elevated risks of ventricular arrhythmias (adjusted HR 1.28; 95% CI 1.11–1.48; P<0.001) and bradyarrhythmias (adjusted HR 1.37; 95% CI 1.14–1.66; P=0.02) compared with those without ERP. Those with J wave amplitudes of at least 0.2 mV faced even higher risks, with adjusted HRs of 1.69 (95% CI 1.21–2.35; P<0.001) for ventricular arrhythmias, 2.18 (95% CI 1.42–3.35; P<0.001) for bradyarrhythmias, and 1.58 (95% CI 1.11–2.24; P=0.02) for AF (Table 2). The Kaplan-Meier curve in Figure 1 illustrates the relationship between J wave amplitude in the ERP and the examined arrhythmias.

Figure 1.

Kaplan-Meier curves for (A) ventricular arrhythmias, (B) bradyarrhythmias, and (C) atrial fibrillation in subjects with an early repolarization pattern (ERP).

Associations Between J-Wave Territory in ERP and Incident Arrhythmias

Those with anterior leads in ERP exhibited increased risks of developing ventricular arrhythmias (adjusted HR 1.32; 95% CI 1.15–1.53), bradyarrhythmias (adjusted HR 1.51; 95% CI 1.25–1.82), and AF (adjusted HR 1.15; 95% CI 1.01–1.31). However, for those with lateral and inferior leads in ERP, the risks for these arrhythmias did not differ significantly from risks in those without ERP, as shown in the multivariable adjusted model (Table 2).

Incident Arrhythmias Associated With Time-Updated and Time-Varying ERP

Subjects with time-varying ERP had a higher risk of 3 types of arrhythmias than those without ERP. After adjusting for various factors, the risk ratios were 1.47 (95% CI 1.26–1.72) for ventricular arrhythmia, 1.58 (95% CI 1.31–1.92) for bradyarrhythmias, and 1.17 (95% CI 1.02–1.34) for AF. When considering the time differences, the duration between the baseline ERP and the time-varying ERP was 7.34 years (95% CI 6.48–8.21 years) for ventricular arrhythmias, 8.44 years (95% CI 7.60–9.28 years) for bradyarrhythmias, and 7.48 years (95% CI 6.77–8.20 years) for AF. The risk for time-varying ERP was higher than for baseline ERP, and patients with time-updated ERP had significantly elevated risks compared with those with baseline or time-varying ERP. Further adjustments revealed HRs of 1.55 (95% CI 1.35–1.77) for ventricular arrhythmias, 1.76 (95% CI 1.48–2.08) for bradyarrhythmias, and 1.25 (95% CI 1.10–1.43) for AF (Table 3).

Table 3.

Unadjusted and Adjusted HRs of Incident Arrhythmias Associated With Time-Updated and Time-Varying ERP

	Time-updated ERP			Time-varying ERP
	No ERP (n=12,833)	ERP (n=1,846)	P value	No ERP (n=12,757)	ERP (n=1,922)	P value
Ventricular arrhythmias
No. events	1,226	324		1,252	298
Unadjusted HR (95% CI)	1.00	1.97 (1.74–2.23)	<0.001	1.00	1.96 (1.69–2.27)	<0.001
Age-, sex-, and race-adjusted HR (95% CI)	1.00	1.52 (1.33–1.73)	<0.001	1.00	1.48 (1.27–1.72)	<0.001
Multivariate adjustedA HR (95% CI)	1.00	1.55 (1.35–1.77)	<0.001	1.00	1.47 (1.26–1.72)	<0.001
Bradyarrhythmias
No. events	864	187		890	161
Unadjusted HR (95% CI)	1.00	1.62 (1.39–1.90)	<0.001	1.00	1.64 (1.36–1.97)	<0.001
Age-, sex-, and race-adjusted HR (95% CI)	1.00	1.62 (1.37–1.91)	<0.001	1.00	1.50 (1.24–1.81)	<0.001
Multivariate adjustedA HR (95% CI)	1.00	1.76 (1.48–2.08)	<0.001	1.00	1.58 (1.31–1.92)	<0.001
Atrial fibrillation
No. events	2,186	332		2,202	316
Unadjusted HR (95% CI)	1.00	1.20 (1.06–1.36)	0.004	1.00	1.17 (1.02–1.33)	0.02
Age-, sex-, and race-adjusted HR (95% CI)	1.00	1.18 (1.03–1.43)	0.01	1.00	1.11 (0.97–1.28)	0.13
Multivariate adjustedA HR (95% CI)	1.00	1.25 (1.10–1.43)	0.001	1.00	1.17 (1.02–1.34)	0.03

