Ventricular premature complexes (VPCs) and atrial premature complexes (APCs) are common findings on routinely obtained electrocardiograms. Despite their common occurrence, the significance of these irregular beats is unclear, especially with regard to risk of sudden cardiac death (SCD). In this study, we examined the prospective relation between baseline VPCs or APCs and SCD, myocardial infarction, and fatal coronary heart disease (CHD) in a population-based sample of subjects from the Atherosclerosis Risk in Communities (ARIC) study excluding participants with known history of CHD or stroke. Baseline examination was conducted from 1987 to 1989, with follow-up data regarding clinical cardiac events collected until December 2002. The total study population was 14,574 subjects. Kaplan-Meier curves and computed univariate and multivariate Cox proportional hazard models were employed to estimate the effect of VPC and APC occurrences on incident cardiac events. During the follow-up period, there were 130 incident cases of SCD, 1,657 incident cases of CHD cases, and 288 cases of fatal CHD. Participants with VPC were 2 times as likely to have SCD (hazard ratio [HR] 2.09, 95% confidence interval [CI] 1.22 to 3.56) compared to those without VPC. Presence of APC was not significantly associated with SCD (HR 1.15, 95% CI 0.56 to 2.39). Compared to subjects without VPC and APC, risk of SCD in subjects with VPC and APC was significantly increased (HR 6.39, 95% CI 2.58 to 15.84). In conclusion, our study shows that subjects with VPCs are significantly more likely to die from SCD, despite not having any known history of cardiovascular disease. This effect appears to be additive when APCs occur concurrently.
Sudden cardiac death (SCD) is an important public health issue with an estimated global prevalence in the range of 4 to 5 million deaths per year. Ventricular tachyarrhythmia is the most common proximate cause of SCD. It is not uncommon to document ventricular premature complexes (VPCs) and atrial premature complexes (APCs) in apparently healthy subjects on baseline or routine electrocardiograms. Epidemiologic studies have shown that VPCs are present in >6% of the population. Age-adjusted prevalences of VPC were 8.2% in African-American men, 6.7% in Caucasian men, 6.9% in African-American women, and 4.7% in Caucasian women. Approximately 10% to 73% of young patients and 21% to 100% of elderly patients have APCs. The prognostic importance of VPCs and APCs has been studied for decades. The relation between ischemic heart disease and VPCs is well documented. Patients with VPCs are at a significantly higher risk for development of ischemic heart disease and death. It is commonly believed that VPCs and APCs are less harmful when seen in subjects with normal left ventricular function and without coronary artery disease. The objective of this study was to investigate the prospective relation between baseline ectopy and clinical cardiac events, including SCD, myocardial infarction, and fatal coronary heart disease (CHD) in a population-based sample of subjects without any history of cardiac disease or stroke.
Methods
We used public-use data from the Atherosclerosis Risk in Communities (ARIC) study, an ongoing prospective study of the cause and natural history of atherosclerosis funded by the National Heart, Lung, and Blood Institute (NHLBI). Four communities took part in the investigation: Forsyth County, North Carolina; Jackson, Mississippi; suburban Minneapolis, Minnesota; and Washington County, Maryland. The total population of the study, conducted from 1987 to 1989, was 15,732 subjects 45 to 64 years of age at their baseline examination. The ARIC study participants underwent follow-up clinical examinations every 3 years and were contacted by telephone annually to ascertain their general health and any hospitalization events. This report was based on baseline examination and follow-up data on clinical cardiac events up to December 2002, with >14 years of follow-up of ARIC study participants.
A standard supine 12-lead electrocardiogram at rest and a 2-minute 3-lead (leads II, V 1 , and V 5 ) rhythm strip were recorded during baseline examination using a Macintosh personal cardiograph (Marquette Electronics, Inc., Milwaukee, Wisconsin). Participants were asked to fast for 12 hours to stop drinking coffee or smoking 1 hour before electrocardiography. Independent trained coders then classified the rhythm strip 3 times after blinding for different arrhythmias such as supraventricular and ventricular complexes, ventricular reruns, bigeminy, trigeminy, and multiform complexes. The Minnesota Coding System was used to code arrhythmias. All electrocardiographic (ECG) records with computer-generated Minnesota codes were visually coded at a coding center using the full Minnesota code. Then, a senior coder adjudicated the final code. The ECG central principal investigator or the coding supervisor adjudicated disagreements concerning electrocardiograms. VPCs and APCs were determined from rhythm strips.
A yearly telephone call was made to participants of the ARIC study to obtain their vital status, hospitalizations, and visits to medical practitioners. Potential cardiovascular and cerebrovascular events were identified and validated through hospital records and death certificates, physician questionnaires, and interviews with next of kin. ARIC study identification procedures have been previously reported.
In this study, we analyzed ARIC officially confirmed cardiac events. SCD was defined as death within 1 hour of acute cardiac symptoms. An incident cardiac event was defined as a definite or probable myocardial infarction, fatal CHD, cardiac revascularization procedure, or a silent myocardial infarction documented by electrocardiogram during follow-up clinical examinations, which included a major or minor Q wave with ischemic ST-T changes. Fatal CHD was defined as death with a confirmed primary cause of myocardial infarction or coronary disease.
Baseline data on the following covariables are used in this report: age, race, gender, smoking status, body mass index, serum potassium, serum magnesium, serum low-density lipoprotein/high-density lipoprotein ratio, heart rate, education, hypertension, any use of blood pressure medications, and heart rhythm medications.
