Background
High-intensity training (HIT) and high-intensity interval training or aerobic interval training (HIIT or AIT) are commonly accepted stimuli for improving anaerobic [
1,
2] and aerobic [
3‐
7] functions. Although the terms HIT and HIIT are sometimes used interchangeably, HIT usually refers to near-maximal or supramaximal exercise intensities (>90 % peak rate of oxygen uptake,
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \)) [
6], whereas HIIT or AIT is often used in the context of exercise intensities between 80 % and 90 %
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \) [
3,
7,
8].
Acute microvascular responses to exercise (e.g., enhanced blood flow and local blood supply of small vessels in active tissue [
9,
10]) have also been reported as long-term adaptations following HIT (e.g., by increased vasodilatory capacity [
9] and augmented capillarization [
5,
11]). Fu et al. showed these effects to be superior to those of moderate continuous training in patients with heart failure [
7]. Furthermore, several studies have suggested that aerobic adaptations are facilitated by the above-mentioned microvascular mechanisms following HIIT/AIT [
3,
4,
7].
Although the literature contains substantial knowledge on the acute effects of exercise on local muscle perfusion in general [
10,
12‐
15], the acute post-exercise effects of different exercise durations on local muscle oxygen availability and blood supply have not yet been sufficiently examined. Most studies have focused on the influence of exercise intensity [
16‐
19] or examined the effect of complete training sessions with only two different work-interval durations [
20].
Relative changes in local, total hemoglobin concentration (ΔTHb) is a widely used parameter for blood-volume changes in terms of capillary filling [
21‐
23] and vasodilation [
24] and can be noninvasively monitored using near-infrared spectroscopy (NIRS) [
22]. With continuous-wave spectrometers, relative changes in the THb and deoxygenated (HHb) and oxygenated Hb (O
2Hb) concentrations can be observed because of the distinct relative transparency of HHb and O
2Hb for specific, near-infrared light wavelengths.
The potential of exercise to prolong augmentation of the local blood supply could be an important determinant for metabolic adaptations. The aim of our study was to identify the association between durations of HIIT/AIT exercise bouts and relative changes in post-exercise concentrations of THb, O2Hb, and HHb to examine (1) the post-exercise blood supply and (2) local oxygen availability.
Discussion
The data of the present study show that cycling exercise at 80 %
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \) triggers post-exercise hyperemia, which is indicated by an overshoot in ΔTHb following exercise bouts in relation to pre-exercise values. ΔTHb has been used as an indirect measure for blood volume previously [
21‐
23]. When we considered the five different exercise durations, the 30-s and 60-s exercise bouts evoked significantly lower overshoot values than the longer exercise bouts.
Because the relationship of post-exercise overshoots of ΔTHb and ΔO2Hb evolved similarly, we suggested that the arterial blood supply primarily accounts for the increased post-exercise ΔTHb. A reduced outflow would be expected to emerge as a concomitant increase in post-exercise ΔHHb at the point of OS. Like OS_THb, OS_O2Hb increased stepwise from 30 to 60-s and from 60 to 90-s exercise, indicating an increasing availability of oxygen in the area of interrogation up to exercise durations ≥90 s.
To the best of our knowledge, the present study is the first to investigate post-exercise muscle reoxygenation in relation to single exercise bouts of different durations as they are used in interval-training regimens. In general, a number of studies have reported on post-exercise hyperemia following single exercise bouts of similar durations as they were applied in the present study [
16,
17,
19]. Furthermore, Danduran et al. [
18] observed hyperemia using NIRS following a graded exercise test design in healthy children. Post-exercise reoxygenation was analyzed in various studies. A significant dependency of post-exercise Hb and muscle oxygenation recovery kinetics on exercise intensity (in terms of a forced reoxygenation following higher exercise intensities) has been reported previously [
16,
17,
19]. Despite this dependence, reoxygenation time is apparently not influenced by exercise intensity [
19,
29]. When Belfry et al. [
36] compared interval-training regimens with constant-load exercise at equal intensity, they found a better matching of O
2 delivery to O
2 utilization during exercise for the interval regimens than that in constant-load exercise. O
2 delivery during interval training was enhanced when recovery was applied in the moderate intensity domain in comparison to “low-intensity” active recovery. Besides the muscle-pump effect, which was presumably increased during interval training with “moderate intensity” recovery, the authors suggested enhanced local vasodilation to be an important determinant for those results.
