- Email: [email protected]
Effects of endurance exercise training on left ventricular systolic performance and ventriculoarterial coupling in patients with coronary artery disease
Effects of endurance exercise training on left ventricular systolic performance and ventriculoarterial coupling in patients with coronary artery disease Morton R. Rinder, MD, Tom R. Miller, MD, PhD, and Ali A. Ehsani, MD St Louis, Mo
Background Endurance exercise training can increase left ventricular (LV) ejection fraction during dynamic exercise in coronary artery disease. This adaptation may be mediated by altered cardiac loading conditions rather than an improvement in intrinsic LV systolic function. To minimize these confounding effects, we used isometric handgrip exercise to assess the training-induced changes in LV systolic function and ventriculoarterial coupling. Methods Twenty-six patients (52 ± 2 years of age) trained for 12 months. LV function was assessed with radionuclide ventriculograpy.
Results LV systolic reserve (the change in LV ejection fraction from rest to handgrip exercise) increased from –7.32 ± 1.2 to –3.4 ± 1.1 (P = .033) without acute changes in end-diastolic volume or the effective arterial load. LV end-systolic elastance increased 37% (P = .039) during handgrip exercise. Resting end-diastolic volume increased and the effective arterial load decreased after training. Conclusions Data suggest that in coronary artery disease adaptations to exercise training include a lower effective arterial load and an increase in EDV at rest, with an improvement in LV systolic function detectable only during afterload stress. (Am Heart J 1999;138:169-74.)
Endurance exercise training can improve endurance, exercise capacity, and the minimal work rate required to induce myocardial ischemia in patients with coronary artery disease (CAD).1-4 Adaptations in skeletal muscle and the autonomic nervous system were generally thought to be responsible for these observations.2,3 However, previous studies that involved a progressively increased exercise intensity demonstrated that in addition to the peripheral adaptations, an improvement in left ventricular (LV) ejection fraction (EF) and stroke volume is also attainable by endurance exercise training in patients with CAD.1,4,6 EF is affected not only by changes in contractile state, but also by alterations in cardiac loading conditions. Therefore the possibility of a larger preload or a lower LV wall stress contributing to a higher EF From the Division of Gerontology and Geriatrics, and the Cardiovascular Division, Department of Medicine, and the Division of Nuclear Medicine, Mallinckrodt Institute of Radiology, Washington University School of Medicine. Supported by NHLBI Grant HL17646 (SCOR in Ischemic Heart Disease), Claude D. Pepper Older American Independence Center AG-13629, and AG-12822. Morton R. Rinder, M.D. was supported by NIH Institutional Training Grant 5 T32 HL07081-22. Submitted January 26, 1998; accepted December 17, 1998. Reprint requests: Ali A. Ehsani, MD, Washington University School of Medicine, 660 S Euclid Ave, Campus Box 8113, St Louis, MO 63110. E-mail: [email protected] Copyright © 1999 by Mosby, Inc. 0002-8703/99/$8.00 + 0 4/1/96758
reported in these studies1,4 cannot be excluded despite a higher systolic blood pressure during peak exercise in the trained state.1 Virtually all previous studies evaluating the cardiovascular adaptations to endurance exercise training have used trained muscle groups during the exercise stimulus that can induce a greater vasodilatation in the trained state and, in turn, a lower systolic load accounting for an improved EF during dynamic exercise.1-4,6 Therefore it may be argued that the changes in ventricular loading conditions caused by adaptations in the skeletal muscle and autonomic nervous system were, at least in part, responsible for the observed improvement in EF after training in previous studies.1,4 The effect of endurance exercise training on the interaction between the ventricular and arterial system, that is, “ventriculoarterial coupling,”7 has not been previously reported in patients with CAD. The adaptive changes in the ventriculoarterial coupling in response to endurance exercise training should provide useful information regarding the influence of the arterial system (effective arterial elastance) on systolic function (LV elastance).7 Under physiologic conditions at a given preload, the interaction between these 2 components provides an optimal level of stroke volume or stroke work to meet the metabolic demands of the body and to maintain hemodynamic homeostasis. Therefore the purpose of this study was to assess the adaptive changes in myocar-
American Heart Journal July 1999
170 Rinder, Miller, and Ehsani
Table I. Adaptations in hemodynamics in response to exercise training in CAD Before Training
SBP (mm Hg) DBP (mm Hg) Heart Rate (beats/min) EF (%) LVEDV (mL) LVESV (mL) SV (mL)
After Training
Rest
Handgrip
120.3 ± 3.0 78.6 ± 2.0 66.0 ± 2.3 55.3 ± 2.9 163.3 ± 12.1 72.7 ± 13.8 90.5 ± 3.3
158.5 ± 5.5 107.8 ± 2.7 75.5 ± 2.3 47.9 ± 2.8 184.8 ± 9.8 89.5 ± 13.5 95.2 ± 5.9
Rest 123.6 ± 3.2 77.0 ± 1.7 61.1 ± 2.1* 57.1 ± 3.0 176.0 ± 11.8† 68.7 ± 11.4 107.3 ± 6.8§
Handgrip 163.5 ± 3.7 109.0 ± 2.5 70.9 ± 1.9ll 53.8 ± 3.1* 178.6 ± 12.4 76.2 ± 13.8‡ 102.0 ± 5.1
DBP, Diastolic blood pressure; SBP, systolic blood pressure; SV, stroke volume. Before vs after for the same conditions: *P < .001. †P = .039. ‡P = .054. §P = .011. llP = .017.
dial functional reserve and ventriculoarterial coupling in response to endurance exercise training in patients with CAD during isometric handgrip exercise. We used this type of stress in our study because we believe it should provide a suitable afterload stimulus that is unlikely to be affected by peripheral adaptations and because our training program included only endurance exercise and did not involve upper extremity exercise.
