The physiology of interval training: a new target to HIIT

Given the enormous health benefits of exercise, there is a large body of research aimed at identifying the optimal type and dose of training to maximise physiological and performance adaptations. Among a variety of training modalities, high‐intensity interval training (HIIT) is recognized as a time‐efficient training strategy to induce similar or even superior adaptations compared to traditional moderate‐intensity continuous training according to a number of physiological, performance and health‐related markers (Gibala et al. 2012).

A strong interest in HIIT has been stimulated by the evidence that three sessions per week of low‐volume Wingate‐based HIIT (i.e. repeated bouts of 30‐s all‐out cycling efforts interspersed with 4/4.5 min of recovery) induce similar performance adaptions to those obtained with an almost 10 times greater weekly training volume performed in the traditional continuous training mode (Gibala et al. 2012). Moreover, in the Wingate‐based HIIT studies reviewed by Gibala et al. (2012), the overall time commitment for HIIT was about 3 times shorter than that of traditional continuous training. This is of great importance considering that ‘lack of time’ is commonly cited as one of the main barriers to exercise adherence (Gibala et al. 2012). Interestingly, despite the sprint nature of Wingate‐based HIIT, endurance‐like adaptations were documented after a few weeks (2–6) of training with this HIIT format. These include increased muscle oxidative capacity and glucose transport capacity and therefore improved insulin sensitivity and glycaemic control, along with cardiovascular adaptations (Gibala et al. 2012). The remarkably similar changes observed after HIIT and traditional endurance training suggest that metabolic adaptations to HIIT could be mediated in part through signalling pathways normally associated with endurance training. This is supported by the well‐documented increase in the peroxisome proliferator‐activated receptor‐γ coactivator 1 α (PGC‐1α; i.e. the master regulator of mitochondrial biogenesis in muscle) after HIIT (Gibala et al. 2012). The increase in PGC‐1α further highlights the potential widespread health benefits of HIIT, given the positive effects that an increase in PGC‐1α has on oxidative capacity, glucose uptake, anti‐oxidant defence and resistance to age‐related sarcopenia (Gibala et al. 2012). However, the Wingate‐based HIIT is an extremely demanding exercise format and may not be safe, well‐tolerated or appealing for some individuals (Gibala et al. 2012).

Less extreme and more practical HIIT formats have been extensively used and commonly compared to traditional continuous training on a matched‐work basis or isocaloric basis. When using this method of comparison, the superiority of HIIT over continuous training is even more evident in promoting health benefits in both healthy individuals and diseased populations (Gibala et al. 2012; Weston et al. 2014). A meta‐analysis comparing HIIT and continuous training in patients with cardiometabolic disease found an almost doubled increase in cardiorespiratory fitness (as measured by ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$) after HIIT compared to continuous training (Weston et al. 2014). This should translate into a greater decrease in risks of morbidity and all‐cause mortality. Other health‐related adaptations occurred significantly more in HIIT compared to continuous training, such as reduced blood pressure, improved insulin sensitivity, increased nitric oxide availability, improved lipid metabolism and increased PGC‐1α (Weston et al. 2014). Despite the strong efficacy of HIIT, methodological limitations are commonly reported in HIIT studies (Weston et al. 2014). Among these, it is worth mentioning the possibility of committing a type II statistical error, given the fact that the length of training is often relatively short, the sample size is usually small and a between‐subject design is typically used.

A recent study published in The Journal of Physiology by a group of leading scientists in the physiology of HIIT provided an elegant experimental design to further our understanding of the effect of HIIT on skeletal muscle mitochondrial adaptations as compared to matched‐work continuous training (MacInnis et al. 2016). The authors used a single‐leg cycling within‐subject parallel‐group design, i.e. one leg was trained with HIIT and the other with moderate‐intensity continuous training (MICT). Specifically, each leg was randomly assigned to complete six sessions of work‐ and duration‐matched HIIT (4 × (5 min at 65% of peak power and 2.5 min at 20% of peak power)) or MICT (30 min at 50% of peak power) over 2 weeks. The HIIT and MICT sessions for the two different legs were performed 10 min apart on the same day, and in alternating order across the six training sessions. Ten healthy young men were included in the study. Methodologically, this experimental design allowed for control of individual variability in training responsiveness, increase in statistical power and elimination of the need for washout periods. This is a very important methodological improvement made in order to limit the occurrence of a type II error in HIIT longitudinal studies. Physiologically, single‐leg cycling may induce greater adaptations than those induced by double‐leg cycling, possibly in view of the increased relative exercise intensity (MacInnis et al. 2016). This makes single‐leg cycling an appealing exercise modality when investigating skeletal muscle adaptations following short‐term training programmes. Of course, the applicability of the single‐leg cycling within‐subject parallel‐group design is limited to the investigation of physiological adaptations (e.g. mitochondrial adaptations) that show no evidence of transfer from the trained leg to the non‐trained leg.

