Exercise Capacity in Hypertrophic Cardiomyopathy
Role of Stroke Volume Limitation, Heart Rate, and Diastolic Filling Characteristics
Background We previously showed that exercise capacity in patients with hypertrophic cardiomyopathy (HCM) is related to peak exercise cardiac output. Cardiac output augmentation during exercise is normally dependent on heart rate (HR) response and stroke volume (SV) augmentation by increased left ventricular end-diastolic volume and/or increased contractility. We hypothesized that in contrast to normal subjects, peak exercise capacity in patients with HCM is determined by the diastolic filling characteristics of the left ventricle during exercise, which would in turn determine the degree to which SV is augmented, and that HR is a relatively unimportant determinant of peak exercise capacity.
Methods and Results Twenty-three patients with HCM underwent invasive hemodynamic evaluation and measurement of maximal oxygen consumption (V̇o2max) during erect treadmill exercise to assess the relative importance of changes in HR and SV in determining exercise capacity. Hemodynamic responses to erect and supine exercise were compared in 10 of these patients. In a separate group of 46 patients with HCM, the relation between V̇o2max and exercise diastolic filling indexes was assessed. Peak HR during erect exercise was 92±8% of predicted maximum. V̇o2max was 29.0±6.4 mL · kg−1 · min−1 and was related significantly to peak exercise cardiac index and SV index (r=.71, P<.0001 and r=.66, P=.001, respectively) but not to peak HR, HR deficit, or resting or peak pulmonary capillary wedge pressure. Peak cardiac output during erect exercise was not related to peak HR (r=.13, P=NS). When erect and supine exercise were compared, peak HR was lower in the supine position (153.3±19.9 beats per minute supine versus 172.0±17.6 beats per minute erect, P=.003), but peak exercise cardiac index was similar (7.9±2.6 L · min−1 · m−2 supine versus 7.5±2.8 L · min−1 · m−2 erect). Pulmonary capillary wedge pressure was higher at rest in the supine versus erect position (15.3±5.2 versus 8.1±6.1 mm Hg) but was not significantly higher at peak exercise in the supine versus erect position (28.5±8 versus 22.4±11.6 mm Hg erect, P=NS). In the separate group of 46 patients with HCM, V̇o2max was significantly inversely related to time to peak filling at peak exercise (r=−.60, P<.0001) but did not correlate with time to peak filling at rest, resting ejection fraction, peak filling rate, or peak exercise peak filling rate.
Conclusions SV is the major determinant of peak exercise capacity in the erect position in patients with hypertrophic cardiomyopathy. This in turn is determined by the exercise left ventricular diastolic filling characteristics. HR augmentation does not appear to be a major determinant of peak cardiac output in the erect position.
A limited exercise capacity is common in patients with hypertrophic cardiomyopathy. We previously demonstrated that erect exercise capacity is strongly related to the increase in cardiac index but not to resting cardiac index or to resting or peak exercise pulmonary capillary wedge pressure (PCWP).1 There was also an inadequate heart rate response in some of those patients. This raised the question of the relative importance of heart rate versus stroke volume limitation as determinants of peak exercise capacity in patients with hypertrophic cardiomyopathy. In normal subjects, peak exercise heart rate is the principal determinant of maximal oxygen consumption, although stroke volume is also important.2 In patients with heart failure, heart rate also appears to be an important determinant of exercise capacity, since chronotropic incompetence is associated with markedly reduced exercise capacity.3 We hypothesized that in hypertrophic cardiomyopathy, because of the associated diastolic dysfunction, the ability to augment stroke volume may be the principal determinant of peak exercise capacity.
We therefore compared changes in stroke volume, heart rate, and PCWP during supine and erect exercise in patients with hypertrophic cardiomyopathy, focusing on the importance of limitation in stroke volume and heart rate in determining cardiac output and exercise capacity. We also assessed the relation between exercise capacity and resting and exercise indexes of diastolic filling in patients with hypertrophic cardiomyopathy.
Two separate protocols were performed. The first was an invasive hemodynamic study and the second a study of diastolic filling during exercise. The protocols were approved by the ethics committees of the Royal Brisbane Hospital and the St George’s Hospital London, and written informed consent was given by all subjects.
