Effect of High-Intensity Interval Training Versus Sprint Interval Training on Time-Trial Performance: A Systematic Review and Meta-Analysis

Abstract

Background: Two forms of interval training commonly discussed in the literature are high-intensity interval training (HIIT) and sprint interval training (SIT). HIIT consists of repeated bouts of exercise that occur at a power output or velocity between the second ventilatory threshold and maximal oxygen consumption (VO2max). SIT is performed at a power output or velocity above those associated with VO2max

Objective: The primary objective of this study is to systematically review published randomized and pair- matched trials to determine which mode of interval training, HIIT versus SIT, leads to a greater improvement in TT performance in active and trained individuals. The second objective of this review is to perform a subgroup analysis to determine if there is a distinction between HIIT programs that differ in work-bout duration.

Data Sources: SPORTDiscus (1800 - present) and Medline with Full Text (1946 - present) were used to conduct a systematic literature search

Study Selection: Studies were selected for the review if they met the following criteria: 1. individuals (males and females) who were considered at least moderately-trained (~3-hours per week of activity) as specified by the authors of the included studies; 2. between the ages of 18 and 45 years; 3. randomized or pair-matched trials that included a HIIT and a SIT group; 4. provided detailed information about the interval training program; 5. were at least 2-weeks in duration; 6. included a TT test that required participants to complete a set distance.

Results: A total of 6 articles met the inclusion criteria for the subjective and objective analysis. The pooled analysis was based on a random-effects model. There was no difference in the change in TT performance when comparing all HIIT versus SIT (0.9%; 90% CI: -0.1% to 1.9%, p = 0.18). However, subgroup analysis based on duration of work interval indicated a 2% greater improvement in TT performance following long-HIIT ( 4 min) when compared to SIT. There was no difference in change in VO2max/peak oxygen consumption (VO2peak) between groups. There was a moderate effect (ES = 0.70) in favour of HIIT over SIT in maximal aerobic power (MAP) or maximal aerobic velocity (MAV).

Conclusion: The results of the meta-analysis indicate that long-HIIT may be the optimal form of interval training to augment TT performance. Additional research that directly compares HIIT exercise differing in work-bout duration would strengthen these results and provide further insight into the mechanisms behind the observed benefits of Long- HIIT.

Full text

Vol.:(0123456789)
Sports Medicine (2020) 50:1145–1161
https://doi.org/10.1007/s40279-020-01264-1
SYSTEMATIC REVIEW
Eect ofHigh‑Intensity Interval Training Versus Sprint Interval Training
onTime‑Trial Performance: ASystematic Review andMeta‑analysis
MichaelA.Rosenblat1,3 · AndrewS.Perrotta2· ScottG.Thomas1,3
Published online: 7 February 2020
© Springer Nature Switzerland AG 2020
Abstract
Background Two forms of interval training commonly discussed in the literature are high-intensity interval training (HIIT)
and sprint interval training (SIT). HIIT consists of repeated bouts of exercise that occur at a power output or velocity between
the second ventilatory threshold and maximal oxygen consumption (VO2max). SIT is performed at a power output or velocity
above those associated with VO2max.
Objective The primary objective of this study is to systematically review published randomized and pair-matched trials to
determine which mode of interval training, HIIT versus SIT, leads to a greater improvement in TT performance in active and
trained individuals. The second objective of this review is to perform a subgroup analysis to determine if there is a distinction
between HIIT programs that differ in work-bout duration.
Data Sources SPORTDiscus (1800–present) and Medline with Full Text (1946–present) were used to conduct a systematic
literature search.
Study Selection Studies were selected for the review if they met the following criteria: (1) individuals (males and females)
who were considered at least moderately trained (~ 3-h per week of activity) as specified by the authors of the included stud-
ies; (2) between the ages of 18 and 45years; (3) randomized or pair-matched trials that included a HIIT and a SIT group;
(4) provided detailed information about the interval training program; (5) were at least 2weeks in duration; (6) included a
TT test that required participants to complete a set distance.
Results A total of 6 articles met the inclusion criteria for the subjective and objective analysis. The pooled analysis was
based on a random-effects model. There was no difference in the change in TT performance when comparing all HIIT versus
SIT (0.9%; 90% CI −1.2–1.9%, p = 0.18). However, subgroup analysis based on duration of work interval indicated a 2%
greater improvement in TT performance following long-HIIT (≥4min) when compared to SIT. There was no difference in
change in VO2max/peak oxygen consumption (VO2peak) between groups. There was a moderate effect (ES = 0.70) in favor of
HIIT over SIT in maximal aerobic power (MAP) or maximal aerobic velocity (MAV).
Conclusion The results of the meta-analysis indicate that long-HIIT may be the optimal form of interval training to augment
TT performance. Additional research that directly compares HIIT exercise differing in work-bout duration would strengthen
these results and provide further insight into the mechanisms behind the observed benefits of long-HIIT.
1 Introduction
Endurance training programs should be optimized to
improve athletic performance while limiting the develop-
ment of fatigue or risk of injury. One of the most impor-
tant variables to consider when prescribing exercise is the
intensity at which an athlete trains as this metric strongly
influences physiological and performance adaptations [1].
The distribution of exercise intensity within a training pro-
gram has provoked great interest over the past decade [2].
The results of a recent meta-analysis of randomized con-
trolled trials examining intensity distribution suggests that
* Michael A. Rosenblat
michael@evidencebasedcoaching.ca
1 Department ofExercise Science, Faculty ofKinesiology
andPhysical Education, University ofToronto, Toronto, ON,
Canada
2 Department ofKinesiology, Langara College, Vancouver,
BC, Canada
3 Training andPerformance Laboratory, Goldring Centre
forHigh Performance Sport, Department ofExercise
Science, Faculty ofKinesiology andPhysical Education,
University ofToronto, Toronto, ON, Canada

1146 M.A.Rosenblat et al.
a polarized training model, which includes approximately
20% of total training volume in the high-intensity exercise
domain and 80% in the low-intensity domain, may lead to a
greater improvement in endurance sport performance when
compared to other intensity distribution models [3].
However, there remains equivocal evidence regarding
the best method to program high-intensity training sessions
and, in particular, the variables defining interval training
stimuli. Interval training consists of repeated bouts of exer-
cise followed by rest or low-intensity exercise, each of which
can last from seconds to minutes in duration. Prescribing
interval training can be quite complex since performance
improvements may be influenced through the manipulation
of a number of programming variables including exercise
mode, duration, intensity, recovery, number of intervals, and
the frequency and distribution of interval training [4, 5]. In
addition to the variables required for programming a specific
exercise session, population characteristics such as age, sex,
training status and background can also impact performance
gains [4].
Two forms of interval training commonly discussed in
the literature are high-intensity interval training (HIIT) and
sprint interval training (SIT). HIIT consists of repeated bouts
of exercise that occur at a power output or velocity within
the severe-intensity domain [6], which occurs between the
second ventilatory threshold (VT2) and maximal oxygen
consumption (VO2max) [7]. In the case when an individual’s
VO2max cannot be determined through exercise testing, peak
oxygen consumption (VO2peak) is used to indicate the upper
border of the domain. SIT is performed at a power output or
velocity above those associated with VO2max [8]. As such, it
can be considered to be completed in the extreme exercise
domain.
Gaps in our understanding of the effects of interval train-
ing may remain in part due to the lack of standardization
for developing HIIT and SIT protocols. This concern was
addressed in the review by Viana etal., where the authors
explain that it may be difficult to generalize the results of
interval training programs due to inconsistent exercise pro-
tocols [9]. There are a number of interval training studies
that use HIIT programs that more closely represent SIT exer-
cise [10–15]. Those programs include work-bouts that are
between 15s and twomin in duration. Due to the short work-
bout duration, a large portion of total energy production is
through anaerobic energy sources [16]. Previous investiga-
tions examining SIT protocols including short-interval rest
periods (e.g., 15s) have demonstrated similar responses
to that of HIIT, requiring a greater proportion of aerobic
metabolism [17–20]. While these SIT protocols have been
shown to improve exercise performance, they may be less
effective for improving VO2max than HIIT protocols incor-
porating longer rest intervals [19]. In addition, confounding
evidence may arise as a consequence of SIT protocols that
incorporate a one-to-one work–rest ratio, whereby power
or velocity is decreased over multiple intervals, yet heart
rate remains elevated. As such, by definition, these intervals
digress to a typically HIIT protocol, with power or velocity
falling below VO2max.
Both HIIT and SIT produce adaptations that are beneficial
for endurance performance. A meta-analysis by Milanović
etal. shows that interval training can lead to improvements
in VO2max and can do so to a greater extent than moderate-
intensity continuous training (MICT) [21]. However, that
analysis did not differentiate between modes of interval train-
ing, including HIIT and SIT in the same analysis group. In
addition, most reviews that address aerobic performance use
VO2max as the primary outcome measure. Although VO2max
has been correlated with race performance [22], strong evi-
dence suggest other variables may positively influence per-
formance outcomes [23, 24]. An alternative measure, time-
trial (TT) performance, has demonstrated a high correlation
with endurance performance, and may directly simulate the
physiological responses required during competition [25,
26]. Time-to-exhaustion (TTE) tests have also been used as
substitute measures for VO2max. However, TTE tests have a
wider variability in results when compared to TT tests [27].
Previous reviews have compared interval training (HIIT,
SIT or combined) with either a non-exercising control or
MICT [21, 28]. There is sufficient evidence that interval
training can enhance performance to a greater extent than
other modes of endurance training. Currently, there remains
a paucity of reviews that compare the effects of HIIT ver-
sus SIT on markers of endurance sport performance. As
such, the primary objective of this study is to systematically
review published randomized and pair-matched trials to
determine which mode of interval training, HIIT versus SIT,
leads to a greater improvement in TT performance in active
Key Points
There was approximately a 2% greater improvement in
time-trial performance following long-duration high-
intensity interval training (HIIT) that consisted of work-
bouts that are 4minutes or greater when compared to
sprint interval training (SIT).
There was no difference in change in maximal or peak
oxygen consumption between HIIT and SIT.
There was a moderate effect (ES = 0.70) in favor of HIIT
over SIT in maximal aerobic power/velocity, with long-
duration HIIT producing the greatest increase in perfor-
mance, a 4% greater change when compared to SIT.

