| |

Received: 9 November 2021    Revised: 18 January 2022    Accepted: 25 January 2022

DOI: 10.1111/sms.14133

REVIEW

Effects of short sprint interval training on aerobic and

anaerobic indices: A systematic review and meta-­analysis

Daniel Boullosa1,2,3   | Boris Dragutinovic4   | Joshua F. Feuerbacher4  |

Stefano Benítez-­Flores5  | Edward F. Coyle6   | Moritz Schumann4

1
Integrated Institute of Health, Federal
University of Mato Grosso do Sul, The effects of short sprint interval training (sSIT) with efforts of ≤10 s on maxi-
Campo Grande, Brazil mal oxygen consumption (V̇O2max), aerobic and anaerobic performances remain
2
College of Healthcare Sciences, James unknown. To verify the effectiveness of sSIT in physically active adults and ath-
Cook University, Townsville, Australia
3
letes, a systematic literature search was conducted according to the Preferred
Research and Development
Department, iLOAD Solutions, Campo Reporting Items for Systematic Reviews and Meta-­Analyses (PRISMA). The da-
Grande, Brazil tabases PubMed/MEDLINE, ISI Web of Science, and SPORTDiscus were system-
4
Department of Molecular and Cellular atically searched on May 9, 2020, and updated on September 14, 2021. Inclusion
Sports Medicine, German Sport
criteria were based on PICO and included healthy athletes and active adults of
University, Cologne, Germany
5
Department of Physical Education and
any sex (≤40 years), performing supervised sSIT (≤10 s of “all-­out” and non-­“all-­
Health, Higher Institute of Physical out” efforts) of at least 2 weeks, with a minimum of 6 sessions. As a comparator,
Education, University of the Republic, a non-­sSIT control group, another high-­intensity interval training (HIIT) group,
Montevideo, Uruguay
6
or a continuous training (CT) group were required. A total of 18  studies were
Human Performance Laboratory,
University of Texas at Austin, Austin, deemed eligible. The estimated SMDs based on the random-­effects model were
Texas, USA −0.56 (95% CI: −0.79, −0.33, p < 0.001) for V̇O2max, −0.43 (95% CI: −0.67, −0.20,
p < 0.001) for aerobic performance, and −0.44 (95% CI: −0.70, −0.18, p < 0.001)
Correspondence
Moritz Schumann, Department for anaerobic performance after sSIT vs. no exercise/usual training. However,
of Molecular and Cellular Sports there were no significant differences (p > 0.05) for all outcomes when comparing
Medicine, German Sport University,
Am Sportpark Müngersdorf 6, 50933
sSIT vs. HIIT/CT. Our findings indicate a very high effectiveness of sSIT protocols
Cologne, Germany. in different exercise modes (e.g., cycling, running, paddling, and punching) to im-
Email: [email protected] prove V̇O2max, aerobic, and anaerobic performances in physically active young
healthy adults and athletes.

KEYWORDS
aerobic fitness, anaerobic fitness, high-­intensity interval training, human performance, sprint
interval training

1  |  I N T RO DU CT ION with active or passive rest intervals. The main purpose of

HIIT is to complete a greater amount of work at a high-­
Traditionally, high-­intensity interval training (HIIT) con- intensity when compared to a single continuous bout at
sists of bouts of cyclic endurance exercises at intensities the same intensity until exhaustion.1 This training modal-
above the lactate threshold or critical power, interspersed ity was initially developed for endurance runners,2 but,

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided
the original work is properly cited.
© 2022 The Authors. Scandinavian Journal of Medicine & Science In Sports published by John Wiley & Sons Ltd.

|
810    wileyonlinelibrary.com/journal/sms
 Scand J Med Sci Sports. 2022;32:810–820.
BOULLOSA et al.      |  811

