Although it is known that the electrocardiographic pattern of early repolarization (ER) occurs most commonly in healthy young bradycardic men, its natural history is uncertain. We considered initial electrocardiograms (ECGs) at rest from 29,281 ambulatory patients recorded from 1987 through 1999 at Veterans Affairs Palo Alto Hospital. With PR interval as the isoelectric line and amplitude criterion as >0.1 mV ER was identified when any of the following fulfilled the amplitude criterion: ST-segment elevation at the end of the QRS duration, J waves as an upward deflection, and slurs as delay on the R wave downstroke. The first 250 ECGs with the greatest ER increase were selected and the database was searched for an ECG >5 months later. Of the 250 patients selected with the greatest amplitude of ER 6 were excluded for electrocardiographic abnormalities, leaving 244 subjects, of whom 122 had another ECG ≥5 months later. Their average age was 42 ± 10 years and average time from the first to second ECG was 10 years. Of the 122 patients 47 (38%) retained ER, whereas most (62%) no longer fulfilled the amplitude criterion. There were no significant differences in heart rate or time interval between ECGs. In conclusion, the electrocardiographic pattern of ER was lost over 10 years in more than half of this young clinical cohort and the loss was not caused by higher heart rate, longer time between ECGs, decrease in R-wave amplitude, death, acute disease, or alterations in electrocardiographic diagnostic characteristics.
ST-segment elevation in the absence of acute infarction was first reported on electrocardiograms (ECGs) of healthy subjects in 1947 and was termed “early repolarization” (ER) in 1951. This “normal RS-T segment elevation variant” was characterized as ST-segment elevation with a distinct notch (J wave) or slur on the downslope of the R wave by Wasserburger and Alt in 1961. In 1976 Kambara and Phillips reported that ER was more common in young bradycardic men and in those of African ethnicity and appeared most commonly in the lateral leads. Although ER is more common in the young, suggesting that it must recede with age, there have been limited studies of its prognosis or serial ECGs in subjects with ER. To answer the question of what happens to this electrocardiographic pattern over time, we selected 250 patients exhibiting the greatest amplitude of ER from a large clinical population of ambulatory patients and examined our clinical electrocardiographic database to see who had ECGs ≥5 months later and compared them.
Methods
In total 45,829 unique inpatient and outpatient ECGs were recorded for clinical indications from March 1987 to December 1999 at the Veterans Affairs Palo Alto Health Care System. All patients were seen at the main Veterans Affairs facility or its satellite clinics, and ECGs were ordered by health care providers usually to screen for occult disease and to obtain a baseline when initiating care. Because clinical diagnostic codes were not available, we excluded those with inpatient status (n = 12,319) to eliminate ECGs possibly associated with acute coronary syndromes and other acute processes. Furthermore, ECGs exhibiting atrial fibrillation or flutter (n = 1,253), ventricular rates >100 beats/min (n = 2,799), QRS durations >120 ms (n = 3,141), paced rhythms (n = 290), ventricular preexcitation (n = 42), and acute myocardial infarction (n = 29) were excluded, leaving 29,281 patients for analysis. Race was determined by self-report at time of electrocardiographic acquisition.
With PR interval as the isoelectric line and the amplitude criterion for ER as ≥0.1 mV (100 μV or 1 mm when 10 mm = 1 mV) above the isoelectric line in lateral lead V 5 , ER was identified when any of the following were seen to fulfill the amplitude criterion: ST-segment elevation at the end of the QRS duration, J waves as an upward deflection, and slurs as a conduction delay on the QRS downstroke. Computer-measured amplitude of ER was sorted from greatest to lowest amplitude and the first 250 ECGs with the greatest ER amplitude were selected and visually confirmed, analyzed, and coded by 3 independent interpreters with conflicts resolved by the senior author.