1147
Effect of Interval Training on Time-Trial Performance
and trained individuals. Various studies have employed HIIT
work-bout durations ranging from 1min to 6min [10, 29].
In addition to the limited research comparing HIIT and SIT,
there are no reviews that compare the effects of HIIT exer-
cise protocols which differ in work-bout duration with SIT
protocols on endurance performance. Therefore, the second
objective of this review is to perform a subgroup analysis to
determine if there is a distinction between HIIT programs
that differ in work-bout duration.
2 Methods
2.1 Protocol andRegistration
The Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA) was used as the protocol for the
design of the review [30]. The PRISMA guidelines include a
27-item checklist considered to improve reporting transpar-
ency, limiting the risk of publication and selection bias [30].
2.2 Eligibility Criteria
2.2.1 Inclusion Criteria
Studies were selected for the review if they met the follow-
ing criteria: (1) Individuals (males and females) who were
considered at least moderately trained (~ 3h per week of
activity) as specified by the authors of the included studies;
(2) between the ages of 18 and 45years; (3) randomized or
pair-matched trials that included a HIIT and a SIT group;
(4) provided detailed information about the interval train-
ing program; (5) were at least 2weeks in duration; and (6)
included a TT test that required participants to complete a
set distance.
2.2.2 Exclusion Criteria
Studies were excluded if participants had pathology or if
interventions included the following: (1) nutritional inter-
ventions (supplements, hydration, fed state, etc.), (2) change
in environmental conditions (heat/cold, altitude, hypoxia/
hyperoxia, etc.), (3) inclusion of modalities (cryotherapy,
compression garments, etc.), and (4) pharmacological
agents,
2.3 Information Sources
An electronic search was conducted that included all publi-
cation years (up to and including December 2018). To mini-
mize selection bias and to perform a comprehensive search,
two databases were used to conduct a systematic literature
search and included SPORTDiscus (1800–present) and Med-
line with Full Text (1946–present).
2.4 Search
2.4.1 Search String
Key search terms that were produced from reviewing previ-
ous literature and using a number of synonyms of the dif-
ferent forms of interval training were grouped and searched
within the article title and abstract, and keywords using the
search conjunction ‘OR’. Combinations of the following
terms were used as search terms: ‘interval training’, ‘inter-
val exercise’ ‘anaerobic interval*’ ‘aerobic interval*’ ‘high
intensity interval*’ ‘sprint interval*’ ‘intermittent exercise’
‘intermittent training’ ‘repeated sprint*’.
2.4.2 Search Limits
To provide a more accurate search, the following limits were
selected: (1) English language, (2) humans, and (3) journal
article, all publications up to and including the year 2018.
2.5 Study Selection
The titles and abstracts of the search results were inde-
pendently assessed for suitability by two authors. Full-text
articles were retrieved if the titles or abstracts met the eli-
gibility criteria or if there was uncertainty. Disagreements
were resolved through a discussion between the two authors,
with a third to be consulted if the first two authors could not
reach agreement. The rationale for excluding articles was
documented.
2.6 Data Collection Process
A data collection form was created using the Cochrane Data
Extraction and Assessment Form template. One author was
responsible for collecting the data and the second author
checked the extracted data. Disagreements were discussed
between the two authors, with a third to be consulted if the
first two authors could not reach agreement.
2.7 Data Items
The following data were extracted from each of the articles
that were included in the review: study methodology (study
design and duration); the participant characteristics (sex,
age, height, mass, VO2max/VO2peak); intervention description

1148 M.A.Rosenblat et al.
(exercise mode, training program duration, interval sessions
performed each week, interval work-bout duration, interval
work-bout intensity (expressed as a percentage of the power
or velocity associated with VO2max/VO2peak); and outcomes
measures (VO2max/VO2peak, power at maximal oxygen uptake
(MAP), velocity at maximal oxygen uptake (MAV), and TT
performance). The correction factor used by Granata etal.
was used to standardized exercise intensity obtained from
testing protocols that exceeded 12min in duration [31].
2.8 Risk ofBias ofIndividual Studies
Two reviewers used the PEDro scale to assess the quality
of the studies included in the review. The PEDro scale is a
10-point ordinal scale used to determine the internal validity
of a study. The specific methodological components assessed
include: (1) randomization, (2) concealed allocation, (3)
baseline comparison, (4) blind participants, (5) blind thera-
pists, (6) blind assessors, (7) adequate follow-up, (8) inten-
tion-to-treat analysis, (9) between group comparisons, and
(10) point estimates and variability [32]. Participant eligibil-
ity is also a component of the PEDro scale; however, it is not
included in the final 10-point score.
2.9 Summary ofMeasures
The primary outcome assessed in this review is TT
performance. Secondary outcome measures include
VO2max/VO2peak and MAP/MAV.
2.10 Synthesis ofResults
Group data are reported as means and standard deviations
with pooled data reported as the standardized mean differ-
ence and its 90 percent confidence intervals. The standard-
ized mean difference, adjusted to account for small sample
size bias, was calculated to establish an effect size, (Hedges’
adjusted g) [33]. Effect size values of 0.2, 0.6 and 1.2 were
interpreted as small, moderate and large effect sizes, respec-
tively [34].
The authors of the included studies were contacted for
data that were not presented in their publications (e.g., pre-
and post-test data). Data expressed using the standard error
of the mean (SEM) were converted to the standard deviation
(SD) using the following formula:
SD
=SEM
√
n . The SD
was estimated using the p value in instances, where the SEM
or SD was not available using the following formula:
SD
=

