Author + information
- Received September 5, 2003
- Revision received October 20, 2003
- Accepted October 28, 2003
- Published online April 7, 2004.
- T.Jared Bunch, MD*,†,
- Krishnaswamy Chandrasekaran, MD*,†,
- Bernard J Gersh, MD, PhD*,†,
- Stephen C Hammill, MD*,†,
- David O Hodge, MS‡,
- Akbar H Khan, MD*,
- Douglas L Packer, MD*,† and
- Patricia A Pellikka, MD*,†,* ()
- ↵*Reprint requests and correspondence:
Dr. Patricia A. Pellikka, Division of Cardiovascular Diseases, Department of Internal Medicine, Mayo Clinic, Mayo Foundation, 200 First Street SW, Rochester, Minnesota 55905, USA.
Objectives The purpose of the study was to determine if atrial ectopy (AE) or atrial arrhythmias during exercise are predictive of an increased risk of cardiac events and death.
Background Although stress-induced atrial arrhythmias are common during exercise testing, there is a paucity of data regarding the correlation with underlying heart disease and cardiovascular outcomes. Atrial arrhythmias may reflect underlying left atrial enlargement and diastolic dysfunction, which are prognostic of mortality. We hypothesized that these stress-induced arrhythmias are associated with long-term adverse cardiac events.
Methods Exercise echocardiography was performed in 5,375 patients (age 61 ± 12 years) with known or suspected coronary artery disease. An abnormal result was defined as exercise-induced atrial fibrillation (AF)/atrial flutter, supraventricular tachycardia (SVT), or AE.
Results A total of 311 (5.8%) patients died (132 [2.5%] from cardiac causes) over a period of 3.1 ± 1.7 years. In addition, 193 (3.6%) patients experienced a myocardial infarction (MI) and 531 (9.9%) patients required revascularization. During exercise testing, 1,272 (24%) patients developed AE, 185 (3.4%) developed SVT, and 43 (0.8%) developed AF. The five-year cardiac death rate was not statistically different between groups (none [3.8%], AE [4.3%], SVT [3.7%], AF [0%], p = 0.43). The five-year rate of MI was significantly different between groups (none [5.7%], AE [8.3%], SVT [0%], AF [9.0%], p = 0.005). The five-year rate of revascularization between groups was not significantly different (none [14.2%], AE [17.0%], SVT [11.8%], AF [14.8%], p = 0.50). A composite of all five-year adverse end points was similar between groups (none [22.7%], AE [27.8%], SVT [17.7%], AF [25.7%], p = 0.10). In stepwise multivariate analysis, AE was not predictive of myocardial infarction when taking into account traditional clinical variables and exercise test results.
Conclusions In this large cohort of patients, the occurrence of AE was predictive of an increased risk of MI. However, the association did not persist after adjustment for clinical and exercise variables known to predict adverse long-term cardiovascular outcomes. The rate of long-term cardiac death or revascularization was not influenced by the development of stress-induced atrial arrhythmias.
Despite extensive evaluation of exercise-induced ventricular arrhythmias, the long-term significance of exercise-induced atrial arrhythmias is unknown. During exercise, ectopic atrial pacemakers, sinus arrhythmia, paroxysms of atrial tachycardia, and junctional tachyarrhythmias are common (1). Paroxysmal atrial fibrillation (AF) and supraventricular tachycardia (SVT) during exercise are rare (1,2). Of these rhythms, premature atrial complexes are the most common, with a general incidence ranging from 2.5% to 10% (3–5). In general, exercise-induced atrial arrhythmias increase with age (6,7).
Although exercise-induced atrial ectopy (AE) is common, there is a paucity of data regarding the correlation with underlying heart disease and cardiovascular outcomes. In a small study, Whinnery et al. (5)found supraventricular premature beats in 40% of 20 patients with cardiac disease versus 10% in 20 healthy controls. In addition, Master (3)reported 24 patients with atrial premature complexes, 15 (63%) of whom had underlying organic cardiac disease. Despite these findings, there have been no subsequent long-term studies to determine the prognostic utility of exercise-induced AE in ascertainment of long-term cardiovascular risk. Accordingly, we studied the hypothesis that AE during exercise would be a predictor of increased risk of cardiovascular disease and death. In addition, we studied the prognostic significance of exercise-induced AF and SVT.