Patients were then queried to determine who had a subsequent ECG. One patient with diagnostic Q waves on the first ECG and 5 patients with abnormalities on the last ECG (1 each with a diagnostic Q wave, atrial fibrillation, and sinus tachycardia and 2 with supraventricular tachycardia) were excluded. Thus, 122 patients were identified to have had a second ECG >5 months later. When multiple ECGs were available the ECG at the longest interval was chosen for comparison, paired, and printed for coding by 3 trained subjects. Chart review of patients with a second ECG was performed to determine if an acute medical condition necessitated this ECG. These patients were then compared according to whether they maintained or lost ER. For example, the patients were sorted by the greatest change in ER amplitude (i.e., lost ER) and the least change (i.e., maintained ER).
NCSS 2007 (NCSS, Kayesville, Utah) was used for all statistical analyses. Unpaired t tests were used for comparisons of continuous variables and chi-square tests were used to compare dichotomous variables between groups. A significance level of 0.01 was used.
Results
Of the total database of 29,281 patients, 479 patients (1.6%) were coded as manifesting ER in the lateral leads. When sorted by ER amplitude, of the 250 ECGs exhibiting the greatest amplitude 244 had otherwise normal ECGs and were queried for a repeat ECG >5 months later. Of these 122 (50%) had a subsequent ECG for comparison. On chart review 5 patients developed heart disease over the interval but no acute symptoms were present at the time of the second ECG.
Table 1 presents demographics and annual cardiovascular mortality of the total population and its subgroups according to selection criteria. Although there were significant differences between the total population and the ER subgroups, none were found between the 2 subgroups. There were 1,995 cardiovascular deaths (7%) over a mean follow-up of 7.6 ± 3.8 years in the total population with a significantly lower mortality (p <0.001) in the ER subgroups (3.0% and 3.0%). Patients with ER were significantly younger (42 ± 13 vs 55 ± 15 years, p <0.001) and had a lower heart rate (63 ± 12 vs 71 ± 13 beats/min, p <0.001). There was a higher prevalence of men (99% vs 87%, p <0.001) and African-Americans (44% vs 13%, p <0.001) in the ER subgroups compared to the total population.
| Characteristics | All Subjects (n = 29,281) | Early Repolarization in Lateral Leads | p Value ⁎ | Most Early Repolarization in Lead V 5 | p Value † |
|---|---|---|---|---|---|
| (n = 479, 1.6%) | (n = 244, 0.8%) | ||||
| Age (years) | 55 ± 15 | 42 ± 13 | <0.001 | 40.7 ± 12.4 | 0.2 |
| Men | 25,544 (87%) | 472 (98.5%) | <0.001 | 241 (99%) | 0.5 |
| African-American | 3,885 (13%) | 210 (44%) | <0.001 | 102 (42%) | 0.2 |
| Body mass index (kg/m 2 ) | 27.3 ± 5.5 | 26.0 ± 4.1 | <0.001 | 25.6 ± 4.0 | <0.1 |
| Heart rate (beats/min) | 70.6 ± 12.6 | 63.2 ± 12.3 | <0.001 | 63 ± 11.9 | 0.8 |
| Cardiovascular deaths | 1,995 (7%) | 15 (3%) | <0.001 | 7 (3%) | 0.95 |
⁎ Between all subjects and those with all early repolarization in lateral leads.
† Between subgroup with early repolarization (n = 479) and most early repolarization in lead V 5 (n = 244).
Table 2 presents the greatest amplitude ER group (n = 244) divided into those who had a second ECG ≥5 months later and a comparison between those who maintained ER to those who lost it. The total ER group had an average ER amplitude of 140 ± 25 μV (1.4 ± 0.25 mm) with 12 patients (5%) exhibiting a heart rate >85 beats/min. J waves or slurs occurred in 76% with J waves being 1 1/2 times more common than the slurs, whereas only 24% of subjects (59 of 244) had only ST-segment elevation. Most ECGs with J waves and slurs also exhibited ST elevation measured at the mid point of the ST segment. On the second ECG (n = 122) average ER amplitude was 78 ± 46 μV with an average of 9.9 ± 5.9 years between the 2 ECGs. A significantly higher percentage of the second ECGs showed a heart rate >85 beats/min (14% vs 6%, p <0.001) and a lower prevalence of J waves and slurs (22% vs 77%, p <0.001) than the first ECG.