n

̄x1−̄x2
t

. A p value expressed using an inequality
(e.g. ‘<’) was discussed as an equality (e.g., ‘=’), providing
a more conservative estimate of the SD. The mean value for
a training load characteristic in the respective subgroup was
used to provide an objective value where only a subjective
description was given. Where possible, between-group com-
parisons were made using the difference of means with the
standard error expressed as a 90 percent confidence
interval.
Individual study results were combined using Review
Manager 5.3 with a random-effect meta-analysis model. This
method considers both within- and between-study variabil-
ity and was used to accommodate for the differences in the
interventions in the individual studies [35]. The consistency
of the meta-analysis was assessed to determine the variabil-
ity in excess of that due to chance. A Chi-squared statistic
(Cochrane Q) was used to evaluate the level of heterogeneity.
The I2 statistic was used to determine the percentage of the
total variation in the estimated effect across studies.
To perform a sub-group analysis, studies were divided
into HIIT groups that differed in work-bout duration based
on oxygen uptake kinetics. Short-HIIT was defined as inter-
val bouts less than 2min in duration to coincide with the
approximate time to reach peak oxygen consumption [36].
Long-HIIT was considered bouts 4min or greater to ensure
that at least 50% of the total work-bout was completed at
VO2max. Medium-HIIT would fall between the subgroups
with work-bouts between 2min and less than 4min. Session
external work was defined as the product of interval inten-
sity, interval work-bout duration, and interval repetitions.
Total external work was defined as the product of session
external work, sessions per week and number of weeks. Both
measures of external work were described in arbitrary units
(a.u.)
2.11 Risk ofBias Across all Studies
The relationship between the effect size and the sample size
was determined visually using a funnel plot. Egger’s test
was used to quantitatively assess for small sample size bias.
3 Results
3.1 Study Selection
The literature search was conducted on December 28, 2018.
The databases SPORTDiscus and Medline were used to
perform the search which yielded a total of 6994 results.
Following the removal of 1678 duplicates, 5316 titles and
abstracts were screened. A total of 28 full-text articles were
screened for eligibility. Six studies met the inclusion criteria
for the qualitative and quantitative analysis (Fig.1).

1149
Effect of Interval Training on Time-Trial Performance
3.2 Study Characteristics
The studies included only male participants with a mean
range of 19–32years of age. Five of the studies included
participants who were endurance-trained individuals (V
O2max/VO2peak = 51.6–64.5 mL·kg−1·min−1) who partici-
pated in sports such as cycling, rowing, running and tri-
athlon [29, 37–40] and one moderately trained individual
(VO2peak = 46.0mL·kg−1·min−1) [41]. The full details of
study characteristics can be found in Table1.
All six studies included a HIIT and a SIT group. The
HIIT group had interval bouts ranging from 1 to 6min in
duration and at intensities between 73% and 100% MAP. The
SIT groups consisted of 30-s all-out sprints and ranged from
114% to 175% MAP. Two of the studies included more than
one HIIT group [39, 40]. Overall, there was one short-HIIT
interval group [40], 5 that included medium-HIIT intervals
[37–40], and 3 that comprised of long-HIIT bouts [29, 40,
41]. See Table2 for additional details.
3.3 Risk ofBias Within Studies
Two studies scored a 4 on the PEDro scale and four scored a
5, with a mean score of 4.7 out of 10 (Table3). There were
no studies that included subject blinding or assessor blind-
ing. In addition, only one study included concealed alloca-
tion. See Table3 for full details.
3.4 Results ofIndividual Studies
Three of the studies found a significantly greater improve-
ment in TT performance following HIIT when compared
Records identified through database
searching
(n = 6994)
ScreeningIncluded Eligibility Identification
Additional records identified through
other sources
(n = 0)
Records after duplicates removed
(n = 5316)
Records screened
(n = 5316)
Records excluded
(n = 5288)
Full-text articles assessed
for eligibility
(n = 28)
Full-text articles excluded,
with reasons
(n = 22)
•Precise details of
intervention unavailable: 2
•Did not include time-
trial: 19
•Time-trial not to set
distance: 1
Studies included in
qualitative synthesis
(n = 6)
Studies included in
quantitative synthesis
(meta-analysis)
(n = 6)
Fig. 1 PRISMA diagram

1150 M.A.Rosenblat et al.
Table 1 Study characteristics
Study Study design Participant characteristics
(mean ± SD)
Intervention Outcome
Akca and Aras [37] Randomized controlled trial,
matched (4weeks)
Rowers (national level),
sex = male, age (years) = 21.8
± 2.4, body mass
(kg) = 79.3 ± 9.1, height
(cm) = 178.4 ± 6.0, VO2peak
(mL·kg−1·min−1) = 56.6 ± 5.7
Medium-HIIT (n = 10) 8 repetitions of 2.5min at 75%
MAP with 3min of active
recovery, 3 times per week
2-km rowing TT, VO2peak, MAP
SIT (n = 10) 10 repetitions of 30s at 125%
MAP with 4min of active
recovery, 3 times per week
Esfarjani and Laursen [38] Controlled trial, matched
(10weeks)
Runners (moderately trained),
sex = male, age (years) = 19.0
± 2.0, body mass (kg) = 73
± 3, height (cm) = 172 ± 4,
VO2max (mL·kg−1·min−1) = 51.6
± 2.7
Medium-HIIT (n = 6) 8 repetitions of intervals for
60% of time at 87% MAV with
3.5min of active recovery, 2
times per week
3-km running TT, VO2max, MAV
SIT (n = 6) 12 repetitions of 30s at 114%
MAV with 4.5min of active
recovery, 2 times per week
Control (n = 5) 60min of continuous exercise, 4
times per week
Granata etal. [41] Randomized controlled trial,
matched (4weeks)
Active (moderately trained),
sex = male, age (years) = 20.9
± 2.0, body mass
(kg) = 82.1 ± 16.2, height
(cm) = 180.2 ± 9.1, VO2peak
(mL·kg−1·min−1) = 46.0 ± 7.3
Long-HIIT (n = 11) 4–7 repetitions of 4min at 73%
MAP with 2min of active
recovery, 3 times per week
20-km cycling TT, VO2peak, MAP
SIT (n = 10) 4–10 repetitions of 30s at 168%
MAP with 4min of passive
recovery, 3 times per week
Control (n = 10) 20–36min of continuous exer-
cise between 92 and 97% of
the second lactic threshold, 3
times per week
Inoue etal. [29] Randomized controlled trial
(6weeks)
Cyclists (trained), sex = male,
age (years) = 32.1 ± 6.5, body
mass (kg) = 69.1 ± 5.3, height
(cm) = 175.6 ± 5.4, VO2max
(mL·kg−1·min−1) = 63.4 ± 4.5
Long-HIIT (n = 7) 7–10 repetitions of 4–6min at
maximal sustained effort with
4min of active recovery, 3
times per week
40-km cycling TT, VO2max, MAP
SIT (n = 9) 8–12 repetitions of 30-s all-out
sprints with 4min of active
recovery, 3 times per week

1151
Effect of Interval Training on Time-Trial Performance
High-intensity interval training (HIIT), maximal aerobic power (MAP), maximal aerobic velocity (MAV), moderate-intensity interval training, not available (n/a), peak oxygen uptake (VO2peak),
sprint interval training (SIT), time-trial (TT)
Table 1 (continued)
Study Study design Participant characteristics
(mean ± SD)
Intervention Outcome
Laursen etal. [39] Controlled trial, matched
(4weeks)
Endurance athletes (highly
trained), sex = male, age
(years) = 25 ± 6, body
mass (kg) = 75 ± 7, height
(cm) = 180 ± 5, VO2peak
(mL·kg−1·min−1) = 64.5 ± 5.2
Medium-HIIT-1 (n = 10) 8 repetitions at 60% of time at
MAP, with 1:2 work-active-
recovery ratio, 2 times per
week
40-km cycling TT, VO2peak, MAP
Medium-HIIT-2 (n = 10) 8 repetitions at 60% of time at
MAP, with active recovery
based on heart rate recovery, 2
times per week
SIT (n = 10) 12 repetitions of 30-s at 185%
MAP with 4.0min of active
recovery, 2 times per week
Control (n = 11) Regular, low-intensity base
training program
Stepto etal. [40] Randomized controlled trial
(3weeks)
Cyclists (trained), sex = male,
age (years) = 26.3 ± 4.6, body
mass (kg) = 76.1 ± 11.1,
height (cm) = n/a, VO2peak
(mL·kg−1·min−1) = 63.1 ± 6.6
Short-HIIT (n = 4) 12 repetitions of 1.0min at
100% MAP with 4min of
active recovery, 2 times per
week
40-km cycling TT, VO2peak, MAP
Medium-HIIT (n = 4) 12 repetitions of 2min at 90%
MAP with 3min of active
recovery, 2 times per week
Long-HIIT (n = 4) 8 repetitions of 4min at 85%
MAP with 1.5min of active
recovery, 2 times per week
MIIT (n = 4) 4 repetitions of 8.0min at 80%
MAP with 1.0min of active
recovery, 2 times per week
SIT (n = 4) 12 repetitions of 30s at 175%
MAP with 4.5min of active
recovery, 2 times per week