This study was approved by the Mayo Clinic Institutional Review Board. We retrospectively studied 6,444 patients who underwent clinically indicated exercise echocardiography at Mayo Clinic (Rochester, Minnesota) from January 1990 through December 1995. Patients were excluded if they refused participation in research or were lost to follow-up. Patients with AF on the baseline electrocardiogram (ECG) were also excluded. A total of 5,375 patients met criteria for inclusion. Patient demographics were obtained at the time of the treadmill test through a patient interview and chart review. Data collected included age, gender, cardiovascular risk factors, medications, prior cardiac procedures, and the reason for pursuing an exercise test. The ejection fraction at rest was measured using a modification of the method of Quinones et al. (8)or by visual estimation (9), and at exercise by visual estimation.
The patients underwent a symptom-limited treadmill exercise test according to the Bruce protocol (88%), Naughton protocol (6%), modified Bruce protocol (3%), or other protocol (3%). Standard blood pressure and 12-channel ECG monitoring were performed. During each stage of exercise, data on the heart rate, blood pressure, and ECG changes were recorded. Those persons recording these data were unaware of the hypothesis of this study.
Information regarding atrial arrhythmias was recorded during each stage of the protocol and upon completion of the study. Atrial arrhythmias were defined in four categories: no arrhythmia, AE, AF or atrial flutter, and SVT. We considered three or more consecutive beats at an accelerated rate (accelerated relative to the sinus tachycardia that is expected during exercise testing) to be a tachycardia. Isolated narrow complex beats and pairs were classified as AE. The exercise ECG was considered positive for ischemia if there was horizontal or downsloping ST-segment depression of at least 1 mm at 80 ms after the J-point, nondiagnostic if the baseline ST-segment was abnormal, and negative for ischemia in the absence of these criteria. Exercise echocardiography results were defined as abnormal if ischemia or fixed wall motion abnormalities were present.
Follow-up data were obtained from mailed questionnaires and scripted telephone interviews. Events were verified by contacting the patients' primary physicians and reviewing medical records and death certificates. The end points were myocardial infarction (MI), revascularization (percutaneous transluminal coronary angioplasty or coronary artery bypass grafting), cardiac death, overall death, and major adverse cardiac events (MACE) (composite of MI, revascularization, and cardiac death). Sudden unexpected death occurring without another explanation was included as cardiac death. Stroke felt to be resultant from a cardioembolic source was included as a cardiac death. Coronary revascularization procedures during the follow-up period were also included. This end point was analyzed because the results of the exercise stress tests were available to treating physicians and may have been used in decisions to perform revascularization, thus altering long-term survival rates as well as representing underlying coronary artery disease.
Continuous variables were reported as mean ± SD and comparisons between groups were based on the Wilcoxon rank-sum test. Categorical variables were summarized as percentages, and group comparisons were based on the chi-square test. Survival free of the end point of interest was estimated by the Kaplan-Meier method. Univariate associations of clinical and exercise variables with the end points were assessed in a Cox proportional-hazards model. A multivariate model was established initially with only clinical baseline data, including rest echocardiographic variables. Exercise ECG, hemodynamic variables, and exercise echocardiographic variables were added in a stepwise forward selection manner to the clinical model. These models did not take into consideration the location of wall motion abnormalities. The significance of adding additional variables to previous modeling steps was based on the change in model-based likelihood statistics with degrees of freedom equal to the number of additional variables.
