ArticlePDF AvailableLiterature Review

The Acute Effect of High-Intensity Exercise on Executive Function: A Meta-Analysis


Abstract and Figures

High-intensity exercise has recently emerged as a potent alternative to aerobic regimens, with ramifications for health and brain function. As part of this trend, single sessions of intense exercise have been proposed as powerful, noninvasive means for transiently enhancing cognition. However, findings in this field remain mixed, and a thorough synthesis of the evidence is lacking. Here, we synthesized the literature in a meta-analysis of the acute effect of high-intensity exercise on executive function. We included a total of 1,177 participants and 147 effect sizes across 28 studies and found a small facilitating effect ( d = 0.24) of high-intensity exercise on executive function. However, this effect was significant only compared with rest ( d = 0.34); it was not significant when high-intensity exercise was compared with low-to-moderate intensity exercise ( d = 0.07). This suggests that intense and moderate exercise affect executive function in a comparable manner. We tested a number of moderators that together explained a significant proportion of the between-studies variance. Overall, our findings indicate that high-intensity cardiovascular exercise might be a viable alternative for eliciting acute cognitive gains. We discuss the potential of this line of research, identify a number of challenges and limitations it faces, and propose applications to individuals, society, and policies.
Content may be subject to copyright.
Perspectives on Psychological Science
1 –31
© The Author(s) 2019
Article reuse guidelines:
DOI: 10.1177/1745691619850568
Our modern societies are becoming increasingly inac-
tive. On average, people spend 5 to 9 hr of their waking
day engaged in sedentary behaviors (Al-Nakeeb etal.,
2012; Matthews etal., 2008; Ruiz etal., 2011), with
detrimental effects on physiological and psychological
health (Iannotti etal., 2009; Tremblay, Colley, Saunders,
Healy, & Owen, 2010). In contrast, physical exercise
has a positive impact on the body (Warburton & Bredin,
2017) and the brain (Gomez-Pinilla & Hillman, 2013),
including several well-documented neurobiological
effects (for a review, see Moreau & Conway, 2013).
Despite the sizeable benefits associated with active
lifestyles, many individuals do not exercise regularly,
the most common reasons being related to a lack of
motivation or time constraints (Anderson, 2003). In an
effort to propose exercise regimens that are more effi-
cient and engaging, researchers have explored different
ways to shorten exercise regimens. Among these, high-
intensity regimens have recently gained popularity
because of their potential to elicit health benefits
comparable with, and sometimes even surpassing, those
elicited by longer workouts (Gibala, Little, Macdonald,
& Hawley, 2012; Weston, Wisløff, & Coombes, 2014).
These regimens come in various forms but are typically
based on intense cardiovascular exercise interleaved
with rest periods.
The long-term impact of physical exercise, including
the chronic effects of high-intensity regimens, has been
studied extensively (for a general review, see Hillman,
Erickson, & Kramer, 2008). This body of research has
demonstrated that chronic exercise exerts a positive
influence on a wide range of variables, including those
related to cognitive performance (Moreau & Conway,
2013; Moreau, Kirk, & Waldie, 2017). Although long-term
850568PPSXXX10.1177/1745691619850568Moreau, ChouExercise and Executive Function
Corresponding Author:
David Moreau, Science Centre, School of Psychology, 23 Symonds St.,
Auckland 1010, New Zealand
The Acute Effect of High-Intensity
Exercise on Executive Function:
A Meta-Analysis
David Moreau1,2 and Edward Chou1
1School of Psychology, The University of Auckland; and 2Centre for Brain Research,
The University of Auckland
High-intensity exercise has recently emerged as a potent alternative to aerobic regimens, with ramifications for
health and brain function. As part of this trend, single sessions of intense exercise have been proposed as powerful,
noninvasive means for transiently enhancing cognition. However, findings in this field remain mixed, and a thorough
synthesis of the evidence is lacking. Here, we synthesized the literature in a meta-analysis of the acute effect of high-
intensity exercise on executive function. We included a total of 1,177 participants and 147 effect sizes across 28 studies
and found a small facilitating effect (d = 0.24) of high-intensity exercise on executive function. However, this effect
was significant only compared with rest (d = 0.34); it was not significant when high-intensity exercise was compared
with low-to-moderate intensity exercise (d = 0.07). This suggests that intense and moderate exercise affect executive
function in a comparable manner. We tested a number of moderators that together explained a significant proportion
of the between-studies variance. Overall, our findings indicate that high-intensity cardiovascular exercise might be a
viable alternative for eliciting acute cognitive gains. We discuss the potential of this line of research, identify a number
of challenges and limitations it faces, and propose applications to individuals, society, and policies.
cognitive control, inhibitory control, cognitive flexibility, attention, working memory, acute physical exercise, HIT
2 Moreau, Chou
commitment to exercise is key in inducing profound,
lasting changes, studying the short-term effects of exer-
cise remains important in further understanding the
mechanisms subserving change at different time scales.
In this dynamic, several theories have attempted to explain
the effects of acute exercise on cognition (Tomporowski,
2003), emphasizing in the process the fundamental dis-
crepancies between short-term effects and long-term, or
chronic, outcomes. For example, the reticular-activating
hypofrontality model (Audiffren, 2016) posits that acute
exercise forces the brain to shift metabolic resources away
from specific regions such as the prefrontal cortex to
instead favor structures that support exercise, such as the
reticular formation and motor cortices. This process typi-
cally facilitates performance on sensory and motor tasks,
whereas the associated hypofrontality is thought to tem-
porarily impair executive function (Audiffren, 2016;
Dietrich & Audiffren, 2011). A number of electroencepha-
lography studies support this view, reporting postexercise
alterations in the form of shifts in the amplitude or latency
of event-related potential components thought to reflect
an increased allocation of cognitive resources after exer-
cise (Hillman, Snook, & Jerome, 2003; Kamijo etal., 2009;
O’Leary, Pontifex, Scudder, Brown, & Hillman, 2011).
Exercise also leads to acute changestypically
increasesin the concentration of several neurochemi-
cals in the brain (for a review, see Moreau & Conway,
2013). Examples of these neurochemicals include the
catecholamines (noradrenaline, dopamine), cortisol,
brain-derived neurotrophic factor, and possibly sero-
tonin (McMorris, Turner, Hale, & Sproule, 2016). These
neurochemicals influence brain function in a complex
manner: Whereas moderate increases in catecholamine
concentrations appear to facilitate performance on most
cognitive tasks, excessive concentrations (from high-
intensity and/or prolonged exercise) can inhibit cogni-
tion instead (McMorris etal., 2016), consistent with the
inverted-U hypothesis (Yerkes & Dodson, 1908). In addi-
tion, physical and emotional fatigue (Barnes & Van Dyne,
2009) may impair cognitive performance at higher inten-
sities of exercise.
Although neurochemical accounts and the reticular-
activating hypofrontality model can help explain the
impact of exercise on cognition while exercising, they
might tell us very little about postexercise effects. This
question has practical importance given that many indi-
viduals may be motivated to exercise during breaks if
it has beneficial effects on both physical health and
cognitive performance. Yet compared with that of low-
to-moderate-intensity exercise (e.g., Ludyga, Gerber,
Brand, Holsboer-Trachsler, & Pühse, 2016), the acute
effect of high-intensity exercise on subsequent cogni-
tive performance is not well understood (Browne etal.,
2017; McMorris, 2016). Exercise-induced impairments
in cognitive performance associated with high-intensity
training could be temporary and thus not representative
of subsequent effects after the bout. For example, one
might posit that detrimental effects would rapidly sub-
side to allow general cognitive improvements similar to,
or even surpassing, those observed after aerobic exer-
cise (Kao, Westfall, Soneson, Gurd, & Hillman, 2017;
Samuel etal., 2017). This would be consistent with the
profound physiological effects but rapid physical recov-
ery typically observed following these types of regi-
mens (MacInnis & Gibala, 2017). In fact, a number of
studies have found facilitating effects of high-intensity
exercise on postexercise cognitive performance (Kao
etal., 2017) despite the well-established impairments
reported during high-intensity exercise (Samuel etal.,
2017). On the other hand, it is also possible that the
debilitating effects of high intensity on cognition remain
after the bout (Ludyga, Pühse, Lucchi, Marti, & Gerber,
2019; Mekari etal., 2015) or that the effect of high-intensity
exercise on cognitive performance depends on modera-
tors such as fitness level or age (Browne etal., 2017). In
line with this idea, several high-intensity exercise studies
have failed to find facilitating effects (Browne etal., 2017;
McMorris & Hale, 2012).
Moreover, more fine-grained data might be required
to truly understand the dynamics of exercise-induced
effects on cognition. For example, a meta-analysis by
Chang and colleagues found impairing effects up to 1
min after the bout and facilitating effects beyond 1 min
after exercise (Y. K. Chang, Labban, Gapin, & Etnier,
2012). To complicate things further, methodological dif-
ferences between studies have possibly contributed to
these discrepancies; for example, the threshold for
high-intensity exercise varies substantially across pro-
tocols (McMorris, 2015). Specifically, the following
thresholds, which are based on either maximum heart
rate (HRmax) or maximum power output (Wmax), are
common in the literature: (a) 77% HRmax (American
College of Sports Medicine, 2010; Y. K. Chang etal.,
2012) and (b) 80% Wmax (Borer, 2003; Browne etal.,
2017; McMorris, Hale, Corbett, Robertson, & Hodgson,
2015). Using conversion formulas (Arts & Kuipers, 1994;
Lounana, Campion, Noakes, & Medelli, 2007), (a) can
be converted to approximately 59.5% Wmax, and (b)
corresponds to 88.6% HRmax. In addition, a wide range
of populations is included in these studies; for example,
the age range spans from children to older adults,
whereas fitness levels range from sedentary to highly
active, or even professionally trained, individuals. These
differences in thresholds, units of measurements, and
demographics may account for some of the inconsisten-
cies between studies.
The aforementioned inconsistencies are best
addressed via meta-analytic investigations, yet previous
Exercise and Executive Function 3
reviews and meta-analyses have included only a rela-
tively low number of studies. The most recent literature
search for a meta-analysis on the acute effect of exer-
cise on cognition was conducted in 20101 (Y. K. Chang
etal., 2012). In the 9 years since, many studies have
been published, including several that examined inter-
mittent forms of high-intensity exercise typically shorter
than traditional aerobic exercise (Klika & Jordan, 2013)
and often associated with higher self-reported enjoy-
ment (Thum, Parsons, Whittle, & Astorino, 2017). In
addition, several potential moderators of the influence
of exercise on cognition have been identified (Y. K.
Chang etal., 2012; Etnier, Nowell, Landers, & Sibley,
2006; Etnier etal., 1997; Lambourne & Tomporowski,
2010; McMorris & Hale, 2012; Sibley & Etnier, 2003;
Tomporowski, 2003), providing the basis for a deeper
understanding at the mechanistic level.
Present Study
In the present meta-analysis, we focused on the effect
of cardiovascular high-intensity exercise on executive
function. Although restrictive, our emphasis on execu-
tive performance was motivated by three distinct fac-
tors. First, executive function is a central component of
cognition known to influence many cognitive processes
and with ecological relevance to various domains, rang-
ing from academic to professional (e.g., Diamond,
2013). Second, intervention studies have demonstrated
repeatedly that executive performance is malleable and
that it can be improved given adequate regimens or
suitable environments (e.g., Diamond & Lee, 2011).
Finally—and perhaps most importantly—there are clear
theoretical predictions, generated on the basis of previ-
ous literature (e.g., Audiffren, 2016; Dietrich & Audiffren,
2011), about the effects of exercise on executive func-
tion. In our view, predictions related to other cognitive
domains are less well informed theoretically and thus
more prone to spurious findings.
We aimed to answer two research questions. First,
what is the effect of a single bout of high-intensity
exercise on executive performance? This question is
important given the mixed findings reported in previ-
ous studies. Second, how is the effect of high-intensity
exercise on executive performance moderated by the
characteristics of exercise, cognitive tasks, research pro-
tocols, and participants? We predicted an effect of high-
intensity exercise on executive function but did not
predict a direction because of inconsistencies in previ-
ous studies (Y. K. Chang etal., 2012; Ludyga etal., 2016;
Verburgh, Königs, Scherder, & Oosterlaan, 2014). We
also postulated larger nondirectional effects on the
basis of previous findings for longer exercise durations
(Browne etal., 2017; Y. K. Chang etal., 2012; Ludyga
etal., 2016; Miller, Hanson, Tennyck, & Plantz, 2019)
and a greater facilitating effect for cycling compared
with running protocols (Lambourne & Tomporowski,
2010). Because of the relative scarcity of studies exam-
ining subcategories of high-intensity exercise and exer-
cise rhythm, we did not make predictions for these
moderators. We also hypothesized a greater facilitating
effect for cognitive tasks administered after a minute
following exercise (Browne etal., 2017; Y. K. Chang
et al., 2012). Finally, we predicted larger effects in
within-subjects (vs. between-groups) studies (Y. K.
Chang etal., 2012) and lower-quality (vs. higher-quality)
studies (Etnier etal., 1997), larger effects on executive
function when contrasting high-intensity exercise with
rest rather than with low-intensity exercise (Ludyga
etal., 2016), and a greater facilitating effect for high-
fitness participants and for individuals 14 years of age
or older on the basis of previous literature (Y. K. Chang
etal., 2012). In addition, we ran exploratory analyses
to test potentially meaningful interactions (e.g., trade-
offs, additive effects, confounds) between moderators
thought to be informative given previously unexplored
relationships or prior mixed findings.
This meta-analysis followed guidelines from the Pre-
ferred Reporting Items for Systematic Review and Meta-
Analysis Protocols (PRISMA-P) 2015 statement (Moher
etal., 2015; Shamseer etal., 2015).
Eligibility criteria
Participant, intervention, comparator, and outcome
(PICO) criteria were used to determine eligibility (Moher
etal., 2015). Specifically, studies had to include an acute
bout of high-intensity exercise as an independent variable
and performance on at least one standardized test of
executive function as a dependent variable.2 For a study
to be included, the executive-function task(s) also needed
to be administered at least once after the exercise bout.
Studies also needed to include a control/comparison
group or condition, with random allocation to groups/
conditions, and needed to be in English (see Fig. 1).
High-intensity threshold. Two thresholds have been
used in the past to define high-intensity exercise: 77%
HRmax (American College of Sports Medicine, 2010; Y. K.
Chang etal., 2012) or 80% Wmax (Borer, 2003; Browne
etal., 2017; McMorris etal., 2015). When exercise inten-
sity was expressed in terms of Wmax, maximal oxygen
uptake (VO2max), or other units, we converted those val-
ues into HRmax equivalents. Details of this conversion are
presented later in the Data Preprocessing section.
4 Moreau, Chou
Search Features
(through January 25, 2018)
Searching electronic databases (Scopus, PubMed, SPORTDiscus, PsycINFO,
Web of Science, ProQuest, ScienceDirect, Google Scholar), and Google,
using combinations of the following search terms:
• For high-intensity exercise: HI*, (for HIT, HIE, HIIT, HICT, HIIE),
High intensity*, physical activity, strength, strenuous, intense,
sprint interval, exercise intensity, anaerobic exercise
• For cognition: stroop, cogniti* (for cognitive, cognition), memory, learning,
attention, language, “executive func*” (for executive function),
intelligence, “reaction time”, rt, expertise, recall,
mental, processing, perception, psychomotor, “decision making”
• For short-term interventions: acute, short term, immediate, burst,
bout, short duration, single
Search Features
(outside of Databases)
• Scanning reference lists of previous
reviews and meta-analyses (e.g.,
Browne et al., 2017; Chang et al.,
2012; McMorris and Hale, 2012;
Lambourne and Tomporowski, 2010)
• Scanning reference lists of relevant
Records Identified Through
Database Searching
(n = 2,320)
Additional Records Identified
(n = 383)
Criteria for Study Inclusion
• Must include a high-intensity exercise bout as an independent variable
• Must include a standardized cognitive test as a dependent variable
• High-intensity exercise bout must be performed within 1 day
• Cognitive testing must be performed after the exercise bout
• Must include a comparison group or condition at rest and/or performing exercise at a different level of intensity
• Must have randomized assignment for between-groups experiments
• Must have full-text in English and be published in a peer-reviewed journal
• Must have healthy human participants
Inclusion Criteria
Abstracts Screened
(n = 2,320)
Abstracts Screened
(n = 383)
Abstracts Excluded
(n = 2,224)
Abstracts Excluded
(n = 333)
Full-Text Articles Assessed
for Eligibility
(n = 96)
Full-Text Articles Assessed
for Eligibility
(n = 50)
Full-Text Articles Excluded
(n = 118)
• Exercise not intense enough
• Exercise not on a single day
• No cognitive test after exercise bout
• Data not provided in mean (SD ) format
• No adequate comparison group
Studies Included in Meta-Analysis (n = 28)
• Within-Subjects Studies (n = 18)
• Between-Groups Studies (n = 10)
49 Independent Samples
147 Effect Sizes
Total N = 1,177
Fig. 1. Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) flow diagram of literature search and study
Exercise and Executive Function 5
We chose our intensity thresholds (high, very high,
and maximal) by integrating the two definitions.3 High
intensity corresponded to 77% to 88.5% HRmax, 59.5% to
79.9% Wmax, or an equivalent—similar to the first thresh-
old. Very high intensity corresponded to 88.6% to 99.9%
HRmax, 80% to 99.9% Wmax, or an equivalent—similar to
the second threshold. Maximal intensity corresponded
to 100% HRmax, 100% Wmax, or an equivalent.