^AVariables that were included in the multivariate analyses were age, race, sex, education, income, hypertension, smoking status, alcohol drinking, diabetes, physical activity, body mass index, systolic blood pressure, diastolic blood pressure, total cholesterol, LDL-C, HDL-C, triglycerides, fasting blood glucose, heart rate, QTc duration, presence of LVH on ECG, and the use of cardiac medications. Abbreviations as in Tables 1,2.

Incident Arrhythmias Associated With Temporal Changes in ERP Status

Multiple ECG measurements were made in 13,357 participants. Among these individuals, 11,359 (85.0%) consistently exhibited normal ECG results, 477 (3.6%) exhibited transient ERP, 470 (3.5%) developed new-onset ERP, and 1,051 (7.9%) consistently showed ERP results in the last 2 assessments.

Individuals with new-onset ERP demonstrated a heightened risk of ventricular arrhythmias (adjusted HR 1.37; 95% CI 1.03–1.82), bradyarrhythmias (adjusted HR 1.81; 95% CI 1.34–2.45), and AF (adjusted HR 1.28; 95% CI 1.01–1.62), compared with those with consistently normal ECG results. In contrast, subjects with transient ERP did not show significantly elevated risks for these outcomes in the multivariable-adjusted model. Notably, compared with patients with persistently normal ECG results, participants with transient ERP were generally younger (β −0.70; 95% CI −0.17, −1.22; P=0.009) and had lower heart rates (β −2.09; 95% CI −1.07, −3.11; P<0.001). Furthermore, individuals with consistent ERP also faced an increased risk of ventricular arrhythmias (adjusted HR 1.59; 95% CI 1.35–1.87), bradyarrhythmias (adjusted HR 1.78; 95% CI 1.46–2.17), and AF (adjusted HR 1.26; 95% CI 1.07–1.48; Table 4).

Table 4.

Unadjusted and Adjusted HR of Incident Arrhythmias Associated With Temporal Changes in ERP Status

Variable	Normal-normal (n=11,359)	ERP-normal (n=477)	P value	Normal-ERP (n=470)	P value	ERP-ERP (n=1,051)	P value
Ventricular arrhythmias
No. events	1,015	45		51		206
Unadjusted HR (95% CI)	1.00	1.48 (1.10–1.99)	0.01	1.48 (1.11–1.99)	0.007	2.18 (1.87–2.53)	<0.001
Age- and sex-adjusted HR (95% CI)	1.00	1.21 (0.89–1.63)	0.22	1.26 (0.95–1.67)	0.11	1.61 (1.37–1.88)	<0.001
Multivariate adjustedA HR (95% CI)	1.00	1.28 (0.95–1.74)	0.11	1.37 (1.03–1.82)	0.03	1.59 (1.35–1.87)	<0.001
Bradyarrhythmias
No. of events	802	21		45		128
Unadjusted HR (95% CI)	1.00	0.95 (0.62–1.47)	0.83	1.84 (1.36–2.49)	<0.001	1.68 (1.39–2.02)	<0.001
No. events	1.00	0.96 (0.62–1.48)	0.84	1.70 (1.25–2.30)	0.001	1.63 (1.34–1.98)	<0.001
Unadjusted HR (95% CI)	1.00	1.03 (0.66–1.59)	0.90	1.81 (1.34–2.45)	<0.001	1.78 (1.46–2.17)	<0.001
Atrial fibrillation
No. events	2,001	63		91		213
Unadjusted HR (95% CI)	1.00	0.95 (0.73–1.25)	0.74	1.24 (0.99–1.57)	0.07	1.23 (1.06–1.43)	0.008
Age- and sex-adjusted HR (95% CI)	1.00	0.93 (0.71–1.23)	0.63	1.20 (0.95–1.51)	0.13	1.18 (1.01–1.39)	0.04
Multivariate adjustedA HR (95% CI)	1.00	1.01 (0.76–1.33)	0.97	1.28 (1.01–1.62)	0.04	1.26 (1.07–1.48)	0.005