to date, it is commonly used in both individual and team non-­“all-­out” efforts will aid in verifying the effectiveness
sports,3 and in clinical practice,4 to enhance both aerobic of short efforts during different HIIT schemes for physical
and anaerobic fitness components.5 Although athletes fitness development. This information is very important to
typically combine HIIT with other training modalities for elucidate the chronic adaptations of sSIT when compared
fitness development, there is the consensus that it allows to other HIIT/SIT and continuous training (CT) protocols,
rapid metabolic and neuromuscular adaptations already while expanding the understanding of the loading factors
after a few brief sessions when compared to continuous (e.g., intensity and work-­to-­rest ratio) associated with the
endurance training methods.6 For this reason, HIIT has more efficient adaptations after different sSIT schemes.
become a very popular training modality, not only in rec- Thus, the aim of this systematic review with meta-­
reational and elite sport but also for health purposes in the analysis was to identify controlled (CTs) and randomized
general population. controlled trials (RCTs) that used very short efforts ≤10 s
Among the HIIT loading parameters that can be ma- (8) over a minimum of 2 weeks, which is the minimum
nipulated, intensity appears to be most important as the time required to induce stable adaptations (9), and to ver-
different HIIT modalities are directly linked to manip- ify the effects of these training regimes on measures of
ulation of internal (e.g., HR) and external (e.g., power) aerobic and anaerobic fitness and performance. To avoid
loading parameters.1,3,7 Thus, in the supramaximal zone the confounding effect of training history (i.e., to be sed-
(i.e., above the maximal oxygen consumption; V̇O2max) of entary), age (maturational factors or aging), and clinical
HIIT intensities, we may refer to the intermittent meth- conditions (e.g., obese and cardiovascular disease), we de-
ods in the classic terminology,1 which includes HIIT with cided to only include healthy physically active adults and
short intervals. Training at “all-­out” maximal effort using athletes.
repeated-­sprint training (RST) and sprint interval train-
ing (SIT) are other HIIT modalities following the classi-
fication by Buchheit and Laursen.3 However, there is no 2  |  METHODS
consensus regarding the HIIT and SIT definitions in the
literature.1 While RST (≤10  s) and SIT (≤45  s) may refer 2.1  |  Systematic literature search
to “all-­out” efforts of varying duration, HIIT with short
intervals is commonly performed at non-­“all-­out” efforts A systematic literature search was conducted according to
of short duration (≤10 s).3 For this reason, HIIT protocols the Preferred Reporting Items for Systematic Reviews and
as RST and SIT do not require identification of metabolic Meta-­Analyses (PRISMA) and was previously registered
parameters and power output for training prescription, with the international database of prospectively registered
while HIIT with short intervals requires, at least, the iden- systematic reviews in health and social care (PROSPERO:
tification of a parameter associated with aerobic power. CRD42020188226).
Recently, it has been suggested that sessions with less The databases PubMed/MEDLINE, ISI Web of Science,
and shorter (i.e., 4–­20-­s) sprints are a more time-­efficient and SPORTDiscus were systematically searched using a
HIIT modality than Wingate-­based SIT.8,9 Particularly, search string that was specifically adapted to the search re-
several SIT protocols with repeated short (≤10  s) efforts quirements for each database (see online Supplementary
(sSIT) were shown to exhibit similar aerobic and anaer- Table 1).
obic adaptations but better perceptual and enjoyment re- The search was conducted on May 9, 2020, and up-
sponses (i.e., “less pain, same gain”) than Wingate-­based dated on September 14, 2021. The literature search process
SIT.10–­13 The greater efficiency of these sSIT protocols is was performed independently by two researchers and in-
related to the fact that the highest mechanical responses cluded saving the online search, removing duplicates and
are achieved during the first seconds of sprinting bouts,9 screening titles, abstracts, and full texts. Possible conflicts
while the reduced glycolytic activity would result in less were solved by consulting a third author. In addition, a
peripheral fatigue14 because of the more reliance on the gray literature search was performed by screening Google
ATP-­PCr pathway during the first 10 s of effort15 Moreover, Scholar and the reference lists of previously identified eli-
the acute responses of different sSIT schemes have been gible full texts. A flow chart of the search process and the
described with respect to physical,10,14 physiological,10,16,17 study selection is displayed in Figure 1.
and perceptual10,12 responses. However, there are only a
few recent studies examining the physical and physiolog-
ical adaptations after a number of sessions over only a 2.2  |  Eligibility criteria
few weeks, in cycling and running sprints with promising
results.18,19 Therefore, a systematic search of longitudinal Inclusion criteria were defined based on the PICOS cri-
studies including any sSIT protocol with both “all-­out” and teria.20 The population included healthy young and
 |
812       BOULLOSA et al.

F I G U R E 1   Flow chart of the search process and the study selection

middle-­aged (18–­40 years) athletes and active adults with or maximal oxygen uptake (V̇O2max) were considered.
no restrictions in terms of sex. The intervention had to be Aerobic performance was defined as the maximal power
comprised of supervised sSIT of at least 2 weeks, with a achieved in a graded exercise test (Pmax), velocity asso-
minimum of 6  sessions. Only CTs and RCTs including ciated with V̇O2max in graded exercise tests (vV̇O2max),
a group performing sprints of ≤10  s of duration (or an completed time in graded exercise tests (GXT time),
equivalent distance-­based [m] sprint) were included. As a completed shuttles or distance in aerobic shuttle run
comparator, eligible studies needed to include a non-­sSIT tests, time trial performances, and the mean power out-
control group (either no exercise or continuing their usual put during a 3-­minute all-­out test (MPO). For anaerobic
training without sSIT) or an endurance training group performance, derived measures of Wingate performance
performing other exercise training regimens such as HIIT, (peak power output [PPO]), sprint performance, repeated
SIT, or CT. The outcomes of interest were defined as aero- sprint performance (mean sprint time), and anaerobic
bic capacity, aerobic performance, and anaerobic perfor- shuttle run performance were considered. Exclusion cri-
mance. For aerobic capacity, measures of peak (V̇O2peak) teria included language other than English and German,
BOULLOSA et al.      |  813

non-­peer-­reviewed articles, abstracts and thesis, cross-­ correlation test29 and the regression test,30 using the stan-
sectional studies assessing only acute exercise responses, dard error of the observed outcomes as predictor, were
and observational studies. used to check for funnel plot asymmetry.
Effect sizes from studies with more than two inter-
vention or control groups were combined in accordance
2.3  |  Data extraction with the recommendations of the Cochrane handbook.31
In the case of multiple measurements for the same out-
Data extraction was performed independently by two come, only one measure was included in the analysis. For
authors. The following data were extracted from each in- aerobic performance, this was based on the following hi-
cluded study: (1) the general characteristics (e.g., author(s), erarchy: Pmax, vV̇O2max, GXT time, completed shuttles
year of publication and aim of the study); (2) participants or distance in aerobic shuttle-­run tests, time trial perfor-
information (i.e., sample size, sex, training status, and mance, and mean power output (MPO). For anaerobic
age); (3) intervention data for all groups (i.e., interven- performance, it was based on the following hierarchy:
tion duration, types of interventions, and training loads); Wingate performance (PPO), repeated sprint performance
and (4) specific outcomes (e.g., measures of V̇O2max and (mean sprint time), sprint performance, and completed
PPO). If the mean and standard deviation of the respec- anaerobic shuttle runs.
tive groups were not reported, authors of primary studies When ≥3  studies were available, subgroup analyses
were contacted to provide baseline and postintervention were conducted for exercise intensity (“all-­out” vs. non-­
data. In case data were presented within a graphic and no “all-­out”). For specific justification of exclusion of indi-
additional data were provided upon request, mean and vidual studies, please refer to the online Supplementary
standard deviation were extracted using WebPlotDigitizer Table 2.
(Pacifica, California, USA, Version: 4.4).21