1152 M.A.Rosenblat et al.
Table 2 Interval training program description
All values represent the mean value for each category; session external work is a product of interval repetitions, work duration and work intensity; total external work is a product of session
external work, program duration and session frequency
a The average training intensity for the respective subgroup was used due to missing data
Study Group Interval type Program
duration
(weeks)
Session
frequency
Interval
repeti-
tions
Work
duration
(min)
Work
intensity
(%)
Recovery mode Recovery
duration
(min)
Session
external work
(a.u.)
Total external
work (a.u.)
Akca and Aras [37] Medium-HIIT HIIT 4 2.0 8.0 2.5 75 Active 3.0 1500 12,000
SIT SIT 4 2.0 10.0 0.5 125 Active 4.0 625 5000
Esfarjani and Laursen [38] Medium-HIIT HIIT 10 2.0 8.0 3.3 87 Active 3.5 2296 45,936
SIT SIT 10 2.0 12.0 0.5 114 Active 4.5 684 13,680
Granata etal. [41] Long-HIIT HIIT 4 3.0 5.5 4.0 73 Active 2.0 1606 19,272
SIT SIT 4 3.0 6.6 0.5 168 Passive 4.0 554 6653
Inoue etal. [29] Long-HIIT HIIT 6 2.8 6.3 4.7 79aActive 4.0 2339 39,298
SIT SIT 6 2.8 7.5 0.5 153aActive 4.0 575 9664
Laursen etal. [39] Medium-HIIT-1 HIIT 4 2.0 8.0 2.4 100 Passive 4.8 1920 15,360
Medium-HIIT-2 HIIT 4 2.0 8.0 2.6 100 Passive 4.0 2640 16,640
SIT SIT 4 2.0 12.0 0.5 185 Passive 4.5 1110 8880
Stepto etal. [40] Short-HIIT HIIT 3 2.0 12.0 1.0 100 Active 4.0 2720 7200
Medium-HIIT HIIT 3 2.0 12.0 2.0 90 Active 3.0 2160 12,960
Long-HIIT HIIT 3 2.0 8.0 4.0 85 Active 1.5 1200 16,320
SIT SIT 3 2.0 12.0 0.5 175 Active 4.5 1050 6300

1153
Effect of Interval Training on Time-Trial Performance
to SIT (Table4). There was no significant difference in
VO2max/VO2peak between the HIIT and SIT groups in any of
the studies (Table5). With respect to MAP/MAV, there was
a significantly greater improvement following HIIT when
compared to SIT in 4 of the subgroups (Table6).
3.5 Synthesis ofResults
3.5.1 Training Load
Three of the studies used incremental tests that were greater
than 12min to determine VO2max/VO2peak [37, 38, 41];
therefore, the correction factor was applied to standardize
exercise intensity. The average session external work was
significantly different between the HIIT and SIT groups
(p < 0.0001), with average values of 1980 ± 475 (a.u.) and
766 ± 248 (a.u.), respectively. However, there was no statis-
tically significant difference in average total external work
between HIIT and SIT, with 20,554 ± 13,070 (a.u.) and 8363
± 3122 (a.u.). The average intensity performed by the HIIT
groups was 88% ± 11% MAP/MAV with an average inter-
val work duration of 2.9 ± 1.2min per work-bout. The SIT
group’s average training intensity was 153% ± 28 MAP/
MAV with an average work-bout duration of 30s. The total
external work was 7200 ± 0 (a.u.), 20,579 ± 13,904 (a.u.),
and 24,963 ± 12,502 (a.u.), for the short-HIIT, medium-
HIIT and long-HIIT groups, respectively. There was no
difference in session external work or total external work
between the medium-HIIT and long-HIIT subgroups.
3.5.2 Time‑Trial
There was no difference in the change in TT performance
when comparing HIIT versus SIT (0.9%; 90% CI −0.2%
to 1.9%, p = 0.18) (Fig. 2). The subgroup analysis indi-
cates that there was approximately a 2% greater improve-
ment following long-HIIT when compared to SIT (2.0%;
90% CI: 0.7% to 3.3%, p = 0.01), producing a large effect
(ES = 0.88). There was a significant difference between
subgroups (p = 0.009), with longer HIIT bouts producing a
greater improvement in performance.
3.5.3 Maximal Oxygen Consumption/Peak Oxygen
Consumption
There was no difference in change in VO2max/VO2peak
between any of the groups, including subgroups (Fig.3).
3.5.4 Maximal Aerobic Power/Maximal Aerobic Velocity
There was a moderate effect (ES = 0.70) in favor of HIIT
over SIT in MAP/MAV. This equates to a 2.4% greater
improvement following HIIT (2.4%; 90% CI 1.3–3.6%,
Table 3 Risk of bias of individual studies
Eligibility is not included in the final 10-point score
Study Eligibility Random
allocation
Concealed
allocation
Baseline
compari-
son
Blind subjects Blind therapists Blind assessors Adequate
follow-up
Inten-
tion-to-
treat
Between-
group com-
parison
Point esti-
mates/vari-
ability
Score
Akca and Aras [37] 0 1 0 1 0 0 0 0 0 1 1 4
Esfarjani and Laursen [38] 0 0 0 1 0 0 0 1 1 1 1 5
Granata etal. [41] 0 1 0 1 0 0 0 1 0 1 1 5
Inoue etal. [29] 0 1 1 1 0 0 0 0 0 1 1 5
Laursen etal. [39] 0 0 0 1 0 0 0 1 0 1 1 4
Stepto etal. [40] 0 1 0 1 0 0 0 1 0 1 1 5
Mean score 4.7