Subjects for the invasive hemodynamic study were selected from 35 consecutive patients with hypertrophic cardiomyopathy according to standard diagnostic criteria.4 Subjects for the exercise diastolic filling study were selected subsequently from 66 consecutive patients with hypertrophic cardiomyopathy. Patients were considered eligible for study if the underlying rhythm was sinus and if they were able to exercise on a treadmill and a supine exercise cycle and were limited only by breathlessness or fatigue. Exclusion criteria included significant respiratory disease, severe mitral regurgitation (grade 3 to 4), inadvisability of withdrawal of cardioactive medications, and inability to acquire informed consent. Twenty-three of the first cohort of 35 patients and 56 of the second cohort of 66 patients satisfied the above criteria and were enrolled in the two protocols. Twenty age- and sex-matched subjects without evidence of cardiac disease formed a control group for the exercise diastolic filling studies.
Eighteen of the 23 patients in the invasive hemodynamic study were newly diagnosed and had not previously taken cardioactive medications. In 4 patients, calcium antagonists were withdrawn at least 3 days before the study. One patient had been taking propranolol, which was withdrawn 5 days before the study. Thirty of the 56 patients in the diastolic filling study were newly diagnosed and had not previously taken cardioactive medications. In the remaining 26, all cardioactive medications were withdrawn at least 5 days before the study. Before the study, the patients underwent practice sessions to accustom them to the techniques of exercise and respiratory gas analysis and had demonstrated <10% difference in maximal oxygen consumption (V̇o2max) on at least two consecutive tests.
Invasive hemodynamic studies. We assessed changes in hemodynamics during both erect and supine exercise. In this group of patients with hypothesized diastolic dysfunction, differences in preload, heart rate, and blood pressure would be expected between these two forms of exercise. The first 10 patients arrived fasting at 8 am. A 20-gauge cannula was inserted under local anesthesia in the brachial artery of the nondominant forearm. A 7F Swan-Ganz flow-directed catheter was advanced to the left or right pulmonary artery via the subclavian vein under local anesthesia. After resting for 1 hour, the subjects underwent symptom-limited supine cycle exercise using a Colin’s Cycle Ergometer starting at 25 W and increasing by 12.5 W every 3 minutes. Systolic blood pressure, PCWP, and cardiac output (thermodilution method) were measured at rest, after each 2 minutes of exercise, and at peak exercise. Pressures were measured with Gould-Statham transducers referenced to atmosphere at midchest level and recorded with a multichannel recorder (Mingograph 7, Siemens-Elema). The patients were given a light lunch and then rested for a further 3 hours. In the afternoon, the subjects underwent symptom-limited erect treadmill exercise using the Bruce protocol5 with simultaneous respiratory gas analysis using an Airspec 200 MGA mass spectrometer linked via an analog-to-digital converter to a BBC microcomputer by an established technique.6 Sampling of mixed expired gases was performed every second, and data were expressed as 10-second means. A printout of minute ventilation, oxygen consumption, carbon dioxide production, and respiratory quotient was obtained. Maximal oxygen consumption was defined as the mean of the highest two values of oxygen consumption obtained during exercise. Blood pressure was monitored continuously, and PCWP and cardiac output (direct Fick method) were measured every 3 minutes and at peak exercise.
The remaining 13 patients underwent insertion of brachial arterial and pulmonary arterial lines at 11 am, were given a light lunch, and were studied 3 hours later. They underwent only erect symptom-limited treadmill exercise, and pressures and cardiac output were measured as described above.
Heart rate deficit during erect exercise was calculated by the equations7 maximum predicted heart rate=220−age (y); heart rate deficit=maximum predicted heart rate−that achieved.
Exercise diastolic filling studies. All patients and control subjects were studied in the fasting state. Respiratory gas analysis was performed in the patients during erect treadmill exercise as described above.