Methods Patients We studied 26 patients (24 men and 2 women, 52 ± 2 years of age) with CAD. All provided written consent, and the study protocol was approved by the Human Studies Committee of Washington University School of Medicine. This study is an extension of a previous study and includes 18 patients whose radionuclide data during dynamic exercise have been previously reported.1 We recruited 8 additional patients to have a sufficient sample size needed for detection of the training-induced changes in LV systolic performance during an afterload stress. Nineteen patients had a myocardial infarction. The interval between myocardial infarction and enrollment in the study was at least 3 months and averaged 14.5 ± 4.8 months. Three patients had coronary artery bypass graft surgery at least 3 months before their participation in the study. Ten patients had chronic stable effort angina, and 18 had exercise-induced myocardial ischemia detected by electrocardiographic or scintigraphic criteria during dynamic exercise. Of the 15 patients in whom coronary angiographic data were available, 3 had significant (≥70% luminal stenosis) single-vessel disease, 8 had 2-vessel disease, and 4 had 3-vessel disease. None of the patients had atrial fibrillation or valvular heart disease, and none underwent revascularization procedures during the study.
Cardiac medications All medications were prescribed by the patients’ personal physicians. Sixteen patients were taking nonselective β-blockers, 9 were taking long-acting nitrates, and 4 were taking cal-
cium-channel blockers. There were no changes in medications during the training program except in one patient who was treated with propranolol after enrollment in the study. In this patient the β-blockade was stopped 48 hours before the final radionuclide study. The interval between the last dose of cardiac medications and the radionuclide ventriculography averaged 18.2 ± 3.1 hours before and 19.0 ± 3.2 hours after training (P = .50).
Treadmill exercise test and maximal oxygen uptake . capacity (VO2 per max) Maximal symptom-limited exercise testing was performed on a treadmill according to the Bruce protocol with a repeat treadmill test . 2 weeks later for measurement of the highest attainable VO. 2 with a modified protocol as previously described.1 V O2max was defined as the mean of the 2 highest consecutive 30-second oxygen measurements .that met the following criteria: (1) attainment of a plateau of VO2 with increasing exercise intensity and/or (2) a respiratory exchange ratio exceeding 1.15. It was possible to obtain . VO2max .in most patients without angina. In the. others, the highest VO2 recorded was designated as peak V O2.
LV function at rest and during handgrip isometric exercise LVEF and end-diastolic volume (LVEDV) were determined with the use of the electrocardiographically gated cardiac blood pool imaging technique as previously described.1,8,9 After completion of resting images, the patients performed isometric exercise at 30% of the previously determined maximal voluntary contraction for 3 minutes. Image acquisition was repeated between minutes 2 and 3 of isometric exercise with measurement of cuff blood pressure in the contralateral arm. The same protocol was repeated after 12 months of training. Images were analyzed in a blinded fashion. LV volumes data were available in 12 of the patients who underwent imaging before and after exercise training. The LV end-systolic volume (LVESV) and stroke volume were derived from the LVEDV and EF.1 We estimated the LV end-systolic pressure
American Heart Journal Volume 138, Number 1, Part 1
Rinder, Miller, and Ehsani 171
Table II. Effects of exercise training on ventriculoarterial coupling in CAD Before Training Rest Ees (mm Hg/mL) Ea (mm Hg/mL) Ea/Ees
2.03 ± 0.34 1.21 ± 0.05 0.85 ± 0.19
After Training
Handgrip 1.90 ± 0.27 1.54 ± 0.12 1.07 ± 0.24
Rest 2.03 ± 0.31 1.05 ± 0.09† 0.70 ± 0.15
Handgrip 2.60 ± 0.42* 1.47 ± 0.09 0.82 ± 0.20‡
Before vs after for the same condition: *P = .039. †P = .009. ‡P = .013.
(ESP) by the following formula: ESP = [(2 × SBP) + DBP]/3 as reported by Kelly et al.10 LV contractile function was evaluated by LV elastance, the ratio of ESP/LVESV designated as Ees, that represents the slope of the LV pressure/volume curve at end ejection and provides a good estimate of LV contractility. The effective arterial load (Ea) was estimated as ESP/stroke volume (arterial elastance). Ventriculoarterial coupling was expressed as the ratio of Ea/Ees.
Exercise program The duration of the supervised endurance exercise training program was 12 months, as previously described. 1 Patients were encouraged to exercise for 60 minutes 3 days a week for the first 3 months and 5 times per week thereafter. The training intensity was 60% to 70% of the maximal heart rate for the first 3 months and was progressively increased to 80% to 90% of maximal heart rate in the last 3 months of training.
Statistical analysis Student t test for paired observations was used for comparison of the data before and after training. Chi-square and leastsquared linear regression analyses were performed when appropriate. General linear model was used to handle the missing data points. Data are expressed as mean ± SE.
Results Improvement in functional capacity and clinical status The patients exercised 4.1 ± 0.13 days per week for the last 9 months of the training program. The intensity of training averaged 85.3% ± 1.5% of the maximal attainable heart rate in the last 6 months of the training program. Training induced a weight loss from 80.7 ± 2.3 kg to 76.9 ± 2.3 kg (P = .003). Eight of the 10 patients with chronic effort angina became asymptomatic after completion of the exercise program (P = .021). In the remaining 2 patients, the frequency and severity of angina was considerably less in the trained state. One patient was hospitalized because of chest pain after a maximal exercise test but myocardial infarction was ruled out. There were no other major complications or sequelae attributable to exercise.
. Heart rate, blood pressure, and peak VO2 The systolic and diastolic blood pressures at rest and during handgrip exercise did not change with training (Table I). Heart rate increased in response to isometric handgrip exercise both before and after training (Table I). However, resting and isometric exercise heart rates were lower in the trained state. The lower exercise heart rate was almost entirely from the training-induced resting bradycardia. The rate-pressure product during submaximal handgrip isometric exercise (30% of maximal voluntary contraction) was 12.19 × 10 3 ± 0.68 × 103 before and 11.64 × 103 ± 0.47 × 10 3 after training (P = .25). . Peak VO2 in absolute terms (1.88 ± 0.06 L/min vs 2.42 ± 0.09 L/min, P < .001) or when normalized for body weight (23.8 ± 0.8 vs 32.2 ± 1.4 mL/kg/min, P < .001) increased significantly in response to endurance exercise training. The maximal dynamic exercise (treadmill) heart rate (144.4 ± 3.7 beats/min vs 150.7 ± 3.4 beats/min, P = .027), systolic blood pressure (154.3 ± 6.9 mm Hg vs 166.8 ± 6.3 mm Hg, P = .016), and the rate-pressure product (22.19 × 10 3 ± 1.27 × 103 vs 25.21 × 103 ± 1.21 × 10 3, P = .002) were higher after training.