1154 M.A.Rosenblat et al.
p = 0.0007) (Fig.4). There was a trend in HIIT subgroup
duration and change in MAP/MAV, similar to that observed
in TT performance, indicating a greater improvement in
MAP/MAV with longer-duration interval bouts (p = 0.0003).
Long-HIIT produced the greatest increase in MAP/MAV,
with a 4% (p < 0.00001) higher change when compared to
SIT.
3.6 Risk ofBias Across Studies
A funnel plot of the standard difference in mean versus
standard error for TT results indicates that there is no evi-
dence of publication bias (p = 0.16) regarding the studies
included in the meta-analysis (Fig.5).
Table 4 Time-trial results
Negative between-group difference favours SIT, positive between-group difference favours HIIT
Study Measurement Group nPre (sec ± SD) Post (sec ± SD) Within-
group change
(% ± SD)
Between-group
difference (%; 90%
CI)
Cohen’s d
Akca and Aras [37] 2-km rowing Medium-HIIT 10 411.6 ± 7.5 406.6 ± 7.0 1.2 ± 1.2 −0.2; −1.1 to 0.8 −0.13
SIT 10 412.0 ± 7.7 406.3 ± 7.1 1.4 ± 1.4
Esfarjani and
Laursen [38]
3-km running Medium-HIIT 6 679.0 ± 38.5 –7.4 ± 2.6 4.0; 1.7 to 6.2 1.55
SIT 6 679.0 ± 32.0 –3.4 ± 2.1
Granata etal. [41] 20-km cycling Long-HIIT 11 2247.7 ± 147.5 2138.1 ± 90.7 4.9 ± 3.2 3.5; 1.4 to 5.6 1.13
SIT 9 2162.3 ± 143.1 2131.9 ± 165.1 1.4 ± 2.5
Inoue etal. [29] 40-km cycling Long-HIIT 7 6091.1 ± 478.3 5785.4 ± 387.3 5.0 ± 2.6 2.1; 0.2 to 4.0 0.9
SIT 9 6143.1 ± 445.7 5960.7 ± 417.0 3.0 ± 1.7
Laursen etal. [39] 40-km cycling Medium-HIIT-1 8 3419.5 ± 188.0 3259.9 ± 211.2 4.8 ± 2.8 0.5; −1.8 to 2.9 0.17
Medium-HIIT-2 9 3491.0 ± 202.7 3299.8 ± 267.3 5.5 ± 2.2 −1.2; −0.8 to 3.3 0.44
SIT 10 3451.0 ± 228.6 3304.3 ± 162.5 4.3 ± 3.2
Stepto etal. [40] 40-km cycling Short-HIIT 3 3618.4 ± 301.7 3608.2 ± 283.0 0.3 ± 0.7 −2.1; −3.8 to
−0.4
1.12
Medium-HIIT 4 3181.7 ± 39.3 3138.45 ± 106.0 1.4 ± 2.2 −0.9; −3.1 to 1.2 −0.44
Long-HIIT 4 3356.4 ± 156.5 3258.75 ± 123.9 2.9 ± 1.5 0.6; −1.4 to 2.6 0.29
SIT 4 3434.9 ± 209.7 3354.6 ± 165.0 2.3 ± 1.9
Table 5 Maximal oxygen consumption and peak oxygen consumption results
Negative between-group difference favours SIT, positive between-group difference favours HIIT
Study Group nPre (mL·min−1 ± SD) Post (mL·min−1 ± SD) Within-
group change
(% ± SD)
Between-group dif-
ference (%; 90% CI)
Cohen’s d
Akca and Aras [37] Medium-HIIT 10 4100.0 ± 650.0 4290.0 ± 630.0 4.6 ± 4.5 −0.8; −4.4 to 2.8 −0.15
SIT 10 4080.0 ± 660.0 4300.0 ± 610.0 5.4 ± 5.3
Granata etal. [41] Long-HIIT 11 3540.0 ± 298.0 3687.0 ± 348.0 4.2 ± 4.9 −2.2; −5.0 to 0.7 −0.51
SIT 9 3937.0 ± 718.0 4185.0 ± 707.0 6.3 ± 2.7
Laursen etal. [39] Medium-HIIT-1 8 4916.0 ± 485.0 5213.0 ± 470.0 6.0 ± 3.2 2.3; −0.7 to 5.2 0.55
Medium-HIIT-2 9 4982.0 ± 341.0 5242.0 ± 217.0 5.2 ± 2.8 1.4; −1.3 to 4.2 0.37
SIT 10 4776.0 ± 287.0 4956.0 ± 433.0 3.8 ± 4.4
Stepto etal. [40] Short-HIIT 3 4519.0 ± 1373.0 4430.0 ± 1228.0 −2.0 ± 4.3 −2.2; −8.0 to 3.6 −0.39
Medium-HIIT 4 5189.0 ± 501.0 5226.0 ± 356.0 0.7 ± 5.2 0.5; −5.5 to 6.5 0.08
Long-HIIT 4 4896.0 ± 248.0 5257.0 ± 491.0 7.4 ± 8.0 7.1; −0.6 to 14.9 0.93
SIT 4 4698.0 ± 381.0 4709.0 ± 563.0 0.2 ± 5.1

1155
Effect of Interval Training on Time-Trial Performance
4 Discussion
4.1 Summary ofEvidence
This is the first systematic review to measure changes in
TT performance following an interval training program.
Previous meta-analyses that studied the effects of inter-
val training have focused solely on VO2max as the primary
outcome [21, 28, 42–45]. Performance outcomes such as
VO2max may not account for individual physiological dif-
ferences at submaximal levels [46]. In fact, in a group of
athletes with the same VO2max, TT results are up to 10%
Table 6 Maximal aerobic power and maximal aerobic velocity results
Maximal aerobic power (MAP) is measured in watts, maximal aerobic velocity is measured in km·h−1, negative between-group difference
favours SIT, positive between-group difference favours HIIT
Study Measurement Group nPre
(mean ± SD)
Post (mean ± SD) Within-
group change
(% ± SD)
Between-group
difference (%; 90%
CI)
Cohen’s d
Akca and Aras [37] MAP HIIT 10 336.0 ± 20.0 351.0 ± 21.0 4.5 ± 4.3 −0.9; −4.5 to 2.7 −0.18
SIT 10 335.0 ± 24.0 353.0 ± 26.0 5.4 ± 5.2
Esfarjani and Laursen
[38]
MAV Medium-HIIT 6 15.6 ± 0.7 16.6 ± 0.6 6.4 ± 6.1 −1.4; −7.8 to 5.1 −0.19
SIT 6 15.4 ± 0.5 16.6 ± 0.6 7.8 ± 7.4
Granata etal. [41] MAP Long-HIIT 11 264.1 ± 37.4 293.2 ± 34.3 11.0 ± 3.6 6.6; 2.7 to 10.4 1.27
SIT 9 280.8 ± 48.2 293.3 ± 51.5 4.5 ± 6.3
Inoue etal. [29] MAP Long-HIIT 7 299.8 ± 24.6 323.1 ± 24.0 7.8 ± 3.3 2.7; −1.3 to 6.6 0.49
SIT 9 294.8 ± 22.9 310.0 ± 22.7 5.1 ± 6.1
Laursen etal. [39] MAP Medium-HIIT-1 8 439.0 ± 28.9 459.6 ± 37.4 4.7 ± 3.1 1.7; −0.7 to 4.1 0.51
Medium-HIIT-2 9 429.3 ± 23.7 459.1 ± 27.2 6.9 ± 2.1 3.9; 1.9 to 6.0 1.37
SIT 10 425.5 ± 32.4 438.3 ± 36.1 3.0 ± 3.2
Stepto etal. [40] MAP Short-HIIT 3 349.7 ± 95.2 354.7 ± 91.6 1.4 ± 1.3 1.0; −0.3 to 2.3 0.94
Medium-HIIT 4 403.3 ± 21.0 411.0 ± 25.6 1.9 ± 1.5 1.5; 0.2 to 2.9 1.17
Long-HIIT 4 389.8 ± 24.3 407.5 ± 26.0 4.6 ± 0.7 4.2; 3.4 to 4.9 5.89
SIT 4 371.8 ± 28.6 373.3 ± 30.0 0.4 ± 0.5
Fig. 2 Forest plot of time-trial results

1156 M.A.Rosenblat et al.
faster in those athletes with higher relative VTs [47]. In a
TT test, athletes are expected to complete a set distance in
the shortest possible time. This approach may provide a
similar experience to endurance events where power output
can fluctuate much like a TT assessment, thereby increasing
the external validity of the measurement [48]. TT results
have been shown to be a reliable measurement (ICC = 0.99)
and are highly correlated with cycling (r = 0.99, p < 0.001)
and running (r = 0.95, p = 0.001) race performance [25, 26].
However, the physiological demands of TT tests may make
them impractical for coaches to perform regularly since
the potential for athlete fatigue may require alterations in
training programs. Nevertheless, in endurance sport science
research, these tests may be the best method of evaluating
Fig. 3 Forest plot of maximal oxygen consumption and peak oxygen consumption
Fig. 4 Forest plot of maximal aerobic power and maximal aerobic velocity

1157
Effect of Interval Training on Time-Trial Performance
the effectiveness of training programs. These testing meth-
ods can then be translated into actual practice.
The results of this meta-analysis show that there is no
significant difference in TT performance change between
HIIT and SIT. There is some degree of commonality among
the various training programs in the studies making up the
meta-analysis. Specifically, all studies compared HIIT with
a SIT program that consisted of 30-s work-bouts followed by
a recovery period that would allow for full recovery between
bouts. However, the studies incorporated HIIT programs that
ranged in work-bout duration from 1 to 5min, making it dif-
ficult to generalize the results. Therefore, a subgroup analy-
sis was completed to analyze the results of HIIT programs of
similar work-bout duration. This subgroup analysis indicates
that longer-duration intervals may have a greater ability to
augment TT performance and MAP, when compared to SIT.
VO2max/VO2peak improved following both HIIT and SIT,
with no difference between interval subgroups. There were
conflicting results concerning the changes in VO2max and
TT performance between the studies by Stepto etal. and
Granata etal. [40, 41]. The distinction is likely due to the
design of the incremental test used to determine VO2max.
The Stepto study used 2.5-min stages, whereas the Gra-
nata study used 4-min stages. Tests that incorporate longer
stages may be more indicative of submaximal performance.
Time-to-exhaustion (TTE) at MAP from tests that use 2-min
stages is approximately 3.7min in highly trained cyclists
[49]; whereas, TTE at MAP following longer-duration tests
(3-min increments) can be as long as 6.8min in highly
trained cyclists [50]. This may be why there was a greater
improvement in TT and MAP with a lesser improvement in
VO2max in the study by Granata etal. when compared to the
study by Stepto etal.
4.2 Classication System forHigh‑Intensity Interval
Training
While exercise intensity is the main component that can be
used to classify an interval training program as either HIIT
or SIT, the duration of the interval itself is a very important
consideration. Previous inquiries into interval training have
manipulated the work–rest ratio to optimize time spent near
or at VO2max [17, 19, 20, 51]. These types of protocols can
produce a significant acute increase in oxygen uptake during
exercise but to a lesser extent than longer HIIT bouts [17,
20, 52]. With respect to training adaptations, HIIT that con-
sists of very short work-bouts (15–30s) may not appear to
be as effective at improving exercise economy as programs
that utilize longer HIIT bouts [19, 51]. Therefore, it may be
beneficial to program HIIT intervals using longer-duration
work-bouts to optimize endurance performance.
There have been a few attempts to classify HIIT exercise
based on the duration of the interval work-bout [4, 28]. In a
previous review, short-HIIT was considered to be work-bouts
under 30s, medium-HIIT as 30s–2min, and long-HIIT as
2–4min in duration [28]. However, there remain inconclu-
sive physiological justifications for choosing these ranges. It
may be appropriate to identify subgroups of interval training
by considering the relative contributions of energy system
(anaerobic vs. aerobic) components. Providing an accurate
Fig. 5 Funnel plot of standard-
ized mean difference versus
standard error of time-trial
performance

