BRAWNER, CLINTON A.; KETEYIAN, STEVEN J.; EHRMAN, JONATHAN K.
BRAWNER, C. A., S. J. KETEYIAN, and J. K. EHRMAN. The relationship of heart rate reserve to V̇O2 reserve in patients with heart disease. Med. Sci. Sports Exerc., Vol. 34, No. 3, pp. 418–422, 2002. Recent reports indicate that among healthy adults, % heart rate reserve (HRR) is more closely related to %V̇O2 reserve (V̇O2R) than it is to %V̇O2max. This finding, in addition to the disparity between %HRR and %V̇O2max which is greater at low intensities and among low fit individuals, led the American College of Sports Medicine to adopt the use of %V̇O2R in place of %V̇O2max when prescribing exercise intensity among healthy adults and persons with heart disease. However, the relationship of %HRR to %V̇O2R among persons with heart disease has not been described.
Purpose: Among patients with a myocardial infarction (MI) and patients with chronic heart failure (HF), test the hypothesis that %V̇O2R is equivalent to %HRR, while %V̇O2peak is not.
Methods: Using a clinical cardiology outcomes database, symptom-limited treadmill tests from 65 patients with MI and 72 patients with HF were identified. Heart rate and V̇O2 data were measured continuously and averaged every 15 s. For each subject, linear regression was used to calculate the slope and y-intercept of %HRR versus %V̇O2R (assuming rest V̇O2 = 3.5 mL·kg−1·min−1) and %HRR versus %V̇O2peak. Mean slope and y-intercept were calculated for each group and compared with the line of identity (slope = 1, y-intercept = 0).
Results: For the MI and HF groups, the slope of %HRR versus %V̇O2R was 0.96 ± 0.02 (±SE;P = NS, slope vs 1) and 0.97 ± 0.02 (P = NS), respectively. And the y-intercept was −1.9 ± 2.1% (P = NS, y-intercept vs 0) and −5.9 ± 2.1% (P < 0.05) for MI and HF, respectively. For both patient groups, the regression of %HRR versus %V̇O2peak resulted in a line that differed (P < 0.001) in both slope and y-intercept from the line of identity.
Conclusions: In patients with heart disease, %HRR is a better estimate of %V̇O2R than %V̇O2peak. This finding does not affect the current recommended use of %HRR. However, when prescribing exercise based on V̇O2, relative intensity should be expressed as %V̇O2R.
Until recently, the linear relationship between % heart rate reserve (HRR) and % maximum oxygen consumption (V̇O2max) led to the use of %HRR as a surrogate for %V̇O2max (3). However, studies involving healthy adults showed that, although %HRR is not equivalent to %V̇O2max, it is equivalent to (9), or more closely related to (10), %V̇O2 reserve (V̇O2R; where V̇O2R = peak V̇O2 − rest V̇O2) during leg ergometry and treadmill exercise, respectively.
In 1998 the American College of Sports Medicine (ACSM) revised its exercise guidelines for healthy adults and adopted the use of %V̇O2R in place of %V̇O2max when prescribing exercise intensity (4,5). This change was based on the 5–15% disparity between %HRR and %V̇O2max, with the largest difference existing at lower intensities; greater disparity between %HRR and %V̇O2max among low-fit individuals; and the ratio between %HRR and %V̇O2R that was newly identified to be equal to a line of identity. In addition, ACSM now recommends the use of %V̇O2R when prescribing exercise intensity for patients enrolled in outpatient cardiac rehabilitation programs (4). However, whether %HRR is more closely related to %V̇O2R in patients with heart disease has not been described.
This study tests the hypothesis that %HRR is equivalent to %V̇O2R, whereas %HRR is not equivalent to %V̇O2peak, during treadmill exercise among patients with heart disease. A group of patients with a history of myocardial infarction (MI) and a group of patients with chronic heart failure (HF) were assessed.