^AVariables that were included in the multivariate analyses were age, race, sex, education, income, hypertension, smoking status, alcohol drinking, diabetes, physical activity, body mass index, systolic blood pressure, diastolic blood pressure, total cholesterol, LDL-C, HDL-C, triglycerides, fasting blood glucose, heart rate, QTc duration, presence of LVH on ECG, and the use of cardiac medications. Abbreviations as in Tables 1,2.

Stratified Analysis for Time-Updated ERP

In a stratified analysis of time-updated ERP, it was found that the risk of ventricular arrhythmias significantly interacts with race, dyslipidemia, and BMI (Figure 2). Specifically, the time-updated ERP in White participants was strongly associated with the risk of ventricular arrhythmias (adjusted HR 1.83; 95% CI 1.54–2.16), whereas the correlation was weaker in Black participants (adjusted HR 1.31; 95% CI 1.06–1.62; P interaction=0.001). Among subjects with dyslipidemia, the risk increased significantly (adjusted HR 1.73; 95% CI 1.50–2.01), whereas no significant changes were observed in those without dyslipidemia (adjusted HR 1.01; 95% CI 0.74–1.37; P interaction=0.003). In addition, the risk was higher in obese subjects compared to those with normal weight (adjusted HR 2.22; 95% CI 1.75–2.81), while the risk was relatively lower in overweight subjects compared to those with normal weight (adjusted HR 1.47; 95% CI 1.20–1.80; P interaction=0.001).

Figure 2.

Stratified analysis of hazard ratios (HRs) for (A) ventricular arrhythmias, (B) bradyarrhythmias, and (C) atrial fibrillation associated with time-updated early repolarization pattern. *Variables that were included in the multivariate analyses were age, race, sex, education, income, hypertension, smoking status, alcohol drinking, diabetes, physical activity, body mass index (BMI), total cholesterol, fasting blood glucose, heart rate, QTc duration, the presence of left ventricular hypertrophy on electrocardiography, and the use of cardiac medications. ^†Participants were diagnosed with dyslipidaemia if they met US National Institutes of Health guidelines for dyslipidaemia: total cholesterol ≥5.7 mmol/L or low-density lipoprotein cholesterol ≥3.6 mmol/L or high-density lipoprotein cholesterol <1.0 mmol/L or triglycerides ≥1.7 mmol/L or pharmacological treatment for dyslipidaemia. CI, confidence interval.

We also examined the relationship between ECG findings and time-updated ERP-related arrhythmic events in obese patients. LVH was linked to an increased risk of ventricular arrhythmias (adjusted HR 2.10; 95% CI 1.47–3.00; P<0.001), bradyarrhythmias (HR 3.29; 95% CI 2.00–5.39; P<0.001), and AF (HR 1.81; 95% CI 1.25–2.61; P<0.001). QTc prolongation was also associated with risks of ventricular arrhythmias (HR 1.10; 95% CI 1.07–1.14; P<0.001), bradyarrhythmias (HR 1.11; 95% CI 1.06–1.16; P<0.001), and AF (HR 1.06; 95% CI 1.02–1.09; P<0.001).

The incidence of LVH was significantly higher in obese than non-obese patients, with an adjusted odds ratio (OR) of 3.02 (95% CI 2.18–3.87; P<0.001). Obese individuals also had a higher incidence of QT prolongation (adjusted OR 1.96; 95% CI 1.47–2.62; P<0.001). Before adjusting for LVH and QT prolongation, the HRs in obese individuals with a time-updated ERP were 2.93 (95% CI 2.16–3.97; P<0.001) for ventricular arrhythmias, 2.73 (95% CI 1.86–4.01; P<0.001) for bradyarrhythmias, and 1.85 (95% CI 1.59–2.15; P<0.001) for AF. After adjustments, the HRs decreased to 2.22 (95% CI 1.75–2.81; P=0.001) for ventricular arrhythmias, 2.44 (95% CI 1.78–3.33; P=0.005) for bradyarrhythmias, and 1.50 (95% CI 1.17–1.91; P=0.02) for AF. These findings indicate that adjusting for LVH and QT prolongation reduced the arrhythmia risk associated with obesity.