1158 M.A.Rosenblat et al.
classification for HIIT subgroups based on anaerobic con-
tributions can be challenging as previous literature suggests
that there is some difficulty in determining the degree of
anaerobic metabolism that takes place during HIIT [5].
Blood lactate levels as well as excess post-exercise oxygen
consumption (EPOC) are common methods that have been
used to determine anaerobic metabolism during interval
exercise although the reliability of such techniques remains
questionable [5].
It may be more productive to base categorization of inter-
vals on known oxygen uptake kinetics. Previous measure-
ments of oxygen uptake during high-intensity exercise indi-
cate that VO2max can be reached in as short a time as 117s
in trained individuals, while taking approximately 209s in
untrained individuals [53]. Since there is variability among
individuals of different training status, it would be appro-
priate to use work-bouts of 120s or less when defining an
anaerobic interval (short-HIIT) for trained individuals, as
the dominant source of total energy is supplied via anaero-
bic metabolism. Conversely, to perform an aerobic interval
(long-HIIT), more than 50% of the total work completed
should consist of exercise at VO2max. Therefore, long-HIIT
should comprise of bouts of exercise that are at least 4min
in duration. Finally, medium-HIIT would incorporate inter-
val exercise that is between 2 and 4min in duration. This
method of classification does not provide an individualized
approach to interval programming. Nevertheless, it does
specify the criteria for athletes and coaches to make an
informed decision concerning their training.
With respect to aerobic metabolism, the measurement
of oxygen uptake can become delayed in the heavy and
severe exercise domains due to the development of the VO2
slow-component which has been described as the continued
increase in oxygen consumption after 3min of constant-load
exercise [54]. Individually determined oxygen kinetics may
be a novel approach to classify HIIT but those data are not
available in the reviewed studies and as a consequence fixed
time points were employed. Future investigations are encour-
aged to examine this approach for evidence-based decision-
making that is specific to sedentary and athletic populations.
4.3 Limitations
There are elements regarding design of the individual studies
that may influence the results of the meta-analysis. Three of
the studies did not randomize participants into their respec-
tive groups [38, 39, 55] and only one included concealed
group allocation.These methodological issues increase the
possibility that the results will be swayed by confounding
variables. In addition, they may influence the decision as to
whether a study participant receives treatment [56]. With
respect to randomization, four of the studies did not match
participants by VO2max/VO2peak or TT results, potentially
producing non-homogeneity in the training groups [39]. The
lack of assessor and participant blinding in all the studies
might introduce significant bias. When subjects are blinded,
it is less likely that the results of treatment are due to a pla-
cebo effect [56]. Blinding assessors prevents their personal
bias from affecting the results [56].
There are a limited number of studies that compare the
effects of HIIT and SIT on TT performance in a healthy,
active population. Six studies met the inclusion criteria for
the qualitative and quantitative analysis (Fig.1). The sub-
group analysis was performed to provide further insight into
the effects of HIIT programs differing in duration. There
were only data from one study for a comparison of the short-
HIIT group, limiting the ability to provide a strong con-
clusion for this analysis. There were 5 comparisons using
medium-HIIT and 3 using long-HIIT (Fig.2). Previous
reviews have included as few as 4 studies with only 2 groups
for subgroup analysis [57]. However, while there are reviews
that include a small sample size, there is a greater risk of
error due to the heterogeneity of the individual studies. In
addition, the small sample size of the individual studies, spe-
cifically the study by Stepto etal. [40], may skew the results
by increasing the variability. Therefore, future investigations
should include larger sample sizes when feasible.
There remains some debate around participant character-
istics that can influence the response to a training interven-
tion. Participant age and fitness level are two variables that,
in theory, can affect the outcome of an intervention. The
studies in the current review included participants between
the ages of 19 and 32years of age. Due to the small range,
it is unlikely that age differences could have affected the
results. A study by Støren etal. directly compared the effects
of age on changes in VO2max following an interval training
program [58]. They included males (n = 72) and females
(n = 22) between the ages of 20 and 70 + years, and found
no difference in relative improvement in VO2max [58].
Training status as determined by baseline VO2max may
affect the magnitude of a training response as there are
physiological limits of oxygen consumption [59]. There
was a broad range of baseline VO2max/VO2peak values in the
studies included in this review, with values ranging from
46.0 to 64.5mL·kg−1·min−1. This is meaningful because
the study by Støren etal. found a significant difference in
individual training response, with those participants who
were inactive demonstrating the greatest improvement in
performance. In contrast, a recent study of well-trained
cyclists (VO2max = 57.9 ± 6.8mL·kg−1·min−1) displayed a
15% improvement in VO2max following an interval training
program, indicating that interval training can elicit a sub-
stantial response in highly trained individuals [60]. However,
changes in VO2max may not appropriately indicate improve-
ments in submaximal endurance performance. The results
of study by Granata etal. show the greatest in improvement