Equilibrium radionuclide ventriculography. Left ventricular ejection fraction and diastolic filling were assessed by equilibrium R-wave gated blood pool scintigraphy using a standard technique at rest and during graded semierect exercise on a cycle ergometer. Ten minutes after the intravenous injection of ≈1.7 mg stannous pyrophosphate, 5 mL of blood was drawn into a heparinized syringe and incubated for 20 minutes with 925 MBq (25 mCi) of 99mTc pertechnetate before reinjection. Studies were acquired on a small-field-of-view gamma camera (GE300A, GE Medical Systems) fitted with a low-energy, general-purpose, parallel-hole collimator and interfaced to a dedicated minicomputer (MaxDelta, Siemens). With the patient on the cycle ergometer, the detector was adjusted for the left anterior oblique view with the best ventricular separation and 10° to 15° of caudal tilt. A 15% tolerance window was set about the patient’s heart rate, and each RR interval was divided into 28 equal frames throughout. A constant number of frames per RR interval ensured constant temporal resolution during diastole at all heart rates. Data from each beat were acquired into a memory buffer in a 64×64 “word” matrix and if accepted, were reformatted with two-thirds forward, one-third backward gating. Four minutes of data was acquired at rest and at each level of exercise after a 30-second period for stabilization of heart rate at the commencement of each stage. The initial workload was 25 W and increased by 12.5-W increments. Exercise was terminated because of patient fatigue, breathlessness, arrhythmia, heart rate >200 beats per minute, or hypotension (systolic blood pressure less than baseline). No patient developed angina during the exercise protocol. The rest and exercise gated blood pool scintigraphs were analyzed by a single operator who was blinded to the patient’s history and exercise performance. The composite cycle derived from each stage was spatially and temporally filtered. Left ventricular counts in each frame were determined by a semiautomated edge-detection algorithm. Left ventricular ejection fraction (percent) was calculated from the background-corrected left ventricular activity-time curve. Stroke counts were calculated as the product of background-corrected end-diastolic counts (corrected for number of cycles accepted and cycle duration) and left ventricular ejection fraction. Peak left ventricular filling rate in terms of end-diastolic volumes per second (EDV/s) and time to peak filling in milliseconds after end systole were calculated from the second derivative of the diastolic activity-time curve. The validity of these radionuclide measures of diastolic filling at high heart rates has been established previously.10 11
Data are expressed as mean±SD. Statistical analysis was performed by paired and unpaired t tests and by linear regression where appropriate. For single comparisons, a value of P<.05 was considered significant. For multiple t tests, the Bonferroni correction was performed to overcome type 1 errors by comparing the probability values, with .05 divided by the number of relevant tests being performed.
The clinical, echocardiographic, radionuclide, and ambulatory ECG features of the study groups are summarized in Table 1⇓.
Erect Treadmill Exercise (23 Patients)
Exercise was terminated because of breathlessness in all cases, and all patients reached anaerobic threshold.11 As shown in Table 2⇓, V̇o2max was 29.0±6.4 mL · kg−1 · min−1 (47% to 120%, with a mean of 76.7±16.1% of predicted) and was less than predicted in all but 1 patient. Anaerobic threshold was 49% to 95%, with a mean of 72% of V̇o2max. Heart rate increased from 88±20 beats per minute at rest to 174±9 beats per minute at peak exercise. The heart rate deficit was 15.2±14.8 beats per minute; maximum heart rate was 91±8% of predicted. Cardiac index increased from 2.4±0.5 L · min−1 · m−2 at rest to 8.1±2.0 L · min−1 ·m−2 at peak exercise, PCWP increased from 6±6 mm Hg at rest to 22±11 mm Hg at peak exercise, and stroke volume index increased from 28.7±7.9 mL/m2 at rest to 47.7±13.0 mL/m2 at peak exercise. Exercise-induced hypotension (defined as a fall in systolic blood pressure of at least 20 mm Hg from the peak value recorded to the value recorded immediately before exercise ceased) was observed in 10 of the 23 patients. Resting and peak heart rate, stroke volume index, and cardiac index were similar in patients with and without exercise hypotension.
