Impaired Chronotropic Response to Exercise Stress Testing as a Predictor of Mortality FREE

An attenuated heart rate response to exercise, known as chronotropic incompetence, has been shown to be predictive of mortality and coronary heart disease risk, even after adjusting for age, physical fitness, standard cardiovascular risk factors, and ST-segment changes with exercise.^1- 2 Chronotropic incompetence is also thought to decrease the accuracy of noninvasive imaging tests, such as stress testing with thallium imaging, regarding diagnosis of significant obstructive epicardial coronary disease.³ Nonetheless, among patients undergoing thallium scintigraphy, chronotropic incompetence has been shown to be related to risk of coronary events.⁴ It is not known whether chronotropic incompetence is independently predictive of all-cause mortality among patients referred for stress testing after accounting for myocardial perfusion defects.

The purpose of this study was to examine the association of chronotropic incompetence with myocardial perfusion defects as assessed by thallium imaging and to determine the ability of chronotropic incompetence to predict all-cause mortality in a low-risk cohort of patients referred for exercise treadmill testing with thallium imaging. We also assessed the separate and combined prognostic relationships of chronotropic incompetence and thallium perfusion defects.

The cohort of patients was derived from adults who were consecutively referred for symptom-limited exercise treadmill stress testing with thallium imaging at the Cleveland Clinic Foundation, Cleveland, Ohio, between September 1990 and December 1993. All patients were undergoing their first treadmill test at the Cleveland Clinic Foundation. Exclusion criteria included history of coronary angiography or invasive procedures, cardiac surgery, congestive heart failure, valvular heart disease, pre-excitation syndrome, or congenital heart disease. Patients taking β-blockers were excluded from analyses. Although patients taking calcium channel blockers were not excluded, we decided to prospectively perform stratified analyses based on use of these drugs. All patients gave oral informed consent prior to undergoing treadmill testing. Research based on the hospital stress testing database was approved by the Cleveland Clinic Institutional Review Board.

Methods of prospective data acquisition in this low-risk cohort have been described in detail elsewhere.^5- 6 Briefly, prior to each treadmill test, a structured interview and chart review were used to obtain data on symptoms, medications, coronary risk factors, prior cardiac events, and cardiac and noncardiac diagnoses. These data were entered prospectively online into a networked computerized database used for all stress tests in our institution. Resting hypertension was defined as a resting systolic blood pressure of more than 140 mm Hg, a resting diastolic blood pressure of more than 90 mm Hg, or treatment with antihypertensive medication.⁷ Assessment of diabetes was based on chart review, interview questions, and medication use. Lipid-related analyses primarily focused on use of lipid-lowering drugs because many patients did not have lipid profiles available at the time of testing. Prior coronary disease events consisted of documented myocardial infarction or hospital admission for unstable angina.

Exercise testing procedures in our laboratory have been described in detail elsewhere.^8- 9 Treadmill testing was carried out according to standard protocols, usually Bruce or modified Bruce.¹⁰ To facilitate estimation of exercise capacity, subjects were not permitted to lean on handrails during exercise. During each stage of exercise, data on symptoms, heart rhythm, heart rate, blood pressure (by indirect arm-cuff sphygmomanometry), estimated workload in METs (1 MET=3.5 mL/kg per minute of oxygen consumption, which corresponds to the resting state), and ST segments were prospectively collected and entered online. Functional capacity in METs was estimated from standard published tables¹⁰ based on protocol and total time completed in the final stage. When ST segments were interpretable, an ischemic response was considered present if there was more than 1 mm of horizontal or downsloping ST-segment depression 80 milliseconds after the J point or more than 1 mm of additional ST-segment elevation in leads without pathologic Q waves. If a patient had more than 1 treadmill test performed during the study period, only the first was considered for analyses.

Chronotropic incompetence was first assessed as failure to achieve 85% of the age-predicted heart rate. Because this method may be confounded by effects of age, physical fitness, and resting heart rate, chronotropic response also was assessed by calculating the ratio of heart rate reserve used to metabolic reserve used at peak exercise; this chronotropic index has been described in detail elsewhere.^1,11 For any given stage of exercise, the percentage of metabolic reserve used is equal to [(MET_stage−MET_rest)/(MET_peak−MET_rest)]×100, where MET refers to metabolic equivalents of oxygen consumption, stage refers to any given stage of exercise, and peak refers to peak exercise. The value of MET_peak refers to actual oxygen consumption noted, not a theoretical peak value. In an analogous fashion, the percentage of heart rate reserve used is equal to [(HR_stage−HR_rest)/(220−age in years−HR_rest)]×100, where HR refers to heart rate.

