Journal Logo


Improvements in Maximal Oxygen Uptake After Sprint-Interval Training Coincide with Increases in Central Hemodynamic Factors


Author Information
Medicine & Science in Sports & Exercise: June 2022 - Volume 54 - Issue 6 - p 944-952
doi: 10.1249/MSS.0000000000002872


As reported already 70 yr ago (1), there is a tight relationship between total blood volume (BV) and maximal oxygen uptake (V̇O2max) (2–4). Both blood and plasma volume (PV) respond rapidly to training, with PV expansion accounting for almost all changes in BV during the first 2–3 wk after training onset (5,6). Later, red blood cell volume also increases until it reaches equilibrium with PV, resulting in a restored hematocrit (Htc) equivalent to pretraining levels (4). Proof-of-concept studies have shown that when BV is manipulated by phlebotomy or artificial BV expansion, V̇O2max is directly affected by altered cardiac diastolic filling, stroke volume (SV), and cardiac output (˙Q) (2,7–9). Thus, collectively, there is strong evidence that improvements in V̇O2max from traditional endurance training (TET) are strongly, but not exclusively, mediated by changes in central hemodynamic factors.

In contrast to the well-characterized adaptations contributing to the improvements in V̇O2max after TET, and despite the fact that both (sub-)maximal high-intensity interval training (HIIT) and supramaximal sprint-interval training (SIT) elicit similar or greater improvements in V̇O2max compared with TET (10,11), little is known about the mechanisms behind the observed increases in V̇O2max after interval training. It has been hypothesized that after TET, an acute reduction in PV during exercise stimulates mechanisms that include an increase in albumin synthesis and albumin content, which in turn leads to expanded PV and decreased Htc, that is, increased BV (12). We have recently reported that SIT leads to a pronounced acute reduction in PV (13), suggesting that a similar mechanism may contribute to the SIT-induced increase in V̇O2max. Nevertheless, central hemodynamic factors such as BV and total hemoglobin mass (tHb) have been somewhat overlooked in previous SIT studies.

In contrast to the sparse literature investigating hemodynamic adaptations to interval training, peripheral processes involved in oxygen utilization, such as mitochondrial biogenesis (14), promotion of skeletal muscle mitochondrial content (15–17), function (18,19), and capillarization (20), have been more thoroughly examined. Consequently, SIT-induced improvements in V̇O2max have been suggested to be mediated by peripheral adaptations, with little or no contribution from central hemodynamic factors (21–23). Although the ability of working skeletal muscle to extract oxygen is undoubtedly important (24–26), we consider it unlikely that improvements in V̇O2max are solely due to peripheral adaptations, given that the respiratory capacity of skeletal muscle exceeds maximal oxygen delivery (27) and central hemodynamic factors such as BV and ˙Q play a central role in explaining exercise-induced improvements in V̇O2max after TET (1,4,7,8,28). A more comprehensive investigation of the underlying biological processes responsible for the improvement in V̇O2max after interval training therefore seems warranted. This would improve our understanding of the key stimuli and mechanisms behind the overall central hemodynamic adaptations and aid in the design of time-efficient training protocols to maximize increases in V̇O2max.

Accordingly, the aim of the current study was to explore the effects of 6 wk of low-volume SIT on V̇O2max and canonical central hemodynamic factors. We hypothesized that 6 wk of SIT would results in a marked improvement in V̇O2max in tandem with improved oxygen delivery capacity, through expanded BV, increased tHb, and improved maximal ˙QQmax).


Ethical approval

All subjects were informed verbally and in writing about the study before providing written informed consent to participate. The study was approved by the Swedish Ethical Review Authority (reference 2019-0449; approved August 26, 2019).


None of the subjects regularly participated in any form of interval training or other structured training programs. During the study intervention, all subjects were asked to refrain from any other training that might affect the outcome of the study. Of the 33 participants recruited, 29 completed the study protocol (13 women and 16 men (baseline V̇O2max, 41.0 ± 7.8 mL·kg−1·min−1; age, 27 ± 5 yr; height, 175.3 ± 8.1 cm; body mass, 72.5 ± 12.0 kg; body mass index, 23.5 ± 2.5 kg·m−2). Three subjects dropped out because of illness (cold or flu-like symptoms), and one subject dropped out because of lack of motivation. Of the 29 subjects included, 1 subject was unable to perform the V̇O2max and ˙Qmax test after the intervention because of flu-like symptoms, 1 subject did not perform the ˙Qmax test after the intervention because he became ill before the test, and we were unable to place a peripheral venous catheter in 1 subject at the time of tHb measurement.