Zafeiridis et al. [
20] were also able to find improved local oxygen delivery during exercise. In contrast to Belfry et al., they found no differences between constant-load exercise and two interval-training regimens with different work-interval durations. The different results in these two studies could likely be attributed to differences in the two study protocols. As we did in the present study, Zafeiridis et al. examined the effect of different work-interval durations. They used work intervals of 30 s with an intensity of 110 % of the power output corresponding to
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2 \max } \) and 120-s intervals with an intensity of 95 %
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2 \max } \) and did not find significantly different local-oxygen delivery during both interval-training regimens. This finding of Zafeiridis et al. is contrary to our findings, which showed significantly higher post-exercise oxygen availability and blood supply following 120 s of exercise than that following 30 s of exercise with 80 %
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \). We cite two possible explanations for this difference. First, the dependency of post-exercise blood supply and oxygen availability on exercise duration, as it is shown in the present study, appears to be valid only within a certain exercise intensity range. In a previous study, we examined the effect of exercise intensity on post-exercise blood supply and oxygen availability and observed a non-linear relationship [
19]. We found no differences between 80 and 90 %
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \), but did not include higher exercise intensities; therefore, it might be possible that our results are not valid for supramaximal exercise. Second, Zafeiridis et al. used different recovery durations between work intervals, which could have contributed to different results. It could be that 30-s work intervals cause hyperemia similar to that associated with 120-s intervals, when recovery periods are adjusted appropriately.
Although the methods used in our study do not provide sufficiently detailed information on the complex regulation of local muscle perfusion, we call attention to a few possible explanations for our results. In general, cardiac output and artery flow increase as a response to dynamic exercise [
12]. In exercising muscles, two antagonistic mechanisms happen: There is global sympathetic mediated vasoconstriction on the one hand whereas acute vasodilation occurs in active tissue secondary to a promoted release of vasoactive substances on the other hand. Because vasodilation superimposes sympathetic mediated vasoconstriction in active muscles, capillary perfusion is increased [
10,
37,
38]. This process is supported mechanically by rhythmic contractions of the muscle during cycle exercise (muscle pump) [
12].
Previous research has shown a biphasic response of muscle blood flow to exercise with an early, rapid response within the first 5–7 s and a prolonged, secondary response starting 15–20 s from the onset of exercise [
15], while a steady state is reached within 1–3 min. Hence, blood-flow adjustments to exercise occur rather quickly. After termination of exercise, blood-flow recovery has been reported to be slower following heavy exercise than following moderate exercise [
13]. This slowing of blood-flow recovery is most likely due to the prolonged presence of vasoactive substances. In rats, it had been shown that acute, vasodilatory responses to exercise can last up to two days or more [
39]. Consequently, prolonged, local vasodilation seems to be dependent on the extent of the release of vasoactive substances during the previous exercise bout.
But which mechanisms are crucial for the increased post-exercise oxygen availability between 30 and 90 s of exercise at 80 %
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \)? We did not measure cardiac output in this study, but our data reveals some basis for speculation. The local oxygen availability was equal from 90 to 240 s of exercise duration (Fig.
4b), while end-exercise heart rate increased continuously. We assumed accordingly that end-exercise cardiac output also increased continuously with exercise duration (otherwise, stroke volume would have declined, which is unlikely). Hence, if the prolonged recovery of cardiac output would have greater implications on post-exercise hyperemia, then post-exercise hyperemia would have been expected to increase continuously following exercise durations >90 s. In a previous study, we observed similar characteristics in the relationship of local blood supply and exercise intensity [
19]. Among local mechanisms, the muscle pump is neglectable as an explanation of our results as it stops with cessation of exercise, which leads to an abrupt drop in muscle blood flow [
13].
Local Vasodilation remains as the most plausible determinant for the hyperemia and increased post-exercise oxygen availability following cycling exercise ≥90 s at 80 %
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \). Factors that trigger vasodilation are various. Among those factors, chemically and flow mediated vasodilation have to be highlighted when discussing exercise induced vasodilation [
40]. In particular, adenosine, acetylcholine, and shear stress are presumed to trigger exercise-induced effects on vasodilation [
41,
42]. It can be assumed that blood flow during exercise and prolonged vasodilation of small vessels in the post-exercise phase are positively related. It is likely that this effect was more pronounced (in our study) following exercise durations ≥90 s compared to shorter exercise bouts. And again, because we did not measure vasodilating substances or cardiac output and blood flow velocity directly, the implications of the above-mentioned mechanisms for our study results remain speculation-based.
We have shown that oxygen availability following cycling exercise is dependent on exercise duration. Our results demonstrate that this dependency is influenced by the relative aerobic fitness (i.e., GET), expressed as a percentage of \( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \). In the GET60+ group, 60 s of exercise were sufficient to evoke a post-exercise overshoot of ΔO2Hb, which was equal to the overshoots following longer exercise bouts. In the GET60− group, 90 s of exercise were needed to obtain those high values. Hence, adjustments in local vasodilation are possibly faster in subjects showing high relative aerobic power. This means that in subjects with higher aerobic fitness, exercise durations of 60 s are sufficient to evoke an increased post-exercise hyperemia and increased oxygen availability.