2.5  |  Assessment of
2.4  |  Data synthesis and analyses methodological quality

Standardized mean differences (SMD) were calculated, The risk-­of-­bias assessment for the included studies was
and an inverse variance-­weighted random-­effects model carried out independently using the PEDro scale by two
was fitted to the effect sizes (ES). Meta-­analyses were per- reviewers. The PEDro scale has previously been rated as
formed using R (3.6.2), RStudio (1.2.5033), and the meta- a valid measure of the methodological quality of rand-
for packages (version 2.4.0).22 Effect sizes were calculated omized trials.32 Studies with scores >6 were considered to
for pre-­test and post-­test control group designs using the be of “high quality,” studies with scores 4 –­ 5 were consid-
raw score standardization recommended previously23,24 ered to be of “medium quality,” and studies that scored less
Furthermore, exact sampling variance of the effect sizes than 4 were considered to be of “low quality.” The follow-
was computed according to previous recommendations.24 ing sources of bias were considered: selection (sequence
Heterogeneity (i.e., τ2) was estimated using the re- generation and allocation concealment), performance
stricted maximum-­likelihood estimator (REML).25 In ad- (blinding of participants/personnel), detection (blinding
dition, in order to complete heterogeneity analyses, the outcome assessors), attrition (incomplete outcome data),
Q-­test for heterogeneity26 and the I2  statistic27 were cal- reporting (selective reporting), and other potential bias
culated. Studentized residuals and Cook's distances were (e.g., recall bias). The risk-­of-­bias assessments for the in-
examined to assess whether studies may be outliers and cluded studies are shown in online Supplementary Table
influential.28 Studies with a studentized residual larger 3. The mean score for the PEDro scale criteria 2 –­ 11 was
than the 100 × (1–­0.05 / (2 × k))th percentile of a standard 3.8/10, that is, medium quality.
normal distribution were declared potential outliers (i.e.,
using a Bonferroni correction with two-­sided α = 0.05 for
k studies included in the meta-­analyses). Studies with a 3  |  RESULTS
Cook's distance larger than the median plus six times the
interquartile range of the Cook's distances were consid- 3.1  |  Study characteristics
ered influential. In case a study was identified as a po-
tential outlier or overly influential, a sensitivity analysis The database search identified 15,837 potentially eligi-
was performed. A trim-­and-­fill-­contour funnel plot was ble articles in the initial search and 4,536 in the updated
provided to estimate the number of studies potentially search. After further screening and eligibility assessment,
missing from the meta-­analysis (Figure S1–­S3). The rank a total of 18  studies were included in the final analyses
 |
814       BOULLOSA et al.

(see Figure 1). The characteristics of studies, participants, τ2 = 0.00, I2 = 0.00%). The regression test indicated funnel
and training interventions are summarized in online plot asymmetry (p = 0.046) but not the rank correlation
Supplementary Table 4. The meta-­analysis included a test (p = 0.381) (Figure S1A).
total of 438 participants, of whom 239 performed super- For the quantitative analysis of differences in V̇O2max
vised sSIT, 107 participants performed no exercise or no of sSIT vs. HIIT/SIT/CT, eight studies were included in
additional sSIT, 49 performed HIIT as control condition, the final analysis. The SMD ranged from −0.64 to 0.39.
45 performed other SIT modality as control condition, and The estimated SMD based on the random-­effects model
13 performed CT as control condition. Of the included was 0.05 (95% CI: −0.19, 0.30, p = 0.676). The forest plot
studies, cycling was the most common mode of exercise showing the observed outcomes and the estimate based
(8 studies),13,19,33–­38 followed by running (6 studies).11,39–­43 on the random-­effects model is shown in Figure  3. The
Additionally, boxing exercise,44 canoe paddling,45 func- Q-­test revealed that the true outcomes are homogenous
tional fitness exercises,46 handcycling,47 and squatting + (Q(7) =4.700, p = 0.696, τ2 = 0.00, I2 = 0.00%). Neither
cycling19 were also assessed by one study each. In 12 of the the rank correlation nor the regression test indicated any
included studies, sSIT was performed at maximal possible funnel plot asymmetry (p = 0.905 and p = 0.537, respec-
intensity (“all-­out”),11,13,19,34,37,38,40,41,44–­47 while 5  stud- tively) (Figure S1B).
ies assessed the effect of non-­“all-­out” high-­intensity ef-
forts,33,36,39,42,43 and one study gave no further description
about the intensity.35 3.3  |  Aerobic performance

Ten studies were included in the quantitative analysis of

3.2  |  V̇O2max sSIT vs. no exercise or usual training. The SMD ranged
from −1.24 to −0.04. The estimated SMD based on the
Twelve studies were included in the quantitative analy- random-­effects model was −0.43 (95% CI: −0.67, −0.20,
sis of sSIT vs. no exercise or usual training. The SMD p < 0.001). The forest plot showing the observed outcomes
ranged from −2.12 to 0.33. The estimated SMD based and the estimate based on the random-­effects model is
on the random-­effects model was −0.56 (95% CI: −0.79, shown in Figure 4. The Q-­test revealed that the true out-
−0.33, p  <  0.001). The forest plot showing the observed comes are homogenous (Q(9) =9.44, p = 0.398, τ2 = 0.00,
outcomes and the estimate based on the random-­effects I2 = 0.00%). Neither the rank correlation nor the regres-
model is shown in Figure 2. The Q-­test revealed that the sion test indicated any funnel plot asymmetry (p = 0.601
true outcomes are homogenous (Q(11) = 14.81, p = 0.192, and p = 0.155, respectively) (Figure S2A).