Demographic, medical history, and exercise test data for all patients who undergo exercise testing in the Preventive Cardiology unit of Henry Ford Hospital are entered into the PREventive Cardiology Outcomes (PRECO) database. Since 1997, data on nearly 3000 patients have been entered. Patients are referred for testing as part of enrollment into cardiac rehabilitation or from the hospital’s general cardiology or heart failure clinics. We identified 1153 patients new to the laboratory for the period January 1997 through December 2000 by using the following criteria: age ≥ 18 yr; sinus rhythm; gas analysis during treadmill testing with peak respiratory exchange ratio ≥ 1.05; and reason for test termination reported as fatigue, exercise-induced hyperpnea, or patient request. Patients whose exercise test was prematurely stopped secondary to a disease-related sign/symptom such as arrhythmia, chest pain, marked dyspnea, or claudication were not included in this analysis.
A subquery was performed to identify three separate groups of patients. These groups were patients whose primary event was a MI (N = 65), those with HF (N = 72), and those with only risk factors (RF) for coronary artery disease (N = 42). Patients in the MI group were without a history of coronary revascularization and did not have evidence of left ventricular dysfunction in their hospital record. The HF group consisted entirely of patients with a resting left ventricular ejection fraction ≤ 35%, as measured by echocardiogram or left ventriculogram during cardiac catheterization. Patients with either an ischemic or nonischemic cardiomyopathy were included. The RF group consisted of patients with no evidence of coronary artery disease in the hospital record and without a history of arrhythmia, valvular surgery, HF, or cardiac transplant. All patients signed informed consent before exercise testing. Use of data from the PRECO database was approved by the hospital’s Institutional Review Board.
Exercise testing protocol.
Patients were prepped using the Mason-Likar electrode placement for recording a 12-lead electrocardiogram (Q3000, Quinton Instruments, Seattle, WA) during exercise. Supine, 3-min standing, and peak heart rates were measured. Expired gases were measured continuously by open-circuit spirometry (CPX/D, MedGraphics, St. Paul, MN) and analyzed in a breath-by-breath mode. Heart rate was electronically transferred from the electrocardiograph to the metabolic cart throughout the test. Tests were performed on a motorized treadmill (Quinton Instruments, Seattle, WA) using 3-min stages and workload increments of approximately 2 METs per stage. Peak exercise measures were identified as the highest value attained during the final minute of exercise.
Data collection and analysis.
Gas exchange and heart rate data were exported from the metabolic cart software (BreezeEx 3.04, MedGraphics) into a computer spreadsheet program (Excel97, Microsoft). Heart rate and gas exchange measures were reported in 15-s intervals. The following data were exported from PRECO for analysis: gender, age, body mass, body mass index, peak V̇O2, standing and peak heart rate, peak rate-pressure product, peak respiratory exchange ratio, risk factors for coronary artery disease, and beta-adrenergic blockade therapy.
Percentages of peak HRR, V̇O2R, and V̇O2 were calculated for each 15-s interval. In the calculation of HRR, the 3-min standing heart rate was used for resting heart rate. In the calculation of V̇O2R, resting V̇O2 was assumed to be 3.5 mL·kg−1·min−1 because this value is commonly used in clinical exercise physiology (4). For each subject, linear regression (Quattro Pro 8, Corel Corp.) using all the 15-s interval data was used to calculate the slope and y-intercept for both %HRR versus %V̇O2R and %HRR versus %V̇O2peak. Mean slope and y-intercept were then calculated for each of the three study groups. Additionally, using each subject’s equations for %HRR versus %V̇O2R and %HRR versus %V̇O2peak, we calculated the mean %V̇O2R and %V̇O2peak, respectively, that would result from exercise intensities prescribed at 50% and 80% HRR for each study group.
Student t-tests (two-tailed) were used to determine whether the mean slope and y-intercept for each group differed from the line of identity (slope = 1 and y-intercept = 0). Mann-Whitney tests were used to compare the resultant mean %V̇O2R and %V̇O2peak calculated from %HRR at both 50% and 80%. Alpha level was set at P ≤ 0.05. A subanalysis was performed within both the MI and HF groups, based on beta-adrenergic blockade therapy.