Determinants of peak exercise capacity. Peak exercise cardiac index was not related to peak exercise heart rate (r=.19, P=NS) but was strongly related to peak exercise stroke volume index (r=.9, P<.0001). As shown in Table 3⇓ and in Figs 1⇓, 2⇓, and 3⇓, exercise capacity (V̇o2max) was related to peak cardiac index (r=.71, P<.0001) and to peak stroke volume index (r=.66, P=.001) but was not related to resting cardiac index, resting stroke volume index, resting or peak heart rate, heart rate deficit, or resting or peak PCWP. Patients with exercise capacity greater than the mean percentage of the age- and sex-predicted V̇o2max for the group (ie, greater than 76.7% of predicted) had higher peak exercise stroke volume index (54.9±9.4 versus 42.6±13.8 mL/m2, P=.02) and higher peak exercise cardiac index (9.3±1.7 versus 7.4±2.5 L · min−1 · m−2, P=.02) compared with patients whose exercise capacity was less than 76.7% of age- and sex-predicted V̇o2max. In patients whose heart rate deficit was ≥20 beats per minute (n=7) compared with those in whom heart rate deficit was <20 beats per minute (n=16), absolute V̇o2max and percentage of age- and sex-predicted V̇o2max were similar (29.2±6.4 versus 28.9±7.3 mL · kg−1 · min−1 and 75.8±25.7% versus 79.8±14.6%, respectively, P=NS). Exercise duration was not related to peak cardiac index (r=.3, P=NS), peak stroke volume index (r=.2, P=NS), or peak heart rate (r=.2, P=NS).
Supine Exercise (10 Patients)
Exercise was terminated because of fatigue in 8 patients and breathlessness in 2. The hemodynamic variables are shown in Table 4⇓. Exercise duration was not related to peak heart rate (r=.14, P=NS), resting or peak PCWP (r=.24, P=NS and r=.21, P=NS, respectively), or resting or peak cardiac index (r=.31, P=NS and r=.23, P=NS, respectively). Peak cardiac index was related to peak stroke volume index (r=.9, P<.001) but not to peak heart rate (r=.14, P=NS).
Comparison Between Erect and Supine Exercise
The hemodynamics at rest and at peak exercise in the supine and erect positions are summarized in Tables 5⇓ and 6⇓, respectively. In the 10 patients studied in both positions, resting heart rate was higher in erect than supine exercise (98.7±16.3 versus 75.8±14.8 beats per minute, P<.0001), and resting stroke volume index and resting PCWP were lower in the erect versus supine position (25.5±7.5 versus 38.4±11.0 mL/m2, P=.005 and 8.1±6.1 versus 15.3±5.2 mm Hg, P<.001, respectively). The peak heart rate was higher in the supine position (172.0±17.6 versus 153.3±19.9 beats per minute, P=.003), and peak exercise stroke volume index and peak exercise PCWP were nonsignificantly lower in the erect and supine positions (43.6±16.5 versus 52.5±17.5 mL/m2, P=.03, and 22.4±11.6 versus 28.5±8.0 mm Hg, P=.03, respectively). There was no significant difference in blood pressure, cardiac index, or systemic vascular resistance at rest or at peak exercise between the two modalities of exercise.
Exercise Diastolic Filling Studies
Erect treadmill exercise. In the 56 patients with hypertrophic cardiomyopathy studied, exercise was terminated because of breathlessness in all cases, and all patients reached anaerobic threshold.11 V̇o2max was 28.5±7.5 mL · kg−1 · min−1 (76.5±19.3% of age- and sex-predicted values). Heart rate increased from 69.8±13.7 beats per minute at rest to 162.7±26.3 beats per minute at peak exercise. Exercise-induced hypotension was observed in 17 patients. Resting left ventricular outflow tract gradients were present in a similar proportion of patients with and without exercise hypotension (4 of 17 patients with exercise hypotension versus 7 of 39 patients with normal blood pressure response, P=NS).