Definitions of acute exercise and executive func-
tion. Acute exercise refers to exercise performed on a
single day (American College of Sports Medicine, 2010).
Only standardized cognitive tasks or analogues were
included. For search purposes, we included cognitive
domains that are either subcomponents of, or related to,
executive function, such as attention, cognitive control,
cognitive flexibility, fluid intelligence, inhibitory control,
planning, and working memory. See Table S2 in the Sup-
plemental Material available online for details on the full
categorization process.
Sources. We searched the following databases: Scopus,
PubMed, PsycINFO, ScienceDirect, Web of Science, Pro-
Quest, SPORTDiscus, and Google Scholar. Searches were
performed between January 10 and January 15, 2018; all
studies published up until the date of the search were con-
sidered for inclusion. We also searched relevant reviews
and meta-analyses (Browne et al., 2017; Y. K. Chang
etal., 2012; Cooper, Dring, & Nevill, 2016; Lambourne &
Tomporowski, 2010; McMorris & Hale, 2012; Stork, Banfield,
Gibala, & Martin Ginis, 2017), together with reference lists,
to identify other relevant articles (see details at http://osf
Strategy. In each database, a minimum of two different
searches were performed. We used the following combi-
nation of key terms or phrases across all databases:
“HIIT,” “HICT,” “HIT,” “High intensity interval,” and “High
intensity intermittent,” plus either “exercise” or “training,
plus any of “stroop,” “cognitive,” “cognition,” “memory,
“learning,” “attention,” “perception,” “language,” and
“executive func*" (* = truncated). In addition to these two
searches, minor variations were performed to find all rel-
evant articles. Search exemplars in the Scopus database
are provided in Table S1 in the Supplemental Material. A
record of all search terms and procedures is available at
Selection. E. Chou screened studies for initial inclusion
on the basis of titles and abstracts, with advice on inclu-
sion criteria from D. Moreau. If an initial decision could
not be made, E. Chou examined the full text for eligibility.
Articles unable to be categorized by E. Chou alone on the
basis of the outlined criteria were referred to D. Moreau,
who decided which articles to include. The full selection
process is presented in Figure. 1. All full-text studies con-
sidered and the rationale behind their inclusion or exclu-
sion are available at
Data collection. E. Chou collected the data. For all
studies, information was collected on a number of vari-
ables. The complete list of these variables is included in
List S1 in the Supplemental Material; full data are avail-
able at E. Chou contacted authors by
e-mail (n = 11) to request unpublished data, with the
deadline for replying set to March 30, 2018. Five authors
provided us with unpublished data; we included data
from four of these authors because they matched all
inclusion criteria.
Key variables identified from previous studies were
included as moderators in all analyses. These variables
included research-protocol characteristics (study design,
comparison group, and study quality), exercise char-
acteristics (duration, intensity, mode, and rhythm),
cognitive-task characteristics (timing of cognitive testing,
cognitive-task domain, and baseline cognitive testing),
and sample characteristics (fitness and age). We provide
further details about each of these variables hereafter.
Research-protocol characteristics. We characterized
whether studies included a within-subjects or between-
groups design. This distinction was thought to be particu-
larly important in the context of exercise interventions—a
meta-analysis by Y. K. Chang etal. (2012) found signifi-
cant improvements in cognitive performance following
exercise in within-subjects, repeated measures studies
(d = 0.11) but not in between-groups, single-measure
studies (d = 0.06), although the difference between the
two types of designs was not itself significant. We also
documented whether studies compared cognitive perfor-
mance following high-intensity exercise with performance
after rest or with performance following lower intensities
of exercise. We included the comparison group as a mod-
erator on the basis of previous findings showing that
light-to-moderate exercise elicits cognitive gains that are
quantitatively different from test-retest effects and because
of current discrepancies regarding the effect of high-
intensity exercise on cognition (Browne etal., 2017; Y. K.
Chang etal., 2012; McMorris & Hale, 2012).
Finally, studies were categorized into three levels of
quality (low, average, and high) on the basis of the
design properties. This is important given that previous
studies have identified an increase in the magnitude of
6 Moreau, Chou
reported effect sizes within a study as the number of
threats to internal validity (i.e., the risk of bias) increases
(Etnier etal., 1997). To mitigate the risk of bias from
individual studies, we assessed study quality following
Cochrane guidelines (Higgins etal., 2011). Study quality
was assessed on the basis of the variables described in
List S1 in the Supplemental Material. E. Chou graded
study quality according to whether specific criteria were
met (no = 0, partial = 1, yes = 2). Items that were not
reported or not applicable were not used in the calcula-
tion of the quality score. Criteria deemed more likely
to affect the risk of bias were given greater weighting
than other criteria (minimum weighting = 1; maximum
weighting = 3). Although the two types of designs
(within subjects/between groups) had different assess-
ment criteria, this difference was reconciled by conver-
sions into comparable percentage scores. These
assessments are available in Tables S3 and S4 in the
Supplemental Material.
Exercise characteristics. Exercise duration was cate-
gorized as follows: 0 to 5 min, 6 to 10 min, 11 to 20 min,
and > 20 min. This categorization was based on the
results of Y. K. Chang etal. (2012), who, across all inten-
sities, found exercise durations of 0 to 10 min to have a
small, negative effect on cognition after exercise and
durations of 11 to 20 min and > 20 min to have small,
positive effects on cognition after exercise. In addition,
we characterized exercise intensity as high, very high,
and maximal on the basis of the aforementioned criteria.
Previous findings have been equivocal; Y. K. Chang etal.
(2012) reported cognitive improvements for both hard
and very hard exercise (when tested at least 1 min after
exercise) and McMorris and Hale (2012) finding moder-
ate—but not what they referred to as heavy—exercise to
improve processing speed during and after exercise.
Finally, we distinguished studies on the basis of both the
modality and rhythm of exercise. The former distinction
was motivated by a previous meta-analysis (Lambourne
& Tomporowski, 2010) that found cycling to lead to bet-
ter cognitive performance after exercise compared with
running. The latter variable (continuous vs. intermittent)
was modeled because of the rising number of studies
using interval-training paradigms. Previous studies have
not examined this variable in detail.
Cognitive-task characteristics. We analyzed the effect
of time of testing using the categories of 0 to 1 min, 1 to
10 min, and > 10 min after exercise. This categorization
was informed by a previous meta-analysis (Y. K. Chang
etal., 2012) that found no effect of high-intensity exercise
on cognitive performance when tests were administered
immediately (0–1 min) after high-intensity exercise and
small-to-moderate facilitating effects when tests were
administered > 1 min after high-intensity exercise. Exec-
utive-function tasks were further categorized relative to
the specific cognitive domains they tapped, namely atten-
tion, cognitive flexibility, inhibitory control, and working
memory. When a task targeted more than one domain, it
was labeled on the basis of its main subdomain. Finally,
studies either involved baseline testing and postexercise
testing or postexercise testing only. We added this mod-
erator variable given that studies with both time points
may be less biased than those for which only posttest
performance was available.
Sample characteristics. Participants were character-
ized as low fitness, moderate fitness, or high fitness. Y. K.
Chang etal. (2012) found high-fitness participants to be
the only group with significant cognitive improvements
both immediately following exercise (0–1 min) and after
a postexercise delay (> 1 min), whereas low-fitness and
moderate-fitness participants showed improvements only
immediately after exercise for the former and only after a
postexercise delay for the latter. However, Etnier etal.
(2006) did not find a clear relationship between fitness
and cognitive performance overall. Like previous research-
ers (see Y. K. Chang etal., 2012), we further categorized
age into the following groups: children (6–13 years), ado-
lescents (14–17 years), young adults (18–30 years), middle-
aged adults (31–60 years), and older adults (> 60 years).
Note that Y. K. Chang et al. (2012) found significant
improvements in cognition following exercise for all age
groups except children ages 6 to 13.
Continuous moderators. When studies had continuous
data available, we also conducted continuous-moderator
analyses for the following moderators: exercise duration at
high intensity, exercise intensity of high-intensity condi-
tion, timing of cognitive testing, study quality, and partici-
pant age. Although continuous-moderator analyses help
to increase statistical power compared with categorical-
moderator analyses (MacCallum, Zhang, Preacher, & Rucker,
2002), we chose to primarily report categorical analyses
because they allowed the inclusion of a larger number of
relevant studies.
The search strategy identified 2,320 records through
database searching and 383 additional records through
other search methods. Of these, 146 full-text articles
were examined for eligibility. Twenty-eight studies were
included in the final meta-analysis (see Fig. 1). For clar-
ity, we present characteristics of within-subjects and
between-groups studies separately; however, these two
types of protocols were analyzed together in our mul-
tilevel model.4 Tables 1 and 2 present an overview of
Table 1. Research Protocol and Sample Characteristics for All Within-Subjects Studies Included in the Meta-Analysis
Participants Condition
Study NMale (%)
Mean age
(years) Mean BMI Fitness 1 (highest intensity) 2 (lowest intensity) 3 4
No. of
effect sizes
Alves etal. (2014) 22 40.91 53.7 25.7 HIIT Stretching 5
Berse etal. (2015) 227 53.2 14.8 HIIT Resting 2
Budde etal. (2012)a46 56.5 23.11 Intermittent
maximal exercise
Resting — 3
Cooper, Bandelow, etal.
44 47.73 12.6 18.9 Sprints Resting 7
Craft (1983)b31 100 8.5 10-min exercise Resting 1-min exercise 5-min exercise 3
Etnier etal. (2016) 16 56.25 23.06 VO2max session Vt −20% session Vt +20% session 4
Kamijo etal. (2004)c12 100 27.5 HIE Resting Low-intensity
Kamijo, Nishihira,
Higashiura, &
Kuroiwa (2007)
12 100 25.7 Hard exercise Resting Light exercise Moderate
Kao, Westfall, Soneson,
Gurd, & Hillman
64 42.19 19.2 23.8 High (VO2max) HIIT Resting Continuous
— 3
Kujach etal. (2018) 25 64 21 Low (sedentary) HIIT Resting 4
Lambrick, Stoner, Grigg,
& Faulkner (2016)
20 45 8.8 18.1 Intermittent exercise Moderate continuous
— 1
Llorens, Sanabria, &
Huertas (2015)d
18 100 22 24.3 Intense exercise Resting 4
Llorens, Sanabria,
Huertas, Molina, &
Bennett (2015)
9 66.67 26 Low (VO2max) Intense exercise Resting 3
Peruyero, Zapata, Pastor,
& Cervelló (2017)
44 52.3 16.39 Mainly vigorous
Resting Mainly light
— 6
I. Ramos etal. (2017) 9 44.44 10.3 16.2 110% of LT Seated drawing 90% of LT 10
Tsukamoto etal. (2016) 12 100 22.9 22.4 High (physically
HIIT Moderate continuous
— 6
Walsh etal. (2018) 22 13.64 20 22.7 HIIT Resting 3
Wohlwend, Olsen,
Håberg, & Palmer
27 50 24.27 High (VO2max) High-intensity
— 4
Note: ADHD = attention-deficit/hyperactivity disorder; BMI = body mass index; HIE = high-intensity exercise; HIIT = high-intensity intermittent exercise; LT = lactate threshold; — = data not provided; VO2max =
maximal oxygen uptake; Vt = ventilatory threshold.
aWhole-group analyses (not by fitness). bExcluded ADHD group from analysis. cOne outlier effect size. dExperiment 1 only.
Table 2. Research Protocol and Sample Characteristics for All Between-Groups Studies Included in the Meta-Analysis
Study k
N Na
Group No. of
1 (highest
2 (lowest
intensity) 3 4
Basso, Shang, Elman,
Karmouta, & Suzuki
8 85 11 40 20.45 23.63 Exercise (four
Rest: video
— 8
H. Chang, Kim, Jung,
& Kato (2017)c
2 18 9 0 22.1 21.6 HIE Resting Moderate-
— 4
Córdova, Silva, Moraes,
Simões, & Nóbrega
4 48 12 0 63.1 24.3 Low (older
110% of AT Resting 60% of AT 90% of AT 18
Etnier, Labban,
Piepmeier, Davis, &
Henning (2014)
2 43 19 34.88 11.5 Moderate
PACER scores)
HIE Resting — 3
Hwang, Brothers, etal.
2 58 29 48.3 22.84 22.63 Moderate
HIE Resting — 6
Hwang, Castelli, &
2 30 15 40 23.4 22.21 Moderate
Exercise +
sham laser
Control (sham
exercise +
sham laser)
— 4
Lemmink & Visschere
2 16 8 100 20.9 High (soccer
HIIT Rest:
— 4
Netz, Tomer, Axelrad,
Argov, & Inbar
3 58 20 30 56.12 Moderate
HIE Rest: movie
— 4
Whyte, Gibbons, Kerr,
& Moran (2015)
2 40 20 100 21.05 High (athletes) HIIT Resting 2
Zimmer etal. (2016) 4 121 30 70 23.9 21.984 High (students
HIE Control (foam
rolling, 35
Note: AT = anaerobic threshold; BMI = body mass index; HIE = high-intensity exercise; HIIT = high-intensity interval training; — = data not provided; PACER = Progressive Aerobic
Cardiovascular Endurance Run; VO2max = maximal oxygen uptake. The Baecke questionnaire is from Baecke, Burema, and Frijters (1982).
aGroup 1 only. bEight independent samples (4 testing delays times 2 conditions). cTwo outlier effect sizes; excluded resistance exercise group. dExcluded laser-therapy groups.
eExcluded results after exercise blocks 2 and 3.
Exercise and Executive Function 9
research protocols and sample characteristics for within-
subjects and between-groups studies, respectively.
Data preprocessing. When exercise intensity was not
expressed in terms of HRmax—for example, %Wmax,
%VO2max, or percentage of heart rate reserve (%HRR,
where HRR = HRmax – HRrest)—it was converted into
HRmax (Arts & Kuipers, 1994; Lounana etal., 2007). In
addition, when average age was available, HRmax was
estimated using the formula HRmax = 208 − (0.7 × Age)
from Tanaka, Monahan, and Seals (2001). Note that based
on these conversion formulas, 77% HRmax corresponded
to approximately 59.5% Wmax. When intensity could not
be converted to an HRmax equivalent, such as with scores
from the Rated Perceived Exertion scale, the article was
examined thoroughly to determine the intensity on the
basis of its description. For example, exercise to fatigue
or exhaustion was coded as maximal intensity. We
excluded articles from our analyses if accurate categori-
zation was not possible. Exercise characteristics are fur-
ther described in Tables 3 and 4.
Studies reported two main types of cognitive test
scores: those with postexercise scores only and those
with both baseline and postexercise scores. For the lat-
ter, we calculated mean-difference (posttest-pretest)
scores. Baseline scores were subtracted from the post-
bout scores for each study, correcting for bias and
adjusting for the direction of each cognitive testhigher
scores indicated better performance in some instances
(e.g., accuracy), whereas lower scores indicated better
performance for other measures (e.g., reaction time).