Subject characteristics are shown in Table 1. For patients in the HF group, resting ejection fraction was 22 ± 8% (mean ± SD), and the number of patients with New York Heart Association functional classification I, II, and III was 11 (15%), 44 (61%), and 17 (24%), respectively.
The results of the regressions of %HRR versus %V̇O2R and %HRR versus %V̇O2peak for the MI, HF, and RF groups are shown in Table 2. For all study groups, the slope of %HRR versus %V̇O2R was not significantly different from the line of identity. In contrast, for all study groups both the slope and y-intercept for %HRR versus %V̇O2peak were significantly different (P < 0.001) from the line of identity. The resultant regression lines are shown in Figure 1.
Table 3 shows a within-group analysis for the MI and HF groups based on beta-adrenergic blockade therapy. Irrespective of this therapy, the slope of %HRR versus %V̇O2R was not significantly different from the line of identity. Among patients with HF on beta-adrenergic blockade therapy, the y-intercept of %HRR versus %V̇O2R was significantly different (P < 0.05) from the line of identity.
Table 4 shows the resultant %V̇O2R and %V̇O2peak based on exercise intensities computed at 50% and 80% of HRR for the three study groups.
We showed that among patients with MI or HF, %HRR is more closely related to %V̇O2R than it is to %V̇O2peak. This finding is important because, until recently, %V̇O2max was believed to be best reflected by %HRR in both healthy persons and those with coronary artery disease (4,6). Based on data from Swain et al. (9,10), when prescribing exercise intensity for apparently healthy subjects, it is more appropriate to associate %HRR with %V̇O2R rather than %V̇O2peak. This change was adopted by ACSM for both healthy people and patients enrolled in outpatient cardiac rehabilitation programs (4,5). Our data support ACSM’s recommendations for the use of %V̇O2R in patients with heart disease.
As expected, within both the MI and HF study groups, differences in peak heart rate were observed when stratified for the use of beta-adrenergic blockade therapy (11). However, in both groups, we demonstrated that the relationship of %HRR to %V̇O2R was not significantly different from the line of identity despite beta-adrenergic blockade therapy. This finding is important because such therapy is being increasingly used in patients after MI and in those with HF (2,7).
Although the slope of %HRR versus %V̇O2R for each of our study groups was not significantly different from the line of identity, the y-intercept was different in the HF and RF groups. This suggests that %HRR is not equal to %V̇O2R. However, as our data show, %HRR is still closer to %V̇O2R, than is the relationship between %HRR and %V̇O2peak. Among patients with HF the y-intercept was −6% for %V̇O2R versus −35% for %V̇O2peak (Table 2, Fig. 1B). Therefore, the observed disparity between %HRR and %V̇O2R at lower exercise intensities (near the y-intercept) is less than what was noted between %HRR and %V̇O2peak.
The applications of our findings are twofold. First, exercise professionals should continue to use %HRR to guide exercise intensity in healthy people and patients with heart disease. However, %HRR still overpredicts %V̇O2R by 5% and 10% across exercise intensities of 50–80% in patients with MI and HF, respectively (Table 4). This disparity is less than what we have shown for %HRR and %V̇O2peak across the same exercise intensity range. Second, when setting training intensity based on a relative V̇O2, such as occurs when conducting a study that assesses exercise response at a given %V̇O2peak, intensity should be expressed as %V̇O2R not %V̇O2peak. This is consistent with Swain et al. (9,10), who state that the use of %V̇O2R lessens discrepancies in relative training intensities between individuals by taking into account the magnitude of each person’s V̇O2R.
A limitation of this study is the absence of a measured V̇O2 at rest. In our data set, it was assumed to be 3.5 mL·kg−1·min−1, which is used by ACSM (4) and is similar to other measured values (8,9,10). However, Ainsworth et al. (1) report standing quietly is equal to 1.2 METs (i.e., V̇O2 = 4.2 mL·kg−1·min−1). To assess the effect of this assumption on the %HRR-%V̇O2R relationship, we recalculated the regressions for the MI group. The resultant slope was 0.91 ± 0.02 (P < 0.001, slope vs 1) and y-intercept was −3.3 ± 2.1% (P = NS, y-intercept vs 0). Although the slope is significantly different, it remains numerically closer (slope = 0.91 ± 0.02) to the line of identity (slope = 1.0) than does the %HRR-%V̇O2peak relationship (slope = 1.20 ± 0.03, Table 2). Ultimately, a measured resting V̇O2 while standing would be best; however, taking the 5–10 min needed to measure a true resting V̇O2 for each patient in a busy clinical exercise testing laboratory may not be necessary or practical.