Although a history of hypertension was associated with the incidence of ERP, its impact on the relationship between ventricular arrhythmias and time-updated ERP was not significant (P interaction=0.23). Furthermore, the interaction between sex and BMI significantly increased the risk of bradyarrhythmias, but no significant interaction was observed with the increased risk of AF.

Discussion

In this community-based cohort of 14,679 middle-aged participants, significant findings have emerged from monitoring the biracial population over a period of 20 years. Initially, individuals with baseline ERP demonstrated significantly elevated risks of ventricular arrhythmias and bradyarrhythmias. In addition, various forms of ERP, including time-varying, time-updated, new-onset, and consistent ERP, are linked to the occurrence of incident arrhythmias. However, for those with transient ERP, the condition appears to be benign. Furthermore, as the amplitude of the J wave and the involvement of anterior leads increased, there was an indication of poorer outcomes for incident arrhythmias. Obese participants exhibited a rising trend in the risk of time-updated ERP linked to incident arrhythmias.

Although the exact mechanism by which ERP increases the risk of arrhythmias is not yet clear, recent experimental evidence indicates heterogeneity in the ERP region, reflecting the dispersion gradient of myocardial refractoriness due to the net outward flow of repolarization currents.²² The suppression of the action potential plateau in the epicardial region leads to transmembrane gradient formation. Although this phenomenon itself may not be arrhythmogenic, the increase in repolarization currents, the loss of the epicardial action potential dome, and the deepening of repolarization dispersion may lower the threshold for early afterdepolarizations, thereby promoting ventricular arrhythmias.^23,24 Furthermore, the heritability of ERP is linked to mutations in cardiac ion channel-encoding genes, such as those encoding sodium, calcium, and potassium channels, which may underlie the development of AF and bradyarrhythmias in ERP patients.^25,26 In addition, studies have shown that mutations in genes encoding potassium channels, such as potassium inwardly rectifying channel subfamily J member 8 (KCNJ8 and KCND3) and potassium voltage-gated channel subfamily D member 3 (KCND3), are related to mechanisms of ventricular and atrial arrhythmias, further increasing the risk of AF in ERP patients.^27,28 Moreover, research by Watanabe et al. indicates that patients with idiopathic ventricular fibrillation carrying sodium voltage-gated channel α subunit 5 (SCN5A) mutations exhibit bradycardia and slowed cardiac conduction, suggesting that some individuals identified as ERP patients may have depolarization abnormalities.²⁹

In early studies, the ERP was considered a persistent feature of the ECG. However, as understanding has grown, it has become increasingly clear that ERP is influenced by various factors, including genetic elements, metabolic indicators, and physical activity.^30,31 Therefore, ERP should be regarded as a dynamic ECG phenomenon that changes over time. Our findings indicate that time-updated ERP has a significantly higher correlation with event-related arrhythmias than baseline ERP and time-varying ERP. Furthermore, both newly emerged and consistent ERP are significantly associated with the occurrence of arrhythmias, suggesting they share similar characteristics with time-updated ERP. Transient ERP appears to be a benign phenomenon, while tracking the amplitude and area of J waves over time may help identify individuals at risk of developing arrhythmias.