Among the 5,375 patients that met inclusion criteria, 1,500 patients developed an exercise-induced atrial arrhythmia (Table 1). Of these patients, 1,272 (23.7%) had AE, 43 (0.8%) had AF or atrial flutter, and 185 (3.4%) had SVT. The clinical and exercise-related characteristics of the patients based upon atrial arrhythmia are shown in Table 2. Patients who developed exercise-induced arrhythmias were older in comparison with patients who did not have atrial arrhythmias (p = 0.001). Patients with stress-induced AF/atrial flutter had a higher mean maximum heart rate in comparison to the other groups (p < 0.0001).
During a mean follow-up of 3.1 ± 1.7 years, 300 patients died. Of these, there were 132 cardiac deaths, including 54 with MI, 28 with congestive heart failure, 19 with sudden death, 14 with ventricular fibrillation cardiac arrest, and 4 with cardioembolic stroke. A total of 195 patients had an MI, and 531 patients underwent revascularization (273 coronary angioplasty, 328 coronary artery bypass graft). The five-year overall death rate was not statistically different between arrhythmia groups (none [9.1%], AE [9.8%], SVT [10%], AF [12.9%], chi-square = 4.8, p = 0.19) (Fig. 1), nor was the five-year cardiac death rate (none [3.8%], AE [4.3%], SVT [3.7%], AF [0%], chi-square = 12.7, p = 0.43) (Fig. 2). The five-year rate of MI was significantly different between groups (none [5.7%], AE [8.3%], SVT [0%], AF [9.0%], chi-square = 2.8, p = 0.005) (Fig. 3). The five-year rate of revascularization between groups was not significantly different (none [14.2%], AE [17.0%], SVT [11.8%], AF [14.8%], chi-square = 2.4, p = 0.5). A composite of all five-year adverse end points was similar between groups (none [22.7%], AE [27.8%], SVT [17.7%], AF [25.7%], chi-square = 6.1, p = 0.1) (Fig. 4).
Predictors of events
The significance of exercise-induced atrial arrhythmias for prediction of MI was further pursued. A multivariate model was constructed with covariates of several clinical variables: age (hazard ratio [HR] 1.02, 95% confidence interval [CI] 1.01 to 1.04, p = 0.001), gender (HR 0.54, 95% CI 0.37 to 0.71, p = 0.0002), diabetes (HR 1.59 95% CI 1.41 to 1.78, p = 0.011), hyperlipidemia (HR 1.44 95% CI 1.29 to 1.59, p = 0.014), previous MI (HR 2.23 95% CI 2.13 to 2.43, p = 0.0001), normal baseline ECG (HR 0.63 95% CI 0.45 to 0.80, p = 0.007), test duration (HR 0.98 95% CI 0.97 to 0.99), p = 0.0002). Atrial ectopy remained predictive of MI risk (HR 1.5, 95% CI 1.1 to 2.1, p = 0.02). However, AF (HR 1.38, 95% CI 1.97 to 0.8, p = 0.58) and SVT (HR 0.0, p = 0.99) were not predictive of MI risk.
A stepwise clinical model was established using four steps to determine the long-term additive prognostic impact of AE on myocardial risk beyond clinical variables. This is shown in Table 2. Stress-induced AE resulted in a HR of 1.32 when added to the model (95% CI 0.95 to 1.84). However, there was no significant difference between AE and no atrial arrhythmias (p = 0.1). Removal of echocardiographic variables from the stepwise model did not alter the significance of AE beyond that of no arrhythmia induced in prediction of long-term risk of MI (p = 0.11).
This is the first study to assess and compare the association of various stress-induced atrial arrhythmias on long-term adverse cardiac events. In a large cohort of patients referred for exercise stress testing, the occurrence of AE was predictive of an increased risk of MI. This association did not persist when taking into account clinical or exercise test parameters. The rates of long-term cardiac death, revascularization, or MACE were not associated with development of stress-induced atrial arrhythmias.