F I G U R E 2   Forest plot showing the differences in effect sizes in V̇O2max of sSIT compared to no sSIT. CI, confidence interval;
RE Model, random-­effects model; SMD, standardized mean difference; sSIT, short sprint interval training; V̇O2max, maximal oxygen
consumption
BOULLOSA et al.      |  815

F I G U R E 3   Forest plot showing the differences in effect sizes in V̇O2max of sSIT compared to HIIT/SIT/CT. CI, confidence interval;
CT, continuous training; HIIT, high-­intensity interval training; RE Model, random-­effects model; SMD, standardized mean difference; sSIT,
short sprint interval training; V̇O2max, maximal oxygen consumption

F I G U R E 4   Forest plot showing the differences in effect sizes in aerobic performance of sSIT compared to no sSIT. CI, confidence
interval; N/A, not available; RE Model, random-­effects model; SMD, standardized mean difference; sSIT, short sprint interval training;
V̇O2max, maximal oxygen consumption

For the quantitative analysis of differences in aerobic funnel plot asymmetry (p = 0.381 and p = 0.320, respec-
performance of sSIT vs. HIIT/SIT/CT, seven studies were tively) (Figure  S2B). The subgroup analysis revealed no
included in the final analysis. The SMD ranged from −0.02 statistical differences (p > 0.05).
to 0.93. The estimated SMD based on the random-­effects
model was 0.15 (95% CI: −0.12, 0.42, p = 0.281). The for-
est plot showing the observed outcomes and the estimate 3.4  |  Anaerobic performance
based on the random-­effects model is shown in Figure 5.
The Q-­test revealed that the true outcomes are homoge- Nine studies were included in the quantitative analysis of
nous (Q(6) =2.86, p = 0.826, τ2 = 0.00, I2 = 0.00%). Neither sSIT vs. no exercise or usual training. The SMD ranged
the rank correlation nor the regression test indicated any from −1.73 to −0.07. The estimated SMD based on the
 |
816       BOULLOSA et al.

F I G U R E 5   Forest plot showing the differences in effect sizes in aerobic performance of sSIT compared to HIIT/SIT/CT. CI, confidence
interval; CT, continuous training; HIIT, high-­intensity interval training; RE Model, random-­effects model; SMD, standardized mean
difference; sSIT, short sprint interval training; V̇O2max, maximal oxygen consumption

F I G U R E 6   Forest plot showing the differences in effect sizes in anaerobic performance of sSIT compared to no sSIT. CI, confidence
interval; RE Model, random-­effects model; SMD, standardized mean difference; sSIT, short sprint interval training; V̇O2max, maximal
oxygen consumption

random-­effects model was −0.44 (95% CI: −0.70, −0.18, were included in the final analysis. The SMD ranged
p < 0.001). The forest plot showing the observed outcomes from −0.64 to 0.74. The estimated SMD based on the
and the estimate based on the random-­effects model is random-­effects model was 0.07 (95% CI: −0.25, 0.39, p =
shown in Figure 6. The Q-­test revealed that the true out- 0.672). The forest plot showing the observed outcomes
comes are heterogeneous (Q(8) =8.21, p = 0.413, τ2 = and the estimate based on the random-­effects model
0.00 I2 = 0.00%). The regression test indicated funnel plot is shown in Figure 7. The Q-­test revealed that the true
asymmetry (p = 0.039) but not the rank correlation test (p outcomes appear to be homogenous, but some heteroge-
= 0.119) (Figure S3A). neity may still be present (Q(7) =10.76, p = 0.149, τ2 =
For the quantitative analysis of differences in anaer- 0.07, I2 = 35.32%). Neither the rank correlation nor the
obic performance of sSIT vs. HIIT/SIT/CT, eight studies regression test indicated any funnel plot asymmetry (p =
BOULLOSA et al.      |  817