Training intervention

The training intervention consisted of three training sessions per week for 6 wk. Each session lasted 10 min and consisted of three 30-s all-out sprints on a mechanically braked cycle ergometer (Monark 894E, Varberg, Sweden) against a braking force equivalent to 7.5% of body weight (Fig. 1). Sprints were separated by 2 min of unloaded cycling, referred to as “rest” in Figure 1. Before the first interval, subjects completed a short warm-up (2.5 min of unloaded cycling). The third interval was followed by 2 min of unloaded cycling, which served as a cool-down. Subjects were instructed to pedal as fast as possible against the inertial resistance of the cycle ergometer. When maximal cadence was reached, braking force was applied manually. Subjects were strongly encouraged to pedal as fast as possible during each 30-s interval. Power output was measured during each training session. Peak power corresponded to the highest 1-s average during each bout, and average power included the entire 30-s bout.

Study overview.

Preintervention and postintervention measurements

Preintervention and postintervention tests were administered the week before and the week after the intervention. The test battery included the following: (i) determination of tHb, (ii) an incremental cycling test to determine V̇O2max, and (iii) an incremental cycling test to determine ˙Qmax (Fig. 1).

The optimized carbon monoxide (CO) rebreathing method was used to determine tHb. Briefly, subjects rested for 15 min before a venous blood sample was drawn from the median cubital vein. The sample was immediately analyzed for baseline carboxyhemoglobin (%HbCO), hemoglobin concentration [Hb], and Hct. End-tidal CO was measured at baseline and after the rebreathing procedure using a CO gas analyzer (Dräger, PAC 700, Lübeck, Germany). Throughout rebreathing, the same gas analyzer was used to check for CO leaks. Subjects breathed a gas mixture of pure CO (0.8 mL·kg−1) and medical oxygen (AGA, Stockholm, Sweden) for 2 min before disconnecting from the spirometer (Blood tec, GmbH, Germany). Two blood samples (1 mL) were collected, one before rebreathing and one 7 min after administration of CO. The samples were then analyzed for %HbCO (ABL 800; Radiometer A/S, Copenhagen, Denmark). tHb was calculated as described previously (29). The coefficient of variation for this method in our laboratory is 1.6%, which is in agreement with previous publications (30,31).

To determine V̇O2max and maximal workload (Wmax), subjects performed an incremental cycling test to volitional fatigue on an electronically braked ergometer (Lode, Groningen, the Netherlands) with identical protocols before and after the intervention. Using an online gas collection system (Vmax Encore 229, Carefusion, CA), O2 and CO2 concentrations were measured continuously and recorded as breath-by-breath values. Subjects began the test at 50 W for 5 min as a warm-up. Subsequently, resistance was increased by 1 W every 3 s, corresponding to an increase of 20 W·min−1, until subjects reached volitional fatigue. V̇O2max was considered to be the highest 20-s value achieved during the test. Criteria for the test were either a plateau in V̇O2 or RER >1.15, a heart rate within 10 beats of age-predicted maximum, and volitional exhaustion.

˙Qmax was determined using a rebreathing technique in which ˙Qmax is derived from pulmonary uptake of the inert gas nitrous oxide (N2O). This assumes that pulmonary uptake of blood-soluble gas is proportional to pulmonary blood flow (32). This method has been shown to be reliable and is widely used (33); however, it is also known to underestimate absolute ˙Qmax values (34). During the maximal incremental test, subjects inhaled a gas mixture containing 5% blood-soluble N2O, 1% insoluble sulfur hexafluoride (SF6), and 94% O2. The gas mixture was mixed with ambient air before the start of the rebreathing protocol. When subjects reached maximal workload (as determined by the V̇O2max assessment), they were switched from ambient air breathing to a closed-circuit system in which they inhaled the gas mixture while photoacoustic gas analyzers continuously quantified closed-circuit gas concentrations (Innovision, Odense, Denmark). Pulmonary N2O uptake was determined by the decrease in N2O concentration over three consecutive exhalations after complete mixing between the remaining pulmonary air and the gas in the rebreathing bag, as determined by a stable gas fraction of the blood insoluble gas (SF6). Subjects were instructed on how to perform rebreathing, and each subject had two trial runs of the rebreathing protocol before the actual measurement.

Estimation of O2 delivery (QaO2) was calculated as the product of ˙Qmax times arterial O2 content (CaO2), which in turn was calculated as the product of hemoglobin concentration, arterial oxygen saturation (SaO2), and the physiological O2 binding coefficient of hemoglobin (1.34 mL·g−1). Because SaO2 was not measured during the V̇O2max tests, we have assumed that there was no change in SaO2 between preintervention and postintervention.


All data are presented as mean and SD unless otherwise stated. Training effects, that is, comparison of preintervention and postintervention data, were derived using mixed linear models with time as a fixed effect and subject/time as a random effect to account for the repeated-measures design. The Linear and Nonlinear Mixed Effects Models (nlme) library in R version 3.5.5 was used for this purpose. Power output variables from the training sessions were analyzed using a mixed linear model with training session–bout as the fixed effect and subject/session as the random effect. Here, session denotes the training sessions from 1 (first) to 18 (last training session), and bout denotes the sprint intervals 1, 2, and 3 in each session. All dependent variables were treated as independent hypothesis tests, and the correction for multiple hypothesis tests was calculated using the Benjamini–Hochberg false discovery rate, where a false discovery rate of 1% was considered significant.