As mentioned earlier, the generation of ERP may result from complex interactions between gene mutations encoding the channels and molecular mechanisms, affecting the cardiac repolarization process, particularly through changes in potassium currents and fluctuations in sodium and calcium ion levels.^22–24 Therefore, consistent ERP may reflect the clinical phenotype of cardiac diseases or genetic susceptibility, which, in turn, affects the occurrence of arrhythmic events.²⁵ Furthermore, the emergence of new-onset ERP typically indicates significant changes in acute conditions, such as acute myocardial ischemia, acute electrolyte imbalances, and acute metabolic disturbances, all of which may heighten the risk of arrhythmias.^6,32 In addition, new-onset ERP may progress to consistent ERP and subsequently facilitate the occurrence of incident arrhythmias. In contrast, transient ERP may reflect a temporary state indicating electrophysiological instability rather than a consistent risk factor. This phenomenon is particularly evident in young individuals, those with lower heart rates, Black individuals, and men.^8,33 Compared with patients with persistently normal electrocardiographic results, participants with transient ERPs were generally younger and had lower heart rates. Younger age and lower heart rate typically protected the heart from the effects of malignant ventricular arrhythmias, thereby supporting the benign characteristics of transient ERP.

Consistent with previous studies, an increase in J wave amplitude is linked to a higher risk of malignant arrhythmias in ERP patients.³⁴ In those with J wave amplitudes of 0.1–0.19 mV, the difference in risk between ventricular arrhythmias and AF may be attributed to distinct electrophysiological mechanisms. J wave amplitude reflects early repolarization of epicardial action potentials in the ventricles, which may have promoted ventricular arrhythmias, whereas AF is more influenced by autonomic tone and atrial remodeling.^7,23,24 However, in this study we found that only the anterior ERP pattern has significant predictive value, whereas earlier research often associated inferior or inferolateral ERP patterns with arrhythmia occurrence.^3,23 These differences in results may be due to variations in study populations and methodology. In addition, research by Kamakura et al. identified that in patients with inferolateral early repolarization syndrome, the presence of J waves in the anterior leads serves as a critical predictor of poor prognosis.³⁵ These patients exhibit notable differences in attack patterns and diurnal distribution compared with those with J waves confined solely to the inferolateral leads. Furthermore, electrophysiological studies involving 52 ERP patients revealed that the occurrence of ventricular fibrillation is linked to the origin of the J waves. Specifically, most patients with J waves originating from the anterior leads exhibit depolarization abnormalities in the right ventricular epicardium, which may represent the primary electrophysiological abnormality triggering ventricular fibrillation. This finding challenges the conventional perspective on repolarization abnormalities and underscores the significance of right ventricular epicardial depolarization abnormalities in anterior lead ERP.³⁶ These findings somewhat support the relationship between ERP originating from anterior leads and the occurrence of arrhythmias.

Our research shows that the risk of ventricular arrhythmias linked to time-updated ERP is higher in White than Black populations, and women are more susceptible to bradyarrhythmias related to time-updated ERP. This may be due to genetic susceptibility differences influenced by race and sex. In addition, individuals with a low BMI exhibited a higher incidence of ERP in the general population. However, in the time-updated ERP group, obese individuals had a significantly greater risk of arrhythmias, particularly ventricular arrhythmias and bradyarrhythmias, than non-obese individuals. One possible explanation is that obesity may contribute to concentric LVH and QT interval prolongation, which could, to some extent, increase the likelihood of ventricular arrhythmias.³⁷ Furthermore, a large cohort study from China found that obese individuals have a higher risk of cardiac conduction diseases, particularly atrioventricular block, which may be closely related to factors such as inflammation, fibrosis, and oxidative stress.³⁸ However, the specific link between obesity and bradyarrhythmias remains unconfirmed, requiring further research to clarify the underlying pathogenic mechanisms.

Strengths and Limitations

This study is the first prospective cohort study to explore the relationship between different patterns of ERP and their temporal changes with various arrhythmias in a healthy middle-aged population. However, the research has some limitations. First, the sample includes only individuals aged 45–64 years, which restricts the assessment of the effect of ERP across different age groups. Second, due to the purely epidemiological nature of this study, we did not delve into the pathophysiological and etiological factors related to ERP. In addition, our reliance on automated detection methods means we are unable to provide individual ECG records of ERP patients to events. The complexity of interpreting ECGs inherently carries a risk of false negatives, particularly in the ECG presentations of patients with atypical Brugada syndrome. Moreover, not every subject underwent an ECG at each follow-up visit, which may result in incomplete records of ERP status. The ERP status on the ECG for some patients may also change during subsequent follow-up, potentially influencing the accuracy of the grouping. Furthermore, although previous studies have validated the effectiveness of diagnostic codes from different sources in classifying arrhythmia outcomes, these codes may lack sufficient sensitivity, potentially leading to the underdiagnosis or misclassification of certain arrhythmias.

