High-Intensity Aerobic Interval Exercise in Chronic Heart Failure
- First Online:
- Cite this article as:
- Meyer, P., Gayda, M., Juneau, M. et al. Curr Heart Fail Rep (2013) 10: 130. doi:10.1007/s11897-013-0130-3
- 1.5k Views
Aerobic exercise training is strongly recommended in patients with heart failure (HF) and reduced left ventricular ejection fraction (LVEF) to improve symptoms and quality of life. Moderate-intensity aerobic continuous exercise (MICE) is the best established training modality in HF patients. For about a decade, however, another training modality, high-intensity aerobic interval exercise (HIIE), has aroused considerable interest in cardiac rehabilitation. Originally used by athletes, HIIE consists of repeated bouts of high-intensity exercise interspersed with recovery periods. The rationale for its use is to increase exercise time spent in high-intensity zones, thereby increasing the training stimulus. Several studies have demonstrated that HIIE is more effective than MICE, notably for improving exercise capacity in patients with HF. The aim of the present review is to describe the general principles of HIIE prescription, the acute physiological effects, the longer-term training effects, and finally the future perspectives of HIIE in patients with HF.
KeywordsIntermittent exerciseInterval trainingCardiac rehabilitationHeart failure
Introduction and Historical Perspective
Until the late 1980s, heart failure (HF) was widely regarded as a classical contraindication to exercise training in the belief that patients with a severely reduced left ventricular ejection fraction (LVEF) had an excessive risk for exercise-related morbidity and mortality. In the 3rd edition of Braunwald’s heart disease textbook published in 1988, one could read: “Reduced physical activity is critical in the care of patients with HF throughout their entire course” . This fear was supported by a non-randomized study performed in the pre-angiotensin converting enzyme inhibitor and beta-blocker eras, which suggested adverse cardiac remodeling in patients with recent anterior myocardial infarction after a 12-week low-level exercise training program . Since the initial landmark trial conducted by Sullivan et al. , many randomized trials demonstrated that regular exercise is safe and provides many benefits in the care of patients with chronic HF and reduced LVEF. Meta-analyses of these studies reported an average increase of peak oxygen uptake (VO2peak) of 15 to 20 % following exercise training, with an improvement of symptoms and quality of life, although the impact on morbidity and mortality remained unclear [4–6]. Finally, the largest multicenter randomized, controlled trial of aerobic training in HF, HF-ACTION, was published in 2009. This trial demonstrated modest but significant benefits related to aerobic training after adjustment for pre-specified prognostic factors, in the primary composite endpoint of all-cause mortality or all-cause hospitalization and did not reveal safety issues compared to the non-exercise group . The training modality used in the vast majority of these studies, including HF-ACTION, was moderate-intensity aerobic continuous exercise (MICE), which therefore forms the basis of training therapies in HF patients. For the first time in 2012, the European Society of Cardiology acknowledged aerobic exercise as a class I, level of evidence A recommendation in patients with HF, which constitutes a complete paradigm shift in less than 25 years .
In more recent years, another aerobic training modality, high-intensity aerobic interval (or intermittent) exercise (HIIE), has emerged and appears to be more beneficial in patients with HF [ [9•], [10, 11], [12•]]. HIIE consists of repeated periods of high-intensity exercise in alternation with periods of low-intensity exercise or rest. First described in 1959 by Reindell and Roskamm in a scientific journal , HIIE has been used since the beginning of the 20th century by elite athletes and definitely popularized by the long-distance runner Emil Zatopek, triple Olympic Champion at the Helsinki Olympic Games in 1952 . Since the early 1990s, Meyer et al. began studying HIIE in patients with coronary artery disease , and later in patients with HF , but a real interest in this training modality only arose a decade ago.