0.548 and p = 0.407, respectively (Figure S3B). The sub- induce significant improvements of V̇O2max after 436 and
group analysis did not reveal a statistically significant 6 weeks43 of training. However, the lower intensity during
difference (p > 0.05). non-­“all-­out” efforts was associated with a higher volume
including 40 –­ 48 efforts during training.36,43 Meanwhile,
the other studies with “all-­out” efforts only completed
4  |  DI S C USSION 6—­36 efforts of diverse duration and work-­to-­rest ratios.
11,13,19,38,40,41,44-­46
Therefore, while the high time efficiency
This is the first meta-­analysis evaluating the effects of sSIT (i.e., rapid V̇O2max improvements after a few sessions) of
(≤ 10-­s sprints) on V̇O2max and measures of aerobic and different protocols of sSIT has been confirmed, the most
anaerobic performances in different exercise modes in optimal dose response in terms of intensity (non-­“all-­out”
healthy adults and athletes. From the current results, it vs. “all-­out”), volume (number of sprints and sessions),
can be suggested that sSIT is an excellent means to develop and work-­to-­rest ratios in terms of V̇O2max improvements
V̇O2max and both aerobic and anaerobic performances in are yet to be defined.
physically active individuals and athletes after short train- Regarding aerobic performances, a number of sSIT
ing periods of ≥2  weeks and that the effects of sSIT are protocols appeared to improve several performance
similar to other continuous or HIIT/SIT protocols. parameters in studies using incremental33,36,39,44,45 and
Our results confirm a high time efficiency of sSIT to time-­trial tests,11,40 while others did not find significant
increase V̇O2max in different populations after a short improvements.34,42,47 Therefore, the positive effects of
training period of ≥2 weeks, when compared to nonexer- sSIT protocols on improving V̇O2max are accompanied
cise or usual training regimens of physically active adults by an increased endurance performance. However, the
and athletes. Importantly, this effectiveness is similar to SMD was slightly greater for HIIT/SIT/CT protocols.
other more time-­consuming endurance training methods Similar to V̇O2max improvements, the variety of proto-
including CT, HIIT, or traditional SIT in diverse popula- cols and exercise modes used limits our interpretation.
tions (95% CI: −0.28, 0.26, p = 0.951). This finding further In this regard, it is interesting to note that the most effec-
expands our current knowledge on the high effectiveness tive protocols for endurance performance enhancements
of low-­volume traditional SIT to improve V̇O2max after involved athletes of different sports such as boxing,44
only a few weeks of training.8,9,48,49 Interestingly, the wrestling,40 and canoe polo,45 who completed very short
exercise modes used included punching,44 paddling,45 (3 –­ 6 s) “all-­out” efforts with reduced recovery intervals
and functional fitness exercises,46 apart from the more (i.e., 10 s), totaling 12–­36 repetitions per session. In addi-
typically used cycling13,19,33-­38 and running.11,39-­43 Of tion, the included studies used different exercise modes
note, although with a limited number of studies, it ap- than cycling and running, including punching,44 pad-
pears that non-­“all-­out” efforts may also be sufficient to dling,45 handcycling,47 and functional exercises46 with

F I G U R E 7   Forest plot showing the differences in effect sizes in anaerobic performance of sSIT compared to HIIT/SIT/CT. CI,
confidence interval; CT, continuous training; HIIT, high-­intensity interval training; RE Model, random-­effects model; SMD, standardized
mean difference; sSIT, short sprint interval training, V̇O2max, maximal oxygen consumption
 |
818       BOULLOSA et al.

diverse results. For instance, the study by La Monica 4.2  |  Perspective

et al.47 included 2 and 4  minutes of recovery between
the 10-­s “all-­out” hand cycling bouts and did not exhibit This systematic review clearly shows the high effi-
significant improvements in aerobic performance, thus ciency of sSIT of “all-­out” and non-­“all-­out” efforts to
reinforcing the need for short recovery intervals to favor improve V̇O2max and both aerobic and anaerobic per-
aerobic adaptations. Therefore, it is suggested that the formances in different exercises. Therefore, for a better
limited effectiveness of sSIT protocols to improve endur- understanding of what factors are related to these per-
ance performances may be more related to the duration formance enhancements, further comparisons should
of recovery intervals rather than to the exercise mode be made with different sSIT and HIIT/SIT protocols of
used and probably to the physical status of participants. equated loads but differing in sprint duration (4–­10 s)19
Further studies should assess other protocol designs to and work-­to-­rest ratios36,47 in different exercise modes.
simultaneously improve V̇O2max and endurance perfor- In this regard, the combination of different exercises as
mances in different tests and populations with different in the concurrent training group (cycling +squatting)
training statuses. of the study by Benítez-­Flores et al.19 should be further
As expected, from the current meta-­analysis, it can explored. In addition, we only included healthy active
be suggested that sSIT protocols are excellent means to adults and athletes; therefore, the applicability of sSIT
improve anaerobic performances when compared to no protocols to other populations such as elderly people51
exercise/usual training regimens. Furthermore, the dif- and clinical populations52 requires more research. Of
ferences with other HIIT/SIT protocols are negligible. Of note, to allow appropriate comparisons, further studies
note, these similar anaerobic performances between sSIT should better elaborate on participants’ characteristics
and HIIT/SIT protocols were evident, despite half (4/8) of and training background to verify whether training sta-
the included mSIT studies were performed with non-­“all-­ tus also affects performance outcomes. Furthermore,
out” efforts.36,39,42,43 Meanwhile, this high effectiveness for although non-­“all-­out” efforts seem to be also effective,
the enhancement of anaerobic performances reinforces the supposedly superiority of “all-­out” efforts to opti-
the superiority of sSIT protocols when compared to tradi- mize the dose-­response and, thus, the efficiency of this
tional SIT protocols of longer sprints, as similar outcomes training method for rapid physiological adaptations
can be achieved with less effort in terms of psychologi- needs to be confirmed. Meanwhile, the use of non-­“all-­
cal50 and physiological8  strains (i.e., “same gain with out” efforts prior to commencing with “all-­out” efforts
less pain”), probably as a result of the lower glycolytic in a periodized fashion would seem an appropriate
activation.19 As observed with V̇O2max and aerobic per- strategy for nonathletic populations. In this regard, at-
formance gains, the improvements in anaerobic perfor- tention should be paid to individual responses53 to iden-
mances also occurred after training with different exercise tify what factors are related to heterogeneity in these
modes. Therefore, sSIT protocols can be confidently used responses.
to enhance anaerobic performances in different exercise
modes.
5  |  CONC LUSIONS