Pooled standard deviation values (SDpooled) were calcu-
lated for each mean-difference score. Where available,
moderator-variable data related to each mean-difference/
raw score were recorded in additional columns. Note
that when studies included more than one posttest ses-
sion, only the testing session closest in time to the exer-
cise bout was analyzed, given the potential for practice
effects to contaminate subsequent testing.
Because different cognitive tests measured perfor-
mance on different scales, and because the cognitive
test scores within each group or condition were avail-
able in either mean-difference or raw-score format, we
calculated bias-corrected Cohen’s d (Cohen, 1992) to
standardize the differences in mean scores between
groups or conditions. For between-groups studies, we
compared the scores between different groups under
different conditions; for within-subjects studies, we
compared the scores between the same group under
different conditions. Cognitive-task characteristics are
further described in Tables 5 and 6.
Main analyses. We conducted all analyses in the R soft-
ware environment (Version 3.5.1; R Core Team, 2018)
using the metafor (Version 2.1; Viechtbauer, 2010) and
multcomp (Version 1.4; Hothorn etal., 2013) packages. We
ran restricted maximum-likelihood, multilevel, random-
effects meta-analyses to estimate overall effects as well as
the heterogeneity between effect sizes. In addition, we con-
ducted mixed-effects meta-analyses to investigate whether
the observed heterogeneity could be explained by the
effects of moderator variables. The R code for our analyses
is available online at
We conducted a multilevel analysis to account for the
dependency between effect sizes (Assink & Wibbelink,
2016). Multilevel analyses group effect sizes on the basis
of higher-level clustering variables (Konstantopoulos,
2011) to prevent inflation and overconfidence in meta-
analytic estimates (Van den Noortgate, López-López,
Marín-Martínez, & Sánchez-Meca, 2013). We used a pro-
cedure that adds random effects at each level of possible
dependency to reduce inflation while preserving valu-
able information provided by studies that report mul-
tiple effect sizes (Viechtbauer, 2010).
Our multilevel analysis had three clustering variables
(i.e., levels). First, we modeled dependency at the
study-design level (i.e., within-subjects and between-
groups designs). Second, we modeled dependency at
the sample levelwithin samples reporting results in
multiple tests and/or samples with two or more com-
parison groups. For example, a group undergoing three
sessions (resting, low-intensity exercise, and high-
intensity exercise) included dependency across all dif-
ference scores given that the high-intensity session
score was compared with both the resting and low-
intensity session scores. Finally, we modeled depen-
dency within the cognitive domains for which multiple
test scores were reported (e.g., response time and accu-
racy in a Stroop task) within individual samples.
Outliers. Effect sizes whose residuals were more than
three standard deviations from the mean were considered
outliers and excluded from our main analysis. There were
three effect sizes meeting this criteria (see Fig. S1 in the
Supplemental Material). In addition, Cook’s distance was
used as an exploratory estimate to detect outliers. This
method takes into account the relative influence of each
effect size on the overall estimate. Effect sizes with a Cook’s
distance more than three times the mean Cook’s distance
were labeled as possible outliers, although they were not
excluded from our main analyses. There were 16 effects
meeting this criteria (see Fig. S2 in the Supplemental Mate-
rial); however, because this criterion was not defined a
priori, we provided the full data set before outlier removal,
together with the R code, at
Interaction analyses. To test whether the moderating
effects of individual categorical moderators were dependent
10 Moreau, Chou
on other moderator variables, we conducted interaction anal-
yses. We limited the scope of interactions to two moderators
at a time and chose to include the following six moderators
in the interaction analyses: comparison group, exercise
rhythm, exercise intensity, exercise duration, timing of cogni-
tive testing, and cognitive domain. These moderators were
chosen because they were thought to relate to each other in
terms of their influence on the effect of high-intensity exer-
cise on executive function. As these analyses involved multi-
ple pairwise comparisons between each level of each
moderator, we controlled the familywise error rate using the
Holm method (Holm, 1979).
Heterogeneity and model comparisons. We estimated
heterogeneity (between-studies variability in effect sizes)
in two ways. First, we calculated the I2 statistic, which esti-
mates the percentage of between-studies variation that is
due to true differences (e.g., differences in research-
protocol characteristics, differences in exercise intensity)
rather than error. Second, using Cochran’s Q statistic, we
tested whether the moderators could account for the
between-studies heterogeneity. We tested whether a
model including the significant moderator reduced the
model’s residual heterogeneity to nonsignificance. If
residual heterogeneity was still significant, we combined
Table 3. Exercise Characteristics for All Within-Subjects Studies Included in the Meta-Analysis
description Modality Rhythm ƛ (%HRmax) Intensity Ttotal (min) THI (min)
Alves etal. (2014) HIIT Cycling Int 85.5 High 25 10
Berse etal. (2015) HIIT Cycling Int 100 Maximal 12 Variable
Budde etal. (2012) Int maximal
Running Int 100 Maximal 6 6
Cooper, Bandelow, etal.
Sprints Running Int 90.87 Very high 10 1.67
Craft (1983) 10-min exercise Cycling Cont 84.1 High 10 10
Etnier etal. (2016) VO2max
Running Cont 100 Maximal 30 30
Kamijo etal. (2004) HIE Cycling Cont 100 Maximal Variable Variable
Kamijo, Nishihira,
Higashiura, & Kuroiwa
Hard exercise Cycling Cont 78.6 High 22 20
Kao, Westfall, Soneson,
Gurd, & Hillman (2017)
HIIT Running Int 85 High 9 4.5
Kujach etal. (2018) HIIT Cycling Int 78.3 High 6 4
Lambrick, Stoner, Grigg,
& Faulkner (2016)
Int exercise Running Int 100 Maximala15 5.5
Llorens, Sanabria, &
Huertas (2015) Exp. 1
Intense exercise Cycling Cont 79 High 13 5
Llorens, Sanabria,
Huertas, Molina, &
Bennett (2015)
Intense exercise Cycling Cont 78.84 High 24 15
Peruyero, Zapata, Pastor,
& Cervelló (2017)
Mainly vigorous
Zumba dancingbInt cHigh 24 Variable
I. Ramos etal. (2017) 110% of LT Running Cont dHigh 10 10
Tsukamoto etal. (2016) HIIT Cycling Int 95.32 Very high 36 16
Walsh etal. (2018) HIIT Burpees, jumping
jacks, mountain
climbing, squat
Int 94 Very high 11 5.5
Wohlwend, Olsen,
Håberg, & Palmer
HIE Running Int 91 Very high 38 16
Note: HIE = high-intensity exercise; HIIT = high-intensity interval training; Cont = continuous; Int = intermittent; HRmax = maximum heart rate;
LT = lactate threshold; — = data not provided THI = exercise duration at high intensity; Ttotal = total exercise duration; VO2max = maximal oxygen
uptake; ƛ = maximum intensity.
aThis was described as moderate intermittent exercise but included 4.5 min of heavy exercise and 1 min of severe exercise. bThese modalities
were analyzed under “running” in moderation analyses. cVigorous exercise measured with an accelerometer. dNo equation available.
Exercise and Executive Function 11
all moderators in our model (including the nonsignifi-
cant ones) to account for any residual heterogeneity. We
evaluated whether the three models (no moderator, sig-
nificant moderators only, all moderators) were different
from each other in terms of goodness of fit and parsi-
mony using the Akaike information criterion–corrected
(AICc) and the Bayesian information criterion (BIC).
Across 28 studies with a range of protocols, as well as
exercise, cognitive task, and sample characteristics,
acute high-intensity exercise had a small, significant,
facilitating effect on executive function after exercise
compared with other conditions (lower-intensity exer-
cise and rest combined). The meta-analytic estimate of
this effect was d = 0.24, 95% confidence interval (CI) =
[0.13, 0.35]. Our main measure of heterogeneity (I2 =
49.08% across all studies) indicated that about half the
variability between effect sizes was potentially due to
real differences between studies, that is, differences that
are not related to random error. Figure 2 provides a
general overview of our findings.
Moderator analyses
Categorical-moderator analyses. Comparison group
was a significant moderator (p = .002; see Table 7 for
details); that is, high-intensity exercise had a facilitating
effect on executive function compared with rest (d = 0.34)
but not significantly different from that of lower-intensity
exercise (d = 0.07). Study quality was a marginally signifi-
cant moderator (p = .08), with the facilitating effect
decreasing in magnitude as study quality increased. Other
moderator variables did not yield significant moderating
effects despite effect sizes being significantly different
from zero at specific levels of those moderators. All details
are reported in Table 7.
Because executive function is an umbrella term that
encompasses many different subdomains, we also
examined the subcomponents attention, cognitive flex-
ibility, inhibitory control, and working memory sepa-
rately. High-intensity exercise had a small, facilitating
effect on cognitive performance for each of these four
subdomains, a finding that was consistent with our
overall analysis. Figure 3 shows forest plots for each of
these subdomains.
Table 4. Exercise Characteristics for All Between-Groups Studies Included in the Meta-Analysis
description Modality Rhythm ƛ (%HRmax) Intensity Ttotal (min) THI (min)
Basso, Shang, Elman,
Karmouta, & Suzuki
Vigorous Cycling Cont 85 High 60 50
H. Chang, Kim, Jung, &
Kato (2017)
HIE Running Cont 85.5 High 40 30
Córdova, Silva, Moraes,
Simões, & Nóbrega
110% of AT Cycling Cont 85.6 High 25 20
Etnier, Labban,
Piepmeier, Davis, &
Henning (2014)
PACER task Running Int 100 Maximal Variable Variable
Hwang, Brothers, etal.
HIE Running Cont 87.9 High 20 10
Hwang, Castelli, &
Gonzalez-Lima (2016)
HIE Running Cont 90.98 Very high 20 10
Lemmink & Visscher
Int Cycling Int 86.8 High 27 16
Netz, Tomer, Axelrad,
Argov, & Inbar (2007)
Moderate (60%
Running Cont 78.7 High 44 35
Whyte, Gibbons, Kerr,
& Moran (2015)
HIIP Sprinting/jumping/
shuffling circuita
Int 94.6 Very high Variable
(> 10 min)
(> 5 min)
Zimmer etal. (2016) HIE Cycling Cont 85 High 35 30
Note: AT = anaerobic threshold; HIE = high-intensity exercise; HIIP = high-intensity intermittent-exercise protocol; Cont = continuous; Int =
intermittent; HRmax = maximum heart rate; PACER = Progressive Aerobic Cardiovascular Endurance Run; THI = exercise duration at high intensity;
Ttotal = total exercise duration; ƛ = maximum intensity.
aThis modality was analyzed under “running” in the moderation analyses.
12 Moreau, Chou
Continuous-moderator analyses. None of the planned
continuous moderators yielded a significant effect (see
Table 8 for details); that is, when examined separately,
the effect of high-intensity exercise on executive function
did not change significantly as a function of unit increases
or decreases in any of the moderators. The effect of an
exploratory moderator, percentage of male participants,
was marginally significant (p = .058), with the facilitating
effect of high-intensity exercise on executive function
decreasing as the percentage of male participants increased.
Interaction analyses. Of the 15 pairs of moderators
tested, two potentially meaningful interactions emerged.
The first was the interaction between comparison group
and exercise intensity, and the second was the interaction
between exercise rhythm and timing of cognitive testing.
When lower-intensity exercise was the comparison
group, participants undergoing very-high-intensity exer-
cise (three studies) experienced significantly larger
improvements in executive performance compared with
those undergoing high-intensity exercise (eight studies;
Δd = 0.35, p = .002). When exercise was intermittent,
participants tested 0 to 1 min after exercise (10 studies)
experienced significantly larger improvements in execu-
tive performance compared with those tested > 10 min
after exercise (three studies; Δd = 0.77, p = .009). Most
interaction analyses for which both moderators had three
or more levels (i.e., both moderators belonged to any of
exercise intensity, exercise duration, timing of cognitive
testing, or cognitive domain) yielded nonsignificant pair-
wise comparisons as a result of the low number of stud-
ies for each level and because of the Holm correction.
These analyses are available at
Heterogeneity and model comparisons. The total
heter ogeneity present in the model without moderators was
significant, Cochran’s Q(146) = 264.151, p < .0001. Only
the categorical-moderator-comparison group significantly
reduced heterogeneity (p < .0001), whereas the effects of
the categorical moderator study quality and the continu-
ous moderator percentage of male participants were mar-
ginally significant (p = .082 and p = .058, respectively).
Although the inclusion of comparison group in the
model significantly reduced heterogeneity, the remain-
ing heterogeneity was still significant, QE(145) =
263.449, p < .0001. The inclusion of additional modera-
tors (i.e., study design, comparison group, cognitive-
task domain, categorical study quality, categorical
exercise intensity, baseline cognitive testing, and per-
centage male) available for most effect sizes decreased
Table 5. Cognitive-Task Characteristics for All Within-Subjects Studies Included in the Meta-Analysis
Study Cognitive test(s)aBaselinebNct (min)
Alves etal. (2014) Victoria Stroop Test, digit-span test Yes 1 0
Berse etal. (2015) Shifting task No 1 0
Budde etal. (2012) d2 Test of Attention No 1 0
Cooper, Bandelow, etal. (2016) Stroop interference test, digit-symbol
substitution task, Corsi blocks
Yes 2 0
Craft (1983) WISC-R digit span, WISC-R coding B, ITPA
visual sequential memory
No 1 0
Etnier etal. (2016) Rey Auditory Verbal Learning Task No 3 0
Kamijo etal. (2004) S1-S2 reaction-time task No 1 3
Kamijo, Nishihira, Higashiura, & Kuroiwa (2007) Modified flanker task No 1 3
Kao, Westfall, Soneson, Gurd, & Hillman (2017) Modified flanker task No 1 20
Kujach etal. (2018) Stroop Color and Word Test Yes 1 15
Lambrick, Stoner, Grigg, & Faulkner (2016) Stroop interference test Yes 3 1
Llorens, Sanabria, & Huertas (2015), Exp. 1 Exogenous spatial-attention task No 2 0
Llorens, Sanabria, Huertas, Molina, & Bennett (2015) Visual-search task No 1 0
Peruyero, Zapata, Pastor, & Cervelló (2017) Stroop Yes 1 5
I. Ramos etal. (2017) Modified flanker task, free-recall test, Trail
Making Test parts A and B, tangram puzzle
No 1 20
Tsukamoto etal. (2016) Stroop Color and Word Test Yes 4 0
Walsh etal. (2018) d2 Test of Attention No 1 10
Wohlwend, Olsen, Håberg, & Palmer (2017) Connors Continuous Performance Test Yes 1 0
Note: N = number of postexercise cognitive test repetitions; t = approximate timing of first postexercise cognitive test; WISC-R, Wechsler Intelligence
Scale for Children–Revised (Wechsler, 1974); ITPA = Illinois Test of Psycholinguistic Abilities.
aIncludes only the tests analyzed in our meta-analysis. bDefined as whether the study has a preexercise cognitive test. cBarring one study on
long-term memory (Etnier etal., 2016), for which we analyzed all timings, we analyzed only the first of these postexercise cognitive tests to
minimize the contribution of test–retest effects (minutes after exercise/control).
Exercise and Executive Function 13
heterogeneity substantially, although it remained sig-
nificant, QE (132) = 230.5262, p < .0001.
Both AICc and BIC criteria provided converging evi-
dence in their assessments of model fit and parsimony.
The best model was the model with comparison group
as the moderator (AICc = 151.7, BIC = 166.2), followed
by the model without moderators (AICc = 158.8, BIC =
170.5). Meanwhile, the model incorporating the most
moderators was least preferred (AICc = 169, BIC =
215.1). Together, these results suggested that the non-
significant moderators did not account for enough
residual heterogeneity to justify their inclusion in the
meta-analytic model.
Assessment of metabias
We assessed bias in three different ways. First, we plot-
ted funnel plots to detect possible publication bias.
Funnel plots help visualize the relationship between
each effect size and their corresponding standard error.