The study was primarily designed to detect changes in V̇O2max and PV based on the following power and effect-size calculations: To detect a change in V̇O2max of 300 mL or 8% with an interindividual variance of 10% in response and a power of 0.8 and P < 0.01, 22 independent observations were sufficient. For PV, with a variance of 18%, 25 subjects were required to detect a change of >7%, and 30 subjects for a change of 9% with a power of 0.8 and P < 0.05. Based on this, 30 subjects were included.

Correlations between change scores in the main outcome variable (V̇O2max), potential moderators (tHb, BV, Wmax, and ˙Qmax), and baseline characteristics (sex, body surface area (BSA), and baseline V̇O2max) were explored using principal component analysis and then modeled with mixed linear models, where the change score of V̇O2max served as the dependent variable and was first analyzed with the potential moderators in univariate models and then controlled for baseline characteristics as fixed effects. Principal component analysis was performed using Factominer, and mixed linear models were performed using LME.


At baseline, V̇O2max was 3.0 ± 0.8 L·min−1 and increased by 10.3% to 3.3 ± 0.9 L·min−1 (P < 0.001) after 6 wk of training (Fig. 2). Wmax was 235.6 ± 45.0 W at baseline and increased by 11.3% (26.6 ± 12.2 W; P < 0.001). SV and ˙Qmax were 65 ± 13 mL and 12.5 ± 2.5 L·min−1 at baseline (Fig. 2) and increased by 9.0% (5.3 ± 5.1 mL) and 8.5% (1.0 ± 0.9 L·min−1), respectively (P < 0.001; Fig. 2). QaO2 increased by 5.5% from 2.4 ± 0.7 L·min−1 at baseline to 2.5 ± 0.6L·min−1 after the intervention (P = 0.018). Maximum heart rate (HRmax) did not change significantly from baseline values of 193 ± 7 bpm. PV increased by 8.1% (276 ± 234 mL) from 3374 ± 396 mL at baseline (P < 0.001). BV and tHb were 5512 ± 788 mL and 706 ± 165 g, respectively, at baseline and increased by 6.8% (382 ± 325 mL) and 5.7% (42 ± 41 g; P < 0.001). [Hb] and Htc decreased by 2.6% (P < 0.001) and 2.3% (P = 0.002), respectively (Fig. 2). Body mass did not change from baseline (72.5 ± 12.0 vs 73.1 ± 11.0 kg). Mean and SD presented in a table format can be found in Supplemental Table 1, Supplemental Digital Content 1,

Group mean and individual data with directional changes in W max, V̇O2max, SV, ˙Q max, PV, BV, tHb, [Hb], and Htc. ****P < 0.0001, ***P < 0.001, **P < 0.01.

At baseline, the average power output was 5.79 ± 0.92 W·kg−1 during bout 1. In bouts 2 and 3, it decreased to 4.49 ± 0.86 and 3.94 ± 0.78 W·kg−1, representing a decrease of 22% and 31%, respectively. Average power in bout 1 remained stable throughout the training sessions and was 5.82 ± 0.99 W·kg−1 in the last training session (P > 0.05). The average power in bouts 2 and 3 increased by 0.013 ± 0.002 W·kg−1 per session (P < 0.001) and was 4.79 ± 0.89 and 4.42 ± 0.79 W·kg−1 during the last training session, respectively (Fig. 3).

Average power output in watts per kilogram of body weight (W·kg−1) during each training session, grouped by sprint bouts throughout the 6 wk of training. Values are presented as mean (O) and SD (whiskers). Significant changes in each bout across the 18 training sessions are denoted. **P < 0.001.

Potential modifiers of changes in V̇O2max were first explored using a principal component analysis (Fig. 4), in which the change score of V̇O2max was examined along with changes in measures of tHb, BV and PV, ˙Qmax, and Wmax, as well as baseline BSA and baseline V̇O2max. This yielded a covariance of ~48% using the first two principal components, with baseline V̇O2max, change in tHb, and baseline BSA being the variables with the highest covariance with changes in V̇O2max. Based on the results of the PCA, it was decided to test the potential modifiers of ∆ V̇O2max while controlling for baseline V̇O2max and sex. A mixed linear model analysis was then performed testing the potential moderators as independent variables with the change in V̇O2max as the dependent outcome variable. To allow for effect-size comparisons between variables, the independent variables were scaled to z-scores (Table 1).