1159
Effect of Interval Training on Time-Trial Performance
in MAP and TT performance following long-HIIT when
compared to the other studies. Baseline measurements of
VO2max for the participants in that study was significantly
lower than in the other studies, indicating that initial train-
ing status can influence the magnitude of improvements in
TT performance. However, there was no significant differ-
ence in the results when the data were removed from the
meta-analysis.
The studies included in this review incorporated three
different modes of exercise including cycling [29, 39–41],
rowing [37] and running [38]. Exercise mode may influence
the acute physiological responses that occur during interval
exercise, producing differences in long-term adaptations.
Cycling and running have been shown to elicit different
acute responses during exercise such as variations in oxygen
uptake kinetics, peak oxygen consumption, skeletal muscle
oxidative capacity and neuromuscular responses [61]. There
is also literature that demonstrates similar physiological dif-
ferences between cycling and rowing [62]. To the best of our
knowledge, there are no studies that compare the effects of
an interval training program using different modes of exer-
cise on measurements of endurance performance. Due to the
differences in the acute responses, it is likely that there is a
difference in the magnitude of the response between exercise
modes, potentially limiting the generalizability of the results.
However, further investigations are required to determine if
variations in training adaptations occur.
There was a significant difference in training load between
the HIIT and SIT groups. However, the variance in training
load may not influence performance. A recent meta-analysis
showed an inverse relationship between training load and
performance following SIT, with a decline in improvement
for every two SIT bouts above 4 repetitions [45]. Therefore,
it is unlikely that the difference in performance was as a
result of variations in external work. There was no differ-
ence in total external work between the medium-HIIT and
long-HIIT groups. We believe that the greater performance
improvement following long-HIIT is likely due to the higher
aerobic contribution of a longer interval bout. There was
no difference in changes in VO2max/VO2peak between any of
the groups. However, there was a greater increase in MAP/
MAV following long-HIIT. Since MAP/MAV is an indica-
tor of exercise economy [63], the greater degree of aerobic
metabolism that occurs during long-HIIT likely would pro-
duce improvements that contribute to submaximal exercise.
The differences in the distance for the TT measures used
to assess performance may influence the results of the meta-
analysis. Previous inquiry suggests that there is a greater
anaerobic contribution to the TTs of shorter duration [64].
The majority of the TTs included in this review incorpo-
rated longer-duration tests (between 20 and 40km) [39–41],
which require a greater aerobic contribution [64]. Therefore,
it could be postulated that the results of the review would
favor long-HIIT due to the aerobic nature of the interven-
tion. Two studies in the meta-analysis included TTs that
were between 6 and 12min in duration [37, 38]. However,
there was no significant change in the results when they are
removed from the analysis.
4.4 Practical Application
The results of the subgroup analysis show long-HIIT to be
the most beneficial form of interval training to augment
performance. While exercise intensity is one of the most
important variables to consider when programming interval
training [1], work-bout duration may be essential to opti-
mize changes in performance. Interval training that includes
work-bouts between 4 and 6min in duration at an intensity
between VT2 and VO2max with 2–4min of recovery would
be best to improve TT performance.
4.5 Future Direction
This is the first analysis to directly compare HIIT and SIT
with markers of endurance sport performance. The results of
the individual studies indicate that SIT may provide benefits
for shorter-duration TTs and HIIT for longer TTs. It may
be beneficial to investigate this hypothesis through a direct
comparison of HIIT and SIT on short- and long-duration
TTs. The subgroup analysis shows that different adaptations
may occur as a result of manipulating the interval work-
bout duration. The study by Stepto etal. directly compared
short-, medium- and long-HIIT groups on TT performance;
however, due to small sample size,a statistically significant
change in the performance variables was not detected [40].
Future studies with larger sample sizes that directly com-
pare medium-HIIT and long-HIIT could provide additional
insight into the benefit of these programs.
5 Conclusion
The subgroup analysis indicates that there was a large effect
in TT performance favoring long-HIIT over SIT. There-
fore, the results based on the currently available literature
suggest that longer-duration HIIT bouts may provide opti-
mal performance adaptation and should be incorporated in
an endurance training program. Additional research that
directly compares HIIT exercise differing in work-bout
duration would strengthen these results and provide further
insight into the mechanisms behind the observed benefits
of long-HIIT.

1160 M.A.Rosenblat et al.
Compliance with Ethical Standards
Funding No sources of funding were used to assist in the preparation
of this article.
Conflicts of interest Michael Rosenblat, Andrew Perrotta and Scott
Thomas declare that they have no conflicts of interest relevant to the
content of this review.
Data availability statement All data supporting the results in this man-
uscript are available within the results sections or in the cited articles.
References
1. Londeree BR. Effect of training on lactate/ventilatory thresholds:
a meta-analysis. Med Sci Sports Exerc. 1997;29(6):837–43.
2. Seiler S, Kjerland GO. Quantifying training intensity distribu-
tion in elite endurance athletes: is there evidence for an “optimal”
distribution? Scand J Med Sci Sports. 2006;16(1):49–56.
3. Rosenblat MA, Perrotta AS, Vicenzino B. Polarized vs. threshold
training intensity distribution on endurance sport performance: a
systematic review and meta-analysis of randomized controlled
trials. J Strength Cond Res. 2019;33(12):3491–500.
4. Buchheit M, Laursen PB. High-intensity interval training, solu-
tions to the programming puzzle: part I: cardiopulmonary empha-
sis. Sports Med. 2013;43(5):313–38.
5. Buchheit M, Laursen PB. High-intensity interval training, solu-
tions to the programming puzzle. Part II: anaerobic energy,
neuromuscular load and practical applications. Sports Med.
2013;43(10):927–54.
6. Gaesser GA, Poole DC. The slow component of oxygen uptake
kinetics in humans. Exerc Sports Sci Rev. 1996;24:35–71.
7. Billat LV. Interval training for performance: a scientific and
empirical practice. Special recommendations for middle- and
long-distance running. Part I: aerobic interval training. Sports
Med. 2001;31(1):13–31.
8. Gibala MJ, Little JP, Macdonald MJ, Hawley JA. Physiological
adaptations to low-volume, high-intensity interval training in
health and disease. J Physiol. 2012;590(5):1077–84.
9. Viana RB, de Lira CAB, Naves JPA, Coswig VS, Del Vecchio FB,
Ramirez-Campillo R, etal. Can we draw general conclusions from
interval training studies? Sports Med. 2018;48(9):2001–9.
10. Edgett BA, Foster WS, Hankinson PB, Simpson CA, Little JP,
Graham RB, etal. Dissociation of increases in PGC-1alpha and its
regulators from exercise intensity and muscle activation following
acute exercise. PLoS One. 2013;8(8):e71623.
11. Kaikkonen P, Hynynen E, Mann T, Rusko H, Nummela A. Heart
rate variability is related to training load variables in interval run-
ning exercises. Eur J Appl Physiol. 2012;112(3):829–38.
12. Olney N, Wertz T, LaPorta Z, Mora A, Serbas J, Astorino TA.
Comparison of acute physiological and psychological responses
between moderate-intensity continuous exercise and three
regimes of high-intensity interval training. J Strength Cond Res.
2018;32(8):2130–8.
13. Paquette M, Le Blanc O, Lucas SJ, Thibault G, Bailey DM, Bras-
sard P. Effects of submaximal and supramaximal interval training
on determinants of endurance performance in endurance athletes.
Scand J Med Sci Sports. 2017;27(3):318–26.
14. Raleigh JP, Giles MD, Scribbans TD, Edgett BA, Sawula LJ, Bon-
afiglia JT, etal. The impact of work-matched interval training on
VO2peak and VO2 kinetics: diminishing returns with increasing
intensity. Appl Physiol Nutr Metab. 2016;41(7):706–13.
15. Wood KM, Olive B, LaValle K, Thompson H, Greer K, Astorino
TA. Dissimilar physiological and perceptual responses between
sprint interval training and high-intensity interval training. J
Strength Cond Res. 2016;30(1):244–50.
16. Tiidus PM, Tupling AR, Houston ME. Biochemistry primer for
exercise science. 4th ed. Windsor: Human Kinetics; 2012.
17. Ballor DL, Volovsek AJ. Effect of exercise to rest ratio on plasma
lactate concentration at work rates above and below maximum
oxygen uptake. Eur J Appl Physiol. 1992;65(4):365–9.
18. Billat LV, Slawinksi J, Bocquet V, Chassaing P, Demarle A,
Koralsztein JP. Very short (15s–15s) interval-training around
the critical velocity allows middle-aged runners to maintain VO2
max for 14 minutes. Int J Sports Med. 2001;22(3):201–8.
19. Helgerud J, Høydal K, Wang E, Karlsen T, Berg P, Bjerkaas M,
etal. Aerobic high-intensity intervals improve VO2max more than
moderate training. Med Sci Sports Exerc. 2007;39(4):665–71.
20. Zafeiridis A, Kounoupis A, Dipla K, Kyparos A, Nikolaidis M,
Smilios I, etal. Oxygen delivery and muscle deoxygenation dur-
ing continuous, long- and short-interval exercise. Int J Sports
Med. 2015;36(11):872–80.
21. Milanović Z, Sporis G, Weston M. Effectiveness of high-inten-
sity interval training (HIT) and continuous endurance training for
VO2max improvements: a systematic review and meta-analysis of
controlled trials. Sports Med. 2015;45(10):1469–81.
22. Foster C. VO2max and training indices as determinants of competi-
tive running performance. J Sports Sci. 1983;1(1):13–22.
23. Coyle EF, Coggan AR, Hopper MK, Walters TJ. Determi-
nants of endurance in well-trained cyclists. J Appl Physiol.
1988;64(6):2622–30.
24. Joyner MJ, Coyle EF. Endurance exercise performance: the physi-
ology of champions. J Physiol. 2008;586(1):35–44.
25. Palmer GS, Dennis SC, Noakes TD, Hawley JA. Assessment of
the reproducibility of performance testing on an air-braked cycle
ergometer. Int J Sports Med. 1996;17(4):293–8.
26. Russell RD, Redmann SM, Ravussin E, Hunter GR, Larson-
Meyer DE. Reproducibility of endurance performance on a
treadmill using a preloaded time trial. Med Sci Sports Exerc.
2004;36(4):717–24.
27. Laursen PB, Francis GT, Abbiss CR, Newton MJ, Nosaka K.
Reliability of time-to-exhaustion versus time-trial running tests
in runners. Med Sci Sports Exerc. 2007;39(8):1374–9.
28. Wen D, Utesch T, Wu J, Robertson S, Liu J, Hu G, etal. Effects
of different protocols of high intensity interval training for VO2max
improvements in adults: a meta-analysis of randomised controlled
trials. J Sci Med Sport. 2019;22(8):941–7.
29. Inoue A, Impellizzeri FM, Pires FO, Pompeu FA, Deslandes AC,
Santos TM. Effects of sprint versus high-intensity aerobic interval
training on cross-country mountain biking performance: a rand-
omized controlled trial. PLoS One. 2016;11(1):e0145298.
30. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioan-
nidis JP, etal. The PRISMA statement for reporting systematic
reviews and meta-analyses of studies that evaluate health care
interventions: explanation and elaboration. J Clin Epidemiol.
2009;62(10):e1–34.
31. Granata C, Jamnick NA, Bishop DJ. Principles of exercise pre-
scriptoin and how they influence exercise-induced changes of tran-
scription factors and other regulators of mitochondrial biogenesis.
Sports Med. 2018;48(7):1541–59.
32. Kamper SJ, Moseley AM, Herbert RD, Maher CG, Elkins MR,
Sherrington C. 15 years of tracking physiotherapy evidence on
PEDro, where are we now? Br J Sports Med. 2015;49(14):907–9.
33. Hedges L, Olkin I. Statistical methods for meta-analysis. New
York: Academic Press; 1981.
34. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive
statistics for studies in sports medicine and exercise science. Med
Sci Sports Exerc. 2009;41(1):3–12.