LV size, systolic function, and ventriculoarterial coupling Resting LVEDV was significantly larger after endurance exercise training (Table I). However, LVEDV during isometric handgrip exercise was similar before and after training. LVEDV increased acutely in response to isometric exercise before but not after training (∆LVEDV 21.5 ± 3.2 mL before and 2.6 ± 6.3 mL after, P = .044). LVESV at rest did not change but tended to be smaller during isometric handgrip exercise after training (P = .054, Table I). Stroke volume increased significantly (19%, P = .011) at rest but not during isometric handgrip exercise in response to training. EF at rest did not change but was higher during isometric exercise after training (Table I). There was a significant improvement of LV systolic reserve capacity, defined as the difference between resting EF and EF during isometric exercise (∆EF), in response to training
American Heart Journal July 1999
172 Rinder, Miller, and Ehsani
Figure 1
Figure 3
Effect of endurance exercise training on left ventricular systolic function. Decline in LVEF in response to an acute afterload stress induced by isometric handgrip exercise was considerably less (P = .033) after training, suggestive of an improvement in LV systolic function.
Figure 2
The adaptive changes in LV Ees in response to exercise training in CAD. In contrast to acute reduction in Ees in response to isometric exercise before training, patients exhibited an increase in Ees during isometric exercise in trained state (P = .009), suggestive of a significant improvement in LV systolic performance.
Inverse relation between adaptive changes in ventriculoarterial coupling, designated as Ea/Ees ratio, and stroke volume during isometric handgrip exercise. Patients who had better systolic function during acute increase in effective arterial load in trained state, as reflected in lower Ea/Ees ratio, were likely to have higher stroke volume.
trained state (Table II). The Ea/Ees ratio, used as a marker of the ventriculoarterial coupling, was significantly smaller during isometric exercise after training (Table II). During isometric exercise, there was an inverse relation between the training-induced changes in the ventriculoarterial coupling (Ea/Ees) and stroke volume; the patients who maintained a better systolic performance in response to an arterial load challenge, as reflected in the lower Ea/Ees ratio, exhibited a higher stroke volume in the trained state (Figure 3). The training-induced increase in stroke volume at rest correlated directly with changes in end-diastolic volume (Figure 4, A) and inversely with alterations in the effective arterial load (Figure 4, B).
Effects of training on myocardial ischemia characterized by a smaller decrease in EF during isometric exercise after training (–7.5 ± 1.2 vs –3.4 ± 1.06; P = .033) (Figure 1). The LV Ees under basal conditions did not change but was significantly higher (37%) during isometric handgrip exercise after training (P = .039, Table II). The LV contractile reserve, as reflected by the change Ees from rest to isometric exercise, improved with training (∆Ees: –0.125 ± 0.1 mm Hg/mL before and 0.563 ± 0.15 mm Hg/mL after, P = .009, Figure 2). In contrast to Ees, Ea was only significantly lower at rest and did not change during isometric exercise in the
Electrocardiographic changes. In the 11 patients who had ischemic ST-segment responses (ie, >0.1 mV horizontal or downsloping ST-segment depression) during dynamic exercise, the magnitude of ST-segment depression decreased from 0.15 ± 0.02 mV to 0.075 ± 0.02 mV (P = .002) at a given absolute exercise intensity after endurance exercise training. Rate-pressure product tended to be lower at these work rates (18.5 × 103 ± 1.7 × 103 vs 15.1 × 103 ± 1.03 × 103; n = 11; P = .064). ST-segment depression at maximal exercise did not change (0.14 ± 0.02 mV vs 0.159 ± 0.03 mV, P = .6) Regional LV wall motion disorders. Before training,
American Heart Journal Volume 138, Number 1, Part 1
Rinder, Miller, and Ehsani 173
Figure 4
A, Direct relation between training-induced changes in LVEDV and stroke volume at rest. Patients with greater LVEDV tended to have larger stroke volume. B, Inverse relation between training-induced changes in effective arterial load and stroke volume at rest. Patients with lower arterial load tended to have higher stroke volume.
9 of the 26 patients exhibited regional LV wall motion abnormalities during handgrip isometric exercise. After training, however, only 2 patients showed significant regional wall motion disorders during handgrip exercise (P = .042).