Exercise radionuclide ventriculography. All 56 patients and 20 control subjects performed symptom-limited semierect exercise. In 10 patients, cavity obliteration made it impossible to quantify diastolic filling at peak exercise, and these patients were therefore excluded from analysis. The results are summarized in Table 7⇓. Mean exercise duration was 7.8±2.4 minutes in patients and 16.0±4.0 minutes in control subjects. Heart rate was similar at rest in control subjects and patients but was higher at peak exercise in the control subjects (148.0±18.0 versus 114.2±17.6 beats per minute, P<.0001). Systolic blood pressure was similar at rest and at peak exercise in the control subjects and patients. Left ventricular ejection fraction was higher at rest in the patients than in the control subjects (67.2±9.5% versus 58±4.5%, respectively, P=.001) but was similar at peak exercise in both groups (69.8±12.8% versus 66.0±9.5%, respectively, P=NS). Peak filling rate was similar at rest (3.0±1.0 versus 3.1±0.7 EDV/s, P=NS) but was lower at peak exercise in patients compared with control subjects (6.0±1.7 versus 7.8±1.2 EDV/s, P=.0001). Time to peak filling was longer both at rest and at peak exercise in patients compared with control subjects (201.2±61.2 versus 122.0±68.0 ms at rest, P<.0001, and 127.0±43.2 versus 56.0±28.0 ms at peak exercise, P<.0001). In control subjects, there was a significant inverse relation between the change in time to peak filling during exercise and the change in heart rate (r=−.7, P=.001), whereas no such relation was present in patients (r=−.09, P=NS). In patients with and without exercise hypotension, resting and peak exercise ejection fraction and diastolic filling parameters were similar (resting ejection fraction, 66.7±8.8% versus 67.5±10.0%, P=NS; peak ejection fraction, 67.1±13.9% versus 69.4±12.1%, P=NS; resting peak filling rate, 3.1±1.2 versus 3.0±.9 EDV/s, P=NS; peak peak filling rate, 5.9±1.7 versus 6.1±1.7 EDV/s, P=NS; resting time to peak filling, 188.6±71.2 versus 208.6±54.5 ms, P=NS; and peak time to peak filling, 122.3±43.6 versus 129.7±43.5 ms, P=NS, respectively).
Relation of exercise capacity to resting and exercise indexes of systolic and diastolic function. As shown in Table 8⇓ and Fig 4⇓, in the 46 patients with hypertrophic cardiomyopathy in whom analysis was possible, V̇o2max was inversely related to peak exercise time to peak filling (r=−.60, P<.0001) but not to resting time to peak filling, resting or peak exercise ejection fraction, or peak filling rate. As in the invasive hemodynamic studies, V̇o2max was not related to peak exercise heart rate.
In this study, we were able to demonstrate several important characteristics of the hemodynamic responses to exercise in patients with hypertrophic cardiomyopathy. First, augmentation of stroke volume during exercise is an important determinant of peak exercise capacity, whereas peak heart rate response is relatively unimportant. Second, peak exercise capacity is strongly related to time to peak filling of the left ventricle. Third, important differences exist between exercise hemodynamics during erect versus supine exercise. We previously demonstrated that erect exercise capacity in hypertrophic cardiomyopathy is strongly associated with peak cardiac index but not with resting cardiac index or resting or peak PCWP.1 In the present study, we confirmed the above observations. This study also confirms that the increase in heart rate during exercise is less than predicted but that this limitation is mild.