Principal component analysis and biplot of ∆ V̇O2max, change scores in hemodynamic parameters, and baseline characteristics. The size of each individual observation denotes the magnitude of ∆ V̇O2max where a larger dot indicates a more substantial response. Fifty-two percent of the total variance across all variables was retained using two principal components, indicative of a high covariance across all change scores. The variable loadings indicate that baseline V̇O2, BSA, and sex along with changes in hemoglobin mass and BV are the most important moderators of ∆ V̇O2max.
TABLE 1 - Beta coefficients with corresponding confidence interval and P values of linear models with ∆ V̇O2max (in milliliters) as the outcome variable, adjusting for sex and baseline V̇O2max.
Variable Name Coefficient (mL/SD) 95% Confidence Interval P
Δ ˙Q max 60.1 −7 to 128 0.78
Δ BV 438 379 to 498 0.028
Δ PV 302 240 to 364 0.13
Δ tHb 421 363 to 480 0.031
Δ W max −64.9 −131 to 1 0.76
Δ SV 48.2 −19 to 115 0.82
Δ HRmax −15.3 −86 to 56 0.95
Δ Hb −17.5 −83 to 48 0.93
Δ Hct −18.0 −82 to 46 0.93
BSA 709 623 to 795 0.015
Coefficients are standardized to z-score for ease of comparison, and each coefficient thus denotes the change in ∆ V̇O2max in milliliters per SD.

Based on a low model contribution and the fact that there was no significant correlation between sex and change in V̇O2max (P = 0.074 univariate and P = 0.82 in the model with BSA and baseline V̇O2max), sex was removed and the final model was based on baseline V̇O2max, BSA, and ∆ tHb. This model had better performance than a model with sex, baseline V̇O2max, and ∆ tHb (Akaike information criterion [AIC] of 2.9 and adjusted R2 of 0.44 vs AIC of 9.8 and R2 of 0.29) and was also superior to a model with sex added in addition to BSA, baseline V̇O2max, and ∆ tHb (AIC of 4.9, R2 of 0.44; Table 2).

TABLE 2 - Beta coefficients with corresponding confidence interval and P values of the final model with ∆ V̇O2max (in milliliters) as the outcome variable.
Variable Name Coefficient (mL/SD) 95% Confidence Interval P
(Intercept) 11.5 −36 to 59 0.94
∆ tHb 359 308 to 410 0.035
BSA 654 588 to 719 0.004
Baseline V̇O2max −418 −482 to −354 0.049
Coefficients are standardized to z-score for ease of comparison, and each coefficient thus denotes the change in ∆ V̇O2max in milliliters per SD. The model had an R2 of 0.44, which corresponds to an explanatory value of 44% of the total variance of ∆ V̇O2max.

The effect sizes in the final model were an increase of 2.5 ± 1.1 mL V̇O2 per gram increase in tHb, 934 ± 293 mL V̇O2 per m2 increase in BSA at baseline, and a decrease of −150 ± 73 mL V̇O2 per liter V̇O2max at baseline. The model had an R2 of 0.44, which corresponds to an explanatory value of 44% of the total variance of ∆ V̇O2max. Adding the tHb change score to the baseline parameters improved model performance from AIC of 6.3 and R2 of 0.33 to AIC of 3.9 and R2 of 0.44 with a likelihood ratio of 5.4 (P = 0.06; Fig. 5).

Model performance. A linear model considering BSA, baseline V̇O2max, and change score in tHb achieved a R 2 of 0.44 corresponding to 44% of the total variance in ∆ V̇O2max. The effect sizes in the final model were an increase of 2.5 ± 1.1 mL V̇O2 per gram increase in tHb, 934 ± 293 mL V̇O2 per m2 increase in BSA at baseline, and a decrease of −150 ± 73 mL V̇O2 per liter V̇O2max at baseline. Standardized effect sizes were an increase of 359 ± 161 mL V̇O2 per SD increase in tHb, 654 ± 206 mL V̇O2 per SD increase in BSA at baseline, and a decrease of −418 ± 201 mL V̇O2 per SD increase in V̇O2max at baseline.


The current study investigated whether BV and other central hemodynamic factors were affected by 6 wk of low-volume SIT in parallel with the expected increase in V̇O2max. Indeed, robust improvements in V̇O2max were observed with concomitant increases in SV, ˙Qmax, and total BV, but the increase in V̇O2max was larger than the changes in these central hemodynamic components, indicating that peripheral adaptations were also of importance, as has been shown in previous studies (15–17). Although tHb also increased, BV expansion was driven by increased PV, which was reflected in a decrease in Htc and [Hb].

The observed increase in V̇O2max is consistent with previous publications in which comparable training was performed. Similar increases have been demonstrated after 4 wk of repeated 15- to 30-s bouts of SIT (35), after 6 wk of combined HIIT and SIT (36), and after a variety of other SIT protocols (10,37). An interesting comparison of the elicited changes in V̇O2max in the present study with some of the aforementioned reports is that, despite the lower total number of bouts of sprints in the current training protocol, the increase in V̇O2max between interventions was similar. This is consistent with previous studies reporting that 2–3 supramaximal intervals in a session are sufficient to produce robust improvements in V̇O2max in previously untrained subjects (37–39).