Statistical analysis was conducted using an effective sample of 14,574 subjects. Kaplan-Meier estimates were used to analyze overall survival of participants with and without VPCs and APCs. Log-rank test was used to test for equality of survival distributions stratified by presence of VPCs and APCs. Cox proportional hazards models were employed to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) of various cardiac events associated with baseline 2-minute ECG rhythm strips, which defined VPCs and APCs separately. For this analytic objective, 3 sets of models were repeated: unadjusted models; models adjusted for major cardiovascular disease risk factors including age, race, gender, education, smoking body mass index, low-density lipoprotein/high-density lipoprotein ratio, hypertension, and diabetes; and fully adjusted models. Fully adjusted models, in addition to the listed variables, included serum potassium, serum magnesium, serum low-density lipoprotein/high-density lipoprotein ratio, heart rate, and any use of heart rhythm medications. We then examined the interaction between VPCs and APCs on incident SCD, incident CHD, and fatal CHD events. All statistical analyses were performed using SAS 9.1 (SAS Institute, Cary, North Carolina).
Results
The ARIC pubic-use data included 15,732 participants. We excluded 1,158 subjects who had prevalent CHD or stroke at baseline or whose baseline 2-minute electrocardiogram was of poor quality. As a result, our analysis consisted of 14,574 middle-aged Americans free of CHD and stroke at baseline. They were followed to December 31, 2002 for incident cardiac events including SCD, CHD, and fatal CHD, an average of 14 years. Median follow-up time of the study was 14 years (interquartile range 1.5). Of these 14,574 subjects, 707 had only VPCs, 640 had only APCs, and 82 had VPCs and APCs at baseline. During follow-up, cardiac events were identified and classified as (1) 130 incident cases of SCD (0.89%) defined as onset of symptoms to death ≤1 hour; (2) 1,657 incident cases of CHD (11.40%) defined as fatal CHD, hospitalization for myocardial infarction detected on electrocardiogram during cohort re-examinations, and revascularization procedures; and (3) 288 cases of fatal CHD (1.98%).
Table 1 lists baseline characteristics of the population. From Cox proportional hazard models ( Table 2 ), presence of VPCs was consistently and significantly associated with SCD, incident CHD, and fatal CHD. For instance, participants with VPCs were >2 times as likely to have SCD (HR 2.80, 95% CI 1.68 to 4.67) compared to those without VPC. After adjusting for traditional cardiovascular risk factors, effect of VPC on SCD remained statistically significant (HR 2.09, 95% CI 1.22 to 3.56). VPC, incident CHD, and fatal CHD relations were similar to that of SCD, but the magnitude for incident CHD was generally smaller. As presented in Table 2 , presence of APCs was not significantly associated with incident CHD and fatal CHD, but magnitudes of the associations were smaller than that of VPC. Kaplan-Meier survival curves for subjects with versus those without VPC are shown in Figure 1 . As indicated in the curve, participants with VPCs at baseline had a significantly lower survival rate during the follow-up period (p <0.0001, log-rank test). However, no significant discrepancy was found when comparing subjects with to those without APCs (data not shown).
| Variable | VPCs | APCs | ||||
|---|---|---|---|---|---|---|
| No (n = 13,748) | Yes (n = 788) | p Value | No (n = 13,685) | Yes (n = 716) | p Value | |
| Age (years) | 53.85 ± 5.7 | 55.97 ± 5.6 | <0.0001 | 53.84 ± 5.7 | 56.18 ± 5.8 | <0.0001 |
| Women | 57.08% | 50.75% | 0.0004 | 56.79% | 55.68% | 0.57 |
| Black | 26.70% | 31.81% | 0.0001 | 26.99% | 26.73% | 0.78 |
| Education (higher than high school) | 77.29% | 69.71% | <0.0001 | 77.23% | 70.08% | <0.0001 |
| Smoking | 25.89% | 25.73% | 0.96 | 25.69% | 29.64% | 0.02 |
| Hypertension | 32.87% | 44.64% | <0.001 | 33.49% | 35.84% | 0.18 |
| Diabetes | 10.69% | 13.67% | 0.004 | 10.92% | 9.50% | 0.23 |
| Body mass index (kg/m 2 ) | 27.61 ± 5.3 | 28.45 ± 5.6 | <0.0001 | 27.66 ± 5.3 | 27.46 ± 5.8 | 0.27 |
| Low-density lipoprotein/high-density lipoprotein ratio | 2.91 ± 1.3 | 2.94 ± 1.3 | 0.55 | 2.92 ± 1.3 | 2.76 ± 1.2 | 0.001 |
| Potassium (mEq/L) | 4.42 ± 0.5 | 4.38 ± 0.5 | 0.02 | 4.42 ± 0.5 | 4.43 ± 0.5 | 0.35 |
| Magnesium (mEq/L) | 1.63 ± 0.2 | 1.62 ± 0.2 | 0.001 | 1.63 ± 0.2 | 1.63 ± 0.2 | 0.93 |
| Heart rate | 68 ± 10.7 | 69 ± 10.5 | <0.0001 | 68 ± 10 | 69 ± 11.3 | 0.16 |
| Heart rhythm medications | 3.75% | 7.14% | <0.001 | 3.73% | 7.86% | <0.0001 |
| Sudden death | 0.83% | 2.26% | <0.0001 | 0.89% | 1.38% | 0.17 |
| Coronary heart disease | 11.19% | 15.91% | <0.0001 | 11.31% | 14.07% | 0.02 |
| Fatal coronary heart disease | 1.83% | 5.14% | <0.0001 | 1.92% | 3.72% | 0.0007 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
