Exercise Programmes for Patients with Chronic Heart Failure
- First Online:
- Cite this article as:
- Meyer, T., Kindermann, M. & Kindermann, W. Sports Med (2004) 34: 939. doi:10.2165/00007256-200434140-00001
- 68 Views
The safety and efficacy of exercise training in patients with chronic heart failure (CHF) have been reported in a large number of scientific studies, with endurance training representing the most frequently applied training stimulus. Beneath the common continuous method of endurance training, the interval method (short bouts of intense exercise interspersed with pre-scheduled rest intervals), was also applied in some studies. Ergometric testing is a prerequisite for all individualised training prescription and is an appropriate method of efficacy documentation. However, there is a surprisingly large range of exercise intensities being prescribed to patients with CHF. Most of the prescription models refer to maximal ergometric measurements. Submaximal references from lactate and ventilatory curves represent an alternative method in measuring accuracy and efficacy of training. The course of heart rate during submaximal incremental exercise can be reliably used to indicate endurance gains in CHF. Some positive reports exist for carefully executed strength endurance training for patients with CHF and there are convincing arguments for the use of coordination and flexibility exercises; however, substantial scientific evidence is lacking.
Over the past 15 years, a large number of training studies have been conducted in patients with chronic heart failure (CHF). The results from these studies, involving more than 700 patients, led authorities from the fields of cardiology and sports medicine to conclude that planned physical activity is of sufficient value to be recommended as a therapeutic agent for patients with CHF.[1,2]The cost-effectiveness of planned physical activity in patients with CHF has also been studied.However, there are still some issues of concern when considering the practical application of exercise prescriptions, training programmes and ergometric efficacy documentations that have been investigated so far. Researchers used a surprisingly large range of ‘dosages’ for their exercise prescriptions. Training programmes ranged from interval exercise on a cycle ergometer to structured programmes in rehabilitation clinics. Resistance training has also been studied.
A critical evaluation of currently available training studies, therefore, seems worthwhile and a synthesis of results from different working groups from a sports medicine point of view could be of considerable value. This article discusses different approaches to training as well as the appropriate application of exercise testing for precise evaluation, management and monitoring of patients with CHF who participate in such programmes. The text challenges some long-held practices and beliefs in cardiac rehabilitation and points to some options for innovative research. For the purposes of this review, the term ‘efficacy’ is primarily confined to functional capacity. Section 3 refers to quality of life and prognosis, which represent other important dimensions of efficacy. However, for the most part, they are beyond the scope of this review.
1. General Considerations
The New York Heart Association (NYHA) classes II and III represent the most frequently investigated subgroups of trained CHF patients.Symptom-free NYHA I patients were rarely recruited but there is no reason to exclude them from participation in training programmes. NYHA IV patients could benefit from training of small muscle groups without relevant cardiovascular load.However, it is commonly suggested that resting dyspnoea has to be terminated by sufficient drug or other non-exercise therapy before training can commence. The vast majority of studied patients were male and there is only one study exclusively studying females.Endurance gains were of similar magnitude to other studies; however, the subjects had a very low initial exercise capacity and the training regimen (knee extensor exercises) was hardly comparable to other investigations applying mainly endurance stimuli.
The lowest reported ejection fraction in the relevant studies was 13%.More severely depressed left ventricular function (together with proof of ischaemia during pre-testing) prevented the training effect in one study,but was a prerequisite for training gains in another.Left ventricular ejection fraction itself does not seem to be the key factor for the decision if a training programme can be successfully executed. A clear effect on the trainability of CHF patients is not evident.
In the reviewed studies (table I, table II, table III), CHF developed primarily from coronary artery disease (CAD) or dilated cardiomyopathy (DCM), but it is not evident that similar results can be expected for other aetiologies. However, there is an overlap between CAD and hypertension and it can be assumed that several non-atherosclerotic aetiologies (e.g. toxic and valve dysfunctions) were subsumed under the term DCM. Kavanagh et al.reported inclusion of a few patients with valvular and toxic aetiologies. There have been no published studies comparing aetiologies with regard to the patients’ response to a training stimulus. Also, neither the numerous investigations conducted on selected patients with CAD only[8,9,11–19]nor the two studies in DCM populations[20,21]clearly differ from the average training effect. Unpublished results from our working group indicate that there are in fact no differences in trainability between CAD and DCM with regard to anaerobic threshold, submaximal heart rate and lactate curve. These considerations could be relevant regarding the influence of age on training effects because DCM patients tend to be younger than CAD patients. Older patients with CHF do not seem to be less trainable, although their initial fitness levels are lower. However, sufficient data from studies comparing age groups are lacking.