A key finding of the current study was that low-volume SIT resulted in an increase in V̇O2max with concomitant increases in BV, PV, and tHb. This is consistent with previous studies reporting central hemodynamic adaptations associated with increased V̇O2max after TET (4,28). The results are also corroborated by studies using a higher volume of interval training and showing that vascular volume and ˙Qmax increased after 6 and 12 wk of training in parallel with the increase in V̇O2max (40,41). This suggests that similar central hemodynamic factors mediate the increase in V̇O2max after SIT and TET. The observed relationship between central hemodynamic factors and V̇O2max supports this assumption, especially given the previously verified causal relationship between ˙Q and V̇O2max (1,2,7).

The observed increase in BV was likely driven by the expanding PV as reflected by the decrease in [Hb] and Htc. This hemodilution reflects the typical kinetics of exercise-induced hypervolemia, in which the PV rapidly expands followed by an increase in erythrocyte volume until equilibrium is reached and Htc is restored (4,40,42,43). The decreased Htc and [Hb] in our data show that this equilibrium was not reached after 6 wk SIT. The increase in BV, PV, and tHb is in contrast to a previous study that reported no effect on the aforementioned variables after a 2-wk HIIT intervention (23). Although it is not surprising that 2 wk of training did not elicit an increase in tHb, it was somewhat unexpected that PV remained unchanged, as this has been demonstrated after a wide range of training protocols with varying intensities and durations (40,42).

The importance of increased BV relates to the resulting increased venous return, SV, and ˙Q. Consistent with the findings reported here, studies using both short (≤4 wk) (44) and longer (>6 wk) training interventions have reported pronounced effects on ˙Q and SV (36,37,41). However, in addition to the aforementioned study with unchanged PV, another study reported unchanged ˙Qmax after cycling-based SIT (22,23). The reason for this discrepancy is unclear, but most likely multifactorial, including differences in methodology, exercise intensity, intervention time, baseline fitness status of subjects, and outcome variables assessed. Collectively, however, it seems that interval intensity, rather than total work volume, is the key modifier of V̇O2max. This is supported by previous work reporting that adding training sessions at the expense of exercise intensity may reduce the training effect (45) and that exercise with a relatively high workload for a longer duration has a smaller effect on muscle volume expansion and PV drop (46). In this regard, it should be noted that in the two studies in which no increase in central hemodynamic factors was found, the intensity in each sprint bout was lower than in the present study. Based on our recent publication (13), it seems that training with a high workload for a shorter duration actually has a greater effect on muscle volume expansion and PV drop than a lower workload for a longer duration. It is plausible that this leads to a greater hypovolemic stimulus, which has greater effects on vascular volumes. This is also supported by reports showing that glycogenolysis (and thus metabolite accumulation) is maximally activated during the first 15 s of the first exercise bout (47). Although more detailed information is needed to explore these plausible mechanisms and how they relate to an increase in albumin synthesis, the present data support the hypothesis of a central hemodynamic adaptation through an increase in BV with SIT.

The contribution of the baseline characteristics of the subjects and the change values of the physiological parameters were analyzed in relation to the improvement of V̇O2max. Based on the initial exploratory PCA analysis, baseline V̇O2max, BSA, and sex were examined as potential moderators of the training response. In the final and best-performing model, BSA made a significant contribution with ∆ V̇O2max increasing by 600 mL per SD increase in BSA, or 934 mL V̇O2·m−2 increase in BSA. The model with BSA outperformed the models with sex as a moderator, and a model that included sex in addition to the BSA was not significantly better at predicting ∆ V̇O2max. Based on the inherited mutual correlation of BSA and sex, the present data set is arguably too small to infer possible independent effects, but nevertheless, it suggests that BSA is a more important moderator of the training effect of this type of training intervention than sex. This is consistent with earlier reports showing that there is no difference in V̇O2max gain between men and woman when subjected to interval-training regimes (11). However, it cannot be ruled out that there is residual variance in ∆ V̇O2max related to sex that could have been quantified in a larger cohort than the present study or that scaling exaggerated the BSA to ∆ V̇O2max relationship. Nevertheless, it is tempting to speculate that the mechanism underlying the positive relationship between BSA and ∆ V̇O2max is due to greater metabolic and hemodynamic stress induced by exercise with greater muscle mass.