Conclusions

Our research indicated a significant correlation between ERP occurrence and both ventricular arrhythmias and bradyarrhythmias. Unlike transient ERP, various forms of ERP, including time-varying, time-updated, new-onset developed, and consistent ERP, were associated with arrhythmia incidence in a middle-aged biracial (Black and White) population. Special attention should be given to individuals with time-updated ERP who exhibit high-amplitude J waves and involvement of the anterior leads. Furthermore, obesity has been found to increase the susceptibility of patients with time-updating ERP to ventricular arrhythmias and bradyarrhythmias.

Acknowledgments

The authors thank the staff and participants of the ARIC study and the Biobank and Data Repository Information Coordinating Center for their contributions.

Sources of Funding

This study was supported by the grants from the National Natural Science Foundation of China (82270333, 81600260), the Natural Science Foundation of Guangdong Province (2024A1515013067), Guangzhou Science and Technology Program (2024B03J1344), High-level Talents Introduction Plan of Guangdong Provincial People’s Hospital (KY012023007), Clinical Research Special Fund of Guangdong Medical Association (2024HY-A6002), and National Science and Technology Innovation Major Project-Research Project on Prevention and Treatment of Cancer, Cardiovascular, Respiratory and Metabolic Diseases (2023ZD0504202, 2023ZD0504204).

Disclosures

The authors declare no conflicts of interest.

Author Contributions

Q.H., Y.-J.C., and X.-H.F. contributed to research design, data collection, analysis, and manuscript preparation. Y.-M.X., Y.-J.L., and J.-J.W. contributed to data acquisition and revision. Y.-L.C., M.-J.H., M.-P.L., and H.-L.Z. contributed to protocol design and revision. Z.-E.C., Q.W., S.-L.L., and S.-L.W. contributed to data interpretation and revision. Q.H., Y.-J.L., J.-J.W., X.-H.F., and Y.-J.C. contributed to protocol preparation and final revision. All authors approved the final version and are accountable for the work’s integrity.

IRB Information

This study was approved by the Ethics Review Committee of Guangdong Provincial People’s Hospital (KY2023-183-03).