In the studies reviewed, the supervision of training was not carried out in a uniform manner. Some investigations had a physician continuously present,[15,24,25,35,43]some had ECG monitoring carried out by experienced staff[8,9,34]and in others, authors merely documented the presence of trainers/physiotherapists.[10,14,16,46]Pre-training control of bodyweight, blood pressure, heart rate and rhythm (also during training) was routinely conducted. In most training centres, access to physicians on duty was guaranteed in emergency cases; however, there were virtually no severe incidents requiring hospitalisation in any study. Only short episodes of angina pectoris,self-limiting ventricular tachycardiasand atrial fibrillation[24,25]were reported. In addition, Hambrecht et al.mentioned the occurrence of a new AV node re-entry tachycardia in one patient without describing the temporal relationship to his training. In the light of studies indicating that physical activity could acutely lead to an increased risk of myocardial infarction and sudden death,[50,51]these reports seem to characterise a relatively low-risk potential of exercise in CHF patients. This is partly in accordance with findings that improved training status reduces the risk of death from cardiac causes during exercise.In addition, it is likely that all investigators were well aware of the severe cardiac disease in their patients and took appropriate measures to avoid health risks as far as possible. These measures included a careful exercise prescription.
1.1 Physiological Basis of Training in Patients with Chronic Heart Failure
In patients with CHF, total muscle mass is reduced,and ultrastructural and biochemical alterations of the muscle cells mimic inactivity atrophy and indicate decreased oxidative capacity.[54–56]However, the changes are not limited to inactive muscle groups.In addition, CHF patients face disproportionate peripheral vasoconstriction due to an impaired endothelial function.During exercise, these changes lead to: (i) an early loss of energy-rich phosphates; (ii) increased blood lactate concentrations for given workloads; and (iii) premature exhaustion.
Neither decreased perfusion nor inactivity alone explain this ‘myopathy syndrome’ sufficiently.[57,60]The neuroendocrine activation and the release of cytokines could represent other important triggering factors.
Lacking correlation between exercise tolerance and left ventricular ejection fractiontogether with a missing positive effect of inotropic agents on functional capacity, despite an increased cardiac output,led to the assumption that with ongoing disease peripheral deteriorations become the limiting factors in CHF. This is substantiated by the finding that during the early postoperative stages after heart transplantation, patients do not show a significant improvement in exercise capacity.In fact, correlations were demonstrated between endurance capacity and the following measures of the degree of myopathy: strength and cross-sectional area of thigh muscles;density of mitochondria;activity of cytochrome-c-oxidase;and kinetics of phosphocreatine.These mechanisms can well explain the main exercise-induced symptoms of CHF even without the presence of diminished cardiac output and increased left ventricular filling pressure.
1.2 Physiological Changes Due to Exercise Training
Exercise training partly reverses muscular changes due to a normalisation of endothelial function and, consequently, an improvement in perfusion reserve.[58,67]The same could be true for the coronary circulation, particularly in CHF patients with ischaemic aetiology.The precise molecular mechanisms are presently unclear but there are indications that an increased expression of endothelial nitric oxide synthase or a decreased nitric oxide degradation are involved.Training-induced lowered inflammatory cytokines could also play a role.In addition, physical training induces an increase in muscle fibre size,a higher density of mitochondria and larger activities of aerobic enzymes.These alterations ultimately lead to decreased lactate production for given workloadsas well as to a delayed decrease in energy-rich phosphates.
Neuroendocrine deteriorations are also partly reversible. Catecholamine levels at rest and during exercise were observed to be lower after training,which can be regarded as an indicator of an improved prognosis.More recently, training-induced decreases in plasma concentrations of angiotensin, aldosterone, antidiuretic hormone and atrial natriuretic peptide were reported.These effects can be interpreted as markers of an increased vagal tone, an improved function of cardiovascular reflexes and as protection against arrhythmias.