These data provide long-term clinical insight into small prior studies that suggest exercise provocation of AE is associated with cardiac disease (3,5). The pathophysiology underlying this association is unclear. During ventricular diastole, the left atrium is exposed to the left ventricular (LV) pressure. Furthermore, the pressure and size of the left atrium are directly influenced by those factors that alter diastolic LV filling (10,11). Deteriorating diastolic function, which is associated with elevated LV diastolic pressures, strongly predicts long-term cardiovascular risk and mortality (11–14). Because increased left atrial size is associated with atrial arrhythmias, this may in part explain this finding. Another potential mechanism is that AE may represent underlying systemic inflammation. Chung et al. (15)reported an increase in C-reactive protein, a risk factor of cardiovascular disease (16), in patients with persistent, lone, and paroxysmal AF as well as AE. Despite this association, in our study, AE did not add to traditional risk factors in long-term ascertainment of MI risk.
These data reveal that AF, AE, and SVT were not predictive of risk of cardiac death, revascularization, and MACE, which support and expand the findings of Mauerer et al. (17). They reported long-term data on 85 patients (6.1% of the total population) who developed stress-induced SVT (≥3 consecutive supraventricular beats at a rate >100 beats/min). In this cohort of patients, there no significant increase in cardial mortality, MI, or syncope in comparison to a control population (17).
All of the exercise-induced atrial arrhythmias increased with age. This finding is consistent with prior reports (6,7,17). Although the reason for the age-related increase of exercised-induced arrhythmias in unknown, it may reflect an age-related increase in left atrial size, exaggerated norepinephrine and epinephrine responses during exercise, or an increase in risk factors of LV dysfunction (17–19).
The data from this study must be viewed in context of the study limitations. First, the data are from an observational study of a cohort of patients from a single center and have the common limitations of nonrandomized studies, including selection biases and uncontrolled confounding. However, to date it is the largest data set to compare exercise-induced atrial arrhythmias and their prognostic impact of long-term cardiovascular risk. Second, the database does not include information regarding pretest arrhythmias. Although baseline AF was excluded, prior occult paroxysmal supraventricular arrhythmias may have been included. Third, data regarding baseline antiarrhythmic drug therapy were not assessed, although we expect this omission to be minimal by excluding individuals with baseline AF. Fourth, the database did not distinguish between low-grade and high-grade AE. However, these data have important clinical implications in both the positive and negative interpretation of exercise tests to risk-stratify individuals at highest risk for adverse cardiovascular outcomes.
Atrial ectopy during exercise was found to be an independent predictor of an increased risk of MI, but not cardiac survival or revascularization. However, in stepwise modeling AE did not add to clinical baseline- or stress-induced risk factors of MI. All other exercise-induced atrial arrhythmias were not independently predictive of cardiac events. These data in a large cohort of patients further validate the value of an exercise test to ascertain both negative and positive risk and prognosis.
- atrial ectopy
- atrial fibrillation
- confidence interval
- hazard ratio
- left ventricular
- major adverse cardiac events
- myocardial infarction
- supraventricular tachycardia
- Received September 5, 2003.
- Revision received October 20, 2003.
- Accepted October 28, 2003.
- American College of Cardiology Foundation
- Dujardin K.S.,
- Enriquez-Sarano M.,
- Rossi A.,
- Bailey K.R.,
- Seward J.B.
- Appleton C.P.,
- Galloway J.M.,
- Gonzalez M.S.,
- Gaballa M.,
- Basnight M.A.
- Moller J.E.,
- Hillis G.S.,
- Oh J.K.,
- et al.
- Dini F.L.,
- Michelassi C.,
- Micheli G.,
- Rovai D.
- Nijland F.,
- Kamp O.,
- Karreman A.J.,
- van Eenige M.J.,
- Visser C.A.
- Moller J.E.,
- Sondergaard E.,
- Poulsen S.H.,
- Egstrup K.
- Chung M.K.,
- Martin D.O.,
- Sprecher D.,
- et al.
- Albert C.M.,
- Ma J.,
- Rifai N.,
- Stampfer M.J.,
- Ridker P.M.
- Fleg J.L.,
- Tzankoff S.P.,
- Lakatta E.G.