Mechanism of Limitation of Peak Exercise Cardiac Output
In both normal subjects and patients with hypertrophic cardiomyopathy, an increase in heart rate during exercise clearly contributes to cardiac output augmentation and hence to exercise capacity. However, in hypertrophic cardiomyopathy, diastolic dysfunction has the potential to limit the increase in stroke volume with exercise. In this study, there was no demonstrable correlation between exercise capacity and peak heart rate. At high heart rates, the reduced time for diastolic filling of the left ventricle might be expected to reduce end-diastolic volume and, by the Frank-Starling mechanism, limit stroke volume. In this study, during erect exercise, both cardiac output and exercise capacity were unrelated to peak exercise heart rate, but both were strongly related to stroke volume. Thus, augmentation of stroke volume appears to be the principal determinant of peak exercise capacity. Although an increase in ejection fraction could contribute to stroke volume augmentation during exercise, the mean resting ejection fraction of 75.4±7.4% in our patients makes it unlikely that an increase in ejection fraction could have accounted for the observed 66% increase in stroke volume during exercise. Therefore, the ability to increase left ventricular end diastolic volume is likely to be a significant factor in cardiac output augmentation. The observation of an inverse relation between peak exercise time to peak filling and V̇o2max and the lack of relation between peak exercise ejection fraction and V̇o2max supports this concept. Although time to peak filling is inversely related to heart rate in control subjects, this inverse relation was absent in the patients studied, which implied that they had various degrees of exercise-induced diastolic filling abnormalities, limiting use of the Frank-Starling mechanism. We have shown a similar inverse relation between peak exercise time to peak filling and V̇o2max in patients with ischemic heart disease.12
Diastolic dysfunction in hypertrophic cardiomyopathy may reflect abnormal compliance as a result of myocyte hypertrophy and fibrosis, or it may be due to an abnormality of active relaxation, perhaps as a consequence of ischemia. However, two caveats should be noted: First, although these patients had some heart rate limitation, in no case was it profound, and it appears likely that in the presence of severe heart rate limitation, cardiac output would be limited to a greater degree. Second, a small subset of patients with hypertrophic cardiomyopathy exhibits a restrictive filling pattern. In this setting, early ventricular filling is rapid but then effectively ceases by mid diastole. In such patients, stroke volume is thus fixed and heart rate response is likely to be the major determinant of cardiac output. None of the patients in this study had evidence of restrictive filling patterns on Doppler echocardiography or resting gated heart pool scans.
All patients in our study were limited by breathlessness during erect exercise, and exercise capacity was less than age- and sex-predicted values in all but one. We discussed in an earlier report the lack of association with PCWP in such patients.1 We accept the limitation of using absolute PCWP measurements (transcapillary alveolar gradient being the crucial measure), but the correlation of exercise capacity with exercise PCWP was so poor that we believe pulmonary juxtacapillary receptor activation13 is unlikely to be the dominant cause of breathlessness during erect exercise in patients with hypertrophic cardiomyopathy.
Because of our previous conflicting observations of the effects of amiodarone on erect versus supine exercise capacity and hemodynamics,14 we felt it important to compare exercise hemodynamics during supine and erect exercise and, in particular, to assess the relative roles of stroke volume limitation and the heart rate response in both positions. A priori, the different loading conditions, particularly in the presence of severe left ventricular diastolic dysfunction, may result in disparate responses to the two forms of exercise. We observed that peak heart rate was lower during supine than erect exercise. Nevertheless, cardiac output was slightly (but not significantly) higher at peak supine compared with erect exercise. This is explained by the nonsignificantly higher stroke volume at peak supine compared with erect exercise. Similar findings have been reported in normal subjects.15 During both supine and erect exercise, the principal determinant of peak exercise cardiac index is peak stroke volume index; the heart rate is unimportant. The absence of a relation between supine and erect exercise duration and peak exercise cardiac index presumably reflects the fact that exercise duration is a rather crude measure of exercise capacity, since in the erect position, the relation between V̇o2max and peak cardiac index was relatively strong (r=.71).
Limitation of Exercise Capacity in Hypertrophic Cardiomyopathy
The mechanism of exercise limitation in hypertrophic cardiomyopathy remains unclear. Extensive work has been carried out in patients with chronic heart failure, and a brief review of relevant data may provide insight into what occurs in hypertrophic cardiomyopathy. In chronic heart failure, exercise capacity is related to peak exercise cardiac output but not to resting cardiac output. Conversely, there is a weak relation to resting but not to peak exercise PCWP.16 17 It has been suggested that reduced cardiac output may lead to increased muscle glycolysis, lactate accumulation, and leg fatigue18 19 and that impaired vasodilator capacity of skeletal muscle blood vessels, perhaps due to vessel wall edema, may also contribute to skeletal muscle hypoperfusion,20 21 although that hypothesis has recently been challenged.22 Ultrastructural and histochemical changes have been demonstrated in limb skeletal muscle that are quite distinct from those associated with disuse, suggesting that skeletal muscle hypoperfusion might be important.23 The role of skeletal muscle hypoperfusion in limiting exercise capacity has never been addressed in patients with hypertrophic cardiomyopathy. Skeletal muscle ultrastructural and histochemical changes have been reported but have been interpreted as being consistent with a primary myogenic myopathy24 rather than being due to hypoperfusion. Notwithstanding the above, peak exercise cardiac output appears to be a major determinant of peak exercise capacity in patients with hypertrophic cardiomyopathy, and this in turn appears to be dependent primarily on stroke volume augmentation, which is determined by exercise diastolic filling characteristics. In these patients with hypertrophic cardiomyopathy, PCWP was higher both at rest and on exercise compared with previous studies in normal subjects, which is almost certainly a reflection of diastolic dysfunction. However, exercise capacity was unrelated to PCWP.