Discussion Our results suggest that long-term endurance exercise training can improve LV systolic function in patients with CAD, which appears to be independent of the vascular and autonomic adaptations that occur with this type of intervention. We used isometric handgrip exercise involving an untrained muscle group to induce an afterload stress that was not affected by the peripheral adaptations, as evidenced by the absence of differences in the arterial elastance and blood pressure responses to isometric exercise between the trained and untrained states. Significant adaptations induced by endurance exercise training in our patients were reflected in a substantial . increase (29%) in the highest attainable VO2 after training. With no apparent significant changes in afterload stress, the reduction in EF observed during isometric handgrip exercise was markedly less pronounced after training, suggesting that either contractile reserve was improved or there was a greater preload reserve preventing the decline in EF in the trained state. However, because the values for LVEDV during isometric exercise were not different before and after training, and the
change in LVEDV from rest to exercise was negligible in the trained state, it is unlikely that the improvement in EF response was mediated primarily by preload. Ees, which reflects the slope of the end-systolic pressure/end-systolic volume relation and is considered a reliable measure of LV contractility,7 significantly improved during isometric exercise in response to training, providing additional evidence that the increase in EF reserve was, at least in part, caused by improvement in LV contractile function. The mechanisms underlying these adaptive myocardial changes are unclear. However, it is likely that this improvement was, in part, the result of amelioration of myocardial ischemia, as reflected in decreases in regional LV wall motion abnormalities and effort angina in the trained state. In the basal resting state, the adaptations to training were limited to a larger LVEDV and a lower effective arterial load that contributed to the training-induced larger resting stroke volume. Although a slower heart rate can be responsible for a larger LVEDV, the magnitude of the reduction in heart rate in our patients was too small to account entirely for the higher LVEDV in the trained state. Some of the previous studies demonstrated no significant improvement in myocardial function with exercise training in patients with CAD.11-13 However, these studies did not use a sufficient exercise stimulus with respect to intensity and duration. It appears that a longer duration of training is needed so that the
American Heart Journal July 1999
174 Rinder, Miller, and Ehsani
patients can exercise safely at a higher intensity to improve LV systolic function.1 In virtually all the studies that showed improved LV function during exercise, the trained muscle group was used to evaluate adaptive changes in LV systolic function in response to stress. Therefore the role for afterload reduction and systolic unloading that can contribute to the observed increased EF could not be excluded.1,4,6 Assessment of LV function during isometric exercise that does not include a trained muscle group should mitigate this shortcoming. Therefore improved LV systolic function that we observed in this study is unlikely caused by decreased afterload. There are potential limitations to our study. (1) Our sample size was small (12 patients had volumetric data) and the study was performed mostly in men (24 of 26). Therefore the results cannot necessarily be extrapolated to women. (2) There was no control group. Thus spontaneous improvements in LV function cannot be excluded. However, this is unlikely because the minimum interval between myocardial infarction or coronary artery bypass grafting and enrollment in the study was 3 months. (3) It is possible that an improvement in myocardial ischemia contributing to amelioration of LV systolic dysfunction during isometric exercise in the trained state is mediated not only by an enhanced myocardial oxygen supply but also by a lower myocardial oxygen demand because of a slower heart rate even though other major variables affecting myocardial oxygen demand—that is, systolic pressure, rate-pressure product, LV systolic function, and LVEDV—were either similar or higher during isometric exercise after training. (4) Our patients, with few exceptions, had normal baseline LV systolic function and were only mildly symptomatic. Therefore our conclusions may not be extrapolated to patients with LV systolic dysfunction or those with extensive coronary atherosclerosis and frequent and severe angina. The improvement in LV systolic function in our patients is likely to be a consequence of a reduction in myocardial ischemia mediated by increased myocardial oxygen supply, lower maximal oxygen demand, or both. It is plausible that enhancement of LV systolic function in response to an afterload stress in our subjects is in part the result of an increase in myocardial oxygen supply, as shown by other investigators.14,15 The results of this study suggest that long-term endurance exercise training induces significant cardiovascular adaptations both in the basal state and during an afterload stress in patients with CAD. In the basal resting state, these adaptations appear to involve exclusively the arterial system and LV size, manifested by a lower effective arterial load and a larger LVEDV, both
contributing to a larger stroke volume at rest. In contrast, during an afterload stress our data suggest that the adaptations to training are confined to an improvement in LV systolic performance without any detectable vascular adaptations or changes in preload. Although the mechanisms underlying these adaptations are unknown, an improvement in myocardial ischemia in response to training is likely to play a significant role.
References 1. Ehsani AA, Biello DR, Schultz J, Sobel BE, Holloszy JO. Improvement of left ventricular contractile function by exercise training in patients with coronary artery disease. Circulation 1986;74:350-8. 2. Mitchell JH. Exercise training in the treatment of coronary heart disease. Adv Intern Med 1975;20:249-72. 3. Clausen JP. Circulatory adjustment to dynamic exercise and effect of exercise training in normal subjects and patients with ischemic heart disease. Prog Cardiovasc Dis 1976;18:459-95. 4. Oberman A, Fletcher GF, Lee J, Nanda N, Fletcher BJ, Jensen N, et al. Efficacy of high-intensity exercise training on left ventricular ejection fraction in men with coronary artery disease. Am J Cardiol 1995;76:643-7. 5. Ehsani AA, Martin WH III, Heath GW, Coyle EF. Cardiac effects of prolonged and intense exercise training in patients with coronary artery disease. Am J Cardiol 1982;50:246-54. 6. Hagberg JM, Ehsani AA, Holloszy JO. Effect of 12 months of intense exercise training on stroke volume in patients with coronary artery disease. Circulation 1983;67:1194-9. 7. Sunagawa K, Maughan WL, Burkoff D, Sagawa K. Left ventricular interaction with arterial load studied in isolated canine ventricle. Am J Physiol 1983;245:H773-80. 8. Biello DR, Sampathkumaran KS, Geltman ED, Briston WA, Scott DH, Grbac RT. Determination of left ventricular ejection fraction: a new method that requires minimal operator training. J Nucl Med Techn 1981;9:77-85. 9. Seldin DW, Esser PD, Nichols AB, Ratner SJ, Anderson PO. Left ventricular volume determined from scintigraphy and digital angiography by a semiautomated geometric method. Radiology 1982;14:800-13. 10. Kelly RP, Ting CT, Yang TM, Liu CP, Maughan WL, Chang MS, et al. Effective arterial elastance as index of arterial vascular load in humans. Circulation 1992;86:513-21. 11. Williams RS, McKinnis RA, Cobb FR, Higginbotham MB, Wallace AG, Coleman RE, et al. Effects of physical conditioning on left ventricular ejection fraction in patients with coronary artery disease. Circulation 1984;70:69-75. 12. Froelicher V, Jensen D, Genter F, Sullivan M, Monckirnan MD, Witzum K, et al. A randomized trial of exercise training in patients with coronary heart disease. JAMA 1984;252:1291-7. 13. Cobb FR, Williams RS, McEwan P, Jones RH, Coleman RE, Wallace AG. Effects of exercise training on ventricular function in patients with recent myocardial infarction. Circulation 1982;66:100-8. 14. Schuler G, Hambrecht R. Regression and non-progression of coronary artery disease with exercise. J Cardiovasc Risk 1996;3:176-82. 15. Niebauer J, Hambrecht R, Schlierf G, Marburger C, Kalberger B, Kubler W, et al. Five years of physical exercise and low fat diet: effects on progression of coronary artery disease. J Cardiovasc Rehab 1995;15:47-64.