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-24-0964

References

• 1.   Macfarlane PW, Antzelevitch C, Haissaguerre M, Huikuri HV, Potse M, Rosso R, et al. The early repolarization pattern: A consensus paper. J Am Coll Cardiol 2015; 66: 470–477.
• 2.   Takase B, Ikeda T, Shimizu W, Abe H, Aiba T, Chinushi M, et al. JCS/JHRS 2022 guideline on diagnosis and risk assessment of arrhythmia. Circ J 2024; 88: 1509–1595.
• 3.   Wu SH, Lin XX, Cheng YJ, Qiang CC, Zhang J. Early repolarization pattern and risk for arrhythmia death: A meta-analysis. J Am Coll Cardiol 2013; 61: 645–650.
• 4.   Junttila MJ, Tikkanen JT, Kenttä T, Anttonen O, Aro AL, Porthan K, et al. Early repolarization as a predictor of arrhythmic and nonarrhythmic cardiac events in middle-aged subjects. Heart Rhythm 2014; 11: 1701–1706.
• 5.   Siebermair J, Sinner MF, Beckmann BM, Laubender RP, Martens E, Sattler S, et al. Early repolarization pattern is the strongest predictor of arrhythmia recurrence in patients with idiopathic ventricular fibrillation: Results from a single centre long-term follow-up over 20 years. Europace 2016; 18: 718–725.
• 6.   Kitamura T, Fukamizu S, Hojo R, Aoyama Y, Komiyama K, Nishizaki M, et al. Early repolarization pattern and its day-to-day dynamic change as markers for ventricular fibrillation in patients with vasospastic angina. Europace 2016; 18: 1252–1258.
• 7.   Hasegawa Y, Watanabe H, Ikami Y, Otsuki S, Iijima K, Yagihara N, et al. Early repolarization and risk of lone atrial fibrillation. J Cardiovasc Electrophysiol 2019; 30: 565–568.
• 8.   Uberoi A, Jain NA, Perez M, Weinkopff A, Ashley E, Hadley D, et al. Early repolarization in an ambulatory clinical population. Circulation 2011; 124: 2208–2214.
• 9.   Serra-Grima R, Doñate M, Álvarez-García J, Barradas-Pires A, Ferrero A, Carballeira L, et al. Long-term follow-up of early repolarization pattern in elite athletes. Am J Med 2015; 128: 192.e1–9.
• 10.   Patton KK, Ellinor PT, Ezekowitz M, Kowey P, Lubitz SA, Perez M, et al. Electrocardiographic early repolarization: A scientific statement from the American Heart Association. Circulation 2016; 133: 1520–1529.
• 11.   Halasz G, Cattaneo M, Piepoli M, Biagi A, Romano S, Biasini V, et al. Early repolarization in pediatric athletes: A dynamic electrocardiographic pattern with benign prognosis. J Am Heart Assoc 2021; 10: e020776.
• 12.   The Atherosclerosis Risk in Communities (ARIC) study: Design and objectives. The ARIC investigators. Am J Epidemiol 1989; 129: 687–702.
• 13.   Wright JD, Folsom AR, Coresh J, Sharrett AR, Couper D, Wagenknecht LE, et al. The ARIC (Atherosclerosis Risk In Communities) study: JACC focus seminar 3/8. J Am Coll Cardiol 2021; 77: 2939–2959.
• 14.   Wang YG, Wu HT, Daubechies I, Li Y, Estes EH, Soliman EZ. Automated J wave detection from digital 12-lead electrocardiogram. J Electrocardiol 2015; 48: 21–28.
• 15.   Devereux RB, Casale PN, Eisenberg RR, Miller DH, Kligfield P. Electrocardiographic detection of left ventricular hypertrophy using echocardiographic determination of left ventricular mass as the reference standard: Comparison of standard criteria, computer diagnosis and physician interpretation. J Am Coll Cardiol 1984; 3: 82–87.
• 16.   Cheng YJ, Deng H, Liao YJ, Fang XH, Liao HT, Liu FZ, et al. Role of ideal cardiovascular health metrics in reducing risk of incident arrhythmias. Eur J Prev Cardiol 2024; 31: 658–666.
• 17.   Zhao D, Post WS, Blasco-Colmenares E, Cheng A, Zhang Y, Deo R, et al. Racial differences in sudden cardiac death. Circulation 2019; 139: 1688–1697.
• 18.   Zhang J, Liu J, Ye M, Zhang M, Yao F, Cheng Y. Incidence and risk factors associated with atrioventricular block in the general population: The atherosclerosis risk in communities study and Cardiovascular Health Study. BMC Cardiovasc Disord 2024; 24: 509.
• 19.   Jensen PN, Gronroos NN, Chen LY, Folsom AR, deFilippi C, Heckbert SR, et al. Incidence of and risk factors for sick sinus syndrome in the general population. J Am Coll Cardiol 2014; 64: 531–538.