| Characteristics | Most ER in Lead V 5 (n = 244, 100%) | Most ER in Lead V 5 on Second ECG | p Value ⁎ | Maintained ER (n = 47, 38.5%) | Lost ER (n = 75, 61.5%) | p Value † <0.001 |
|---|---|---|---|---|---|---|
| (n = 122, 50%) | ||||||
| Age (years) | 40.7 ± 12.4 | 42.4 ± 10.3 | 0.1 | 39.3 ± 8.3 | 44.3 ± 10.9 | 0.008 |
| Age at second electrocardiogram | — | 52.3 ± 11.4 | — | 47.9 ± 10.5 | 55.1 ± 11.2 | <0.001 |
| Men | 241 (99%) | 122 (100%) | 0.2 | 47 (100%) | 75 (100%) | 1.0 |
| African-American | 102 (42%) | 66 (54%) | 0.02 | 29 (62%) | 37 (49%) | 0.2 |
| Body mass index (kg/m 2 ) | 25.6 ± 4.0 | 26.3 ± 4.2 | 0.1 | 26.3 ± 3.9 | 26.3 ± 4.4 | 0.9 |
| Cardiovascular deaths | 7 (3%) | 5 (4%) | 0.5 | 2 (4.25%) | 3 (4%) | 0.9 |
| Electrocardiographic measurements | ||||||
| Computer ST level on first electrocardiogram (μV) | 140.1 ± 24.7 | 138.4 ± 20.4 | 0.6 | 144.3 ± 21.8 | 134.6 ± 18.6 | >0.01 |
| Computer ST level one second electrocardiogram (μV) | — | 77.8 ± 46.0 | — | 122.4 ± 25.2 | 49.8 ± 31.8 | <0.001 |
| Computer ST level difference (μV) | — | 63.9 ± 40.5 | — | 30.9 ± 32.4 | 84.6 ± 29.8 | <0.001 |
| Heart rate | ||||||
| Heart rate on first electrocardiogram (beats/min) | 63 ± 11.9 | 64.7 ± 11.4 | 0.2 | 63.4 ± 11.5 | 65.5 ± 11.3 | 0.3 |
| Heart rate on second electrocardiogram | — | 68.7 ± 14.5 | — | 65.57 ± 13.5 | 70.7 ± 14.9 | 0.1 |
| Change in heart rate (beats/min) | — | 4.0 ± 14.2 | — | 2.2 ± 10.2 | 5.2 ± 16.1 | 0.3 |
| Heart rate >85 beats/min on first electrocardiogram | 12 (5%) | 7 (6%) | 0.7 | 1 (3%) | 6 (8%) | 0.2 |
| Heart rate >85 beats/min on second electrocardiogram | — | 17 (14%) | — | 5 (11%) | 12 (16%) | 0.4 |
| Time factor | ||||||
| Period between electrocardiograms (years) | — | 9.9 ± 5.9 | 0.1 | 8.58 ± 5.7 | 10.79 ± 5.9 | 0.04 |
| Other early repolarization criteria | ||||||
| First electrocardiogram | ||||||
| J wave | 114 (47%) | 55 (45%) | 0.8 | 24 (51%) | 31 (41%) | 0.3 |
| Slur | 71 (29%) | 39 (32%) | 0.5 | 9 (19%) | 30 (40%) | <0.01 |
| J wave or slur | 185 (76%) | 94 (77%) | 0.8 | 33 (70%) | 61 (80%) | 0.2 |
| J wave or Slur + STE | 175 (72%) | 81 (66%) | 0.6 | 31 (66%) | 50 (67%) | 0.8 |
| ST-segment elevation only | 59 (24%) | 28 (23%) | 0.8 | 14 (30%) | 14 (19%) | <0.001 |
| Second electrocardiogram | ||||||
| J wave | — | 17 (14%) | <0.001 † | 17 (36%) | — | — |
| Slur | — | 10 (8%) | <0.001 † | 10 (21%) | — | — |
| J wave or slur | — | 27 (22%) | <0.001 † | 27 (57%) | — | — |
| ST-segment elevation only | — | 17 (14%) | <0.001 † | 17 (36%) | — | — |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