Baseline V̇O2max was negatively correlated with ∆ V̇O2max, with ∆ V̇O2max being 150 mL lower for each liter of baseline V̇O2max. In agreement with previous studies, the influence of baseline V̇O2max on the model, and therefore the predictability of the exercise response by baseline power, is rather small. It is also possible that there is a regression to the mean phenomena with repeated measures in this correlation, which cannot be quantified in the present study because of the absence of a nonexercising control group. It must therefore be considered uncertain whether this is a true biological variation in which individuals with lower baseline V̇O2max are more responsive to this type of training than individuals with higher aerobic capacity. Of the physiological parameters examined as moderators of ∆ V̇O2max, the change in tHb made the strongest contribution. The final model based on baseline BSA, baseline V̇O2max, and change in tHb had an R2 of 0.44, meaning it explained 44% of the total variance in ∆ V̇O2max. Bivariate correlations can be found in Supplemental Table 2, Supplemental Digital Content 2,

Initially, studies examining interval training focused on processes involved in peripheral oxygen handling, such as mitochondrial biogenesis (14) and capillarization (20). Because some previous studies reported little or no contribution from central hemodynamic factors (21–23), it was assumed that the increase in V̇O2max with SIT was mediated mainly by peripheral adaptations. However, if SIT did not increase ˙Qmax, a higher V̇O2max can only be explained by improved blood flow coordination or increased oxygen extraction. This seems unlikely given the limitation in oxygen delivery during maximal work and the excess capacity of mitochondrial respiration (27). Moreover, given the already low venous Hb saturation at peak V̇O2 in the untrained state, it is questionable whether peripheral remodeling alone can explain the observed increases in V̇O2max seen with SIT. Although ˙Qmax and tHb are the major determinants of aerobic capacity during exercise using large muscle groups (4,48), the ability to achieve V̇O2max also depends on efficient peripheral mechanisms for oxygen uptake and utilization (49), as demonstrated by unilateral training models (48). This is also supported by the improvements in sprint performance over the course of the training intervention. No change in Wmax was observed, but a successive increase in average power output during each sprint and an improvement in recovery between each sprint was apparent, that is, the power output of the second and third sprints approached the power output of the first bout over the course of the training intervention (Fig. 3). This is likely due to peripheral adaptations, as our previous work suggests that power output during bouts 2 and 3 reflects endurance capacity, that is, the ability to sustain higher power output for longer periods of time (13). Thus, our findings, together with previously published literature, support the concept of symmorphosis, that is, that both central and peripheral adaptations are involved in the improvements in V̇O2max after a SIT intervention.

Because central and peripheral adaptations are regulated by different mechanisms, it is likely that the design of interval training (e.g., intensity vs duration/volume) will affect these two processes to different degrees, as will the kinetics at which these adaptations occur. To fine-tune SIT programs to the desired outcome, future studies should attempt to tease out the independent and combined effects of the various program design factors on these different adaptations. Furthermore, this spurs future research to explore plausible regulatory mechanisms that may explain how contrasting exercise modalities such as TET and SIT may lead to similar improvements in central hemodynamic factors despite the large differences in exercise stimuli. As mentioned earlier, one of the limitations in the present study is the underestimation of ˙Qmax in our measurements. The limitations of the ˙Qmax measurements are also evident in the PCA (Fig. 4), where, unlike tHb, the variables derived from this measurement were not associated with the changes in V̇O2max, even though the variables are surrogates for oxygen transport. One proposed explanation for the measurement error is the recirculation of nitrous oxide into the pulmonary system during the rebreathing procedure (36). However, the reproducibility and reliability of the method have shown good results (33,50). Nevertheless, one should be cautious when interpreting the absolute values generated by the method, especially if one is attempting to derive other physiological parameters from the existing data on ˙Qmax.


In conclusion, we report that 6 wk of SIT elicits robust improvements in V̇O2max and that increases in central hemodynamic factors together with BSA explain a significant proportion of the variance in improvements in V̇O2max. Thus, in contrast to previous reports suggesting that the training effects on V̇O2max observed with SIT are mainly explained by peripheral adaptations, our data strongly suggest that adaptations in central hemodynamic factors also play a significant role.

The authors thank Nestor Lögdal for his contribution to this study. This work was supported by grants from the Swedish Research Council for Sport Science. The results of this study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The results of the present study do not constitute endorsement by the American College of Sports Medicine.

Conflict of Interest: The authors have no conflict of interest to declare.