Background Endurance exercise training can increase left ventricular (LV) ejection fraction during dynamic exercise in coronary artery disease. This adaptation may be mediated by altered cardiac loading conditions rather than an improvement in intrinsic LV systolic function. To minimize these confounding effects, we used isometric handgrip exercise to assess the training-induced changes in LV systolic function and ventriculoarterial coupling. Methods Twenty-six patients (52 ± 2 years of age) trained for 12 months. LV function was assessed with radionuclide ventriculograpy.
Results LV systolic reserve (the change in LV ejection fraction from rest to handgrip exercise) increased from –7.32 ± 1.2 to –3.4 ± 1.1 (P = .033) without acute changes in end-diastolic volume or the effective arterial load. LV end-systolic elastance increased 37% (P = .039) during handgrip exercise. Resting end-diastolic volume increased and the effective arterial load decreased after training. Conclusions Data suggest that in coronary artery disease adaptations to exercise training include a lower effective arterial load and an increase in EDV at rest, with an improvement in LV systolic function detectable only during afterload stress. (Am Heart J 1999;138:169-74.)
Endurance exercise training can improve endurance, exercise capacity, and the minimal work rate required to induce myocardial ischemia in patients with coronary artery disease (CAD).1-4 Adaptations in skeletal muscle and the autonomic nervous system were generally thought to be responsible for these observations.2,3 However, previous studies that involved a progressively increased exercise intensity demonstrated that in addition to the peripheral adaptations, an improvement in left ventricular (LV) ejection fraction (EF) and stroke volume is also attainable by endurance exercise training in patients with CAD.1,4,6 EF is affected not only by changes in contractile state, but also by alterations in cardiac loading conditions. Therefore the possibility of a larger preload or a lower LV wall stress contributing to a higher EF From the Division of Gerontology and Geriatrics, and the Cardiovascular Division, Department of Medicine, and the Division of Nuclear Medicine, Mallinckrodt Institute of Radiology, Washington University School of Medicine. Supported by NHLBI Grant HL17646 (SCOR in Ischemic Heart Disease), Claude D. Pepper Older American Independence Center AG-13629, and AG-12822. Morton R. Rinder, M.D. was supported by NIH Institutional Training Grant 5 T32 HL07081-22. Submitted January 26, 1998; accepted December 17, 1998. Reprint requests: Ali A. Ehsani, MD, Washington University School of Medicine, 660 S Euclid Ave, Campus Box 8113, St Louis, MO 63110. E-mail: [email protected] Copyright © 1999 by Mosby, Inc. 0002-8703/99/$8.00 + 0 4/1/96758
reported in these studies1,4 cannot be excluded despite a higher systolic blood pressure during peak exercise in the trained state.1 Virtually all previous studies evaluating the cardiovascular adaptations to endurance exercise training have used trained muscle groups during the exercise stimulus that can induce a greater vasodilatation in the trained state and, in turn, a lower systolic load accounting for an improved EF during dynamic exercise.1-4,6 Therefore it may be argued that the changes in ventricular loading conditions caused by adaptations in the skeletal muscle and autonomic nervous system were, at least in part, responsible for the observed improvement in EF after training in previous studies.1,4 The effect of endurance exercise training on the interaction between the ventricular and arterial system, that is, “ventriculoarterial coupling,”7 has not been previously reported in patients with CAD. The adaptive changes in the ventriculoarterial coupling in response to endurance exercise training should provide useful information regarding the influence of the arterial system (effective arterial elastance) on systolic function (LV elastance).7 Under physiologic conditions at a given preload, the interaction between these 2 components provides an optimal level of stroke volume or stroke work to meet the metabolic demands of the body and to maintain hemodynamic homeostasis. Therefore the purpose of this study was to assess the adaptive changes in myocar-
American Heart Journal July 1999
170 Rinder, Miller, and Ehsani
Table I. Adaptations in hemodynamics in response to exercise training in CAD Before Training
SBP (mm Hg) DBP (mm Hg) Heart Rate (beats/min) EF (%) LVEDV (mL) LVESV (mL) SV (mL)
After Training
Rest
Handgrip
120.3 ± 3.0 78.6 ± 2.0 66.0 ± 2.3 55.3 ± 2.9 163.3 ± 12.1 72.7 ± 13.8 90.5 ± 3.3
158.5 ± 5.5 107.8 ± 2.7 75.5 ± 2.3 47.9 ± 2.8 184.8 ± 9.8 89.5 ± 13.5 95.2 ± 5.9
Rest 123.6 ± 3.2 77.0 ± 1.7 61.1 ± 2.1* 57.1 ± 3.0 176.0 ± 11.8† 68.7 ± 11.4 107.3 ± 6.8§
Handgrip 163.5 ± 3.7 109.0 ± 2.5 70.9 ± 1.9ll 53.8 ± 3.1* 178.6 ± 12.4 76.2 ± 13.8‡ 102.0 ± 5.1
DBP, Diastolic blood pressure; SBP, systolic blood pressure; SV, stroke volume. Before vs after for the same conditions: *P < .001. †P = .039. ‡P = .054. §P = .011. llP = .017.
dial functional reserve and ventriculoarterial coupling in response to endurance exercise training in patients with CAD during isometric handgrip exercise. We used this type of stress in our study because we believe it should provide a suitable afterload stimulus that is unlikely to be affected by peripheral adaptations and because our training program included only endurance exercise and did not involve upper extremity exercise.