MacInnis et al. (2016) found that HIIT induces greater increases in mitochondrial content biomarkers in human skeletal muscle compared to MICT. Specifically, citrate synthase maximal activity and mass‐specific oxidative phosphorylation capacities (complex I, and complexes I and II) were greater in HIIT relative to MICT, while whole‐muscle cytochrome c oxidase subunit IV protein content increased similarly between training modalities. Conversely, no change in mitochondrial function (i.e. mitochondria‐specific oxygen flux) was found in either leg. Unlike previous double‐leg cycling HIIT studies, ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ did not increase in either of the exercise modalities, possibly due to the lower cardiorespiratory stimulus associated with single‐leg cycling. These findings led the authors to conclude that HIIT induces superior mitochondrial adaptations compared to MICT despite equal total work and session duration, probably because of the higher intensity and/or the different pattern of contractions (MacInnis et al. 2016). However, ratings of perceived exertion, dyspnoea, heart rate and blood lactate were substantially higher during the HIIT sessions compared to continuous sessions (MacInnis et al. 2016). Taken together, these responses may indicate that the overall effort, and thus relative exercise intensity, was higher during HIIT. Despite the benefits derived from adopting the above‐described experimental design, the potential difference in overall effort between the two training regimens may introduce a confounding factor potentially raising doubts about the superiority of HIIT vs. continuous training. This is a common limitation of the matched‐work approach, which results in HIIT being more demanding than continuous training (Nicolò et al. 2014). Indeed, when the session duration is the same, the practice of matching total work implies that HIIT and continuous training are equalized by absolute exercise intensity (same average workload) instead of relative exercise intensity (similar overall effort, stress and exercise demand). Although exercise is often prescribed according to relative exercise intensity, the adaptations induced by HIIT or continuous training are rarely investigated by balancing the different exercise modalities for the same relative exercise intensity. The comparison of HIIT with matched‐work continuous training is further complicated by the fact that the difference in effort between HIIT and continuous training is influenced by the specific format of HIIT. For instance, this difference increases with the decrease in the work‐to‐rest duration ratio (Nicolò et al. 2014).

Since a considerable number of studies comparing HIIT with continuous training used the matched‐work approach (Gibala et al. 2012; Weston et al. 2014), the superiority of HIIT over continuous training may have been biased at least to some extent by the extensive adoption of this method of comparison (Nicolò et al. 2014). In order to overcome some of the limitations of balancing HIIT and continuous exercise for total work, the use of a method of comparison that guarantees the same overall effort across different exercise modalities has been proposed (Nicolò et al. 2014). Nevertheless, the matched‐work approach is still one of the methods of comparisons prevalently used to evaluate the efficacy of HIIT vs. continuous training, as the study by MacInnis et al. (2016) testifies. What may appear a purely methodological issue is in fact of fundamental physiological and practical relevance. Suffice to consider that one of the main barriers to regular physical activity is the physical effort required (Marcora, 2016). Accordingly, moderate‐intensity exercise may be preferable to vigorous exercise for some individuals as the latter requires more effort and is unpleasant (Marcora, 2016). On the other hand, vigorous exercise has a clear advantage over moderate‐intensity exercise in inducing health‐related adaptations. Therefore, training modalities that reduce effort while maintaining the physiological stimulus for adaptations should be recommended from an exercise‐adherence perspective. With this in mind, the burning question that exercise physiologists still need to answer convincingly is the following: is there an HIIT format that induces superior physiological and performance adaptations compared to other exercise modalities when the overall effort required by different training regimens is the same?

Additional information

Competing interests

None declared.