1. Kjellberg SR, Rudhe U, Sjöstrand T. Increase of the amount of hemoglobin and blood volume in connection with physical training. Acta Physiol Scand. 1949;19(2–3):146–51.
2. Celsing F, Svedenhag J, Pihlstedt P, Ekblom B. Effects of anaemia and stepwise-induced polycythaemia on maximal aerobic power in individuals with high and low haemoglobin concentrations. Acta Physiol Scand. 1987;129(1):47–54.
3. Convertino VA. Blood volume response to physical activity and inactivity. Am J Med Sci. 2007;334(1):72–9.
4. Sawka MN, Convertino VA, Eichner ER, Schnieder SM, Young AJ. Blood volume: importance and adaptations to exercise training, environmental stresses, and trauma/sickness. Med Sci Sports Exerc. 2000;32(2):332–48.
5. Gore CJ, Scroop GC, Marker JD, Catcheside PG. Plasma volume, osmolarity, total protein and electrolytes during treadmill running and cycle ergometer exercise. Eur J Appl Physiol Occup Physiol. 1992;65(4):302–10.
6. Fellmann N. Hormonal and plasma volume alterations following endurance exercise. A brief review. Sports Med. 1992;13(1):37–49.
7. Bonne TC, Doucende G, Flück D, et al. Phlebotomy eliminates the maximal cardiac output response to six weeks of exercise training. Am J Physiol Regul Integr Comp Physiol. 2014;306(10):R752–60.
8. Ekblom B, Goldbarg AN, Gullbring B. Response to exercise after blood loss and reinfusion. J Appl Physiol. 1972;33(2):175–80.
9. Krip B, Gledhill N, Jamnik V, Warburton D. Effect of alterations in blood volume on cardiac function during maximal exercise. Med Sci Sports Exerc. 1997;29(11):1469–76.
10. Gist NH, Fedewa MV, Dishman RK, Cureton KJ. Sprint interval training effects on aerobic capacity: a systematic review and meta-analysis. Sports Med. 2014;44(2):269–79.
11. Phillips BE, Kelly BM, Lilja M, et al. A practical and time-efficient high-intensity interval training program modifies cardio-metabolic risk factors in adults with risk factors for type II diabetes. Front Endocrinol (Lausanne). 2017;8:229.
12. Montero D, Lundby C. Regulation of red blood cell volume with exercise training. Compr Physiol. 2018;9(1):149–64.
13. Mandić M, Forsgren MF, Romu T, et al. Interval-induced metabolic perturbation determines tissue fluid shifts into skeletal muscle. Physiol Rep. 2021;9(7):e14841.
14. Little JP, Safdar A, Bishop D, Tarnopolsky MA, Gibala MJ. An acute bout of high-intensity interval training increases the nuclear abundance of PGC-1α and activates mitochondrial biogenesis in human skeletal muscle. Am J Physiol Regul Integr Comp Physiol. 2011;300(6):R1303–10.
15. Gibala MJ, Little JP, van Essen M, et al. Short-term sprint interval versus traditional endurance training: similar initial adaptations in human skeletal muscle and exercise performance. J Physiol. 2006;575(Pt 3):901–11.
16. Ma J K, Scribbans T D, Edgett B A, et al. Extremely low-volume, high-intensity interval training improves exercise capacity and increases mitochondrial protein content in human skeletal muscle. Open J Mol Integr Physiol. 2013;3(4):202–10.
17. Burgomaster KA, Hughes SC, Heigenhauser GJF, Bradwell SN, Gibala MJ. Six sessions of sprint interval training increases muscle oxidative potential and cycle endurance capacity in humans. J Appl Physiol. 2005;98(6):1985–90.
18. Vincent G, Lamon S, Gant N, et al. Changes in mitochondrial function and mitochondria associated protein expression in response to 2-weeks of high intensity interval training. Front Physiol. 2015;6:51.
19. Larsen FJ, Schiffer TA, Zinner C, et al. Mitochondrial oxygen affinity increases after sprint interval training and is related to the improvement in peak oxygen uptake. Acta Physiol (Oxf). 2020;229(3):e13463.
20. Scribbans TD, Edgett BA, Vorobej K, et al. Fibre-specific responses to endurance and low volume high intensity interval training: striking similarities in acute and chronic adaptation. PLoS One. 2014;9(6):e98119.
21. Macpherson RE, Hazell TJ, Olver TD, Paterson DH, Lemon PW. Run sprint interval training improves aerobic performance but not maximal cardiac output. Med Sci Sports Exerc. 2011;43(1):115–22.
22. Raleigh JP, Giles MD, Islam H, et al. Contribution of central and peripheral adaptations to changes in maximal oxygen uptake following 4 weeks of sprint interval training. Appl Physiol Nutr Metab. 2018;43(10):1059–68.
23. Jacobs RA, Flück D, Bonne TC, et al. Improvements in exercise performance with high-intensity interval training coincide with an increase in skeletal muscle mitochondrial content and function. J Appl Physiol. 2013;115(6):785–93.
24. Mortensen SP, Damsgaard R, Dawson EA, Secher NH, González-Alonso J. Restrictions in systemic and locomotor skeletal muscle perfusion, oxygen supply and VO2 during high-intensity whole-body exercise in humans. J Physiol. 2008;586(10):2621–35.