In studies by Swain et al. (9,10) in which V̇O2 was measured at rest, the regression of %HRR to %V̇O2R yielded slopes of 1.00 ± 0.01 and 1.03 ± 0.01, and y-intercepts of −0.1 ± 0.6 and 1.5 ± 0.6, respectively. Because our data (MI: slope = 0.96 ± 0.02, y-intercept = −1.9 ± 2.1%; HF: slope = 0.97 ± 0.02, y-intercept = −5.9 ± 2.1%) are in general agreement with these reports, we believe the use of an estimated versus a measured resting V̇O2 does not result in a clinically meaningful effect on the HRR-V̇O2R relationship.
Rationale for using %V̇O2R in place of %V̇O2peak when prescribing exercise intensity among healthy adults has been previously shown. However, the relationship of %HRR versus %V̇O2R had not been described in patients with heart disease. This study supports the use of %V̇O2R in patients with an MI or HF, whether or not they are receiving beta-adrenergic blockade medications. Therefore, when prescribing exercise based on V̇O2, intensity should be expressed as %V̇O2R.
Address for correspondence: Clinton A. Brawner, Preventive Cardiology, 6525 Second Avenue, Detroit, MI 48202; E-mail: firstname.lastname@example.org.
1. Ainsworth, B. E., W. L. Haskell, M. C. Whitt, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med. Sci. Sports. Exerc. 32 (Suppl. 9): S498–S516, 2000.
2. American College of Cardiology/American Heart Association Task Force on Practice Guidelines. ACC/AHA practice guidelines for the management of patients with acute myocardial infarction. J. Am. Coll. Cardiol. 28: 1328–1428, 1996.
3. American College of Sports Medicine. ACSM’s Guidelines for Exercise Testing and Prescription, 5th Ed. Baltimore: Williams & Wilkins, 1995, pp. 158–162.
4. American College of Sports Medicine. ACSM’s Guidelines for Exercise Testing & Prescription
, Sixth Ed. Philadelphia: Lippincott Williams & Wilkins, 2000, pp. 145, 170–171, 302, 308.
5. American College of Sports Medicine. The recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in healthy adults. Med. Sci. Sports Exerc. 30: 975–991, 1998.
6. Foss, M. L., and S. J. Keteyian (Eds). Methods of aerobic training and physiologic responses. In:Fox’s Physiological Basis for Exercise and Sport,
Sixth Ed. Boston: McGraw-Hill, 1998, pp. 298–300.
7. Heart Failure Society of America. HFSA guidelines for management of patients with heart failure caused by left ventricular systolic dysfunction. J. Cardiac Failure 5: 358–382, 1999.
8. McGandy, R. B., C. H. Barrows, A. Spanias, A. Meredith, J. L. Stone, and A. H. Norris. Nutrient intakes and energy expenditure in men of different ages. J. Gerontol. 21: 581–587, 1966.
9. Swain, D. P., and B. C. Leuholtz. Heart rate reserve is equivalent to %VO2
reserve, not to %VO2
max. Med. Sci. Sports Exerc. 29: 410–414, 1997.
10. Swain, D. P., B. C. Leuholtz, M. E. King, L. A. Haas, and J. D. Branch. Relationship between % heart rate reserve and % VO2
reserve in treadmill exercise. Med. Sci. Sports Exerc. 30: 318–321, 1998.
11. Wilmore, J. H., B. J. Freund, M. J. Joyner, et al. Acute response to submaximal and maximal exercise consequent to beta-adrenergic blockade: implications for the prescription of exercise. Am. J. Cardiol. 55: 135D–141D, 1985.