In a meta-analysis unaffected by publication bias, effect
sizes from larger samples, which have smaller standard
errors, should cluster around the mean effect. In
contrast, effect sizes from smaller samples, with larger
standard errors, are expected to be more dispersed
around the mean. In addition, all effect sizes are
expected to be represented somewhat equally on both
sides of the mean, forming a symmetric, inverse, funnel-
like shape. Asymmetry in funnel plots may result from
reporting biases, poor methodological quality, true het-
erogeneity, artifacts, or chance (Sterne etal., 2011). The
funnel plot for our overall analysis (Fig. 4) was fairly
symmetrical and evenly distributed, and few effects fell
outside the 99% CI, suggesting that the present meta-
analysis was not substantially affected by publication
Second, we analyzed whether study quality (cate-
gorical) affected the estimated effect. According to our
criteria (Tables S3 and S4 in the Supplemental Material),
there were 23 high-quality studies, 4 average-quality
studies, and 1 low-quality study. Study quality had a
marginally significant moderating effect on the effect
of high-intensity exercise on executive function (p =
.082). Whereas high-quality studies reported a small
facilitating effect (d = 0.19, 95% CI = [0.07, 0.31]),
average-quality studies reported a moderate facilitating
Table 6. Cognitive-Task Characteristics for All Between-Groups Studies Included in the Meta-Analysis
Study Cognitive test(s) BaselineaN t(min)
Basso, Shang, Elman, Karmouta, &
Suzuki (2015)b
Hippocampal measures:
Hopkins Verbal Learning Test–Revised (delayed recall,
retention, recognition), modified Benton Visual Retention Test
Yes 4c30
Prefrontal measures:
Stroop test, Symbol Digits Modalities Test, Hopkins Verbal
Learning Test–Revised (total recall), digit-span test
(forward and backward), Trail Making Test parts A and B
H. Chang, Kim, Jung, & Kato
Stroop test Yes 1 15
Córdova, Silva, Moraes, Simões, &
Nóbrega (2009)
Simple response time, verbal fluency, Tower of Hanoi, Trail
Making Test parts A and B
Yes 1 8
Etnier, Labban, Piepmeier, Davis,
& Henning (2014)
Rey Auditory Verbal Learning Task (lists A and B) No 2d0
Hwang, Brothers, etal. (2016) Stroop test, Trail Making Test parts A and B Yes 1 10
Hwang, Castelli, & Gonzalez-Lima
Psychomotor vigilance task, delayed match-to-sample task Yes 1 13
Lemmink & Visscher (2005) Choice reaction-time test No 1 0
Netz, Tomer, Axelrad, Argov, &
Inbar (2007)
Alternate-uses test, digit-span test (forward) Yes 1 5
Whyte, Gibbons, Kerr, & Moran
Symbol Digits Modalities Test, Stroop interference test Yes 1 0.25
Zimmer etal. (2016) Stroop test Yes 1 0
Note: The cognitive tests listed include only those used in our meta-analysis. N = number of postexercise cognitive test repetitions; t =
approximate timing of first postexercise cognitive test (minutes after exercise/control). The Hopkins Verbal Learning Test is a product of PAR
(Lutz, FL). The Benton Visual Retention Test is described in Benton (1992). The Symbol Digits Modalities test is described in Smith (1982). The
Rey Auditory and Verbal Learning Test is described in Schmidt (1996).
aThis indicates whether the study had a preexercise cognitive test. bIn this study, prefrontal and hippocampal measures were recorded in
summary z scores. cAll test sessions were examined because this study examined test timing as an independent variable. dBoth of these were
examined because the first repetition corresponded to working memory and the second to long-term memory.
14 Moreau, Chou
Ramos et al. (2017), 7
Craft (1983), 2
Ramos et al. (2017), 9
Kamijo et al. (2007), 12
Llorens, Sanabria, & Huertas (2015), 2
Córdova et al. (2009), 12
Córdova et al. (2009), 6
Llorens, Sanabria, & Huertas (2015), 1
Ramos et al. (2017), 3
Kamijo et al. (2007), 3
Craft (1983), 1
Llorens, Sanabria, & Huertas (2015), 3
Córdova et al. (2009), 9
Kamijo et al. (2007), 8
Lemmink & Visscher (2005), 3
Kamijo et al. (2004), 2
Kamijo et al. (2007), 7
Alves et al. (2014), 2
Budde et al. (2012), 2
Whyte et al. (2015), 2
Netz et al. (2007), 4
Ramos et al. (2017), 8
Ramos et al. (2017), 10
Kamijo et al. (2004), 1
Córdova et al. (2009), 3
Córdova et al. (2009), 13
Kamijo et al. (2007), 9
Kamijo et al. (2007), 10
Zimmer et al. (2016), 11
Córdova et al. (2009), 10
Wohlwend et al. (2017), 4
Kujach et al. (2018), 3
Córdova et al. (2009), 7
Tsukamoto et al. (2016), 2
Kujach et al. (2018), 2
Córdova et al. (2009), 14
Córdova et al. (2009), 2
Córdova et al. (2009), 8
Etnier et al. (2014), 1
Córdova et al. (2009), 11
Córdova et al. (2009), 4
Tsukamoto et al. (2016), 4
Cooper et al. (2016), 7
Whyte et al. (2015), 1
Chang et al. (2017), 1
Study Name Effect Size (Cohen’s d) and 95% CI
−2 −1
Fig. 2. (continued on next page)
effect (d = 0.59, 95% CI = [0.26, 0.92]), and there were
too few low-quality studies to reliably estimate an aver-
age effect size. Our finding suggested that studies of
lower quality reported larger effect sizes because of
poor control of confounding variables. This result fits
well with findings from earlier research (Etnier etal.,
1997) and might suggest a biasing effect from average-
quality studies.
To further address metabias, we constructed a
p-curve analysis (Simonsohn, Nelson, & Simmons, 2014)
for significant (p < .05) effect sizes in our studies. The
rationale for the p-curve analysis is that the p-value
Exercise and Executive Function 15
Study Name Effect Size (Cohen’s d) and 95% CI
−1 01
Basso et al. (2015), 6
Wohlwend et al. (2017), 1
Hwang, Brothers, et al. (2016), 5
Zimmer et al. (2016), 7
Hwang, Castelli, et al. (2016), 1
Córdova et al. (2009), 17
Zimmer et al. (2016), 10
Ramos et al. (2017), 2
Tsukamoto et al. (2016), 3
Alves et al. (2014), 5
Basso et al. (2015), 4
Tsukamoto et al. (2016), 5
Budde et al. (2012), 1
Alves et al. (2014), 1
Budde et al. (2012), 3
Kamijo et al. (2007), 4
Kujach et al. (2018), 1
Kujach et al. (2018), 4
Etnier et al. (2014), 2
Kao et al. (2017), 1
Kamijo et al. (2007), 2
Ramos et al. (2017), 5
Zimmer et al. (2016), 3
Zimmer et al. (2016), 1
Wohlwend et al. (2017), 3
Basso et al. (2015), 8
Berse et al. (2015), 1
Kamijo et al. (2007), 1
Zimmer et al. (2016), 9
Netz et al. (2007), 1
Zimmer et al. (2016), 4
Netz et al. (2007), 3
Lemmink & Visscher (2005), 1
Craft (1983), 3
Kamijo et al. (2007), 6
Cooper et al. (2016), 5
Berse et al. (2015), 2
Zimmer et al. (2016), 2
Córdova et al. (2009), 15
Wohlwend et al. (2017), 2
Tsukamoto et al. (2016), 6
Llorens, Sanabria, & Huertas (2015), 4
Cooper et al. (2016), 6
Cooper et al. (2016), 3
Cooper et al. (2016), 2
Fig. 2. (continued on next page)
16 Moreau, Chou
Study Name Effect Size (Cohen’s d) and 95% CI
−1 012
Llorens, Sanabria, Huertas, Molina, & Bennett (2015), 3
Etnier et al. (2016), 3
Hwang, Castelli, et al. (2016), 3
Peruyero et al. (2017), 6
Hwang, Brothers, et al. (2016), 6
Kamijo et al. (2007), 5
Peruyero et al. (2017), 1
Peruyero et al. (2017), 3
Etnier et al. (2016), 2
Basso et al. (2015), 1
Lemmink & Visscher (2005), 4
Etnier et al. (2016), 4
Zimmer et al. (2016), 8
Hwang, Brothers, et al. (2016), 2
Alves et al. (2014), 4
Walsh et al. (2018), 3
Lemmink & Visscher (2005), 2
Chang et al. (2017), 2
Netz et al. (2007), 2
Walsh et al. (2018), 1
Ramos et al. (2017), 4
Peruyero et al. (2017), 5
Walsh et al. (2018), 2
Kamijo et al. (2007), 11
Basso et al. (2015), 3
Córdova et al. (2009), 16
Ramos et al. (2017), 1
Hwang, Brothers, et al. (2016), 3
Lambrick et al. (2016)
Cooper et al. (2016), 1
Hwang, Brothers, et al. (2016), 4
Córdova et al. (2009), 1
Zimmer et al. (2016), 12
Zimmer et al. (2016), 5
Zimmer et al. (2016), 6
Basso et al. (2015), 5
Basso et al. (2015), 2
Tsukamoto et al. (2016), 1
Hwang, Castelli, et al. (2016), 4
Kao et al. (2017), 3
Hwang, Brothers, et al. (2016), 1
Alves et al. (2014), 3
Kao et al. (2017), 2
Ramos et al. (2017), 6
Basso et al. (2015), 7
Córdova et al. (2009), 18
Fig. 2. (continued on next page)
Exercise and Executive Function 17
−2 −1 0123
Llorens, Sanabria, Huertas, Molina, & Bennett (2015), 2
Llorens, Sanabria, Huertas, Molina, & Bennett (2015), 1
Peruyero et al. (2017), 4
Chang et al. (2017), 3
Chang et al. (2017), 4
Cooper et al. (2016), 4
Peruyero et al. (2017), 2
Etnier et al. (2014), 3
Hwang, Castelli, et al. (2016), 2
Etnier et al. (2016), 1
Córdova et al. (2009), 5
Study Name Effect Size (Cohen’s d) and 95% CI
RE Model, d = 0.2377, p < .0001, 95% CI = [.1257, .3496]
Fig. 2. Forest plot of effect sizes for all included studies. The sizes of the squares represent relative sample sizes. The diamond at the bot-
tom represents the overall effect. The dotted vertical line represents an effect size (d ) of zero. CI = confidence interval; RE = random effects.
distribution for a true effect that is not influenced by
reporting bias should contain fewer high (p > .04) than
low (p .01) significant p values (Simonsohn etal.,
2014). To construct a p curve, we first converted Cohen’s
d values to t values, adjusting for study design (within-
subjects or between-groups). We then generated the
curve using an online app (available at http://www The p curve (Fig. 5) was right-
skewed, with fewer large (p > .04) than small (p .01)
p values. This suggested no clear sign of publication
bias for the significant effect sizes analyzed. The power
estimate of the studies examined was relatively high
(74%, 90% CI = [56%, 86%]).
Short but intense bursts of exercise have gained traction
in recent years as a viable substitute to aerobic exercise
to elicit cognitive improvements (e.g., Moreau etal.,
2017). Whether this form of exercisetypically referred
to as high-intensity exerciseis associated with facili-
tating or debilitating effects on cognitive performance
immediately following exercise, however, remains
unclear. The present meta-analysis provides a quantita-
tive answer to this question, together with an analysis
of the specific variables that help explain some of the
past discrepancies. In particular, we focused on the
effect of high-intensity exercise, defined as exercise
above 77% of one’s maximum heart rate, on postexercise
executive performance. We identified 28 studies examin-
ing this relationship, with comparisons to either lower-
intensity exercise or rest, and investigated subdomains
of executive function, including attention, cognitive flex-
ibility, inhibitory control, and working memory.
Overall, high-intensity exercise had a small, signifi-
cant facilitating effect on executive function after exer-
cise. This finding was in line with our prediction of an
effect, although we did not hypothesize a direction.
Importantly, this result was inconsistent with theoretical
predictions of impairing effects, such as the inverted-U
theory (Yerkes & Dodson, 1908), the reticular-activating
hypofrontality model (Audiffren, 2016; Dietrich &
Audiffren, 2011), and neurochemical hypotheses
(McMorris etal., 2016). However, it aligns with a previ-
ous meta-analysis by Y. K. Chang etal. (2012), who
found that high-intensity exercise improved postexer-
cise cognitive performance, and with a meta-analysis
by Ludyga etal. (2016), who reported gains in executive
performance associated with acute bouts of moderate-
intensity exercise. Ludyga etal. (2016) stated that their
meta-analysis “examined effects of aerobic exercise on
executive function based on moderate intensity,” and
further indicate that it therefore “remains unclear how
other exercise intensities influence executive control
and if those effects are further moderated by the
subjects’ characteristics” (p. 1622). Our meta-analysis
fills this gap in the literature, providing corroborating
evidence with more intense but shorter bouts of
The general facilitating effect of high-intensity exer-
cise should not, however, obscure the inherently com-
plex dynamics related to the effect of exercise on
cognition. Because studies typically average perfor-
mance on a given cognitive task, a lot of information
18 Moreau, Chou
Table 7. Categorical Moderator Analyses
Moderator category, moderator variable, and planned
levels N d [95% CI]
Exercise duration at high-intensitya
0–5 min 6 0.18 [−0.03, 0.40]
6–10 min 60.24* [0.04, 0.44]
11–20 min 5 0.21 [−0.04, 0.47]
> 20 min 6 0.29** [0.09, 0.48]
Exercise intensity (HRmax) of high-intensity conditiona
High (~77%–88.5% HRmax) 16 0.27*** [0.13, 0.42]
Very high (~88.6%–99.9% HRmax) 6 0.22 [−0.06, 0.50]
Maximal (100% HRmax) 6 0.15 [−0.11, 0.41]
Exercise modality
Cycling 13 0.15 [0.00, 0.30]
Running 15 0.32*** [0.17, 0.47]
Exercise rhythm
Continuous 14 0.28*** [0.12, 0.44]
Intermittent 14 0.19* [0.03, 0.36]
Cognitive task
Timing of cognitive testinga
0–1 min 14 0.18* [0.01, 0.34]
> 1–10 min 7 0.30** [0.08, 0.52]
> 10 min 8 0.28** [0.07, 0.49]
Cognitive task domain
Attention 22 0.17* [0.03, 0.32]
Cognitive flexibility 50.30* [0.00, 0.59]
General executive function 1 0.39 [−0.21, 0.98]
Long-term memory 2 0.20 [−0.32, 0.72]
Inhibitory control 12 0.27** [0.08, 0.46]
Working memory 80.32** [0.08, 0.55]
Baseline cognitive testing
Has baseline cognitive testing 16 0.24*** [0.10, 0.38]
No baseline cognitive testing 12 0.23* [0.04, 0.42]
Study design
Within-subjects 18 0.23 [−0.10, 0.57]
Between groups 10 0.25 [−0.10, 0.60]
Comparison group typeb
Resting group 21 0.34*** [0.20, 0.47]
Lower-intensity exercise group 14 0.07 [−0.09, 0.23]
Study qualitya
Low 1 0.27 [−0.46, 1.01]
Average 4 0.59*** [0.26, 0.92]
High 23 0.19** [0.07, 0.31]
Participant fitness
Low 3 0.24 [−0.13, 0.61]
Moderate 4 0.35* [0.04, 0.66]
High 6 0.05 [−0.20, 0.31]
Participant agea
6–13 4 0.13 [−0.15, 0.41]
14–18 2 0.51* [0.12, 0.90]
19–30 19 0.26*** [0.12, 0.40]
31–60 2 0.19 [−0.21, 0.60]
> 60 1 −0.04 [−0.55, 0.48]
Note: Boldface type indicates effects that were both significant and had an adequate (> 5) number
of individual studies supplying the effects. HRmax = maximum heart rate; N = number of independent
studies supplying categorical data.
aThese were also analyzed as continuous moderators. bThis was a significant moderator; the effect of
exercise on cognitive performance after exercise was different on at least two different levels of this
p < .10. *p < .05. **p < .01. ***p < .001.