Central haemodynamics seem to be largely unaffected by training.Left ventricular and pulmonary artery pressures remain unchanged.However, maximal cardiac output could increaseand stroke volume tends to be higher after training, whereas ejection fraction shows no relevant alterations.An increase in stroke volume and, consecutively, in cardiac output can be caused by improved arterial compliance and, therefore, lower vascular resistance in the circulation.Investigations in animals point to the possibility that diastolic function could also profit from exercise training.Taken together, changes in central circulation are much smaller than peripheral effects of training. Initial fears that regular physical training could lead to worsening pump function of the left ventriclewere not substantiated in more recent studies.[15,31]
2. Training Methods
Endurance training was applied in most of the training studies because of its well known positive effects on the cardiovascular system in both healthy individuals and patients with cardiovascular diseases other than CHF. In the recent literature, interval training is proposed as an alternative to the traditional form of endurance training that consists of exercise of constant intensity for periods of several minutes up to hours (continuous method). This is based on the assumption that intervals stress the cardiovascular system less than continuous exercise bouts because of their short duration and consecutive regenerative periods. On the other hand, peripheral/muscular training effects are reported to be equal or even enhanced with interval training.[44,76]Whatever training method is chosen, the following descriptors need to be determined prior to the commencement of training: mode of exercise, duration, frequency and intensity. The latter is considered to be most critical because too high intensities could trigger ischaemia (in CAD patients only) and arrhythmias. Therefore, an appropriate pre-test is obligatory for any decision about intensity. The pre-test is usually conducted as an incremental exercise test with simultaneous ECG. This combines diagnosis/determination of disease severity and exercise prescription in the least time-consuming manner.
2.1 Continuous Method
The most extensive experiences with endurance training in cardiac patients exist for the continuous method, i.e. exercise of constant intensity that is carried out over given durations (table I). Interruptions could be necessary if symptoms of cardiac overload or exhaustion occur but they do not follow a pre-determined schedule. Intensity can be prescribed as external load/power (cycling) or velocity (running, walking). Alternatively, measures of the individual subjective (rate of perceived exertion [RPE]) or objective cardiovascular strain (heart rate) could be applied. Due to the imminent cardiac risks from overload and the lacking congruence between objective measures and RPE, subjective measures are considered to be too inaccurate for monitoring the training of these severely ill patients.[77,78]Theoretically, lactate measurements during training could be an alternative means, but their frequent application raises financial and organisational problems.
The continuous method aims at a constant sustainable intensity and the total physiological load is calculated as the product of intensity and duration of exercise. The training effect should be roughly proportional to these parameters. Several prospective studies were able to document positive effects of the continuous method on endurance capacity of CHF patients (table I). The frequency of training was between two and seven sessions per week for periods of 3 weeks (typical duration of a stay in a rehabilitation clinic) to >1 year. The duration of a single training session was usually between 20 and 60 minutes; however, contents of the sessions were not always described in sufficient detail.
There was a surprisingly large range of intensity prescriptions that varied between 40%and 80%of peak oxygen uptake (V̇O2peakR). Prescriptions with reference to maximal heart rate (HRmax) were between 60%[11,13]and 85%,and between 60% and 80% heart rate reserve (HRR) [difference between resting HR and HRmax].[15,34]So far, there have been no trials using submaximal parameters (lactate or gas exchange thresholds) as reference values. Theoretically, the intensity range around the first increase in blood lactate could be an adequate training stimulus without the danger of overloading the patients. It corresponds well to the anaerobic threshold,[79,80]as determined from gas exchange measurements during incremental exercise. Initial results from our working group indicate that this is a suitable parameter for monitoring training as well as efficacy documentation.[81–84]Beyond all other questions of exercise prescription, it is evident that only those intensities that remain free from symptoms during ergometric pre-testing can be prescribed safely.
2.2 Interval Method
The basic principle of the interval method in training CHF patients is to use short exercise bouts interspersed with resting periods of similar or longer duration. In contrast to intense stimuli for the peripheral muscles, the strain to the cardiovascular system remains relatively small due to the delay of central cardiovascular responses. The main intention is an adequate peripheral training stimulus without simultaneous cardiac danger.
Some investigations, mainly from one working group, point to an equal or slightly superior efficacy of the interval method when compared with the continuous method (table II).[37,44–46]Intensity for the intervals is derived from a so-called ‘steep ramp’ test on a cycle ergometer during which the workload is increased by 25W every 10 seconds starting with 0W. This leads to the attainment of relatively high power outputs within short periods of time. Prescription is done by means of the maximal power output being reached: 30 seconds at 50% alternating with 60 seconds at 10W with a total duration of 16 minutes.An alternative way to determine training intensity, e.g. based on common incremental multi-stage or ramp protocols as they are used in routine diagnostics, has not yet been established. One single study used 90-second intervals of cycling at 80% of V̇O2peakR interspersed with 30-second breaks.The authors documented relatively small and insignificant gains in exercise capacity but no adverse events during their training sessions.