In a group of healthy nonhospitalized adults, the ratio of percentage of heart rate reserve used to percentage of metabolic reserve used during exercise was approximately 1.0 (95% confidence interval [CI], 0.8-1.3).¹¹ Thus, chronotropic incompetence can be defined as a ratio of percentage of heart rate reserve used to percentage of metabolic reserve used of less than 0.8; this is referred to as a low chronotropic index. This approach to assess chronotropic response accounts for age, functional capacity, and resting heart rate; it is not merely a reflection of physical fitness or exercise time. The chronotropic index is not related to physical activity or functional capacity, is not elevated among patients with a poor functional capacity, is unaffected by exercise protocol, and is not affected by the stage of exercise used for measurement.^1,11

Except for patients undergoing sophisticated gas-exchange analyses, exercise capacity in METs is estimated, not directly measured. Because all patients in this study underwent symptom-limited testing, we chose to consider the ratio of heart rate reserve used to metabolic reserve used at peak exercise, when, by definition, the proportion of metabolic reserve used has a value of 1.0. Thus, using this approach, the chronotropic index is based entirely on directly measured variables, ie, resting heart rate, peak heart rate, and age.¹² Because the value of the chronotropic index is independent of stage of exercise considered, this measure at least indirectly takes into account effects of functional capacity.¹¹ By using measures obtained at peak exercise and thereby eliminating the variability associated with estimation, rather than direct measurement, of oxygen consumption in METs, the value of chronotropic index we calculated also could be considered the fraction of heart rate reserve used at peak exercise.

Scintigraphic techniques used in our laboratory have been described in detail elsewhere.¹³ Patients were injected with 74 to 111 Bq of thallous chloride Tl 201 one minute prior to termination of the treadmill test; Tl 201 reinjection was not routinely performed. Single-photon emission computed tomography imaging was performed using a standard acquisition method within 10 minutes of stress and repeated 3 to 4 hours later for redistribution imaging. Images were reconstructed using a Hamming filter and projected in short-axis, vertical, and horizontal long-axis views in a side-by-side display. A 12-segment schema was used to describe thallium perfusion defects.⁵ For any given segment, ischemia was interpreted in the presence of more than 20% reversibility and scar by the presence of counts less than 80% of maximum (<70% for the posterior wall). Segments thought to be attenuated because of breast artifact were not considered abnormal. The coding of the thallium data was blinded with regard to clinical and exercise data and the hypothesis of this study. Normal values and prognostic properties of thallium scintigraphy in our laboratory in a similar patient population have been described in detail elsewhere.⁶

The primary end point of this study was all-cause mortality during approximately 2 years of follow-up. Mortality was ascertained using the Social Security Death Master Files (Epidemiology Resources, Newton, Mass).¹⁴ The cause of death was assessed by examination of death certificates and cardiac mortality was considered a secondary end point; probable causes of death were coded by a reviewer who was blinded to clinical, exercise, and thallium data and the study hypothesis.

Subjects were divided into 2 groups based on ability to reach at least 85% of the age-predicted maximum heart rate as assessed by the equation 220 −age in years. Comparisons between groups on continuous variables were made using the t test for normally distributed variables and the Wilcoxon rank sum test for other variables; comparisons regarding categorical variables were performed with the χ² and Fisher exact tests as appropriate. Chronotropic variables were related to thallium perfusion defects by calculation of unadjusted odds ratios (ORs) and CIs.

Chronotropic variables were related to all-cause mortality using Kaplan-Meier plots, log-rank χ² statistics, and stepwise forward Cox proportional hazards analyses.¹⁵ Potential confounders included age; sex; thallium perfusion defects; smoking; diabetes; use of lipid-lowering drugs; hypertension; pathologic Q waves; history of known coronary disease and lung disease; use of heart rate–lowering calcium channel blockers, digoxin, or nitrates; and angina, claudication, or test-terminating blood pressure changes during exercise. In secondary analyses, patients were divided into 4 groups according to thallium perfusion defects (absent or present) and chronotropic response (normal or abnormal).

All analyses were performed using Version 6.12 of the SAS statistical package (SAS Institute, Cary, NC).

A total of 1877 men and 1076 women were eligible for analyses. Of these 2953 patients, 316 (11%) failed to reach at least 85% of their age-predicted maximum heart rate and 762 (26%) had a low chronotropic index (ie, they used less than 80% of the heart rate reserve by the time they reached peak exercise). The relatively high prevalence of chronotropic abnormalities likely reflected the higher risk nature of a population of patients referred for stress thallium studies compared with populations of healthy adult subjects.

Baseline characteristics of the study subjects according to ability to reach at least 85% of the age-predicted maximum heart rate are shown inTable 1. Patients who failed to reach this heart rate were older; were more likely to smoke; had a history of hypertension, diabetes, chronic lung disease, and known coronary disease; and were more likely to be using heart rate–lowering calcium channel blockers, digoxin, or nitrates.