Exercise and Executive Function 19
Córdova et al. (2009), 8
Córdova et al. (2009), 7
Córdova et al. (2009), 9
Craft (1983), 1
Ramos et al. (2017), 3
Craft (1983), 3
Ramos et al. (2017), 5
Etnier et al. (2014), 1
Alves et al. (2014), 1
Basso et al. (2015), 4
Ramos et al. (2017), 2
Hwang, Castelli, et al. (2016), 1
Ramos et al. (2017), 6
Basso et al. (2015), 2
Cooper et al. (2016), 1
Ramos et al. (2017), 1
Basso et al. (2015), 3
Ramos et al. (2017), 4
Basso et al. (2015), 1
Hwang, Castelli, et al. (2016), 2
Etnier et al. (2014), 2
Working Memory
Study Name Effect Size (Cohen’s d) and 95% CI
RE Model, d = 0.3189, p = .0079, 95% CI = [.0835, .5543]
−2 −1
Fig. 3. (continued on next page)
can be lost about the fine-grained relationship between
exercise and cognitive performance (see, e.g., Moreau
& Corballis, 2018). For instance, it is possible that high-
intensity exercise leads to an early dip in performance
as a result of competing physiological resources fol-
lowed by a subsequent increase in performance as a
result of higher levels of activation. If both of these
phases (debilitating effect followed by a facilitating
effect) occur within the course of a cognitive task, aver-
age performance could represent a very coarse measure
of the effect of exercise on cognitive performance. The
lack of systematic analyses investigating cognitive
dynamics after exercise, together with inconsistencies
in defining thresholds of high-intensity, perhaps explain
some of the discrepancies between our results and
those of previous meta-analyses (Y. K. Chang etal.,
2012; Lambourne & Tomporowski, 2010; McMorris &
Hale, 2012). With this caveat in mind, moderator analy-
ses helped to further identify factors influencing the
effect of high-intensity exercise on executive function.
We discuss these factors in detail hereafter.
What characteristics of exercise matter
in eliciting cognitive enhancement?
The type of comparison or control used in the included
studies was a significant moderator. Specifically,
high-intensity exercise was associated with superior
executive performance compared with rest but not with
lower-intensity exercise; in the latter comparison, the
two exercising conditions elicited similar gains in exec-
utive function. These findings are in line with our
hypothesis of a larger effect for comparisons to resting
conditions. In contrast with the predictions made by a
number of models, for example, psychological (Thum
etal., 2017) or neurochemical (McMorris etal., 2016)
theories, high-intensity exercise does not appear to
impair executive function. This is despite some of the
physiological changes induced by exercise (e.g.,
increases in lactate and catecholamines, shifts in blood
flow to cortical regions) being larger following high-
intensity regimens (McMorris etal., 2016), suggesting
that rather than following an inverted-U function, the
facilitating effect of exercise holds at high-exercise
intensities, at least after the bout.
To test this idea, we used common subcategories to
further characterize high-intensity exercise, namely high
(77%–88.5% HRmax), very high (88.6%–99.9% HRmax), and
maximal (100% HRmax). Whereas high-intensity exercise
was associated with a small, facilitating effect, exercise
at very-high and maximal intensities was associated with
nonsignificant effects. However, this moderator did not
significantly affect the relationship between exercise and
executive function in categorical-moderator analyses,
20 Moreau, Chou
Study Name Effect Size (Cohen’s d) and 95% CI
RE Model, d = 0.2734, p = .005, 95% CI = [.0826, .4643]
−2 −1
Chang et al. (2017), 1
Whyte et al. (2015), 1
Tsukamoto et al. (2016), 4
Kujach et al. (2018), 2
Zimmer et al. (2016), 11
Kamijo et al. (2007), 3
Zimmer et al. (2016), 2
Cooper et al. (2016), 5
Kamijo et al. (2007), 6
Kamijo et al. (2007), 1
Zimmer et al. (2016), 1
Zimmer et al. (2016), 3
Kamijo et al. (2007), 2
Kao et al. (2017), 1
Kujach et al. (2018), 1
Kamijo et al. (2007), 4
Tsukamoto et al. (2016), 3
Zimmer et al. (2016), 10
Kao et al. (2017), 2
Kao et al. (2017), 3
Zimmer et al. (2016), 12
Lambrick et al. (2016)
Hwang, Brothers, et al. (2016), 3
Chang et al. (2017), 2
Alves et al. (2014), 4
Hwang, Brothers, et al. (2016), 2
Peruyero et al. (2017), 3
Kamijo et al. (2007), 5
Llorens, Sanabria, Huertas, Molina, & Bennett (2015), 3
Cooper et al. (2016), 4
Peruyero et al. (2017), 4
Llorens, Sanabria, Huertas, Molina, & Bennett (2015), 2
Inhibitory Control
Cognitive Flexibility
Study Name Effect Size (Cohen’s d) and 95% CI
RE Model, d = 0.298, p = .0476, 95% CI = [.0032, .5929]
−2 −1
Córdova et al. (2009), 14
Córdova et al. (2009), 13
Ramos et al. (2017), 10
Ramos et al. (2017), 9
Córdova et al. (2009), 15
Berse et al. (2015), 2
Netz et al. (2007), 1
Berse et al. (2015), 1
Córdova et al. (2009), 17
Córdova et al. (2009), 16
Netz et al. (2007), 2
Hwang, Brothers, et al. (2016), 6
Córdova et al. (2009), 18
Fig. 3. (continued on next page)
Exercise and Executive Function 21
−2 −1
Cooper et al. (2016), 7
Córdova et al. (2009), 4
Córdova et al. (2009), 11
Córdova et al. (2009), 2
Tsukamoto et al. (2016), 2
Kujach et al. (2018), 3
Wohlwend et al. (2017), 4
Córdova et al. (2009), 10
Kamijo et al. (2007), 10
Kamijo et al. (2007), 9
Córdova et al. (2009), 3
Kamijo et al. (2004), 1
Ramos et al. (2017), 8
Netz et al. (2007), 4
Whyte et al. (2015), 2
Budde et al. (2012), 2
Alves et al. (2014), 2
Kamijo et al. (2007), 7
Kamijo et al. (2004), 2
Lemmink & Visscher (2005), 3
Kamijo et al. (2007), 8
Llorens, Sanabria, & Huertas (2015), 3
Llorens, Sanabria, & Huertas (2015), 1
Córdova et al. (2009), 6
Córdova et al. (2009), 12
Llorens, Sanabria, & Huertas (2015), 2
Kamijo et al. (2007), 12
Craft (1983), 2
Ramos et al. (2017), 7
Wohlwend et al. (2017), 2
Tsukamoto et al. (2016), 6
Llorens, Sanabria, & Huertas (2015), 4
Cooper et al. (2016), 6
Cooper et al. (2016), 3
Cooper et al. (2016), 2
Lemmink & Visscher (2005), 1
Netz et al. (2007), 3
Zimmer et al. (2016), 4
Zimmer et al. (2016), 9
Wohlwend et al. (2017), 3
Kujach et al. (2018), 4
Budde et al. (2012), 3
Budde et al. (2012), 1
Tsukamoto et al. (2016), 5
Alves et al. (2014), 5
Zimmer et al. (2016), 7
Hwang, Brothers, et al. (2016), 5
Wohlwend et al. (2017), 1
Alves et al. (2014), 3
Hwang, Brothers, et al. (2016), 1
Hwang, Castelli, et al. (2016), 4
Tsukamoto et al. (2016), 1
Zimmer et al. (2016), 6
Zimmer et al. (2016), 5
Córdova et al. (2009), 1
Hwang, Brothers, et al. (2016), 4
Kamijo et al. (2007), 11
Walsh et al. (2018), 2
Peruyero et al. (2017), 5
Walsh et al. (2018), 1
Lemmink & Visscher (2005), 2
Walsh et al. (2018), 3
Zimmer et al. (2016), 8
Lemmink & Visscher (2005), 4
Peruyero et al. (2017), 1
Peruyero et al. (2017), 6
Hwang, Castelli, et al. (2016), 3
Córdova et al. (2009), 5
Peruyero et al. (2017), 2
Chang et al. (2017), 4
Chang et al. (2017), 3
Llorens, Sanabria, Huertas, Molina, & Bennett (2015), 1
Study Name Effect Size (Cohen’s d) and 95% CI
RE Model, d = 0.1735, p = .0215, 95% CI = [.0255, .3214]
Fig. 3. Forest plots of effect sizes for working memory (a), inhibitory control (b), cognitive flexibility (c), and
attention (d). The sizes of the squares represent within-domain differences in sample size only. The diamond
at the bottom represents the overall effect. The dotted vertical line represents an effect size (d) of 0. CI =
confidence interval; RE = random effects.
22 Moreau, Chou
whereas in continuous-moderator analyses there was
only a small, nonsignificant decrease in facilitating
effects as intensity increased, particularly at the highest
intensities (> 88.5% HRmax). The inconsistencies between
past theoretical (McMorris etal., 2016) and meta-analytical
(Y. K. Chang etal., 2012) studies are perhaps not surpris-
ing given that they relied on different thresholds. Using
a threshold of > 88.5% HRmax, McMorris and Hale (2012)
did not find facilitating effects, whereas McMorris etal.
(2016) predicted detrimental effects; using a threshold
of > 77% HRmax, Y. K. Chang etal. (2012) reported
improved postexercise performance. Our results,
derived using the lower threshold, align with past meta-
analytic literature (Y. K. Chang etal., 2012; McMorris
and Hale, 2012) and support the idea of an optimal
intensity (i.e., 77%–88.5% HRmax) of high-intensity exer-
cise to induce improvements in executive function.
Other exercise characteristics could potentially affect
executive function. Just as higher exercise intensities
are thought to be associated with greater physiological
Effect Size (Cohen’s d)
Standard Error
Fig. 4. Funnel plot of effect sizes for studies in the meta-analysis. Each dot represents an individual effect size and is plotted as a
function of standard error. Light gray and dark gray triangles denote 95% and 99% confidence intervals, respectively, for the effect
sizes, given the absence of publication (or small-study) bias. The vertical line represents the random-effects-model estimate (d = 0.24).
Table 8. Continuous Moderator Analyses
Moderator type and variable N d0 [95% CI] Δ [95% CI] ƛ
at ƛ
Exercise duration at high-intensity 23 0.16* [0.00, 0.32] +0.0039 [−0.0036, 0.0115] 50 0.36
Exercise intensity (%HRmax) of high-intensity condition 26 0.27 [−1.01, 1.55] −0.0007 [−0.0151, 0.0137] 100 0.20
Cognitive task
Timing of cognitive testinga27 0.19* [0.04, 0.33] +0.0065 [−0.0057, 0.0188] 30 0.38
Research protocol/participant
Study quality 28 0.93* [0.07, 1.79] −0.0078 [−0.0173, 0.0018] 100 0.15
Participant age 28 0.31** [0.08, 0.54] −0.0030 [−0.0113, 0.0053] 63.1 0.12
Percentage male (exploratory) 28 0.44*** [0.20, 0.68] −0.0037 [−0.0076, 0.0001] 100 0.07
Note: dproj = projected Cohen’s d; d0 = estimated effect when moderator is 0; N = number of independent studies supplying continuous data; Δ =
change in effect size associated with increasing the value of the continuous moderator by 1 on the effect-size estimate; ƛ = maximum value of the
continuous moderator.
aThis does not include testing for long-term memory.
p < .10. *p < .05. **p < .01. ***p < .001.
Exercise and Executive Function 23
effects, so are longer bouts of exercise (Awopetu, 2014).
We therefore expected greater effects on cognitive per-
formance as exercise duration increased. Although
duration as a continuous moderator trended in this
direction, it was nonetheless nonsignificant. Likewise,
different exercise durations (0–5, 6–10, 11–20, and > 20
min) led to small, facilitating effects (some significant,
some nonsignificant) that were similar in magnitude
to one another. Note that this could be due to the
larger effects associated with higher intensities being
compensated by shorter durations; overall, the cor-
relation between intensity and duration was negative
and marginally significant. Furthermore, whereas the mag-
nitude of the facilitating effect was consistent across
different durations, its latency could potentially vary
independently. For example, some longer exercise
bouts might be associated with a delayed, prolonged
peak of the facilitating effect. We could not examine
this relationship because too few studies examined
duration as an independent variable. Fortunately, the
field is moving toward the continuous measurement of
physiological (e.g., lactate, serum brain-derived neuro-
trophic factor, heart rate) and cognitive changes during
and after exercise, aspects that will help further our
understanding of the complex, dynamic interactions
between physiology and cognition.
We also explored how exercise rhythmeither con-
tinuous or intermittentaffected behavior. Intermittent-
exercise protocols have become increasingly popular,
perhaps because of their cardiorespiratory (Elliott,
Rajopadhyaya, Bentley, Beltrame, & Aromataris, 2015;
Milanovic´, Sporiš, & Weston, 2015; J. S. Ramos, Dalleck,
Tjonna, Beetham, & Coombes, 2015; Weston et al.,
2014) and psychoaffective (Thum etal., 2017) benefits
compared with traditional continuous exercise. We
examined whether this form of exercise had compa-
rable benefits on immediate executive function but did
not find a difference between the effects of intermittent
and continuous exercise compared with rest—both
types of protocols were associated with small, facilitat-
ing effects. This result is consistent with physiological
responses to exercise, which appear to be similar
regardless of whether rest periods are interleaved
throughout the exercise protocol (Arnardóttir, Boman,
Larsson, Hedenström, & Emtner, 2007; Safarimosavi,
Mohebbi, & Rohani, 2018). Note that this could be an
important advantage of intermittent exercisewith less
exercise volume overall but with higher peak intensities
.01 .02 .03.04 .05
p Value
Percentage of Test Results
Observed p Curve
Null of No Effect
Null of 33% Power
Power Estimate: 74%, 90% CI = [56%, 86%]
Tests for Right-Skewness: pFull = .0001, pHalf < .0001
Tests for Flatness: pFull = .9996, pHalf < .9999, pBinomial = .3748
Fig. 5. P curve for all significant (p < .05) effect sizes. The observed p curve includes
35 statistically significant (p < .05) results, of which 23 were p < .025; 112 additional
results were entered but excluded from the p curve because they were p > .05. The 90%
confidence interval is given for the power estimate.
24 Moreau, Chou
compared with continuous exercise, similar effects can
potentially be elicited.
Two caveats need to be acknowledged, however.
First, these short-term similarities may not reflect long-
term equivalence, both in terms of mechanisms and
outcomes. Several studies have investigated the physi-
ological effect of chronic high-intensity exercise
(Maillard, Pereira, & Boisseau, 2018), as well as the
relationship between acute and chronic physiological
effects (Tonoli etal., 2012), yet the extent to which
acute effects translate or even relate to long-term
change at the cognitive level remains unknown. Sec-
ond, the high intensity typically associated with inter-
mittent exercise may not be desirable for specific
populations (e.g., low-fit individuals, older adults). In
this regard, acute exercise at more moderate intensities
(55% to 70% HRmax) might be preferred for individuals
prone to injury or at risk of cardiovascular conditions
(Ludyga etal., 2016). The rhythm and intensity of exercise
remain dissociable, however, and further research will
allow exploring whether lower intensities of intermittent
exercise could elicit similar improvements in executive
function (Kujach etal., 2018; Ludyga etal., 2016).
Finally, an additional characteristic of exercise was
worth investigating in our viewwhen exercising at
high intensity, does it matter whether one runs, cycles,
or swims? In other words, is the modality of exercise
relevant, or should it just be treated as a personal pref-
erence with no direct impact on cognition? Most studies
examined either running or cycling, as these modalities
were relatively easy to measure and control. Because
the two modalities have important physiological differ-
ences (Millet, Vleck, & Bentley, 2009), it was plausible
that these be reflected in terms of differences in execu-
tive function. However, we did not find differences
attributable to modalities of exercise in the present
meta-analysis—both were associated with small facili-
tating effects, with a marginally significant effect in the
case of cycling and a significant effect in the case of
running. This result differed from our prediction: We
hypothesized, on the basis of Lambourne and Tompo-
rowski’s (2010) meta-analysis, that cycling would elicit
a greater facilitating effect on executive function. Note
that in Lambourne and Tomporowski (2010), the dif-
ference between modalities was more pronounced
when cognitive tests were administered during, rather
than after, exercise, which could suggest that the
observed benefit of cycling compared with running in
their study was due to the difficulty of performing cog-
nitive tasks while running rather than true differences
in the effect of running versus cycling. Nevertheless,
the disparity between the findings of the present study
and those reported by Lambourne and Tomporowski
(2010) points to exercise modality as a potential mod-
erator of interest for future investigation.