4.1  |  Strengths and limitations of the Our data suggest the effectiveness of sSIT protocols
meta-­analysis comprised of exercise bouts of ≤10  s, to enhance both
V̇O2max and aerobic and anaerobic performances, mak-
As with the majority of other exercise science-­related ing sSIT a powerful time-­efficient means to enhance
meta-­analyses, the main limitation of the current sys- physical fitness and performance within only few weeks
tematic review with meta-­analysis is the heterogeneity and with a reduced exercise dose. Importantly, this ef-
of studies with respect to sample characteristics and fectiveness has been proven in different exercise modes
training history, exercise modes, loading parameters such as cycling and running, as well as sport-­specific
and protocols, and performance outcomes measures. exercises such as paddling and punching. Further stud-
However, with our inclusion criteria we have limited ies should elaborate on the loading parameters and rest
the sprints to both “all-­out” and non-­“all-­out” efforts periods associated with optimal adaptations in diverse
≤10 s, independently of diverse HIIT and SIT definitions populations.
proposed by different authors, thus providing enough
studies comparing sSIT to other training modalities, ACKNOWLEDGEMENT
therefore, allowing a quantitative analysis of the se- Open access funding enabled and organized by
lected outcomes. ProjektDEAL.
BOULLOSA et al.      |  819

Appl Physiol Nutr Metab. 2018;43(6):595-­601. doi:10.1139/

DATA AVAILABILITY STATEMENT apnm-­2017-­0595
The data that supports the findings of this study are avail- 12. Townsend LK, Islam H, Dunn E, Eys M, Robertson-­Wilson
able in the supplementary material of this article. J, Hazell TJ. Modified sprint interval training protocols.
Part II. Psychological responses. Appl Physiol Nutr Metab.
ORCID 2017;42(4):347-­353. doi:10.1139/apnm-­2016-­0479
Daniel Boullosa  https://orcid. 13. Hazell TJ, MacPherson REK, Gravelle BMR, Lemon PWR. 10
org/0000-0002-8477-127X or 30-­S sprint interval training bouts enhance both aerobic and
anaerobic performance. Eur J Appl Physiol. 2010;110(1):153-­
Boris Dragutinovic  https://orcid.
160. doi:10.1007/s0042​1-­010-­1474-­y
org/0000-0002-1283-9714 14. Boullosa D, Dragutinovic B, Deutsch JP, et al. Acute and de-
Edward F. Coyle  https://orcid. layed effects of time-­matched very short “all out” efforts in con-
org/0000-0002-4329-8287 centric vs. Eccentric cycling. Int J Environ Res Public Health.
Moritz Schumann  https://orcid. 2021;18(15):7968. doi:10.3390/ijerp​h1815​7968
org/0000-0001-9605-3489 15. Bogdanis GC, Nevill ME, Boobis LH, Lakomy HK. Contribution
of phosphocreatine and aerobic metabolism to energy supply
during repeated sprint exercise. J Appl Physiol. 1996;80(3):876-­
REFERENCES
884. doi:10.1152/jappl.1996.80.3.876
1. Tschakert G, Hofmann P. High-­intensity intermittent exercise:
16. Vardarli E, Satiroglu R, Allen JR, Bjellquist-­Ledger R, Burton
methodological and physiological aspects. Int J Sports Physiol
HM, Coyle EF. Physiological responses to maximal 4 s sprint
Perform. 2013;8(6):600-­610. doi:10.1123/ijspp.8.6.600
interval cycling using inertial loading: the influence of inter-­
2. Billat LV. Interval training for performance: a scientific and
sprint recovery duration. Eur J Appl Physiol. 2021;121(8):2295-­
empirical practice. Special recommendations for middle-­ and
2304. doi:10.1007/s0042​1-­021-­04677​-­6
long-­distance running. Part I: aerobic interval training. Sports
17. Islam H, Townsend LK, Hazell TJ. Modified sprint interval
Med. 2001;31(1):13-­31. doi:10.2165/00007​256-­20013​1010-­00002
training protocols. Part I. Physiological responses. Appl Physiol
3. Buchheit M, Laursen PB. High-­intensity interval training,