2.3 Other Methods
The physical constraints of CHF patients are not limited to endurance capacity. Due to the chronic inactivity of the musculoskeletal system, other deficits arise in strength, coordination and flexibility. However, scientific investigations that fulfil accepted standards of health-risk monitoring and efficacy control are confined to strength endurance training.[5–7,47,48]Results indicate that gentle strength endurance training is suitable for CHF patients and leads to clinical improvement. However, more studies are necessary to confirm these promising findings and to refine exercise protocols. Until then, strength endurance training should be limited to patients under close supervision of a physician. A scientific accompaniment for such trials is desirable. The study by Meyer et al.could serve as a clue for exercise prescription: the patients did leg-press exercise at 60% and 80% of the pre-determined one repetition maximum. Twelve repetitions per minute were chosen and monitoring of haemodynamic reactions revealed no acute threatening of these subjects.
At present, even without substantial scientific evidence, training consisting of coordination and flexibility can be expected to be of value in patients with CHF. Everyday activities should benefit from such training effects. Experiences in other populations of chronically ill patients (e.g. CAD without CHF, chronic obstructive lung disease) lead to these assumptions. Callisthenics are part of the routine sessions in the established ambulatory cardiac rehabilitation groups. Clinical gains after mixed training programmes are often attributed to an economisation of daily life activities. In consideration of the low cardiovascular load from coordination and flexibility exercises, health risks beyond those of endurance training are unlikely even in CHF patients.
3. Efficacy Testing
Physicians as well as CHF patients need some form of feedback about the success of their training measures. The results of ergometric testing have relevance in addition to information from patient history or, alternatively, from structured questionnaires. Easily obtainable maximal parameters like maximal power output or maximal oxygen uptake (V̇O2max) as well as submaximal indicators of functional capacity could be subject to pre-post comparisons. However, when using maximal measures, it is imperative to pay attention to comparable degrees of effort.Documentation and evaluation of the HRmax being reached (with unchanged medication) and the reason for cessation of exercise (e.g. dyspnoea, angina pectoris, exhaustion of the leg muscles) is mandatory. Complementary information about the degree of effort comes from parameters such as: maximal blood lactate concentration (capillary blood sampling), maximal respiratory exchange ratio and the presence or absence of levelling-off in oxygen uptake (V̇O2) [both from gas exchange measurements]. Maximal power output and V̇O2max could be particularly helpful in the cross-sectional (initial) assessment of functional capacity because well substantiated normative values exist. Most of the reviewed training studies used maximal measurements as efficacy criteria (tables I–III).
Submaximal parameters of endurance capacity can provide a reliable evaluation of training effects when applied appropriately. Due to their independency from subjective influences, they could even be superior to maximal measurements.Usually submaximal parameters are derived from the lactate and heart rate curve during incremental exercise. A rightward shift of both curves (with unchanged medication) indicates a training effect. Several lactate threshold models can be helpful but they have not yet been broadly applied in CHF patients. Thus, for practical purposes, shifts in the heart rate curve are recommended as primary criterion for the assessment of individual endurance gains.[9,10,12,16]This is even more advisable as the only necessary equipment (ECG apparatus) is present almost ubiquitously and because heart rate indicates the myocardial load quite well. In addition, it should be considered that the most reliable results come from tests closely resembling the exercise mode of training, i.e. cycle ergometry for cycling training and treadmill testing for walking training.
3.1 Effect Size of Endurance Training Gains
CHF patients showed training-induced improvements of about 15–20% in endurance capacity (tables I–III).1 Control subjects, on average, maintained their fitness level. The effect of drug therapy with ACE inhibitors on functional capacity remains comparatively smaller (+10%). The direct relevance of ergometric measurements for the daily life of CHF patients is limited; however, an accompanying significant improvement in quality of life, if investigated, was documented almost unanimously.[6,10,12,16,20,26,33,36,38,40,43,46,49]Although there are preliminary indications that physical training reduces morbidity and mortality,a conclusive statement would be premature at this stage. This study involved only 99 patients of whom 50 trained, >80% of them with an ischaemic aetiology. However, a very recent meta-analysis substantiates these findings in a total of 801 patients.