Exercise characteristics according to ability to reach target heart rate are shown in Table 2. Patients who failed to reach their target heart rate were more likely to experience angina and had a lower exercise capacity. As expected, the chronotropic index was much lower among the patients who failed to reach their target heart rate (0.55 vs 0.97, P<.001). There was a close correlation between chronotropic index and percentage of target heart rate achieved (Figure 1).

The most common primary reason (as many as 3 could be given) for exercise termination was fatigue (95% among patients who reached their target heart rate vs 83% among those who failed to; P<.001). Patients who failed to reach their target rate were more likely to have exercise stopped, either primarily or secondarily, because of blood pressure changes (8.2% vs 3.2%, P<.001), claudication (5.1% vs 0.9%, P<.001), angina (14.6% vs 3.1%, P<.001), or, when they were interpretable, ischemic ST-segment changes (5.7% vs 2.6%, P=.002). Thirty-five patients (1.2%) had their tests stopped because of arrhythmias, but there was no association between development of arrhythmias and the exercise heart rate response.

Thallium perfusion defects in at least 1 of 12 segments were noted in 612 patients (21%); reversible defects in at least 1 segment were noted in 311 patients (11%), of whom 116 (37%) also had fixed defects consistent with scar. Failure to achieve at least 85% of the age-predicted maximum heart rate was associated with the presence of thallium perfusion defects (39% vs 19%; OR, 2.76; 95% CI, 2.15-3.53; P<.001) and with the presence of reversible thallium defects (17% vs 10%; OR, 1.86; 95% CI, 1.35-2.56; P<.001). Similarly, a low chronotropic index was associated with thallium perfusion defects (30% vs 17%; OR, 2.07; 95% CI, 1.71-2.50; P<.001) and with reversible defects (15% vs 9%; OR, 1.75; 95% CI, 1.37-2.25; P<.001).

During 2 years of follow-up, there were 91 deaths, of which 22 were deemed likely to be cardiac. Patients who failed to reach 85% of their age-predicted maximum heart rate were more likely to die (log-rank χ²=21; P<.001) (Figure 2). Similarly, patients who had a low chronotropic index had higher death rates (log-rank χ²=36; P<.001) (Figure 3).

In unadjusted proportional hazards models, failure to reach 85% of the age-predicted maximum heart rate was predictive of death (unadjusted relative risk [RR], 2.88; 95% CI, 1.79-4.62; P<.001; absolute 2-year difference, 4% [95% CI, 0.6%-8.8%]); similarly, the proportion of the maximum heart rate achieved, when considered as a continuous variable, was predictive of death (for a 1-SD decrease, RR, 1.42; 95% CI, 1.19-1.69; P<.001; absolute 2-year difference, 4.5% [95% CI, 1.8%-7.0%]). A low chronotropic index was also predictive of death (unadjusted RR, 3.26; 95% CI, 2.16-4.93; P<.001); similarly, the chronotropic index, when considered as a continuous variable, was predictive of death (for a 1-SD decrease, unadjusted RR, 2.38; 95% CI, 1.55-3.67; P<.001).

Because the exercise heart rate response was strongly associated with rate-lowering calcium channel blocker use (Table 1), stratified analyses were performed to determine whether this variable either confounded or modified the association between impaired chronotropic response and death. Among the 2534 patients who were not taking these calcium channel blockers, there were 74 deaths. Failure to reach 85% of the age-predicted maximum heart rate was predictive of death (RR, 2.34; 95% CI, 1.31-4.19; P=.004), as was a low chronotropic index (RR, 2.98; 95% CI, 1.89-4.70; P<.001). Among the 419 patients who were taking rate-lowering calcium channel blockers, there were 17 deaths. In this group, failure to reach 85% of the age-predicted maximum heart rate was also predictive of death (RR, 4.62; 95% CI, 1.78-11.98; P=.002) as was a low chronotropic index (RR, 5.52; 95% CI, 1.59-19.22; P=.007). Thus, impaired chronotropic response was predictive of death irrespective of calcium channel blocker use. No interactions were noted between calcium channel blocker use and exercise heart rate responses for prediction of death.

Failure to reach 85% of the age-predicted maximum heart rate tended to predict cardiac death (RR, 2.50; 95% CI, 0.92-6.77; P=.07). A low chronotropic index, however, was predictive of cardiac death (RR, 4.21; 95% CI, 1.80-9.85; P<.001).