How were our results influenced
by testing, protocol, and sample
In addition to variables related to exercise itself, we
tested moderators related to the specific measurements
used to assess executive function, for example, whether
the timing of testing had an effect on cognitive perfor-
mance. We did not find a significant effect of timing;
tests conducted in the ranges 0 to 1, 1 to 10, and > 10
min after a bout were all associated with small, facilitat-
ing effects. These results differed from our predictions,
which were made on the basis of Y. K. Chang etal.
(2012), who found tests administered after 1 minbut
not 0 to 1 minafter a bout to elicit facilitating effects.
Several studies administered the same postexercise
tasks repeatedly to further elucidate the temporal char-
acteristics of exercise-induced cognitive change. We
chose not to analyze tests beyond the first repetition
(i.e., posttest) to reduce confounds from practice
effects. Recent studies examining the influence of test
timing have strived to control practice effects; for exam-
ple, Zimmer etal. (2017) randomly assigned partici-
pants to multiple groups with only one posttest session,
which was scheduled at a different time after exercise
(0, 30, 60, and 90 min). They found a moderating effect
of test timing on cognitive performance, with a detri-
mental effect of high-intensity exercise on executive func-
tion at the earlier (0, 30 min) but not the later (60, 90
min) time points. Furthermore, performance decline was
correlated with blood-lactate concentrations, suggesting
that the observed cognitive effects might be mediated by
physiological variables (Zimmer etal., 2017).
Furthermore, we tested whether our findings were
contingent on a comparison against a baseline measure
of executive performance. Contrary to our prediction,
the presence or absence of baseline cognitive testing
did not significantly moderate the results; studies with
baseline testing (i.e., repeated-measure studies) and
studies without baseline testing (i.e., single-measure
studies) both showed small, facilitating effects of exer-
cise on executive performance. Our findings also did
not differ depending on the subdomain of executive
function we focused on (i.e., attention, cognitive flex-
ibility, inhibitory control, or working memory). These
results matched our general prediction and were in
line with findings from Y. K. Chang etal. (2012), who
found an overall facilitating effect of exercise on exec-
utive function when collapsing all intensities and test
Exercise and Executive Function 25
Most studies included in the current meta-analysis
were of high quality, with several average-quality stud-
ies and only one low-quality study. Study quality was
a marginally significant moderator of the influence of
high-intensity exercise on executive function, with a
moderate, facilitating effect for average-quality studies
and a small, facilitating effect for high-quality studies.
This finding was in line with our prediction (although
we did not predict the direction of the effect) as well
as previous research (Etnier etal., 1997). The larger
effects in average-quality studies possibly arose because
of poor control of confounding variables, and low- and
average-quality studies generally had small sample sizes
(see Tables S3 and S4 in the Supplemental Material).
Because of the paucity of lower-quality studies in the
present meta-analysis, care should be taken in interpret-
ing the effects of study quality.
Moreover, study design—whether the experiment
included a within-subjects or between-groups design—
did not moderate the relationship between high-
intensity exercise and executive function. Both types
of design were associated with nonsignificant, small,
facilitating effects of high-intensity exercise on execu-
tive function, with substantial heterogeneity in the
results. This result differed from our prediction (which
was based on the results of Y. K. Chang etal., 2012) of
a larger effect for within-subjects studies. Our results
suggest that most within-subjects studies included in
the present meta-analyses controlled adequately for
confounds, leading to noninflated effects, which may
not have been the case in previous meta-analyses.
Finally, we also tested the influence of sample char-
acteristics on the relationship between exercise and
executive performance. These included fitness level,
age, and sex. None of these moderators showed a sig-
nificant effect despite prior studies pointing to the con-
trary (e.g., Y. K. Chang etal., 2012; Etnier etal., 2006).
Note that for fitness level, there were important caveats
that prevented a full examination of this moderator,
which we detail in the limitations section. With respect
to age, few studies examined participants who were
not young adults, limiting our ability to investigate the
effect of this moderator.
Interactions and model comparisons
We tested a number of interactions between moderators
to examine whether the effect of a moderator variable
was contingent on another one. From the 15 pairs of
moderators defined a priori, we found two significant
interactions after correcting for multiple comparisons.
First, when the comparison group was a lower intensity
exercise group, those individuals undergoing very-high-
intensity exercise experienced significantly larger
facilitating effects than those undergoing high-intensity
exercise. Second, when exercise was intermittent, those
who were tested 0 to 1 min after exercise experienced
significantly larger facilitating effects than those who
were tested > 10 min after exercise. However, given the
low number of studies supplying effects at particular
levels (three effect sizes for very-high- vs. lower-intensity
exercise; three effect sizes for intermittent exercise and
testing > 10 min after a bout), these results should be
interpreted with caution.
To further understand relationships between mod-
erator variables, we compared a number of models that
included single or multiple moderators.5 Of all models,
the model with seven moderators had the lowest resid-
ual heterogeneity, although it remained significantly
larger than zero. Model comparisons showed the model
with one moderator (comparison group) to be a better
fit than the model with the largest number of modera-
tors and the model with no moderators. Thus, including
the one significant moderator of our analyses to the
model was beneficial despite the inherent penalty for
model complexity. In contrast, the addition of other
moderators to the model was detrimental to overall fit
despite the resulting decrease in heterogeneity.
Were our results biased?
Our analyses suggested low bias overall, as illustrated
by the funnel plot shown in Figure 4 and corroborated
with the p-curve analysis shown in Figure 5 that indi-
cated a low probability of p hacking for the significant
effect sizes. Furthermore, analyses of study quality sug-
gested that although most studies were not significantly
biased (i.e., high-quality studies), a few were biased
toward large effects (i.e., low- and average-quality stud-
ies) compared with the overall meta-analytic estimate.
Because the inclusion of lower-quality studies did not
substantially alter funnel-plot symmetry, it is reasonable
to assume that the overall meta-analytic estimate was
not substantially affected by biased studies. Importantly,
the meta-analytic estimate of a subset of the meta-
analytic data composed exclusively of high-quality
studies remained significant, showing a small facilitat-
ing effect of high-intensity exercise. Altogether, these
assessments of quality suggested a low degree of bias;
however, given that we could not obtain all the poten-
tially relevant data we requested, it remains possible
that our findings were affected by publication bias, at
least to a certain extent.
Limitations of the present study
We should point out a few limitations to the present
meta-analysis. First, the study was not preregistered,
26 Moreau, Chou
thus possibly increasing the risk of confirmation bias
and providing weaker control over theory-driven
choices in the selection, coding, and interpretation of
studies (Lakens, Hilgard, & Staaks, 2016). With meta-
analyses, errors may arise in the implementation of
inclusion criteria, extraction of data, and coding of
study characteristics; however, to mitigate bias, we thor-
oughly documented each step of the present meta-
analysis and made this documentation publicly available
to increase transparency and allow reproducibility.
Our meta-analysis was also limited by inherent fea-
tures of the studies we included. For example, a major-
ity of the within-subjects studies, and a few of the
between-groups studies, did not include baseline
scores. This is problematic given that baseline testing
reduces error by providing a reference for each indi-
vidual (Higgins etal., 2011). In several instances, it also
proved difficult to precisely quantify participant fitness
because of missing information and typical lack of indi-
vidual data. Complex relationships such as the one
linking exercise and executive function can be under-
stood only with a detailed assessment of the dynamics
of cognitive performance, that is, of the specific char-
acteristics and their evolution in time. This type of
analysis relies on more than summary statistics and
instead requires individual trial-by-trial data. The stud-
ies we reviewed in this meta-analysis did not include
this information, but we believe future studies should
collect, retain, and ideally share these types of data to
allow more detailed analyses.
In addition, the categorizations we made with regard
to exercise intensities and cognitive domains could
have influenced the results. With respect to the former,
we should note that we did not convert intensity scores
into VO2max, arguably considered the best measure for
exercise testing (Fletcher etal., 2013). Rather, we con-
verted intensity scores into %HRmax using conversion
formulas (Arts & Kuipers, 1994; Lounana etal., 2007)
that were possibly not representative of the samples
included in the current meta-analysis. This was our
preference to allow the inclusion of a larger set of stud-
ies, possibly to the slight detriment of overall quality
the moderator analysis for exercise intensity should be
interpreted with this limitation in mind.
Likewise, our categorization of cognitive tasks under
the umbrella term of executive function helped sim-
plify our analyses and our interpretation of results.
However, there remain important differences in defini-
tions of executive function (see, e.g., Diamond, 2013;
Miyake etal., 2000), and different theoretical accounts
could lead to variations in how the evidence is assessed.
However, we attempted to mitigate this effect by fur-
ther looking into subdomains of executive function to
provide a more detailed analysis at a finer level of
investigation and to increase overall accuracy and
Finally, there were particular characteristics of exer-
cise that remain largely unexplored. For example, few
studies explored the distinction between high (77%–
88.5% HRmax), very-high (88.6%–99.9% HRmax), and
maximal (100% HRmax) intensities. There were very few
studies that examined the adolescent, middle-aged, and
older-adult age groups or that investigated sex differ-
ences in the influence of high-intensity exercise on
executive function. Importantly, these additional ques-
tions do not necessarily require additional studies—
many of these analyses could be carried out if
demographic data were openly available.
Concluding Remarks
In a meta-analysis including 147 effect sizes across 28
studies, we found that high-intensity exercise had a
small, facilitating effect on executive performance and
that this effect was comparable with that of more mod-
erate, but longer, forms of exercise. We believe our
findings have important applications for individuals,
society, and policy. Specifically, they demonstrate the
relevance of high-intensity exercise for those seeking
time-efficient ways to induce immediate cognitive
improvements, such as children who are often con-
strained to sedentary learning environments, profes-
sionals whose work conditions involve long hours
sitting at a desk, or older individuals with limited
opportunities to exercise. In these various contexts,
recognizing the benefits of high-intensity exercise could
promote healthier, more productive lifestyles, with a
positive impact on the community at large.
Action Editor
Laura King served as action editor for this article.
Author Contributions
D. Moreau conceived the study, provided statistical and
research expertise; supervised the literature search, data col-
lection, and data analyses; and provided funding for the proj-
ect. E. Chou performed the literature search, collected the
data, and conducted the analyses. D. Moreau and E. Chou
wrote the manuscript, and both authors approved the final
version for submission. E. Chou is the guarantor.
We thank Beau Gamble for his help with data analysis, and
three anonymous reviewers for their constructive comments.
Declaration of Conflicting Interests
The author(s) declared that there were no conflicts of interest
with respect to the authorship or the publication of this article.
Exercise and Executive Function 27
D. Moreau is supported by the Royal Society of New Zealand
(Marsden) and the Neurological Foundation of New Zealand.
E. Chou was supported by a University of Auckland summer
research scholarship (2017–2018).
Supplemental Material
Additional supporting information can be found at http://
1. Note that a more recent meta-analysis (Ludyga etal., 2016)
investigated the acute effect of moderate aerobic exercise—but
not of higher intensities—on executive function.
2. Note that standardization refers to the paradigm (e.g.,
n-back), not necessarily its specific implementation at the task
level (e.g., type of stimuli, number of trials, duration, data-col-
lection software, system specifications).
3. Note that Y. K. Chang etal. (2012) used similar categories
(hard, very hard, and maximal) but with thresholds of 77% to
93% HRmax, 93% to 100% HRmax, and > 100% HRmax, respectively.
4. For the sake of completeness, separate analyses for within-
subjects and between-groups studies are also reported at
5. Note that this approach does not consider individual mod-
erator levels; rather, it incorporates all levels of each moderator
included in an attempt to account for some of the heterogeneity
present in the meta-analytic model. Because this makes it diffi-
cult to interpret the contributions of individual moderators, only
model-level statistics, such as residual heterogeneity, model fit,
and parsimony, are meaningful.
References marked with an asterisk indicate studies included
in the meta-analysis.
Al-Nakeeb, Y., Lyons, M., Collins, P., Al-Nuaim, A., Al-Hazzaa,
H., Duncan, M. J., & Nevill, A. (2012). Obesity, physical
activity and sedentary behavior amongst British and Saudi
youth: A cross-cultural study. International Journal of
Environmental Research and Public Health, 9, 1490–1506.
*Alves, C. R. R., Tessaro, V. H., Luis, A. C. T., Murakava, K.,
Roschel, H., Gualano, B., & Takito, M. Y. (2014). Influence
of acute high-intensity aerobic interval exercise bout
on selective attention and short-term memory tasks. Per-
ceptual and Motor Skills, 118, 63–72.
American College of Sports Medicine. (2010). ACSM’s guide-
lines for exercise testing and prescription (8th ed.).
Baltimore, MD: Author.
Anderson, C. B. (2003). When more is better: Number of
motives and reasons for quitting as correlates of physical
activity in women. Health Education Research, 18, 525–537.
Arnardóttir, R. H., Boman, G., Larsson, K., Hedenström, H.,
& Emtner, M. (2007). Interval training compared with
continuous training in patients with COPD. Respiratory
Medicine, 101, 1196–1204.
Arts, F. J., & Kuipers, H. (1994). The relation between power
output, oxygen uptake and heart rate in male athletes.
International Journal of Sports Medicine, 15, 228–231.
Assink, M., & Wibbelink, C. J. M. (2016). Fitting three-
level meta-analytic models in R: A step-by-step tutorial.
Tutorials in Quantitative Methods for Psychology, 12,
Audiffren, M. (2016). The reticular-activating hypofrontality
(RAH) model of acute exercise. In T. McMorris (Ed.),
Exercise-cognition interaction (pp. 147–166). San Diego,
CA: Academic Press.
Awopetu, A. R. (2014). A review of the physiological effects
of exercise duration and intensity during walking and jog-
ging. Journal of Emerging Trends in Educational Research
and Policy Studies, 5, 660–667.
Baecke, J. A., Burema, J., & Frijters, J. E. (1982). A short
questionnaire for the measurement of habitual physical
activity in epidemiological studies. American Journal of
Clinical Nutrition, 36, 936–942.
Barnes, C. M., & Van Dyne, L. (2009). ‘I’m tired’: Differential
effects of physical and emotional fatigue on work-
load management strategies. Human Relations; Studies
Towards the Integration of the Social Sciences, 62, 59–92.
*Basso, J. C., Shang, A., Elman, M., Karmouta, R., & Suzuki,
W. A. (2015). Acute exercise improves prefrontal cortex
but not hippocampal function in healthy adults. Journal of
the International Neuropsychological Society, 21, 791–801.
Benton, A. L. (1992). Benton Visual Retention Test (5th ed.).
San Antonio, TX: The Psychological Corporation.
*Berse, T., Rolfes, K., Barenberg, J., Dutke, S., Kuhlenbäumer,
G., Völker, K., . . . Knecht, S. (2015). Acute physical exer-
cise improves shifting in adolescents at school: Evidence
for a dopaminergic contribution. Frontiers in Behavioral
Neuroscience, 9, Article 196. doi:10.3389/fnbeh.2015.00196.
Borer, K. T. (2003). Exercise endocrinology. Champaign, IL:
Human Kinetics.
Browne, S. E., Flynn, M. J., O’Neill, B. V., Howatson, G.,
Bell, P. G., & Haskell-Ramsay, C. F. (2017). Effects of
acute high-intensity exercise on cognitive performance
in trained individuals: A systematic review. Progress in
Brain Research, 234, 161–187.
*Budde, H., Brunelli, A., Machado, S., Velasques, B., Ribeiro,
P., Arias-Carrión, O., & Voelcker-Rehage, C. (2012).
Intermittent maximal exercise improves attentional per-
formance only in physically active students. Archives of
Medical Research, 43, 125–131.
*Chang, H., Kim, K., Jung, Y.-J., & Kato, M. (2017). Effects
of acute high-intensity resistance exercise on cognitive
function and oxygenation in prefrontal cortex. Journal of
Exercise Nutrition & Biochemistry, 21, 1–8.