The patients in the invasive hemodynamic study were selected from 35 consecutive patients with hypertrophic cardiomyopathy. Of these, 1 was excluded because of severe mitral regurgitation, 1 because it was considered unwise to withdraw β-blockers, 2 because of significant respiratory disease, 2 because of atrial fibrillation, and 6 because of refusal to undergo the study. There was therefore some selection bias, but we believe that our results would still be relevant to the majority of patients with this disease, since only a minority were excluded because they were potentially more severely affected than the rest of the study group. In the exercise diastolic filling study, 4 patients were excluded because of significant respiratory disease, 4 because of atrial fibrillation, and 2 because of refusal to undergo the study.
Our exercise data are consistent with the patient group’s being sedentary and unconditioned. Training increases the ability to increase stroke volume. Thus, deconditioning may have contributed to some degree to the inability to increase stroke volume appropriately in these patients, although the cardiomyopathic process appears likely to be much more important.
In 10 patients, peak exercise diastolic filling and stroke volume were not analyzable because of cavity obliteration. These patients all had resting left ventricular outflow tract gradients >40 mm Hg. Thus, the diastolic filling data may not be applicable to patients with severe outflow tract gradients. Our study demonstrates a relation between exercise capacity and a measure of diastolic filling on exercise, ie, time to peak filling, rather than a measure of diastolic function per se. Assessments of diastolic function would require simultaneous measurement of pressure and volume changes. Magorien et al25 showed a relation between peak filling rate and the active phase of diastolic relaxation (maximum −dP/dT, r=−.85) but not with the passive phase of diastolic filling (chamber stiffness, Kd, r=−.08). Other workers have shown an association between the increased diastolic filling during exercise and a decrease in the time constant of relaxation (tau).26 We do not believe, however, that this caveat in any way negates the clinical importance of our observation of an association between an index of diastolic filling (time to peak filling) characteristics and exercise capacity.
It would have been preferable to perform diastolic filling studies during erect treadmill exercise, but it would have been technically impractical. The exercise technique we used involved a semierect cycle that the patients sat on at an angle of 60° to the horizontal. Although we accept this limitation, our observation of an association between peak exercise time to peak filling and maximal oxygen consumption during erect exercise provides a rationale for the importance of stroke volume augmentation in these patients.
The reliable measurement of diastolic filling rates and time to peak filling depends on having a sufficiently high temporal sampling frequency for each cardiac cycle examined—ie, high temporal resolution. The number of frames per synthetic cardiac cycle is a compromise between the requirement for adequate temporal resolution on one hand and the requirement for sufficient counts per frame for statistical precision on the other. Bacharach et al10 showed that, for reproducible measurements of peak filling rate in normal individuals, the frame duration should not exceed 50 ms at rest or 20 ms with exercise. However, using a fixed frame duration throughout an exercise protocol implies that, as heart rate increases with exercise, temporal resolution will become coarser. In the present study, the fixed frame rate we used ensured that temporal resolution remained constant regardless of heart rate, while statistical precision remained comparable, since the shorter frame duration with increasing heart rates was compensated for by the greater number of cardiac cycles accumulated during each 4-minute acquisition period.