Nutr Metab. 2017;42(4):339-­346. doi:10.1139/apnm-­2016-­0478
solutions to the programming puzzle: part I: cardiopulmonary
18. Satiroglu R, Lalande S, Hong S, Nagel MJ, Coyle EF. Four-­
emphasis. Sports Med. 2013;43(5):313-­338. doi:10.1007/s4027​
second power cycling training increases maximal anaerobic
9-­013-­0029-­x
power, peak oxygen consumption, and total blood volume. Med
4. Campbell WW, Kraus WE, Powell KE, et al. High-­intensity
Sci Sports Exerc. 2021;53(12):2536-­2542. doi:10.1249/mss.00000​
interval training for cardiometabolic disease prevention. Med
00000​002748
Sci Sports Exerc. 2019;51(6):1220-­1226. doi:10.1249/MSS.00000​
19. Benítez-­Flores S, Medeiros AR, Voltarelli FA, et al. Combined
00000​001934
effects of very short “all out” efforts during sprint and resis-
5. MacInnis MJ, Gibala MJ. Physiological adaptations to in-
tance training on physical and physiological adaptations after
terval training and the role of exercise intensity. J Physiol.
2 weeks of training. Eur J Appl Physiol. 2019;119(6):1337-­1351.
2017;595(9):2915-­2930. doi:10.1113/JP273196
doi:10.1007/s0042​1-­019-­04125​-­6
6. Gillen JB, Gibala MJ. Is high-­intensity interval training a
20. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement
time-­efficient exercise strategy to improve health and fitness?
for reporting systematic reviews and meta-­analyses of studies
Appl Physiol Nutr Metab. 2014;39(3):409-­412. doi:10.1139/
that evaluate healthcare interventions: explanation and elabo-
apnm-­2013-­0187
ration. BMJ. 2009;339:b2700. doi:10.1136/bmj.b2700
7. Buchheit M, Laursen PB. High-­intensity interval training, solu-
21. Drevon D, Fursa SR, Malcolm AL. Intercoder reliability and
tions to the programming puzzle: part II: anaerobic energy,
validity of webplotdigitizer in extracting graphed data. Behav
neuromuscular load and practical applications. Sports Med.
Modif. 2016;41(2):323-­339. doi:10.1177/01454​45516​673998
2013;43(10):927-­954. doi:10.1007/s4027​9-­013-­0066-­5
22. Viechtbauer W. Conducting meta-­analyses in R with the meta-
8. Vollaard NBJ, Metcalfe RS. Research into the health bene-
for package. J Stat Softw. 2010;36(3):1-­48.
fits of sprint interval training should focus on protocols with
23. Becker BJ. Synthesizing standardized mean-­change mea-
fewer and shorter sprints. Sports Med. 2017;47(12):2443-­2451.
sures. Br J Math Stat Psychol. 1988;41(2):257-­278. doi:10.1111/
doi:10.1007/s4027​9-­017-­0727-­x
j.2044-­8317.1988.tb009​01.x
9. Vollaard NBJ, Metcalfe RS, Williams S. Effect of number of
24. Morris SB. Estimating effect sizes from pretest-­posttest-­
sprints in an SIT session on change in v O2max: a meta-­analysis.
control group designs. Organ Res Methods. 2007;11(2):364-­386.
Med Sci Sports Exerc. 2017;49(6):1147-­1156. doi:10.1249/
doi:10.1177/10944​28106​291059
MSS.00000​00000​001204
25. Viechtbauer W. Bias and efficiency of meta-­analytic variance
10. Benítez-­Flores S, de Sousa AFM, Totó EC, et al. Shorter sprints
estimators in the random-­effects model. J Educ Behav Stat.
elicit greater cardiorespiratory and mechanical responses with
2005;30(3):261-­293. doi:10.3102/10769​98603​0003261
less fatigue during time-­matched sprint interval training (SIT)
26. Cochran WG. The combination of estimates from different ex-
sessions. Kinesiology. 2018;50(2):137-­148. doi:10.26582/​K.50.2.13
periments. Biometrics. 1954;10(1):101-­129. doi:10.2307/3001666
11. McKie GL, Islam H, Townsend LK, Robertson-­Wilson J, Eys
27. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-­
M, Hazell TJ. Modified sprint interval training protocols: phys-
analysis. Stat Med. 2002;21(11):1539-­1558. doi:10.1002/sim.1186
iological and psychological responses to 4 weeks of training.
 |
820       BOULLOSA et al.