Methods Patients We studied 26 patients (24 men and 2 women, 52 ± 2 years of age) with CAD. All provided written consent, and the study protocol was approved by the Human Studies Committee of Washington University School of Medicine. This study is an extension of a previous study and includes 18 patients whose radionuclide data during dynamic exercise have been previously reported.1 We recruited 8 additional patients to have a sufficient sample size needed for detection of the training-induced changes in LV systolic performance during an afterload stress. Nineteen patients had a myocardial infarction. The interval between myocardial infarction and enrollment in the study was at least 3 months and averaged 14.5 ± 4.8 months. Three patients had coronary artery bypass graft surgery at least 3 months before their participation in the study. Ten patients had chronic stable effort angina, and 18 had exercise-induced myocardial ischemia detected by electrocardiographic or scintigraphic criteria during dynamic exercise. Of the 15 patients in whom coronary angiographic data were available, 3 had significant (≥70% luminal stenosis) single-vessel disease, 8 had 2-vessel disease, and 4 had 3-vessel disease. None of the patients had atrial fibrillation or valvular heart disease, and none underwent revascularization procedures during the study.
Cardiac medications All medications were prescribed by the patients’ personal physicians. Sixteen patients were taking nonselective β-blockers, 9 were taking long-acting nitrates, and 4 were taking cal-
cium-channel blockers. There were no changes in medications during the training program except in one patient who was treated with propranolol after enrollment in the study. In this patient the β-blockade was stopped 48 hours before the final radionuclide study. The interval between the last dose of cardiac medications and the radionuclide ventriculography averaged 18.2 ± 3.1 hours before and 19.0 ± 3.2 hours after training (P = .50).
Treadmill exercise test and maximal oxygen uptake . capacity (VO2 per max) Maximal symptom-limited exercise testing was performed on a treadmill according to the Bruce protocol with a repeat treadmill test . 2 weeks later for measurement of the highest attainable VO. 2 with a modified protocol as previously described.1 V O2max was defined as the mean of the 2 highest consecutive 30-second oxygen measurements .that met the following criteria: (1) attainment of a plateau of VO2 with increasing exercise intensity and/or (2) a respiratory exchange ratio exceeding 1.15. It was possible to obtain . VO2max .in most patients without angina. In the. others, the highest VO2 recorded was designated as peak V O2.
LV function at rest and during handgrip isometric exercise LVEF and end-diastolic volume (LVEDV) were determined with the use of the electrocardiographically gated cardiac blood pool imaging technique as previously described.1,8,9 After completion of resting images, the patients performed isometric exercise at 30% of the previously determined maximal voluntary contraction for 3 minutes. Image acquisition was repeated between minutes 2 and 3 of isometric exercise with measurement of cuff blood pressure in the contralateral arm. The same protocol was repeated after 12 months of training. Images were analyzed in a blinded fashion. LV volumes data were available in 12 of the patients who underwent imaging before and after exercise training. The LV end-systolic volume (LVESV) and stroke volume were derived from the LVEDV and EF.1 We estimated the LV end-systolic pressure
American Heart Journal Volume 138, Number 1, Part 1
Rinder, Miller, and Ehsani 171
Table II. Effects of exercise training on ventriculoarterial coupling in CAD Before Training Rest Ees (mm Hg/mL) Ea (mm Hg/mL) Ea/Ees
2.03 ± 0.34 1.21 ± 0.05 0.85 ± 0.19
After Training
Handgrip 1.90 ± 0.27 1.54 ± 0.12 1.07 ± 0.24
Rest 2.03 ± 0.31 1.05 ± 0.09† 0.70 ± 0.15
Handgrip 2.60 ± 0.42* 1.47 ± 0.09 0.82 ± 0.20‡
Before vs after for the same condition: *P = .039. †P = .009. ‡P = .013.
(ESP) by the following formula: ESP = [(2 × SBP) + DBP]/3 as reported by Kelly et al.10 LV contractile function was evaluated by LV elastance, the ratio of ESP/LVESV designated as Ees, that represents the slope of the LV pressure/volume curve at end ejection and provides a good estimate of LV contractility. The effective arterial load (Ea) was estimated as ESP/stroke volume (arterial elastance). Ventriculoarterial coupling was expressed as the ratio of Ea/Ees.
Exercise program The duration of the supervised endurance exercise training program was 12 months, as previously described. 1 Patients were encouraged to exercise for 60 minutes 3 days a week for the first 3 months and 5 times per week thereafter. The training intensity was 60% to 70% of the maximal heart rate for the first 3 months and was progressively increased to 80% to 90% of maximal heart rate in the last 3 months of training.
Statistical analysis Student t test for paired observations was used for comparison of the data before and after training. Chi-square and leastsquared linear regression analyses were performed when appropriate. General linear model was used to handle the missing data points. Data are expressed as mean ± SE.
Results Improvement in functional capacity and clinical status The patients exercised 4.1 ± 0.13 days per week for the last 9 months of the training program. The intensity of training averaged 85.3% ± 1.5% of the maximal attainable heart rate in the last 6 months of the training program. Training induced a weight loss from 80.7 ± 2.3 kg to 76.9 ± 2.3 kg (P = .003). Eight of the 10 patients with chronic effort angina became asymptomatic after completion of the exercise program (P = .021). In the remaining 2 patients, the frequency and severity of angina was considerably less in the trained state. One patient was hospitalized because of chest pain after a maximal exercise test but myocardial infarction was ruled out. There were no other major complications or sequelae attributable to exercise.
. Heart rate, blood pressure, and peak VO2 The systolic and diastolic blood pressures at rest and during handgrip exercise did not change with training (Table I). Heart rate increased in response to isometric handgrip exercise both before and after training (Table I). However, resting and isometric exercise heart rates were lower in the trained state. The lower exercise heart rate was almost entirely from the training-induced resting bradycardia. The rate-pressure product during submaximal handgrip isometric exercise (30% of maximal voluntary contraction) was 12.19 × 10 3 ± 0.68 × 103 before and 11.64 × 103 ± 0.47 × 10 3 after training (P = .25). . Peak VO2 in absolute terms (1.88 ± 0.06 L/min vs 2.42 ± 0.09 L/min, P < .001) or when normalized for body weight (23.8 ± 0.8 vs 32.2 ± 1.4 mL/kg/min, P < .001) increased significantly in response to endurance exercise training. The maximal dynamic exercise (treadmill) heart rate (144.4 ± 3.7 beats/min vs 150.7 ± 3.4 beats/min, P = .027), systolic blood pressure (154.3 ± 6.9 mm Hg vs 166.8 ± 6.3 mm Hg, P = .016), and the rate-pressure product (22.19 × 10 3 ± 1.27 × 103 vs 25.21 × 103 ± 1.21 × 10 3, P = .002) were higher after training.