In patients with hypertrophic cardiomyopathy, peak exercise heart rate achieved is less than predicted, but the heart rate deficit is only mild. Whereas in normal subjects there is a moderate relation between peak exercise heart rate and maximum oxygen consumption and in patients with heart failure there is a strong relation, in patients with hypertrophic cardiomyopathy this is not so. Our data suggest that this is a reflection of the underlying diastolic filling characteristics. Stroke volume is the major determinant of peak exercise capacity, and although this may be determined in part by increased ejection fraction, our data demonstrate that increased diastolic filling is the most important determinant of stroke volume augmentation and hence peak exercise capacity. This is also supported by the correlation between peak exercise time to peak filling and V̇o2max.
- Received June 14, 1993.
- Revision received June 29, 1995.
- Accepted July 7, 1995.
- Copyright © 1995 by American Heart Association
Levy WC, Cerqueira MD, Abrass IB, Schwartz RS, Stratton JR. Endurance exercise training augments diastolic filling at rest and during exercise in healthy young and older men. Circulation. 1993;88:116-126.
Keeling P, Haywood G, Klare E, Jennison S, Bent S, McKenna W. Chronotropic incompetence in chronic heart failure and its association with reduced exercise capacity in patients with dilated cardiomyopathy. Proc 13th Annual Scientific Session of the North American Society of Pacing and Electrophysiology. PACE Pacing Clin Electrophysiol. 1992;15:575. Abstract.
Report of the WHO/ISFC Task Force on the Definition and Classification of Cardiomyopathies. Br Heart J. 1980;44:672-673.
Drinkwater B, Horvarth S, Wells C. Aerobic power of females, ages 10-68. J Gerontol. 1975;30:385-394.
Bacharach SL, Green MV, Borer JS, Epstein SE. Left ventricular peak ejection rate, filling rate and ejection fraction-frame requirements at rest and exercise: concise communication. J Nucl Med. 1979;20:189-193.
Lipkin DP, Perrins J, Poole-Wilson PA. Respiratory gas exchange in the assessment of patients with impaired left ventricular function. Br Heart J. 1985;54:321-328.
Lele S, Macfarlane D, Morrison S, Thomson H, Khafagi S, Frenneaux M. Exercise diastolic function and functional capacity in patients with coronary artery disease and mild left ventricular systolic dysfunction. Eur Heart J. In press.
Frenneaux MP, Counihan PJ, Porter A, Lipkin DP, McKenna WJ. Effects of amiodarone on erect and supine exercise haemodynamics and exercise capacity in patients with hypertrophic cardiomyopathy. Eur Heart J. 1992;13:687-696.
Poliner LR, Dehmer GJ, Lewis SE, Parkey R, Gunnar Blomquist CG, Willerson JT. Left ventricular performance in normal subjects: a comparison of the responses to exercise in the upright and supine positions. Circulation. 1980;62:528-534.
Lipkin DP, Canepa Anson R, Stephens MR, Poole-Wilson PA. Factors determining symptoms in heart failure. Br Heart J. 1986;55:439-445.
Wilson JR, Martin JL, Schwartz D, Ferraro N. Exercise intolerance in patients with chronic heart failure: role of impaired nutritive flow to skeletal muscle. Circulation. 1984;69:1079-1087.
Weiner D, Fink LI, Maris J, Jones RA, Chance B, Wilson JR. Abnormal skeletal muscle bioenergetics during exercise in patients with heart failure: role of reduced muscle blood flow. Circulation. 1986;73:1127-1136.
Massie B, Conway M, Yonge R, Forstick S, Ledingham J, Sleight P, Radda G, Rajagopalan B. Skeletal muscle metabolism in patients with congestive heart failure: relation to clinical severity and blood flow. Circulation. 1987;76:1009-1019.
Cafario A, Rossi B, Risaliti R, Siciliano G, Marchetti A, Augelini C, Crea F, Mariani M, Muratorio A. Type 1 fiber abnormalities in skeletal muscle in patients with hypertrophic and dilated cardiomyopathy: evidence of subclinical myogenic myopathy. J Am Coll Cardiol. 1989;14:1464-1473.
Cheng CP, Igarachi Y, Little WC. Mechanism of augmented rate of left ventricular filling during exercise. Circ Res. 1992;70:9-19.