Chang, Y. K., Labban, J. D., Gapin, J. I., & Etnier, J. L. (2012).
The effects of acute exercise on cognitive performance:
A meta-analysis. Brain Research, 1453, 87–101.
Cohen, J. (1992). A power primer. Psychological Bulletin,
112, 155–159.
*Cooper, S. B., Bandelow, S., Nute, M. L., Dring, K. J.,
Stannard, R. L., Morris, J. G., & Nevill, M. E. (2016).
28 Moreau, Chou
Sprint-based exercise and cognitive function in adoles-
cents. Preventive Medicine Reports, 4, 155–161.
Cooper, S. B., Dring, K. J., & Nevill, M. E. (2016). High-
intensity intermittent exercise: Effect on young people’s
cardiometabolic health and cognition. Current Sports
Medicine Reports, 15, 245–251.
*Córdova, C., Silva, V. C., Moraes, C. F., Simões, H. G., &
Nóbrega, O. T. (2009). Acute exercise performed close
to the anaerobic threshold improves cognitive perfor-
mance in elderly females. Brazilian Journal of Medical
and Biological Research, 42, 458–464.
*Craft, D. H. (1983). Effect of prior exercise on cognitive per-
formance tasks by hyperactive and normal young boys.
Perceptual and Motor Skills, 56, 979–982.
Diamond, A. (2013). Executive functions. Annual Review of
Psychology, 64, 135–168.
Diamond, A., & Lee, K. (2011). Interventions shown to aid
executive function development in children 4 to 12 years
old. Science, 333, 959–964.
Dietrich, A., & Audiffren, M. (2011). The reticular-activat-
ing hypofrontality (RAH) model of acute exercise.
Neuroscience and Biobehavioral Reviews, 35, 1305–1325.
Elliott, A. D., Rajopadhyaya, K., Bentley, D. J., Beltrame, J. F.,
& Aromataris, E. C. (2015). Interval training versus con-
tinuous exercise in patients with coronary artery disease:
A meta-analysis. Heart, Lung & Circulation, 24, 149–157.
*Etnier, J. L., Labban, J. D., Piepmeier, A., Davis, M. E., &
Henning, D. A. (2014). Effects of an acute bout of exer-
cise on memory in 6th grade children. Pediatric Exercise
Science, 26, 250–258.
Etnier, J. L., Nowell, P. M., Landers, D. M., & Sibley, B. A.
(2006). A meta-regression to examine the relationship
between aerobic fitness and cognitive performance. Brain
Research Reviews, 52, 119–130.
Etnier, J. L., Salazar, W., Landers, D. M., Petruzzello, S. J.,
Han, M., & Nowell, P. (1997). The influence of physi-
cal fitness and exercise upon cognitive functioning: A
meta-analysis. Journal of Sport and Exercise Psychology,
19, 249–277.
*Etnier, J. L., Wideman, L., Labban, J. D., Piepmeier, A. T.,
Pendleton, D. M., Dvorak, K. K., & Becofsky, K. (2016).
The effects of acute exercise on memory and brain-
derived neurotrophic factor (BDNF). Journal of Sport and
Exercise Psychology, 38, 331–340.
Fletcher, G. F., Ades, P. A., Kligfield, P., Arena, R., Balady,
G. J., & Bittner, V. A., . . . American Heart Association
Exercise, Cardiac Rehabilitation Prevention, and
Committee of the Council on Clinical Cardiology Council
on Nutrition Physical Activity and Metabolism Council
on Cardiovascular and Stroke Nursing and Council on
Epidemiology and Prevention. (2013). Exercise standards
for testing and training: A scientific statement from the
American Heart Association. Circulation, 128, 873–934.
Gibala, M. J., Little, J. P., Macdonald, M. J., & Hawley, J. A.
(2012). Physiological adaptations to low-volume, high-
intensity interval training in health and disease. The
Journal of Physiology, 590, 1077–1084.
Gomez-Pinilla, F., & Hillman, C. (2013). The influence of
exercise on cognitive abilities. Comprehensive Physiology,
3, 403–428.
Higgins, J. P. T., Altman, D. G., Gotzsche, P. C., Juni, P.,
Moher, D., Oxman, A. D., . . . Cochrane Statistical
Methods Group. (2011). The Cochrane Collaboration’s
tool for assessing risk of bias in randomised trials. BMJ,
343, Article d5928. doi:10.1136/bmj.d5928
Hillman, C. H., Erickson, K. I., & Kramer, A. F. (2008). Be
smart, exercise your heart: Exercise effects on brain and
cognition. Nature Reviews Neuroscience, 9, 58–65.
Hillman, C. H., Snook, E. M., & Jerome, G. J. (2003). Acute
cardiovascular exercise and executive control function.
International Journal of Psychophysiology, 48, 307–314.
Holm, S. (1979). A simple sequentially rejective multiple test
procedure. Scandinavian Journal of Statistics, Theory and
Applications, 6, 65–70.
Hothorn, T., Bretz, F., Westfall, P., Heiberger, R. M.,
Schuetzenmeister, A., & Scheibe, S. (2013). multcomp:
Simultaneous inference in general parametric models
(R package version 1.2-18). Vienna, Austria: Foundation
for Statistical Computing.
*Hwang, J., Brothers, R. M., Castelli, D. M., Glowacki, E. M.,
Chen, Y. T., Salinas, M. M., . . . Calvert, H. (2016). Acute
high-intensity exercise-induced cognitive enhancement
and brain-derived neurotrophic factor in young, healthy
adults. Neuroscience Letters, 630, 247–253.
*Hwang, J., Castelli, D. M., & Gonzalez-Lima, F. (2016).
Cognitive enhancement by transcranial laser stimulation
and acute aerobic exercise. Lasers in Medical Science,
31, 1151–1160.
Iannotti, R. J., Janssen, I., Haug, E., Kololo, H., Annaheim, B.,
Borraccino, A., & HBSC Physical Activity Focus Group.
(2009). Interrelationships of adolescent physical activity,
screen-based sedentary behaviour, and social and psy-
chological health. International Journal of Public Health,
54(Suppl. 2), 191–198.
Kamijo, K., Hayashi, Y., Sakai, T., Yahiro, T., Tanaka, K.,
& Nishihira, Y. (2009). Acute effects of aerobic exer-
cise on cognitive function in older adults. The Journals
of Gerontology, Series B: Psychological Sciences & Social
Sciences, 64, 356–363.
*Kamijo, K., Nishihira, Y., Hatta, A., Kaneda, T., Kida,
T., Higashiura, T., & Kuroiwa, K. (2004). Changes in
arousal level by differential exercise intensity. Clinical
Neurophysiology, 115, 2693–2698.
*Kamijo, K., Nishihira, Y., Higashiura, T., & Kuroiwa, K.
(2007). The interactive effect of exercise intensity and task
difficulty on human cognitive processing. International
Journal of Psychophysiology, 65, 114–121.
*Kao, S.-C., Westfall, D. R., Soneson, J., Gurd, B., & Hillman,
C. H. (2017). Comparison of the acute effects of high-
intensity interval training and continuous aerobic walking
on inhibitory control. Psychophysiology, 54, 1335–1345.
Klika, B., & Jordan, C. (2013). High-intensity circuit train-
ing using body weight. ACSM’s Health & Fitness Journal,
17(3), 8–13.
Konstantopoulos, S. (2011). Fixed effects and variance com-
ponents estimation in three-level meta-analysis. Research
Synthesis Methods, 2, 61–76.
*Kujach, S., Byun, K., Hyodo, K., Suwabe, K., Fukuie,
T., Laskowski, R., . . . Soya, H. (2018). A transferable
high-intensity intermittent exercise improves executive
Exercise and Executive Function 29
performance in association with dorsolateral prefrontal
activation in young adults. NeuroImage, 169, 117–125.
Lakens, D., Hilgard, J., & Staaks, J. (2016). On the reproduc-
ibility of meta-analyses: Six practical recommendations.
BMC Psychology, 4(1), Article 24. doi:10.1186/s40359-016-
Lambourne, K. A. P. M., & Tomporowski, P. (2010). The effect
of exercise-induced arousal on cognitive task performance:
A meta-regression analysis. Brain Research, 1341, 12–24.
*Lambrick, D., Stoner, L., Grigg, R., & Faulkner, J. (2016).
Effects of continuous and intermittent exercise on execu-
tive function in children aged 8–10 years. Psychophysiology,
53, 1335–1342.
*Lemmink, K., & Visscher, C. (2005). Effect of intermittent
exercise on multiple-choice reaction times of soccer play-
ers. Perceptual and Motor Skills, 100, 85–95.
*Llorens, F., Sanabria, D., & Huertas, F. (2015). The influence
of acute intense exercise on exogenous spatial atten-
tion depends on physical fitness level. Experimental
Psychology, 62, 20–29.
*Llorens, F., Sanabria, D., Huertas, F., Molina, E., & Bennett,
S. (2015). Intense physical exercise reduces overt atten-
tional capture. Journal of Sport and Exercise Psychology,
37, 559–564.
Lounana, J., Campion, F., Noakes, T. D., & Medelli, J. (2007).
Relationship between %HRmax, %HR reserve, %VO2max,
and %VO2 reserve in elite cyclists. Medicine and Science
in Sports and Exercise, 39, 350–357.
Ludyga, S., Gerber, M., Brand, S., Holsboer-Trachsler, E., &
Pühse, U. (2016). Acute effects of moderate aerobic exer-
cise on specific aspects of executive function in different
age and fitness groups: A meta-analysis. Psychophysiology,
53, 1611–1626.
Ludyga, S., Pühse, U., Lucchi, S., Marti, J., & Gerber, M.
(2019). Immediate and sustained effects of intermittent
exercise on inhibitory control and task-related heart rate
variability in adolescents. Journal of Science and Medicine
in Sport/Sports Medicine Australia, 22, 96–100.
MacCallum, R. C., Zhang, S., Preacher, K. J., & Rucker, D. D.
(2002). On the practice of dichotomization of quantitative
variables. Psychological Methods, 7, 19–40.
MacInnis, M. J., & Gibala, M. J. (2017). Physiological adapta-
tions to interval training and the role of exercise intensity.
The Journal of Physiology, 595, 2915–2930.
Maillard, F., Pereira, B., & Boisseau, N. (2018). Effect of high-
intensity interval training on total, abdominal and visceral
fat mass: A meta-analysis. Sports Medicine, 48, 269–288.
Matthews, C. E., Chen, K. Y., Freedson, P. S., Buchowski,
M. S., Beech, B. M., Pate, R. R., & Troiano, R. P. (2008).
Amount of time spent in sedentary behaviors in the United
States, 2003-2004. American Journal of Epidemiology, 167,
McMorris, T. (2015). Exercise-cognition interaction: Neuro-
science perspectives. San Diego, CA: Academic Press.
McMorris, T. (2016). History of research into the acute
exercise-cognition interaction: A cognitive psychology
approach. In T. McMorris (Ed.), Exercise-cognition inter-
action: Neuroscience perspectives (pp. 1–28). San Diego,
CA: Academic Press.
McMorris, T., & Hale, B. J. (2012). Differential effects of dif-
fering intensities of acute exercise on speed and accuracy
of cognition: A meta-analytical investigation. Brain and
Cognition, 80, 338–351.
McMorris, T., Hale, B. J., Corbett, J., Robertson, K., & Hodgson,
C. I. (2015). Does acute exercise affect the performance
of whole-body, psychomotor skills in an inverted-U fash-
ion? A meta-analytic investigation. Physiology & Behavior,
141, 180–189.
McMorris, T., Turner, A., Hale, B. J., & Sproule, J. (2016).
Beyond the catecholamines hypothesis for an acute
exercise–cognition interaction: A neurochemical perspec-
tive. In T. McMorris (Ed.), Exercise-cognition interaction:
Neuroscience perspectives (pp. 65–103). San Diego, CA:
Academic Press.
Mekari, S., Fraser, S., Bosquet, L., Bonnéry, C., Labelle, V.,
Pouliot, P., . . . Bherer, L. (2015). The relationship
between exercise intensity, cerebral oxygenation and cog-
nitive performance in young adults. European Journal of
Applied Physiology, 115, 2189–2197.
Milanovic´, Z., Sporiš, G., & Weston, M. (2015). Effectiveness
of high-intensity interval training (HIT) and continuous
endurance training for VO2max improvements: A system-
atic review and meta-analysis of controlled trials. Sports
Medicine, 45, 1469–1481.
Miller, M. G., Hanson, N., Tennyck, J., & Plantz, K. (2019).
A comparison of high-intensity interval training (HIIT)
volumes on cognitive performance. Journal of Cognitive
Enhancement, 3, 168–173.
Millet, G. P., Vleck, V. E., & Bentley, D. J. (2009). Physiological
differences between cycling and running: Lessons from
triathletes. Sports Medicine, 39, 179–206.
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H.,
Howerter, A., & Wager, T. D. (2000). The unity and
diversity of executive functions and their contributions
to complex “frontal lobe” tasks: A latent variable analysis.
Cognitive Psychology, 41, 49–100.
Moher, D., Shamseer, L., Clarke, M., Ghersi, D., Liberati, A.,
Petticrew, M., . . . PRISMA-P Group. (2015). Preferred
reporting items for systematic review and meta-analysis
protocols (PRISMA-P) 2015 statement. Systematic Reviews,
4, 1.
Moreau, D., & Conway, A. R. A. (2013). Cognitive enhance-
ment: A comparative review of computerized and ath-
letic training programs. International Review of Sport and
Exercise Psychology, 6, 155–183.
Moreau, D., & Corballis, M. C. (2018). When averaging goes
wrong: The case for mixture model estimation in psy-
chological science. Journal of Experimental Psychology:
General. Advance online publication. doi:10.1037/
Moreau, D., Kirk, I. J., & Waldie, K. E. (2017). High-intensity
training enhances executive function in children in a ran-
domized, placebo-controlled trial. eLife, 6, Article e25062.
*Netz, Y., Tomer, R., Axelrad, S., Argov, E., & Inbar, O. (2007).
The effect of a single aerobic training session on cogni-
tive flexibility in late middle-aged adults. International
Journal of Sports Medicine, 28, 82–87.
30 Moreau, Chou
O’Leary, K. C., Pontifex, M. B., Scudder, M. R., Brown, M. L.,
& Hillman, C. H. (2011). The effects of single bouts of aer-
obic exercise, exergaming, and videogame play on cogni-
tive control. Clinical Neurophysiology, 122, 1518–1525.
*Peruyero, F., Zapata, J., Pastor, D., & Cervelló, E. (2017). The
acute effects of exercise intensity on inhibitory cognitive
control in adolescents. Frontiers in Psychology, 8, Article
921. doi:10.3389/fpsyg.2017.00921
*Ramos, I., Browne, R., Machado, D., Sales, M., Pereira, R., &
Campbell, C. (2017). Ten minutes of exercise performed
above lactate threshold improves executive control in chil-
dren. Journal of Exercise Physiology Online 20(2), 73–83.
Ramos, J. S., Dalleck, L. C., Tjonna, A. E., Beetham, K. S.,
& Coombes, J. S. (2015). The impact of high-intensity
interval training versus moderate-intensity continuous
training on vascular function: A systematic review and
meta-analysis. Sports Medicine, 45, 679–692.
R Core Team. (2018). R: A language and environment for
statistical computing. Vienna, Austria: R Foundation for
Statistical Computing.
Ruiz, J. R., Ortega, F. B., Martínez-Gómez, D., Labayen, I.,
Moreno, L. A., & De Bourdeaudhuij, I., . . . HELENA Study
Group. (2011). Objectively measured physical activity and
sedentary time in European adolescents: The HELENA
study. American Journal of Epidemiology, 174, 173–184.
Safarimosavi, S., Mohebbi, H., & Rohani, H. (2018). High-
intensity interval vs. continuous endurance training:
Preventive effects on hormonal changes and physiological
adaptations in prediabetes patients. Journal of Strength
and Conditioning Research. Advance online publication.
Samuel, R. D., Zavdy, O., Levav, M., Reuveny, R., Katz, U., &
Dubnov-Raz, G. (2017). The effects of maximal intensity
exercise on cognitive performance in children. Journal
of Human Kinetics, 57, 85–96.