1161
Effect of Interval Training on Time-Trial Performance
35. Lau J, Ioannidis JP, Schmid CH. Summing up evidence: one
answer is not always enough. Lancet. 1998;351(9096):123–7.
36. Caputo F, Mello MT, Denadai BS. Oxygen uptake kinetics
and time to exhaustion in cycling and running: a comparison
between trained and untrained subjects. Arch Physiol Biochem.
2003;111(5):461–6.
37. Akca F, Aras D. Comparison of rowing performance improve-
ments following various high-intensity interval training. J Strength
Cond Res. 2015;29(8):2249–54.
38. Esfarjani F, Laursen PB. Manipulating high-intensity interval
training: effects on VO2max, the lactate threshold and 3000m run-
ning performance in moderately trained males. J Sci Med Sport.
2007;10(1):27–35.
39. Laursen PB, Shing CM, Peake JM, Coombes JS, Jenkins DG.
Interval training program optimization in highly trained endurance
cyclists. Med Sci Sports Exerc. 2002;34(11):1801–7.
40. Stepto NK, Hawley JA, Dennis SC, Hopkins WG. Effects of differ-
ent interval-training programs on cycling time-trial performance.
Med Sci Sports Exerc. 1999;31(5):736–41.
41. Granata C, Oliveira RS, Little JP, Renner K, Bishop DJ. Train-
ing intensity modulates changes in PGC-1alpha and p53 pro-
tein content and mitochondrial respiration, but not markers of
mitochondrial content in human skeletal muscle. FASEB J.
2016;30(2):959–70.
42. Bacon AP, Carter RE, Ogle EA, Joyner MJ. VO2max trainability
and high intensity interval training in humans: a meta-analysis.
PLoS One. 2013;8(9):e73182.
43. 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.
44. Sloth M, Sloth D, Overgaard K, Dalgas U. Effects of sprint
interval training on VO2max and aerobic exercise performance:
a systematic review and meta-analysis. Scand J Med Sci Sports.
2013;23(6):e341–52.
45. Vollaard NB, Metcalfe RS, Williams S. Effect of number of sprints
in an SIT session on change in VO2max: a meta-analysis. Med Sci
Sports Exerc. 2017;49(6):1147–56.
46. Seiler S. What is best practice for training intensity and duration
distribution in endurance athletes? Int J Sports Physiol Perform.
2010;5(3):276–91.
47. Coyle EF, Feltner ME, Kautz SA, Hamilton MT, Montain SJ, Bay-
lor AM, etal. Physiological and biomechanical factors associated
with elite endurance cycling performance. Med Sci Sports Exerc.
1991;23(1):93–107.
48. Karsten B, Baker J, Naclerio F, Klose A, Bianco A, Nimmerichter
A. Time trials versus time-to-exhaustion tests: effects on critical
power, W’, and oxygen-uptake kinetics. Int J Sports Physiol Per-
form. 2018;13(2):183–8.
49. de Lucas RD, de Souza KM, Costa VP, Grossl T, Guglielmo LG.
Time to exhaustion at and above critical power in trained cyclists:
the relationship between heavy and severe intensity domains. Sci
Sports. 2013;28(1):e9–14.
50. Billat V, Faina M, Sardella F, Marini C, Fanton F, Lupo S,
etal. A comparison of time to exhaustion at VO2 max in elite
cyclists, kayak paddlers, swimmers and runners. Ergonomics.
1996;39(2):267–77.
51. Rønnestad BR, Hansen J, Vegge G, Tonnessen E, Slettalokken
G. Short intervals induce superior training adaptations compared
with long intervals in cyclists—an effort-matched approach.
Scand J Med Sci Sports. 2015;25(2):143–51.
52. Naves JPA, Rebelo ACS, Silva L, Silva MS, Ramirez-Campillo
R, Ramirez-Velez R, et al. Cardiorespiratory and perceptual
responses of two interval training and a continuous training pro-
tocol in healthy young men. Eur J Sport Sci. 2019;19(5):653–60.
53. Caputo F, Denadai BS. The highest intensity and the shortest
duration permitting attainment of maximal oxygen uptake during
cycling: effects of different methods and aerobic fitness level. Eur
J Appl Physiol. 2008;103(1):47–57.
54. Poole DC, Jones AM. Oxygen uptake kinetics. Compr Physiol.
2012;2(2):933–96.
55. Matsuo T, Saotome K, Seino S, Shimojo N, Matsushita A,
Iemitsu M, etal. Effects of a low-volume aerobic-type interval
exercise on VO2max and cardiac mass. Med Sci Sports Exerc.
2014;46(1):42–50.
56. PEDro Scale. 1999 [cited 1999 February 26, 2019]; Physiotherapy
Evidence Database]. https ://www.pedro .org.au/engli sh/downl
oads/pedro -scale /.
57. Heerey JJ, Kemp JL, Mosler AB, Jones DM, Pizzari T, Scholes
MJ, etal. What is the prevalence of hip intra-articular pathologies
and osteoarthritis in active athletes with hip and groin pain com-
pared with those without? A systematic review and meta-analysis.
Sports Med. 2019;49(6):951–72.
58. Støren O, Helgerud J, Saebo M, Støa EM, Bratland-Sanda
S, Unhjem RJ, etal. The effect of age on the VO2max response
to high-intensity interval training. Med Sci Sports Exerc.
2017;49(1):78–85.
59. Bassett DR, Howley ET. Limiting factors for maximum oxygen
uptake and determinants of endurance performance. Med Sci
Sports Exerc. 2000;32(1):70–84.
60. Hebisz P, Hebisz R, Zaton M, Ochmann B, Mielnik N. Concomi-
tant application of sprint and high-intensity interval training on
maximal oxygen uptake and work output in well-trained cyclists.
Eur J Appl Physiol. 2016;116(8):1495–502.
61. Millet GP, Bentley DJ. Physiological differences between
cycling and running: lessons from triathletes. Sports Med.
2009;39(3):179–206.
62. Lindenthaler JR, Rice AJ, Versey NG, McKune AJ, Welvaert M.
Differences in physiological responses during rowing and cycle
ergometry in elite male rowers. Front Physiol. 2018;9:1010.
63. Billat LV, Koralsztein JP. Significance of the velocity at
VO2max and time to exhaustion at this velocity. Sports Med.
1996;22(2):90–108.
64. Borszcz FK, Tramontin AF, de Souza KM, Carminatti LJ,
Costa VP. Physiological correlations with short, medium,
and long cycling time-trial performance. Res Q Exerc Sport.
2018;89(1):120–5.