LV size, systolic function, and ventriculoarterial coupling Resting LVEDV was significantly larger after endurance exercise training (Table I). However, LVEDV during isometric handgrip exercise was similar before and after training. LVEDV increased acutely in response to isometric exercise before but not after training (∆LVEDV 21.5 ± 3.2 mL before and 2.6 ± 6.3 mL after, P = .044). LVESV at rest did not change but tended to be smaller during isometric handgrip exercise after training (P = .054, Table I). Stroke volume increased significantly (19%, P = .011) at rest but not during isometric handgrip exercise in response to training. EF at rest did not change but was higher during isometric exercise after training (Table I). There was a significant improvement of LV systolic reserve capacity, defined as the difference between resting EF and EF during isometric exercise (∆EF), in response to training
American Heart Journal July 1999
172 Rinder, Miller, and Ehsani
Figure 1
Figure 3
Effect of endurance exercise training on left ventricular systolic function. Decline in LVEF in response to an acute afterload stress induced by isometric handgrip exercise was considerably less (P = .033) after training, suggestive of an improvement in LV systolic function.
Figure 2
The adaptive changes in LV Ees in response to exercise training in CAD. In contrast to acute reduction in Ees in response to isometric exercise before training, patients exhibited an increase in Ees during isometric exercise in trained state (P = .009), suggestive of a significant improvement in LV systolic performance.
Inverse relation between adaptive changes in ventriculoarterial coupling, designated as Ea/Ees ratio, and stroke volume during isometric handgrip exercise. Patients who had better systolic function during acute increase in effective arterial load in trained state, as reflected in lower Ea/Ees ratio, were likely to have higher stroke volume.
trained state (Table II). The Ea/Ees ratio, used as a marker of the ventriculoarterial coupling, was significantly smaller during isometric exercise after training (Table II). During isometric exercise, there was an inverse relation between the training-induced changes in the ventriculoarterial coupling (Ea/Ees) and stroke volume; the patients who maintained a better systolic performance in response to an arterial load challenge, as reflected in the lower Ea/Ees ratio, exhibited a higher stroke volume in the trained state (Figure 3). The training-induced increase in stroke volume at rest correlated directly with changes in end-diastolic volume (Figure 4, A) and inversely with alterations in the effective arterial load (Figure 4, B).
Effects of training on myocardial ischemia characterized by a smaller decrease in EF during isometric exercise after training (–7.5 ± 1.2 vs –3.4 ± 1.06; P = .033) (Figure 1). The LV Ees under basal conditions did not change but was significantly higher (37%) during isometric handgrip exercise after training (P = .039, Table II). The LV contractile reserve, as reflected by the change Ees from rest to isometric exercise, improved with training (∆Ees: –0.125 ± 0.1 mm Hg/mL before and 0.563 ± 0.15 mm Hg/mL after, P = .009, Figure 2). In contrast to Ees, Ea was only significantly lower at rest and did not change during isometric exercise in the
Electrocardiographic changes. In the 11 patients who had ischemic ST-segment responses (ie, >0.1 mV horizontal or downsloping ST-segment depression) during dynamic exercise, the magnitude of ST-segment depression decreased from 0.15 ± 0.02 mV to 0.075 ± 0.02 mV (P = .002) at a given absolute exercise intensity after endurance exercise training. Rate-pressure product tended to be lower at these work rates (18.5 × 103 ± 1.7 × 103 vs 15.1 × 103 ± 1.03 × 103; n = 11; P = .064). ST-segment depression at maximal exercise did not change (0.14 ± 0.02 mV vs 0.159 ± 0.03 mV, P = .6) Regional LV wall motion disorders. Before training,
American Heart Journal Volume 138, Number 1, Part 1
Rinder, Miller, and Ehsani 173
Figure 4
A, Direct relation between training-induced changes in LVEDV and stroke volume at rest. Patients with greater LVEDV tended to have larger stroke volume. B, Inverse relation between training-induced changes in effective arterial load and stroke volume at rest. Patients with lower arterial load tended to have higher stroke volume.
9 of the 26 patients exhibited regional LV wall motion abnormalities during handgrip isometric exercise. After training, however, only 2 patients showed significant regional wall motion disorders during handgrip exercise (P = .042).