• 20.   Rooney MR, Soliman EZ, Lutsey PL, Norby FL, Loehr LR, Mosley TH, et al. Prevalence and characteristics of subclinical atrial fibrillation in a community-dwelling elderly population: The ARIC study. Circ Arrhythm Electrophysiol 2019; 12: e007390.
• 21.   Mou L, Norby FL, Chen LY, O’Neal WT, Lewis TT, Loehr LR, et al. Lifetime risk of atrial fibrillation by race and socioeconomic status: ARIC study (Atherosclerosis Risk in Communities). Circ Arrhythm Electrophysiol 2018; 11: e006350.
• 22.   Antzelevitch C, Yan GX. J wave syndromes. Heart Rhythm 2010; 7: 549–558.
• 23.   Tikkanen JT, Anttonen O, Junttila MJ, Aro AL, Kerola T, Rissanen HA, et al. Long-term outcome associated with early repolarization on electrocardiography. N Engl J Med 2009; 361: 2529–2537.
• 24.   Barra S, Providência R, Paiva L, Nascimento J. Early repolarization patterns and the role of additional proarrhythmic triggers. Europace 2013; 15: 482–485.
• 25.   Sinner MF, Porthan K, Noseworthy PA, Havulinna AS, Tikkanen JT, Müller-Nurasyid M, et al. A meta-analysis of genome-wide association studies of the electrocardiographic early repolarization pattern. Heart Rhythm 2012; 9: 1627–1634.
• 26.   Imai Y, Kusano K, Aiba T, Ako J, Asano Y, Harada-Shiba M, et al. JCS/JCC/JSPCCS 2024 guideline on genetic testing and counseling in cardiovascular disease. Circ J 2024; 88: 2022–2099.
• 27.   Delaney JT, Muhammad R, Blair MA, Kor K, Fish FA, Roden DM, et al. A KCNJ8 mutation associated with early repolarization and atrial fibrillation. Europace 2012; 14: 1428–1432.
• 28.   Teumer A, Trenkwalder T, Kessler T, Jamshidi Y, van den Berg ME, Kaess B, et al. KCND3 potassium channel gene variant confers susceptibility to electrocardiographic early repolarization pattern. JCI Insight 2019; 4: e131156.
• 29.   Watanabe H, Nogami A, Ohkubo K, Kawata H, Hayashi Y, Ishikawa T, et al. Electrocardiographic characteristics and SCN5A mutations in idiopathic ventricular fibrillation associated with early repolarization. Circ Arrhythm Electrophysiol 2011; 4: 874–881.
• 30.   Walsh JA 3rd, Ilkhanoff L, Soliman EZ, Prineas R, Liu K, Ning H, et al. Natural history of the early repolarization pattern in a biracial cohort: CARDIA (Coronary Artery Risk Development in Young Adults) study. J Am Coll Cardiol 2013; 61: 863–869.
• 31.   Junttila MJ, Tikkanen JT, Porthan K, Oikarinen L, Jula A, Kenttä T, et al. Relationship between testosterone level and early repolarization on 12-lead electrocardiograms in men. J Am Coll Cardiol 2013; 62: 1633–1634.
• 32.   Fukuda T, Shinohara T, Yonezu K, Mitarai K, Hirota K, Kondo H, et al. Vagal response is involved in the occurrence of ventricular fibrillation in patients with early repolarization syndrome. Heart Rhythm 2023; 20: 879–885.
• 33.   Ilkhanoff L, Soliman EZ, Prineas RJ, Walsh JA 3rd, Ning H, Liu K, et al. Clinical characteristics and outcomes associated with the natural history of early repolarization in a young, biracial cohort followed to middle age: The Coronary Artery Risk Development in Young Adults (CARDIA) study. Circ Arrhythm Electrophysiol 2014; 7: 392–399.
• 34.   Olson KA, Viera AJ, Soliman EZ, Crow RS, Rosamond WD. Long-term prognosis associated with J-point elevation in a large middle-aged biracial cohort: The ARIC study. Eur Heart J 2011; 32: 3098–3106.
• 35.   Kamakura T, Wada M, Ishibashi K, Inoue YY, Miyamoto K, Okamura H, et al. Differences in the onset mode of ventricular tachyarrhythmia between patients with J wave in anterior leads and those with J wave in inferolateral leads. Heart Rhythm 2017; 14: 553–561.
• 36.   Nademanee K, Haissaguerre M, Hocini M, Nogami A, Cheniti G, Duchateau J, et al. Mapping and ablation of ventricular fibrillation associated with early repolarization syndrome. Circulation 2019; 140: 1477–1490.
• 37.   Sabbag A, Sidi Y, Kivity S, Beinart R, Glikson M, Segev S, et al. Obesity and exercise-induced ectopic ventricular arrhythmias in apparently healthy middle aged adults. Eur J Prev Cardiol 2016; 23: 511–517.
• 38.   Liu P, Wang Y, Zhang X, Zhang Z, Zhao N, Ou W, et al. Obesity and cardiac conduction block disease in China. JAMA Netw Open 2023; 6: e2342831.