Principles of HIIE Prescription
The fundamental principle of HIIE is that periods of high-intensity exercise are interspersed with recovery periods, which allows the individual to re-engage in high-intensity exercise multiple times during the same session. The rationale is to accumulate more time in high-intensity zones compared to a continuous exercise where exhaustion would occur more prematurely, and therefore to produce a stronger stimulus for cardiovascular and muscular adaptations [10, 17]. A parameter frequently used to quantify the overall aerobic stimulus induced by a HIIE training protocol is the time spent at a high percentage of VO2peak [12•, 18–25]. HIIE can take the form of cycling, walking, rowing, swimming or other activities, although stationary cycling and treadmill walking have been the most utilized. There are an unlimited number of possible HIIE protocols using different intensities and durations of work/recovery intervals. Different interval combinations have been tested in various populations and induce profound variations in acute physiological responses to exercise that must be taken into consideration during exercise prescription [14, 24–27]. In sports training, three different categories of HIIE have been described [12•]: (1) long intervals: 3 to 15 minutes at an intensity corresponding to 85–90 % of maximal oxygen uptake (VO2max); (2) moderate intervals: 1 to 3 minutes at 95–100 % of VO2max; and (3) short intervals: 10 seconds to 1 minute at 100–120 % of VO2max.
HIIE protocols investigated in different randomized training studies
Weeks of training
Dimopoulos 2006 
100 % of PPO
Roditis 2007 
Wisloff 2007 
90–95 % of peak HR
50–70 % of peak HR
Smart 2011 [44•]
70 % of PPO
Fu 2011 [49•]
80 % of HRR
40 % of HHR
Iellamo 2012 
75–80 % of HRR
45–50 % of HRR
Freyssin 2012 
Ergometer cycling / treadmill walking
50 to 80 % of MSEC
A considerable body of evidence has been accumulated in recent years with regards to HIIE in HF and other medical conditions thanks to several Norwegian groups [10, 11]. Their HIIE protocol in HF consisted of 4/3 minute work/recovery intervals of uphill treadmill walking that were repeated 4 times, at respective work/recovery intensities of 90–95 % and 50–70 % of peak heart rate . As will be discussed later, very impressive improvements in terms of exercise capacity and cardiac reverse remodeling were demonstrated using this protocol in a small randomized trial in patients with HF . We believe, however, that velocities or power output corresponding to a percentage of VO2peak should be employed to prescribe HIIE intensity because the use of target heart rate zones is often not reliable for the following reasons in HF: (1) frequent chronotropic incompetence due to maximally titrated beta-blockers, (2) a high prevalence of chronic atrial fibrillation, and (3) the constant occurrence of heart rate drift during exercise requiring frequent workload adaptations [33, 34].
In summary, no single approach to HIIE prescription can be recommended at present. The most frequently used interval durations range from 30 seconds to 4 minutes, usually with equal work/recovery duration. The work and recovery intensities range respectively from 80 % to 120 % of VO2peak and from 0 % (passive recovery) to 50 % of VO2peak. Warm-up and cool-down periods must be included at the start and end of HIIE sessions that usually have a total duration of 15 to 45 minutes and may be repeated two to five times weekly.
Acute Effects of HIIE
Several studies have investigated the acute physiological responses to HIIE compared to MICE in HF. Meyer et al. compared on a cycle ergometer HIIE (30/60 seconds work/recovery intervals at intensities of respectively 50 % of MSEC and 15 Watts) with MICE at an intensity of 75 % of VO2peak . HIIE resulted in a higher power output during interval work phase, but with lower rate-pressure product, rating of perceived exertion and plasma catecholamine levels, despite higher blood lactate levels. The authors concluded that HIIE provided a greater peripheral muscular stimulus with a lower central cardiac stimulus in HF patients. Using the same protocol, LVEF measured by radionuclide ventriculography increased significantly and by the same magnitude during HIIE and MICE at the same average power output, both in HF patients and controls [36, 37]. This finding is important given the persistent fear that HIIE may acutely further deteriorate left ventricular function. Recently, Tomczak et al. examined in nine patients with mild HF (NYHA functional class I and II, mean LVEF = 35.8 ± 7.2 %), the acute effects of HIIE using the same protocol as described by Wisloff et al.  on biventricular function assessed by cardiac magnetic resonance, 6 and 30 minutes post-exercise . Overall, they demonstrated an improvement of biventricular systolic function with a 2.4 % absolute increase of LVEF at 30 minutes post-exercise (P <0.05), which adds further evidence about the safety of HIIE in patients with mild HF and reduced LVEF.