Advantages of improved post-exercise O2 availability
There are several advantages of improved local blood and oxygen supply, as it has been observed following exercise durations >90 s at 80 %
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \). First, gas exchange is facilitated when the functional cross-sectional area of capillaries is increased [
9]. Moreover, muscle reoxygenation has been shown to improve with endurance training [
43]. A higher local muscle perfusion is associated with enhanced oxidative metabolism, such as fatty acid oxidation [
44‐
46]. Romijn et al. [
44] showed reduced fatty acid mobilization during strenuous exercise (85 %
\( \overset{.}{\mathrm{V}}{\mathrm{O}}_{2\mathrm{peak}} \)), as well as a strongly increased post-exercise plasma fatty-acid availability following strenuous exercise [
44,
45]. This increased post-exercise, plasma fatty-acid availability may be one explanation for improvements in fatty-acid oxidation capacity as long-term metabolic adaptations that have been observed following HIT in women [
3]. Kimber et al. [
47] found a significant dependency of post-exercise fatty-acid oxidation on fatty-acid availability. Exercise that increases fatty-acid availability is, therefore, likely to enhance fatty-acid oxidation after or between exercise bouts. The improved post-exercise blood supply during recovery or, when active recovery is applied, during the relief interval could therefore augment fatty-acid oxidation in particular. However, we reiterate that these suggestions are speculative because we did not measure fatty-acid oxidation.
Study limitations
When using NIRS, the influence of adipose tissue thickness (ATT) has to be considered [
34]. If skin and subcutaneous tissue thickness is near to or exceeds the penetration depth of the emitted photons, muscular effects will be blunted. Skin and subcutaneous thickness was approximately 3.75 mm (7.5 ± 3.1 mm skinfold thickness, measured with a caliper). Because penetration depth has been shown to be approximately half of the optode distance [
22,
34], it can be concluded that NIRS signal was capable for measuring muscle oxidation in our study as the inter-optode distance was 3.5 cm, enabling a penetration depth of 1.75 cm.
Another point to mention is the 5-min recovery period in-between exercise bouts, which was too short to enable a full recovery of ΔO
2Hb and ΔTHb. It therefore has to be considered, that prior exercise bouts influenced the subsequent exercise bouts [
31] First, this baseline drift in ΔO
2Hb and ΔTHb did not show any relationship to exercise duration. Second, we randomized the order of the exercise bouts and placed an intensive warm-up in front of the first experimental exercise bout. Consequently, this issue should not have influenced our results.
We recruited only male subjects to reduce variability in subject characteristics and because of the minor adipose tissue thickness in males relative to that in females, which apparently affects NIRS signals [
34]. Hence, the relevance of our results on females is limited.
Work intervals within interval-training regimens are described by interval duration and interval intensity. We investigated the influence of exercise intensity on local muscle O
2 availability and blood supply in an earlier study [
19]. The present study aimed to examine the effect of duration, using isolated, “interval-training-associated” exercise bouts. But besides the work interval, several other variables in interval-training prescriptions, such as intensity and duration of the recovery interval and number of repetitions, should be noted. The latter is of particular importance. Because HIT consists of repeated sets of transitions from high to low intensity, work-interval durations <90 s also are potentially effective to increase post-exercise oxygen availability due to a cumulative effect. This “repeated bout” effect could occur when low intervals are too short to provide an adequate recovery. It is therefore conceivable that effects of HIT on local oxygen availability could be similar across different exercise durations, for example when the work-to-rest ratio is kept constant. It has to be highlighted that all intervals in our study had been carried out with the same exercise intensity. Usually, work intervals are prescribed with higher intensity during “short-interval” HIT-regiments [
1]. As a consequence, the low interval has to be extended to provide sufficient recovery to enable the completion of the following work interval. Consequently, the relationship between interval intensity, duration and recovery has to be focused in future research.
Finally, the post-exercise O2 availability and blood supply are two of several determinants causing aerobic adaptations. Investigation of other determinants may require other, optimal “work-interval” durations.
Competing interests
The authors declare that they have no competing interests.
Authors’ contribution
FS was responsible for the present study’s concept and design, as well as for data acquisition, analysis, and interpretation. Furthermore, he was largely responsible for drafting of the manuscript. CvonO collaborated substantially with data acquisition. FKP was deeply involved in drafting and reviewing the manuscript. TS contributed during the preparation of the study’s concept and design and reviewed the manuscript. RO supervised the entire study and manuscript processing, was responsible for the ethical proposal, and gave the final approval for submission of the manuscript.
Stöcker, Fabian: Research associate and lecturer. Main topic: Exercise physiology.
Von Oldershausen, Christoph: Master student.
Paternoster, Florian Kurt: Research associate and lecturer. Main topic: Biomechanics.
Schulz, Thorsten: Phd, Senior Researcher and lecturer. Main topic: Endocrinology.
Oberhoffer, Renate: Professor, MD. Main topic: Cardiology, Pediatrics.