3.1.1 Influence of Training Descriptors on the Effect Size
Duration, mode, frequency and intensity of the training programmes could affect the size of endurance training gains. Studies of short duration[9,33,37,39](i.e. <2 months) demonstrate smaller endurance gains. This simply means that minimum durations are necessary to elicit relevant effects. Usual stays in rehabilitation clinics (about 3 weeks) could be too short and continuing training on an ambulatory basis is warranted. However, there seems to be some kind of ‘ceiling effect’. At least studies of above average duration (i. e. >3 months) do not lead to a clear additional increase in effect size.[10,14,16,17,19,20,32,34,35,41,69]This is substantiated by the only long-term study also reporting results from earlier stages of the programme,which documented the largest effects during the first 8 weeks. This is not surprising since even in healthy individuals training effects are most pronounced during the first months of training. It is possible that obtainable peripheral effects in CHF patients reach their maximum after 2–3 months, whereas the impaired central circulation then becomes the limiting factor.
It is difficult to give a clear statement about the influence of the exercise mode. The large majority of studies applied cycle ergometry supposedly because of its precise adjustability. However, in some studies there were mixed programmes involving walking, arm ergometry and other modes of exercise[8,9,14,15,17–19,21,34–36,38,43,69]as well as cycle ergometry. Obviously, this precludes a differentiation of their respective effects. The only study applying only walkingreported average effects after a very long observation period of 12 months. Investigations with comparisons between different modes of exercise are still lacking.
The choice of exercise frequency was very uniform between the reviewed studies. Almost all researchers applied 3–5 sessions per week. The only two studies with daily training reported training effects above the average.[15,32]Nechwatal et al.,whose patients exercised on 6 days of the week, could not verify this; however, their study duration was only 3 weeks. Thus, more frequent exercise than 3–5 times per week could really elicit larger training gains within programme durations of 3–6 months. However, it remains questionable if CHF patients would really exercise daily outside of very controlled conditions as they are typical for scientific studies or in rehabilitation clinics. There are no published results about comparisons between different training frequencies and it is not known if <3 sessions per week are still of value.
Many authors do not report precise values for intensity prescriptions but use ranges instead (e. g. 60–70% V̇O2peakR). This makes comparisons between intensities difficult. No study has focused on such a comparison so far. When considering ‘extreme’ choices in the lower range of intensities (40% V̇O2peakR[23,25]) it is not evident that endurance gains are smaller. The reported improvements in V̇O2peakR are 17% in both studies[23,25]and those investigators who chose intensities in the upper range (80% HRR and 75% V̇O2peakR[34,41,69]) did not observe larger effects: 16% and 19%, respectively. It appears that a sufficient exercise frequency and programme duration is more important than the choice of exercise mode and intensity with regard to gains in endurance capacity.
4. Areas for Future Research
There are some incongruencies within the present literature that shall be addressed to indicate areas of interest for future research. They refer to exercise testing as well as conducting training programmes.
4.1 Exercise Testing
Presently, there is a large reliance on measurements at maximal effort for ergometric testing of CHF patients. However, during the past decades, alternative ways to evaluate endurance capacity and to prescribe exercise have been developed. Even if maximal testing does not lead to inappropriate risk for CHF patients,[43,88]it does not seem advisable to focus ergometric testing solely on exercise time, V̇O2peakR and HRmax. The attainment of maximal values depends on factors that could be subject to error particularly in CHF, such as: individual patient motivation, day-to-day variation of physical form and attitude of the investigator towards exercising to maximum. In addition, there are other sources of undue variance that apply to all maximal tests, such as biological variation and imprecision of measurements (which are more likely at high intensities).
In training studies, such imponderabilities can lead to substantial overestimations of the training effect. As exercise cannot be carried out in a ‘blind’ manner, there could be a tendency in the members of the training group to spend more effort during re-testing to demonstrate their progress.Also, the physicians who supervise the test often do not remain blind to the randomisation result and are subject to investigator bias if they expect the training programme to be successful. It is acknowledged that remaining blind can be quite difficult during history taking or even during presence in the testing room. The only solutions are strict procedures of blinding the investigator[11,12,89]and the consequent documentation of criteria for the degree of effort (HRmax, maximal respiratory exchange ratio, maximal lactate concentration, levelling off in V̇O2).