Our group performed in 20 patients (90 % men; aged 61 ± 9.9 years) with HF and reduced LVEF (mean 26 ± 7 %) a comparison of acute responses after HIIE (2 × 8 minutes of 30/30 seconds work/recovery intervals at respectively 100 % of PPO and rest) compared to isocaloric MICE (22 minutes at 60 % of PPO) . A greater proportion of subjects completed the prescribed HIIE session compared to MICE (85 % vs. 40 %, P < 0.0075) with similar time spent over 90 % of VO2peak and similar mean cardiopulmonary responses (VO2, ventilation, heart rate and O2 pulse) during MICE and HIIE. However, during HIIE, patients exercised for a shorter effective cycling time (7.5 ± 1.5 vs. 17 ± 5 minutes), at a higher power output (51 ± 21 vs. 46 ± 20 Watts), which translated into superior efficiency (energy expenditure/effort time) of HIIE compared to MICE. Finally, HIIE was well tolerated with ratings of perceived exertion that tended to be lower compared to MICE. We did not observe any significant arrhythmias, abnormal blood pressure responses, or increase in troponin-t, B-type natriuretic peptide (BNP) or high-sensitivity C-reactive protein. Using the same protocol, we further evaluated the central hemodynamic changes by cardiac bioimpedance during HIIE and MICE in 13 patients (age, 59 ± 6 years; LVEF, 27 ± 6 %) . Compared to MICE, HIIE elicited similar mean cardiac output (9.26 ± 1.93 L/min vs 10.06 ± 3.14 L/min), mean stroke volume (96 mL ± 22 vs. 93 ± 21 mL) and mean calculated difference in arterio-venous oxygen content. When examining the kinetics of these variables, our study demonstrated the relative stability of central hemodynamics during HIIE, which is in line with the lack of complications observed in our previous studies. We also compared carbohydrate and lipid oxidation during HIIE and MICE using gas exchange analysis (Frayn equation) in 18 patients with HF and reduced LVEF . HIIE and isocaloric MICE elicited a 4.8- to 5-fold higher carbohydrate oxidation and a 1.22- to 1.42-fold higher lipid oxidation compared to resting values but absolute and relative values of substrate muscle oxidation were similar during both exercise modalities. Finally, with respect to the arrhythmogenic potential of HIIE, Labrunee et al. performed three consecutive 24-hour Holter recordings in 12 patients with HF following an HIIE session, a MICE session or a control period without physical exercise . Interestingly, the number of premature ventricular contractions, ventricular couplets or episodes of non-sustained ventricular tachycardia was clearly and significantly lower after HIIE compared to MICE and no exercise. This novel finding may be due to an improvement in sympatho-vagal balance following HIIE, which was corroborated by a greater reduction of heart rate and better indices of heart rate variability after HIIE compared to MICE.
Overall, these studies indicate that HIIE results in similar mean cardiorespiratory responses but probably superior peripheral effects compared to isocaloric MICE. HIIE appears better tolerated with lower ratings of perceived exertion allowing more patients to complete their planned exercise sessions. When taking into account the effective exercise time without resting intervals, HIIE is a more efficient exercise modality compared to MICE. Finally, in these selected HF patients HIIE appeared safe and did not induce significant arrhythmias, signs of myocardial injury, or acute deteriorations of left or right ventricular function.
HIIE Training Studies
The first HIIE training study in patients with HF was performed by the group of K. Meyer . They performed a random-order crossover study comparing the effects of 3-week HIIE training versus activity restriction in 18 patients with severe chronic HF (mean LVEF 21 ± 1 %, mean VO2peak 12.2 ± 0.7 ml/kg/min). The HIIE protocol consisted of 30/60 second work/recovery intervals at respectively 50 % of MSEC (corresponding to slightly more than 100 % of PPO) and 15 Watts, during 15 minutes, 5 times/week. They demonstrated a 24 % increase in VO2peak  and a 6 5 % improvement in the 6-minute walk test  in the exercise group.