Discussion Our results suggest that long-term endurance exercise training can improve LV systolic function in patients with CAD, which appears to be independent of the vascular and autonomic adaptations that occur with this type of intervention. We used isometric handgrip exercise involving an untrained muscle group to induce an afterload stress that was not affected by the peripheral adaptations, as evidenced by the absence of differences in the arterial elastance and blood pressure responses to isometric exercise between the trained and untrained states. Significant adaptations induced by endurance exercise training in our patients were reflected in a substantial . increase (29%) in the highest attainable VO2 after training. With no apparent significant changes in afterload stress, the reduction in EF observed during isometric handgrip exercise was markedly less pronounced after training, suggesting that either contractile reserve was improved or there was a greater preload reserve preventing the decline in EF in the trained state. However, because the values for LVEDV during isometric exercise were not different before and after training, and the
change in LVEDV from rest to exercise was negligible in the trained state, it is unlikely that the improvement in EF response was mediated primarily by preload. Ees, which reflects the slope of the end-systolic pressure/end-systolic volume relation and is considered a reliable measure of LV contractility,7 significantly improved during isometric exercise in response to training, providing additional evidence that the increase in EF reserve was, at least in part, caused by improvement in LV contractile function. The mechanisms underlying these adaptive myocardial changes are unclear. However, it is likely that this improvement was, in part, the result of amelioration of myocardial ischemia, as reflected in decreases in regional LV wall motion abnormalities and effort angina in the trained state. In the basal resting state, the adaptations to training were limited to a larger LVEDV and a lower effective arterial load that contributed to the training-induced larger resting stroke volume. Although a slower heart rate can be responsible for a larger LVEDV, the magnitude of the reduction in heart rate in our patients was too small to account entirely for the higher LVEDV in the trained state. Some of the previous studies demonstrated no significant improvement in myocardial function with exercise training in patients with CAD.11-13 However, these studies did not use a sufficient exercise stimulus with respect to intensity and duration. It appears that a longer duration of training is needed so that the
American Heart Journal July 1999
174 Rinder, Miller, and Ehsani
patients can exercise safely at a higher intensity to improve LV systolic function.1 In virtually all the studies that showed improved LV function during exercise, the trained muscle group was used to evaluate adaptive changes in LV systolic function in response to stress. Therefore the role for afterload reduction and systolic unloading that can contribute to the observed increased EF could not be excluded.1,4,6 Assessment of LV function during isometric exercise that does not include a trained muscle group should mitigate this shortcoming. Therefore improved LV systolic function that we observed in this study is unlikely caused by decreased afterload. There are potential limitations to our study. (1) Our sample size was small (12 patients had volumetric data) and the study was performed mostly in men (24 of 26). Therefore the results cannot necessarily be extrapolated to women. (2) There was no control group. Thus spontaneous improvements in LV function cannot be excluded. However, this is unlikely because the minimum interval between myocardial infarction or coronary artery bypass grafting and enrollment in the study was 3 months. (3) It is possible that an improvement in myocardial ischemia contributing to amelioration of LV systolic dysfunction during isometric exercise in the trained state is mediated not only by an enhanced myocardial oxygen supply but also by a lower myocardial oxygen demand because of a slower heart rate even though other major variables affecting myocardial oxygen demand—that is, systolic pressure, rate-pressure product, LV systolic function, and LVEDV—were either similar or higher during isometric exercise after training. (4) Our patients, with few exceptions, had normal baseline LV systolic function and were only mildly symptomatic. Therefore our conclusions may not be extrapolated to patients with LV systolic dysfunction or those with extensive coronary atherosclerosis and frequent and severe angina. The improvement in LV systolic function in our patients is likely to be a consequence of a reduction in myocardial ischemia mediated by increased myocardial oxygen supply, lower maximal oxygen demand, or both. It is plausible that enhancement of LV systolic function in response to an afterload stress in our subjects is in part the result of an increase in myocardial oxygen supply, as shown by other investigators.14,15 The results of this study suggest that long-term endurance exercise training induces significant cardiovascular adaptations both in the basal state and during an afterload stress in patients with CAD. In the basal resting state, these adaptations appear to involve exclusively the arterial system and LV size, manifested by a lower effective arterial load and a larger LVEDV, both
contributing to a larger stroke volume at rest. In contrast, during an afterload stress our data suggest that the adaptations to training are confined to an improvement in LV systolic performance without any detectable vascular adaptations or changes in preload. Although the mechanisms underlying these adaptations are unknown, an improvement in myocardial ischemia in response to training is likely to play a significant role.
References 1. Ehsani AA, Biello DR, Schultz J, Sobel BE, Holloszy JO. Improvement of left ventricular contractile function by exercise training in patients with coronary artery disease. Circulation 1986;74:350-8. 2. Mitchell JH. Exercise training in the treatment of coronary heart disease. Adv Intern Med 1975;20:249-72. 3. Clausen JP. Circulatory adjustment to dynamic exercise and effect of exercise training in normal subjects and patients with ischemic heart disease. Prog Cardiovasc Dis 1976;18:459-95. 4. Oberman A, Fletcher GF, Lee J, Nanda N, Fletcher BJ, Jensen N, et al. Efficacy of high-intensity exercise training on left ventricular ejection fraction in men with coronary artery disease. Am J Cardiol 1995;76:643-7. 5. Ehsani AA, Martin WH III, Heath GW, Coyle EF. Cardiac effects of prolonged and intense exercise training in patients with coronary artery disease. Am J Cardiol 1982;50:246-54. 6. Hagberg JM, Ehsani AA, Holloszy JO. Effect of 12 months of intense exercise training on stroke volume in patients with coronary artery disease. Circulation 1983;67:1194-9. 7. Sunagawa K, Maughan WL, Burkoff D, Sagawa K. Left ventricular interaction with arterial load studied in isolated canine ventricle. Am J Physiol 1983;245:H773-80. 8. Biello DR, Sampathkumaran KS, Geltman ED, Briston WA, Scott DH, Grbac RT. Determination of left ventricular ejection fraction: a new method that requires minimal operator training. J Nucl Med Techn 1981;9:77-85. 9. Seldin DW, Esser PD, Nichols AB, Ratner SJ, Anderson PO. Left ventricular volume determined from scintigraphy and digital angiography by a semiautomated geometric method. Radiology 1982;14:800-13. 10. Kelly RP, Ting CT, Yang TM, Liu CP, Maughan WL, Chang MS, et al. Effective arterial elastance as index of arterial vascular load in humans. Circulation 1992;86:513-21. 11. Williams RS, McKinnis RA, Cobb FR, Higginbotham MB, Wallace AG, Coleman RE, et al. Effects of physical conditioning on left ventricular ejection fraction in patients with coronary artery disease. Circulation 1984;70:69-75. 12. Froelicher V, Jensen D, Genter F, Sullivan M, Monckirnan MD, Witzum K, et al. A randomized trial of exercise training in patients with coronary heart disease. JAMA 1984;252:1291-7. 13. Cobb FR, Williams RS, McEwan P, Jones RH, Coleman RE, Wallace AG. Effects of exercise training on ventricular function in patients with recent myocardial infarction. Circulation 1982;66:100-8. 14. Schuler G, Hambrecht R. Regression and non-progression of coronary artery disease with exercise. J Cardiovasc Risk 1996;3:176-82. 15. Niebauer J, Hambrecht R, Schlierf G, Marburger C, Kalberger B, Kubler W, et al. Five years of physical exercise and low fat diet: effects on progression of coronary artery disease. J Cardiovasc Rehab 1995;15:47-64.