In 16 of the 39 training studies being reviewed (tables I–III) not even the HRmax during exercise testing was reported despite the use of maximal parameters as efficacy criterion. Also, in a considerable number of investigations that documented criteria for the degree of effort, the increase in HRmax accounted for up to 50% of the increase in V̇O2peakR.[8,17,19,23,25,32,34]In no study was an appropriate correction procedure applied, nor was the interpretation of maximal data attenuated accordingly.
Considering these shortcomings and the slightly larger effects of endurance training on ventilatory and lactate thresholds (tables I and II), the reliance on submaximal parameters could seem a worthwhile alternative.[59,91,92]Rightward shifts in the (submaximal) heart rate[8,12,13,32,69]and the lactate curveindicate endurance gains, provided medication was constant. To rule out sheer habituation effects, employing a control group is obligatory, of course. Lactate threshold models as well as gas exchange dynamics enable alternative ways of exercise prescription independent of the degree of effort spent.
Altogether, there is much more information to be obtained from exercise testing in CHF patients than just selecting the maximal measurements. This has already been recognised by a number of researchers who analysed their data in more depth. In addition to the use of threshold models as descriptors of endurance capacity, submaximal measurements of the ventilatory equivalent for carbon dioxide[93–95]as well as V̇O2 kinetics[96–98]have been utilised for the assessment of disease severity and prognosis in CHF.
4.2 Training Programme
As mentioned in section 2.1 there is a surprisingly large range of exercise intensities for endurance training that have been prescribed to CHF patients. A part of this variability could be attributed to unequal degrees of effort being spent during pre-testing because maximal measurements are generally used as reference values. Surprisingly large differences in the reported pre-testing values for HRmax[9,16,17,32,48] point to this assumption. Moreover, percentages of V̇O2peakR and HRmax as they are typically used in the CHF training studies (table I) lack sufficient theoretical justification because they are simply transferred from investigations in healthy subjects. Even in healthy subjects and under conditions of sufficient degree of effort, they do not indicate uniform workloads but lead to largely different individual degrees of metabolic and cardiovascular strain.
Metabolically, the intensity range corresponding to the first increase in blood lactate should be an adequate training stimulus without the danger of overloading the patients. It corresponds well to the ‘anaerobic threshold’[79,80]as determined from gas exchange measurements during incremental exercise. Initial results from our working group indicate that this parameter is well suited for monitoring training as well as for efficacy documentation in CHF patients.[81–84]It could be worthwhile to use this indicator or other soundly derived intensities from submaximal reference values for future research in exercise training of CHF patients.
Some conclusive statements can be given regarding exercise programmes in patients with CHF:
Endurance training improves functional capacity (and quality of life) in CHF patients. There are indications that even prognosis is positively affected. Continuous as well as interval programmes have been shown to be efficacious.
Exercise prescription for the continuous method was most often applied with reference to maximal ergometric measurements. There was a wide range of intensities being prescribed but 60–70% of V̇O2max, 70–75% of HRmax and 65–70% of HRR represent ranges that were covered by most studies. However, it seems well justified to alternatively rely on submaximal prescriptors such as anaerobic threshold. Prescriptions for interval programmes were almost uniformly derived from a steep ramp cycle protocol.
Efficacy testing is most efficiently done by means of standard protocols for cardiopulmonary exercise testing. Maximal measurements (V̇O2peakR, peak power) were most frequently used as indicators of fitness changes. However, they have several disadvantages. Thus, reliable submaximal indicators are attractive for this purpose too. Readings of heart rate during incremental exercise as well as anaerobic threshold have been investigated a few times and could be of value. The average endurance gain due to training is 15% of maximal parameters and up to 20% of submaximal ones.
Resistance training could be an effective tool in training therapy of CHF patients. However, there is not enough data to recommend its ubiquitous use. Its potential cardiovascular risks should lead to careful application of relatively low workloads (strength endurance training) and close supervision by physicians.
Other forms of physical training covering flexibility and coordination exercises have not yet been investigated thoroughly. However, their application can be assumed to be safe for theoretical reasons based on experiences in less severely diseased patient populations.
V̇O2peak improved by an arithmetic average of 15.4% (range: 4—33%), whereas the mean increase in anaerobic threshold was 19.5% (range: 7—39%). These approximate calculations are clearly subject to error from differing frequencies, intensities and durations of the exercise programmes; however, there was no difference between controlled studies and those using a crossover design.
The preparation of this manuscript was in part supported by the Deutsche Herzstiftung (German Heart Foundation). The authors have no conflicts of interest that are directly relevant to the content of this review.