Randomized training studies comparing HIIE to MICE
Baseline VO2peak (ml/kg/min)
↑ VO2peak (HIIE vs. MICE)
Dimopoulos 2006  (N = 24)
40 minutes at 50 % of PPO
8 % vs. 6 %
↑ HR recovery at 1 min in MICE only
Roditis 2007  (N = 21)
40 minutes at 50 % of PPO
8 % vs. 9 %
↑ phase II O2 kinetics in MICE only
Wisloff 2007  (N = 27)
47 minutes at 70–75 % of peak HR
46 % vs. 14 %
↑ LVEF, ↑ endothelial function, ↑ quality of life superior in HIIE
Smart 2011 [44•] (N = 23)
30 minutes at 70 % of PPO
21 vs. 13 %
No change in cardiac volumes, LVEF and endothelial function
Fu 2011 [49•] (N = 45)
30 minutes at 60 % of HRR
22.5 % vs. no significant change
↑ ventilatory efficiency, ↑ cardiac output, in HIIE only
Iellamo 2012  (N = 16)
30–45 minutes at 45–60 % of HRR
22 % vs. 22 %
No change in cardiac output, in cardiac dimensions, and in LVEF
Freyssin 2012  (N = 26)
61 minutes at VT1 + aquagym + gymnastics
27 % vs. 2 %
↑ 6-minute walk test by 12 % in HIIE (6 % in MICE)
These impressive findings have generated a wave of enthusiasm in the cardiovascular rehabilitation community. Subsequently, several other groups have demonstrated the benefits of HIIE training in HF patients. Using a similar protocol with 3/3 minutes work/recovery intervals at respectively 80 % and 40 % of PPO, Fu et al. demonstrated a 23 % increase in VO2peak in the HIIE group (p < 0.05) compared to no significant change in the MICE group [49•]. Ventilatory efficiency and cardiac output were also significantly increased in the HIIE group compared to the MICE group. Comparable improvements in VO2peak (respectively 21 % and 13 % in the HIIE and MICE groups) were observed by Smart et al. after 16 weeks of stationary cycling thrice a week. However, no significant changes in ventricular dimensions or systolic or diastolic function were detected over the study period. A particularity of this study was that HIIE and MICE protocols only differed by the addition of 60 seconds of rest each minute in the HIIE group, and otherwise used the same moderate intensity (70 % of VO2peak) for exercise intervals [44•]. Finally, in a short but very intensive intervention (6 times per week during 8 weeks), using 30/60 second work/recovery intervals at intensities of up to 12 0 % of VO2peak for work intervals and employing a passive recovery, Freyssin et al. demonstrated a VO2peak increase of 22 % compared to 2 % only in the MICE group .
However, not all training studies have demonstrated the superiority of HIIE over MICE for improving VO2. Using a protocol with 30/30 seconds work/recovery intervals at high intensity (100 % of PPO) and passive recovery, Dimopoulous et al. only found a modest 8 % increase in VO2peak in the HIIE group (P = 0.01) compared to 6 % in the MICE group (P <0.05) without significant differences between groups . Moreover, examining markers of parasympathetic activity, they reported a significant increase in heart rate recovery at one minute post-exercise only in the MICE group. In a subsequent study using the same protocol, these authors obtained similar results in terms of exercise capacity improvements and they also demonstrated that only MICE training improved phase II O2 kinetics, an indirect index of muscle oxidative capacity . More recently, Iellamo et al. performed a very similar intervention as described by Wisloff et al. but with slightly lower intensities (75–80 % of heart rate reserve versus 90–95 % of peak heart rate) . VO2peak improved by 22 % in both HIIE and MICE groups while neither training modality influenced ventricular remodelling parameters or cardiac output.
In summary, these studies overall suggest that HIIE training elicits larger improvements of VO2peak compared to MICE in patients with HF whereas the impact on cardiac dimensions and function and on endothelial function remains uncertain. Differences in results may be due to differences in patient characteristics and/or in training protocols. However, we believe it is currently not possible to conclude that a specific HIIE protocol is superior to another. Furthemore, all studies were single-centre, and, in the vast majority, investigators were not blinded to outcome assessments, which represent limitations.