25. Jacobs RA, Rasmussen P, Siebenmann C, et al. Determinants of time trial performance and maximal incremental exercise in highly trained endurance athletes. J Appl Physiol (1985). 2011;111(5):1422–30.
26. Richardson RS, Noyszewski EA, Kendrick KF, Leigh JS, Wagner PD. Myoglobin O2 desaturation during exercise. Evidence of limited O2 transport. J Clin Invest. 1995;96(4):1916–26.
27. Boushel R, Gnaiger E, Calbet JAL, et al. Muscle mitochondrial capacity exceeds maximal oxygen delivery in humans. Mitochondrion. 2011;11(2):303–7.
28. Convertino VA. Blood volume: its adaptation to endurance training. Med Sci Sports Exerc. 1991;23(12):1338–48.
29. Prommer N, Schmidt W. Loss of CO from the intravascular bed and its impact on the optimised CO-rebreathing method. Eur J Appl Physiol. 2007;100(4):383–91.
30. Gore CJ, Hopkins WG, Burge CM. Errors of measurement for blood volume parameters: a meta-analysis. J Appl Physiol. 2005;99(5):1745–58.
31. Steiner T, Wehrlin JP. Comparability of haemoglobin mass measured with different carbon monoxide-based rebreathing procedures and calculations. Scand J Clin Lab Invest. 2011;71(1):19–29.
32. Krogh A, Lindhard J. Measurements of the blood flow through the lungs of man. Skand Arch Physiol. 1912;27(2):100–25.
33. Fontana P, Boutellier U, Toigo M. Reliability of measurements with Innocor during exercise. Int J Sports Med. 2009;30(10):747–53.
34. Siebenmann C, Rasmussen P, Sørensen H, et al. Cardiac output during exercise: a comparison of four methods. Scand J Med Sci Sports. 2015;25(1):e20–7.
35. Zelt JGE, Hankinson PB, Foster WS, et al. Reducing the volume of sprint interval training does not diminish maximal and submaximal performance gains in healthy men. Eur J Appl Physiol. 2014;114(11):2427–36.
36. Astorino TA, Edmunds RM, Clark A, et al. High-intensity interval training increases cardiac output and VO2max. Med Sci Sports Exerc. 2017;49(2):265–73.
37. Bostad W, Valentino SE, McCarthy DG, et al. Twelve weeks of sprint interval training increases peak cardiac output in previously untrained individuals. Eur J Appl Physiol. 2021;121(9):2449–58.
38. Metcalfe RS, Babraj JA, Fawkner SG, Vollaard NBJ. Towards the minimal amount of exercise for improving metabolic health: beneficial effects of reduced-exertion high-intensity interval training. Eur J Appl Physiol. 2012;112(7):2767–75.
39. Gillen JB, Martin BJ, MacInnis MJ, Skelly LE, Tarnopolsky MA, Gibala MJ. Twelve weeks of Sprint interval training improves indices of cardiometabolic health similar to traditional endurance training despite a five-fold lower exercise volume and time commitment. PLoS One. 2016;11(4):e0154075.
40. Warburton DE, Haykowsky MJ, Quinney HA, et al. Blood volume expansion and cardiorespiratory function: effects of training modality. Med Sci Sports Exerc. 2004;36(6):991–1000.
41. Daussin FN, Ponsot E, Dufour SP, et al. Improvement of VO2max by cardiac output and oxygen extraction adaptation during intermittent versus continuous endurance training. Eur J Appl Physiol. 2007;101(3):377–83.
42. Convertino VA, Brock PJ, Keil LC, Bernauer EM, Greenleaf JE. Exercise training-induced hypervolemia: role of plasma albumin, renin, and vasopressin. J Appl Physiol Respir Environ Exerc Physiol. 1980;48(4):665–9.
43. Green HJ, Coates G, Sutton JR, Jones S. Early adaptations in gas exchange, cardiac function and haematology to prolonged exercise training in man. Europ J Appl Physiol. 1991;63(1):17–23.
44. Astorino TA, Edmunds RM, Clark A, et al. Increased cardiac output and maximal oxygen uptake in response to ten sessions of high intensity interval training. J Sports Med Phys Fitness. 2018;58(1–2):164–71.
45. Vollaard NBJ, Metcalfe RS, Williams S. Effect of number of sprints in an SIT session on change in V̇O2max: a meta-analysis. Med Sci Sports Exerc. 2017;49(6):1147–56.
46. Convertino VA, Keil LC, Bernauer EM, Greenleaf JE. Plasma volume, osmolality, vasopressin, and renin activity during graded exercise in man. J Appl Physiol Respir Environ Exerc Physiol. 1981;50(1):123–8.
47. Parolin ML, Chesley A, Matsos MP, Spriet LL, Jones NL, Heigenhauser GJ. Regulation of skeletal muscle glycogen phosphorylase and PDH during maximal intermittent exercise. Am J Physiol. 1999;277(5):E890–900.
48. Saltin B. Hemodynamic adaptations to exercise. Am J Cardiol. 1985;55(10):42D–7D.
49. Saltin B, Nazar K, Costill DL, et al. The nature of the training response; peripheral and central adaptations to one-legged exercise. Acta Physiol Scand. 1976;96(3):289–305.
50. del Torto A, Skattebo Ø, Hallén J, Capelli C. Cardiac output with modified cardio-impedance against inert gas rebreathing during sub-maximal and maximal cycling exercise in healthy and fit subjects. Eur J Appl Physiol. 2019;119(1):163–70.


Supplemental Digital Content

Copyright © 2022 by the American College of Sports Medicine