After adjusting for potential confounders (listed in the "Methods" section) in Cox proportional hazards analyses, failure to reach 85% of the age-predicted maximum heart rate remained independently predictive of death (adjusted RR, 1.84; 95% CI, 1.13-3.00;P=.01). In this model, thallium perfusion defects were also independently predictive of death (adjusted RR, 2.10; 95% CI, 1.37-3.22; P<.001). When considered as a continuous variable, percentage of target heart rate achieved was also independently predictive of death (for a 1-SD decrease, adjusted RR, 1.21; 95% CI, 1.02-1.44; P=.03). Analogously, a low chronotropic index was independently predictive of death (adjusted RR, 2.19; 95% CI, 1.43-3.44; P<.001) after adjusting for potential confounders. When considered as a continuous variable, chronotropic index was also independently predictive of death (for a 1-SD decrease, adjusted RR, 1.54; 95% CI, 1.04-2.27; P=.03). In a model that adjusted for thallium perfusion defects, a low chronotropic index was also predictive of cardiac death (adjusted RR, 3.14; 95% CI, 1.33-7.45; P=.009).

In further analyses, patients were stratified according to exercise heart rate response and the presence or absence of either any exercise thallium perfusion defects or reversible defects. Patients who had normal heart rate responses and normal thallium scans were at low risk for death, whereas patients who had one or both abnormalities were at increased risk (Figure 4). The presence of chronotropic incompetence alone carried approximately the same mortality risk as the presence of perfusion defects alone, while the presence of both abnormalities was associated with a markedly increased risk. In age- and sex-adjusted Cox regression analyses (Table 3), the combination of failure to reach 85% of the age-predicted maximum heart rate and thallium perfusion defects was associated with a high risk for death (adjusted RR, 2.72). Patients with both a low chronotropic index (Table 3) and thallium perfusion defects also were at high risk (adjusted RR, 3.31). In separate analyses, no interactions were noted between exercise heart rate responses and thallium perfusion defects regarding prediction of death.

In this cohort of consecutive patients referred for stress testing with thallium imaging, chronotropic incompetence was predictive of all-cause mortality during 2 years of follow-up. Both failure to achieve 85% of the age-predicted maximum heart rate and a low chronotropic index were associated with adverse risk profiles and were associated with thallium perfusion defects; nonetheless, even after adjusting for potential confounders in both stratified and multivariate analyses, chronotropic incompetence was independently predictive of death. The combination of chronotropic incompetence, especially when assessed by a low chronotropic index, and thallium perfusion defects was not rare and was associated with a particularly high mortality risk. Thus, our findings indicate that a poor chronotropic response to exercise does not mean that the test itself is invalid; rather, it provides important additional information to other exercise and imaging data.

Exercise stress testing relies on the development of a sufficient workload to produce evidence of hypoperfusion in the presence of significant obstructive epicardial coronary artery disease.¹⁶ Previous groups have focused on the decreased ability of nuclear techniques to accurately diagnose coronary artery disease in the setting of a submaximal heart rate response.^3,17 Our current study expands on these studies in 3 important respects: (1) chronotropic incompetence was associated with higher likelihoods of total and reversible thallium perfusion defects; (2) despite this association, chronotropic incompetence was identified as an important and independent predictor of mortality; and (3) chronotropic incompetence identified a group of patients with thallium perfusion defects who are at particularly high risk and therefore arguably deserving of aggressive investigation and treatment.

The mechanisms by which chronotropic incompetence predicts mortality risk are unclear. Our current study shows that associated myocardial perfusion defects do not entirely explain the increased risk. Chronotropic incompetence may be reflective of a modulation of autonomic tone that reflects more severe cardiovascular perturbations, as in, for example, patients with moderate-to-severe congestive heart failure.^18- 19 There is a well-known association between autonomic dysfunction and cardiovascular risk.^20- 23

Our study has several limitations. First, the use of a stepped protocol may lead to overestimation of exercise workloads at different stages of exercise²⁴; therefore, in this study, measurement of the chronotropic index was limited to peak exercise. Further studies will be needed among patients undergoing gas-exchange metabolic stress testing to confirm the association between chronotropic index, as measured in submaximal stages of exercise, and risk. Second, we did not have measures of left ventricular function. Although left ventricular cavity dilatation was systematically noted, in this low-risk population in which patients with congestive heart failure were deliberately excluded, there were too few patients with this finding to allow for meaningful analyses. Third, it is beyond the scope of our study to determine whether chronotropic incompetence is a potentially modifiable risk factor.

It has been argued that a poor heart rate response to exercise renders a stress imaging study invalid; some have argued that patients who fail to reach 85% of their age-predicted maximum heart rate should not have a radioisotope injected near peak exercise. Our data indicate that chronotropic incompetence means much more than just an invalid or nondiagnostic study. Furthermore, the additional associations that perfusion defects and chronotropic response have with outcome suggest that nuclear imaging should be performed even among patients who do not mount a normal exercise heart rate response. Thus, although multiple variables were considered using sophisticated multivariate analyses, our results primarily suggest that clinicians should consider chronotropic incompetence to be a prognostic finding that is independent of, but as ominous as, thallium perfusion defects.