Conclusions and Future Directions
When considering the current body of evidence of HIIE in HF summarized in this review, several conclusions can be drawn. First of all, HIIE is more effective than MICE to improve VO2peak in patients with HF and reduced LVEF. This is an important finding given that better functional capacity translates into an improvement of symptoms and quality of life. Secondly, HIIE was shown to be more efficient, meaning that the same energy expenditure can be performed in a shorter time period, which may be an important practical aspect to consider during a rehabilitation program. Thirdly, HIIE acute and training studies performed in selected HF populations, including predominantly men with ischemic heart disease in NYHA functional class II, did not reveal safety issues.
However, more research of HIIE in HF is definitely warranted to further explore several issues.
Will HIIE have an impact on clinical outcomes?
VO2peak being one of the strongest predictors of outcomes in this population , HIIE may potentially contribute, among other therapeutic strategies, in further reducing HF morbidity and mortality. However, to demonstrate such an effect would require many more randomized training studies or a very large “HF-ACTION-like” trial investigating HIIE, which is hardly conceivable at present. In a similar vein, the impact of HIIE on cardiac reverse remodeling parameters and especially LVEF, which is also a known predictor of outcomes in HF, remains uncertain and should also be further investigated.
Is HIIE really safe in HF patients?
Even if HIIE training appeared safe in a recent large observational study in coronary patients , more data are needed in HF patients who are older, more fragile, and often equipped with defibrillators. Future investigations should include a larger sample and a broader representation of HF patients. This will not only confirm the safety of HIIE in this population but also help to better identify predictors of response to that training modality.
What is the optimal HIIE protocol to be used?
As mentioned earlier in this review, no specific HIIE protocol has been proven superior, and therefore no single regimen can be recommended. Future training studies should compare different HIIE protocols to establish the optimal one. However this is a rather complex issue not only because there are countless possible interval combinations but also because protocols should probably be individualized and tailored to patients, especially according to their baseline exercise capacity and training level. As mentioned earlier, very deconditioned patients may benefit from protocols using shorter intervals, lower intensities and passive recoveries whereas patients with higher training levels could better improve with longer intervals, higher intensities and active recoveries.
Will HIIE improve adherence to exercise training in HF patients?
Non-adherence was a major issue in the HF-ACTION study where only 30 to 40 % of patients followed training recommendations after 3 months despite a close follow-up by study personnel . There is no data yet about the adherence to HIIE in HF patients. In coronary patients, the long-term adherence to HIIE was superior to MICE  and it was perceived as a more enjoyable training modality in healthy adults . In our experience, HIIE may be very motivating for HF patients when performed in groups in the same way as the very popular “spinning” sessions proposed by fitness centers.
Some of these issues will be addressed in an ongoing multicenter European randomized trial, SMARTEX-HF . A total of 200 patients with HF and LVEF ≤35 % will be randomized for a 12-week exercise program to one of three different arms: HIIE as proposed by Wisloff et al. , MICE, and regular exercise according to patients’ preferences. The primary endpoint will be cardiac reverse remodeling whereas secondary endpoints will include VO2peak, biomarkers, quality of life and level of physical activity assessed by questionnaires.
In a recent position statement, the European Society of Cardiology has incorporated HIIE as a recommended training modality in patients with HF [64•]. In agreement with this paper, we believe that HIIE should not replace other training modalities in HF, such as MICE, strength and respiratory training but should rather complement them. To be effective, exercise training should be progressive and varied. Therefore, analogous to optimizing pharmacotherapy, combining and tailoring different exercise training modalities according to each patient’s baseline exercise capacity, personal needs, preferences and goals seem the most judicious approach to exercise prescription.
Philippe Meyer declares that he has no conflict of interest.
Mathieu Gayda declares that he has no conflict of interest.
Martin Juneau declares that he has no conflict of interest.
Anil Nigam declares that he has no conflict of interest.