28. Viechtbauer W, Cheung MW-­L. Outlier and influence diagnos- exercise intensity. J Physiol. 2003;550(3):855-­861. doi:10.1113/
tics for meta-­analysis. Res Synth Methods. 2010;1(2):112-­125. jphys​iol.2003.040162
doi:10.1002/jrsm.11 44. Kamandulis S, Bruzas V, Mockus P, Stasiulis A, Snieckus A,
29. Begg CB, Mazumdar M. Operating characteristics of a Venckunas T. Sport-­specific repeated sprint training improves
rank correlation test for publication bias. Biometrics. punching ability and upper-­body aerobic power in experienced
1994;50(4):1088-­1101. amateur boxers. J Strength Cond Res. 2018;32(5):1214-­1221.
30. Sutton AJ, Rothstein H, Borenstein M, eds. Publication Bias in doi:10.1519/JSC.00000​00000​002056
Meta-­Analysis: Prevention, Assessment and Adjustments. Wiley; 45. Sheykhlouvand M, Khalili E, Gharaat M, Arazi H, Khalafi M,
2005. Tarverdizadeh B. Practical model of low-­volume paddling-­based
31. Higgins JPT, Thomas J, Chandler J, et al. Cochrane handbook sprint interval training improves aerobic and anaerobic perfor-
for systematic reviews of interventions. Cochrane Database Syst mances in professional female canoe polo athletes. J Strength Cond
Rev. 2019;5:1-­694. doi:10.1002/97811​19536604 Res. 2018;32(8):2375-­2382. doi:10.1519/JSC.00000​00000​002152
32. Elkins MR, Herbert RD, Moseley AM, Sherrington C, 46. Moghaddam M, Estrada CA, Muddle TWD, Magrini MA,
Maher C. Rating the quality of trials in systematic reviews Jenkins NDM, Jacobson BH. similar anaerobic and aerobic
of physical therapy interventions. Cardiopulm Phys Ther J. adaptations after 2 high-­intensity interval training configura-
2010;21(3):20-­26. tions: 10:5 s vs. 20:10 s work-­to-­rest ratio. J Strength Cond Res.
33. Belfry GR, Paterson DH, Thomas SG. High-­intensity 10-­s work: 2021;35(6):1685-­1692. doi:10.1519/JSC.00000​00000​002939
5-­s recovery intermittent training improves anaerobic and 47. La Monica MB, Fukuda DH, Starling-­Smith TM, et al.
aerobic performances. Res Q Exerc Sport. 2020;91(4):640-­651. Examining work-­to-­rest ratios to optimize upper body sprint
doi:10.1080/02701​367.2019.1696928 interval training. Respir Physiol Neurobiol. 2019;262:12-­19.
34. Camacho-­Cardenosa M, Camacho-­Cardenosa A, Martínez doi:10.1016/j.resp.2019.01.005
Guardado I, Marcos-­Serrano M, Timon R, Olcina G. A new 48. Sultana RN, Sabag A, Keating SE, Johnson NA. The effect of
dose of maximal-­intensity interval training in hypoxia to im- low-­volume high-­intensity interval training on body compo-
prove body composition and hemoglobin and hematocrit lev- sition and cardiorespiratory fitness: a systematic review and
els: a pilot study. J Sports Med Phys Fitness. 2017;57(1–­2):60-­69. meta-­analysis. Sports Med. 2019;49(11):1687-­1721
doi:10.23736/​S0022​-­4707.16.06549​-­X 49. Rosenblat MA, Perrotta AS, Thomas SG. Effect of high-­
35. Lloyd Jones MC, Morris MG, Jakeman JR. Effect of work: intensity interval training versus sprint interval training
rest ratio on cycling performance following sprint inter- on time-­trial performance: a systematic review and meta-­
val training: a randomized control trial. J Strength Cond Res. analysis. Sports Med. 2020;50(6):1145-­1161. doi:10.1007/
2019;33(12):3263-­3268. doi:10.1519/JSC.00000​00000​003381 s4027​9-­020-­01264​-­1
36. Lee CL, Hsu WC, Cheng CF. Physiological adaptations to 50. Haines M, Broom D, Gillibrand W, Stephenson J. Effects of
sprint interval training with matched exercise volume. Med three low-­volume, high-­intensity exercise conditions on affec-
Sci Sports Exerc. 2017;49(1):86-­95. doi:10.1249/MSS.00000​ tive valence. J Sports Sci. 2020;38(2):121-­129. doi:10.1080/02640​
00000​001083 414.2019.1684779
37. Ortenblad N, Lunde PK, Levin K, Andersen JL, Pedersen 51. Allen JR, Satiroglu R, Fico B, et al. Inertial load power cycling
PK. Enhanced sarcoplasmic reticulum Ca(2+) release fol- training increases muscle mass and aerobic power in older
lowing intermittent sprint training. Am J Physiol Regul Integr adults. Med Sci Sports Exerc. 2021;53(6):1188-­1193. doi:10.1249/
Comp Physiol. 2000;279(1):R152-­R160. doi:10.1152/ajpre​ MSS.00000​00000​002588
gu.2000.279.1.R152 52. Tong TK, Zhang H, Shi H, et al. Comparing time efficiency of
38. Tanisho K, Hirakawa K. Training effects on endurance capacity sprint vs. High-­intensity interval training in reducing abdomi-
in maximal intermittent exercise: comparison between continu- nal visceral fat in obese young women: a randomized, controlled
ous and interval training. J Strength Cond Res. 2009;23(8):2405-­ trial. Front Physiol. 2018;9:1-­9. doi:10.3389/fphys.2018.01048
2410. doi:10.1519/JSC.0b013​e3181​bac790 53. Schulhauser KT, Bonafiglia JT, McKie GL, et al. Individual
39. Cavar M, Marsic T, Corluka M, et al. Effects of 6 weeks of differ- patterns of response to traditional and modified sprint interval
ent high-­intensity interval and moderate continuous training training. J Sports Sci. 2021;39(10):1077-­1087. doi:10.1080/02640​
on aerobic and anaerobic performance. J Strength Cond Res. 414.2020.1857507
2019;33(1):44-­56. doi:10.1519/JSC.00000​00000​002798
40. Farzad B, Gharakhanlou R, Agha-­Alinejad H, et al.
Physiological and performance changes from the addition of a SUPPORTING INFORMATION
sprint interval program to wrestling training. J Strength Cond Additional supporting information may be found in the
Res. 2011;25(9):2392-­2399. doi:10.1519/JSC.0b013​e3181​fb4a33 online version of the article at the publisher’s website.
41. Fernandez-­Fernandez J, Zimek R, Wiewelhove T, Ferrauti A.
High-­intensity interval training vs. repeated-­sprint training
in tennis. J Strength Cond Res. 2012;26(1):53-­62. doi:10.1519/
How to cite this article: Boullosa D, Dragutinovic
JSC.0b013​e3182​20b4ff
B, Feuerbacher JF, Benítez-­Flores S, Coyle EF,
42. Mohr M, Krustrup P, Nielsen JJ, et al. Effect of two different
intense training regimens on skeletal muscle ion transport pro- Schumann M. Effects of short sprint interval
teins and fatigue development. Am J Physiol Regul Integr Comp training on aerobic and anaerobic indices: A
Physiol. 2007;292(4):1594-­1602. doi:10.1152/ajpre​gu.00251.2006 systematic review and meta-­analysis. Scand J Med
43. Russell AP, Somm E, Praz M, et al. UCP3 protein regulation in Sci Sports. 2022;32:810–­820. doi:10.1111/sms.14133
human skeletal muscle fibre types I, IIa and IIx is dependent on