Schmidt, M. (1996). Rey Auditory and Verbal Learning Test:
A handbook. Los Angeles, CA: Western Psychological
Shamseer, L., Moher, D., Clarke, M., Ghersi, D., Liberati, A.,
Petticrew, M., . . . PRISMA-P Group. (2015). Preferred
reporting items for systematic review and meta-analysis
protocols (PRISMA-P) 2015: Elaboration and explanation.
BMJ, 350, Article g7647. doi:10.1136/bmj.g7647
Sibley, B. A., & Etnier, J. L. (2003). The relationship between
physical activity and cognition in children: A meta-anal-
ysis. Pediatric Exercise Science, 15, 243–256.
Simonsohn, U., Nelson, L. D., & Simmons, J. P. (2014).
P-curve: A key to the file-drawer. Journal of Experimental
Psychology: General, 143, 534–547.
Smith, A. (1982). Symbol Digit Modalities Test (SDMT) man-
ual (Revised). Los Angeles, CA: Western Psychological
Sterne, J. A. C., Sutton, A. J., Ioannidis, J. P. A., Terrin, N.,
Jones, D. R., Lau, J., . . . Higgins, J. P. T. (2011).
Recommendations for examining and interpreting funnel
plot asymmetry in meta-analyses of randomised controlled
trials. BMJ, 343, Article d4002. doi:10.1136/bmj.d4002.
Stork, M. J., Banfield, L. E., Gibala, M. J., & Martin Ginis, K. A.
(2017). A scoping review of the psychological responses
to interval exercise: Is interval exercise a viable alterna-
tive to traditional exercise? Health Psychology Review, 11,
Tanaka, H., Monahan, K. D., & Seals, D. R. (2001). Age-
predicted maximal heart rate revisited. Journal of the
American College of Cardiology, 37, 153–156.
Thum, J. S., Parsons, G., Whittle, T., & Astorino, T. A. (2017).
High-intensity interval training elicits higher enjoyment
than moderate intensity continuous exercise. PLoS ONE,
12(1), Article e0166299. doi:10.1371/journal.pone.0166299
Tomporowski, P. D. (2003). Effects of acute bouts of exercise
on cognition. Acta Psychologica, 112, 297–324.
Tonoli, C., Heyman, E., Roelands, B., Buyse, L., Cheung, S. S.,
Berthoin, S., & Meeusen, R. (2012). Effects of different types
of acute and chronic (training) exercise on glycaemic control
in type 1 diabetes mellitus. Sports Medicine, 42, 1059–1080.
Tremblay, M. S., Colley, R. C., Saunders, T. J., Healy, G. N.,
& Owen, N. (2010). Physiological and health implications
of a sedentary lifestyle. Applied Physiology, Nutrition, and
Metabolism, 35, 725–740.
*Tsukamoto, H., Suga, T., Takenaka, S., Tanaka, D., Takeuchi,
T., Hamaoka, T., . . . Hashimoto, T. (2016). Greater impact
of acute high-intensity interval exercise on post-exercise
executive function compared to moderate-intensity con-
tinuous exercise. Physiology & Behavior, 155, 224–230.
Van den Noortgate, W., López-López, J. A., Marín-Martínez,
F., & Sánchez-Meca, J. (2013). Three-level meta-analysis
of dependent effect sizes. Behavior Research Methods, 45,
Verburgh, L., Königs, M., Scherder, E. J. A., & Oosterlaan, J.
(2014). Physical exercise and executive functions in pread-
olescent children, adolescents and young adults: A meta-
analysis. British Journal of Sports Medicine, 48, 973–979.
Viechtbauer, W. (2010). Conducting meta-analyses in R with
the metafor package. Journal of Statistical Software, 36(3).
*Walsh, J. J., Dunlap, C., Miranda, J., Thorp, D. B., Kimmerly,
D. S., & Tschakovsky, M. E. (2018). Brief, high-intensity
interval attention in university students exercise improves
selective attention in university students. International
Journal of Exercise Science, 11(5), 152–167.
Warburton, D. E. R., & Bredin, S. S. D. (2017). Health benefits
of physical activity: A systematic review of current system-
atic reviews. Current Opinion in Cardiology, 32, 541–556.
Wechsler, D. (1974). Wechsler intelligence scale for children—
revised. New York, NY: Psychological Corporation.
Weston, K. S., Wisløff, U., & Coombes, J. S. (2014). High-
intensity interval training in patients with lifestyle-induced
cardiometabolic disease: A systematic review and meta-
analysis. British Journal of Sports Medicine, 48, 1227–1234.
*Whyte, E. F., Gibbons, N., Kerr, G., & Moran, K. A. (2015).
Effect of a high-intensity intermittent-exercise protocol on
neurocognitive function in healthy adults: Implications for
return-to-play management after sport-related concussion.
Journal of Sport Rehabilitation, 24(4), Article 2014-0201.
*Wohlwend, M., Olsen, A., Håberg, A. K., & Palmer, H. S.
(2017). Exercise intensity-dependent effects on cognitive
control function during and after acute treadmill running
Exercise and Executive Function 31
in young healthy adults. Frontiers in Psychology, 8, Article
406. doi:10.3389/fpsyg.2017.00406
Yerkes, R. M., & Dodson, J. D. (1908). The relation of strength
of stimulus to rapidity of habit-formation. Journal of
Comparative Neurology and Psychology, 18, 459–482.
Zimmer, P., Binnebößel, S., Bloch, W., Hübner, S. T., Schenk,
A., Predel, H. G., . . . Oberste, M. (2017). Exhaustive
exercise alters thinking times in a Tower of London task
in a time-dependent manner. Frontiers in Physiology, 7,
Article 694. doi:10.3389/fphys.2016.00694
*Zimmer, P., Stritt, C., Bloch, W., Schmidt, F.-P., Hübner, S. T.,
Binnebößel, S., . . . Oberste, M. (2016). The effects of
different aerobic exercise intensities on serum serotonin
concentrations and their association with Stroop task
performance: A randomized controlled trial. European
Journal of Applied Physiology, 116, 2025–2034.
... The relationship between physical exercise and cognition has garnered significant attention from scholars in recent years [1][2][3]. A key interest is the potential synergistic effects of combining physical and cognitive exercises to enhance overall performance, particularly in populations with demanding cognitive and physical requirements, such as military personnel. ...
... experienced by both groups. Conclusions: These results confirm a single session of acute physical exercise can improve several cognitive functions, consistent with the current literature [3]. However, the addition of cognitive tasks concurrently with physical exercise did not alter these cognitive improvements, nor hinder the physical demands experienced during the session. ...
... The capacity for exercise to modify central (brain) and peripheral (autonomic) nervous system functioning provides possible pathways for exercise to alter cognition in the short term [40][41][42]. Indeed, meta-analytic research shows that exercise in controlled settings can have acute positive effects on cognitive performance [43][44][45], perceived energy levels [46], anxiety, mood, self-esteem [47][48][49], attention, and academic performance [43,50]. These promising outcomes have encouraged further research exploring the potential for exercise to enhance learning in schools, with several studies showing that school-based exercise interventions can have similar short-term effects [12,43,[51][52][53][54][55][56][57]. ...
Exercise has transient effects on cognition and mood, however the impact of Physical Education (PE) on cognitive and affective processes across the school day has not been examined. This study used wearables and questionnaires to track student arousal, engagement, and emotion across school days/periods following PE. Skin conductance, heart rate, heart rate variability, and self-reported engagement, arousal, and valence were analyzed for 23 students (age 15–17 years) on days with and without PE. Sympathetic arousal was significantly higher for two hours following PE and there were stronger decreases in arousal across other classes relative to days without PE. On days with PE, engagement decreased, whereas valence increased from morning to afternoon. These findings highlight the importance of considering acute effects of PE on learning across the entire school day, and demonstrates the feasibility of wearables to clarify how the timing of PE could positively or negatively affect self-regulation and learning.
... Currently, HIIT is increasingly prescribed as a potential therapeutic intervention to address a variety of chronic illnesses including cardiovascular disease, cancer, and metabolic syndrome, due to the robust evidence showing significantly enhanced cardiorespiratory fitness [6,[10][11][12][13]. The approach appears to be a viable strategy for fostering mental, psychological, and cognitive health and may reduce the severity of anxiety and depression [14][15][16][17][18][19]. Lastly, HIIT can be undertaken without the need for expensive gym equipment or access to commercial exercise training facilities. ...
Full-text available
Background It is important to consider biological sex as a variable that might influence exercise adaptation in order to optimize exercise prescription for men and women. Objective The aim of this study was to quantify the impact of biological sex on maximal oxygen uptake (V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max) and performance outcomes after high-intensity interval training (HIIT). Methods A systematic search and review was conducted by two independent reviewers up to 8 September 2022 using MEDLINE, SPORTDiscus, and Sports Medicine & Education Index in ProQuest. Trials including healthy adults were included if they presented data for or compared male and female V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max or performance outcomes in response to HIIT. Performance outcomes included measures of exercise performance and concurrently measured physiological adaptations. Where appropriate, a random-effects, pre-post meta-analysis was undertaken. Data were sub-grouped for men and women, baseline training level, mean age, intervention type, and intervention length. Heterogeneity was assessed using Chi², Cochran’s Q, and Higgins I² and sensitivity analyses, where required. Study quality was assessed using the Newcastle–Ottawa Scale and publication bias was assessed through visual inspection of funnel plots. Results Thirty-three references from 28 trials were included in the review (n = 965; 462 women and 503 men). Meta-analyses included 19 studies for V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max, eight for peak power output from V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max testing (PPO), and five for threshold power (powerAT). Meta-analyses revealed similar increases in V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max in women (g = 0.57; 95% CI 0.44–0.69) and men (g = 0.57; 95% CI 0.42–0.72), and powerAT in women (g = 0.38; 95% CI 0.13–0.64) and men (g = 0.38; 95% CI 0.11–0.64). Raw mean differences for change in V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max were Δ 0.32 L·min⁻¹ and 3.50 mL·kg⁻¹·min⁻¹ in men, versus Δ 0.20 L·min⁻¹ and 3.34 mL·kg⁻¹·min⁻¹ for women. No significant sex differences were present for the primary analysis of any outcome. After sub-grouping, significant differences were present for PPO where the effect size was higher for well-trained women (g = 0.37) compared with well-trained men (g = 0.17), and for V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max where interventions with a duration of 4 weeks or less had significantly smaller effect sizes compared with those longer than 4 weeks (p < 0.001). Unweighted mean percentage change in V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max, PPO, and powerAT across studies was 11.16 ± 7.39%, 11.16 ± 5.99%, and 8.07 ± 6.55% for women, and 10.90 ± 5.75%, 8.22 ± 5.09%, and 7.09 ± 7.17% for men, respectively. Significant heterogeneity was present for both V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max and PPO (I², range: 62.06–78.80%). Sub-grouping by baseline training status and intervention length decreased heterogeneity in most groups. A qualitative synthesis of other outcomes indicated similar improvements in fitness and performance for men and women with some evidence suggesting differences in the mechanisms of adaptation. Limitations and Risk of Bias Publication bias is unlikely to have significantly influenced results for V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}O2max or powerAT, but the meta-analysis of PPO could have benefitted from additional study data to strengthen results. The overlap in age categories and sensitivity of the analysis limits the accuracy of the results of the sub-grouping by age. Conclusions Findings indicated no sex-specific differences for any fitness or performance outcomes. Baseline training status and intervention length accounted for most variability in outcomes. PROSPERO registration number: CRD42021272615.
... Some scholars documented positive effects after HIIT sessions compared with single bouts of continuous aerobic [4,11,15] and resistance exercise [4]. In recent systematic reviews and meta-analyses, it has been demonstrated that, on the one hand, acute HIIT interventions with a total time duration between 11 and 20 min or between 21 and 30 min tended to have positive effects on executive function [16][17][18], but, on the other hand, those with a total time of less than 10 min or more than 30 min did not consistently show positive effects. ...
Full-text available
The aim of this study was to assess the effect of acute short- versus long-interval high-intensity interval training (HIIT) on cognitive performance and psychological states in secondary school students. Fifteen secondary school students (nine males and six females: mean age = 16.2 ± 0.4 years, mean Body Mass Index = 21.2 ± 1.5 kg/m2, and maximum oxygen uptake = 42.2 ± 5.9 mL/kg/min) participated in the current study. They performed one of the following three sessions in a randomized order: (i) a long-interval HIIT (LIHIIT), (ii) a short-interval HIIT (SIHIIT), and (iii) a control condition (CC). Cognitive performance and perceived exertion were assessed pre and immediately post each condition using the d2 test and the Rating of Perceived Exertion (RPE) tool, respectively. Mood state was quantified using the Brunel Mood Scale (BRUMS) questionnaire immediately post each condition. The findings reported higher concentration performance in the SIHIIT compared to the LIHIIT condition (p = 0.043) and the CC (p < 0.001) and in the LIHIIT compared to the CC (p = 0.023). Moreover, the total count of errors was higher in the CC than in the LIHIIT (p = 0.01) and in the SIHIIT conditions (p < 0.001) and in the LIHIIT than in the SIHIIT condition (p = 0.03). RPE value was higher in the LIHIIT and SIHIIT conditions than in the CC (both p < 0.001), whereas no statistically significant difference between LIHIIT and SIHIIT conditions (p = 0.24) was found. Regarding the BRUMS, a significant difference between conditions in the fatigue subscale was found, being higher in LIHIIT with respect to SIHIIT (p = 0.03) and CC (p < 0.05). Vigor differed between conditions, with a higher value than in the LIHIIT (p = 0.04) and CC (p < 0.001). All the remaining subscales did not significantly differ between conditions (p > 0.05). Practitioners may implement short-interval HIIT prior to any tasks that require high levels of visual attention.
... Physical exercise, as a healthy and non-invasive intervention, may provide an important target to decrease the obesity of adolescents and restore and improve the poor cognitive function observed in adolescents with obesity (Xie et al., 2017). The improvement of cognitive function by acute exercise has been supported by crosssectional studies (e.g., Hillman et al., 2006), longitudinal studies (e.g., Sofi et al., 2011), and meta-analysis (e.g., Moreau and Chou, 2019). However, the most effective exercise program remains controversial due to the influence of differences in exercise intensity, cognitive testing time points, and individual differences in cardiorespiratory fitness (CF) (Pontifex et al., 2019). ...
Full-text available
Background Adolescent obesity is associated with impaired inhibitory control. Acute exercise can improve executive function. However, due to the influence of exercise intensity, cognitive test timing, and cardiorespiratory fitness (CF) level, the most effective exercise program remains controversial. Methods The current study investigated the time-course effects of moderate-intensity continuous exercise (MICE) and high-intensity interval exercise (HIIE) on inhibitory control (Stroop) and task-related heart rate variability (HRV) in adolescents with different CF. A mixed experimental design of 2 CF levels (high CF, HCF; low CF, LCF) × 3 exercise methods (MICE, HIIE, CON) × 3 test timing (pre, post-0, post-20) was adopted. Heart rate variability (HRV) and Stroop task tests were conducted before exercise (pre), immediately after exercise (post-0), and 20 min after exercise (post-20). Results Individuals with HCF exhibited a positive decrease in Stroop response time immediately and 20 min after MICE and HIIE, compared to pretest response times (RT). Conversely, individuals with LCF showed a slight increase in Stroop task (RT) only immediately after HIIE. All individuals had a slight increase in ACC after MICE and HIIE compared to before exercise. In addition, compared with the control group, the time-domain index (the square root of the mean squared differences of successive NN intervals, RMSSD) of HRV was significantly decreased, the frequency-domain index (the absolute power of the Low-Frequency band/the absolute power of the High-Frequency band ratio, LF/HF) was significantly increased after MICE and HIIE, and the effect of HIIE on RMSSD and LF/HF was significantly greater than that of MICE. Conclusion The current study found that the acute effects of MICE and HIIE on inhibitory control in obese adolescents were influenced by the interaction of cognitive test timing and cardiorespiratory fitness. Individuals with high cardiorespiratory fitness performed better on the Stroop task than individuals with low cardiorespiratory fitness. The inhibitory control of HIIE in high-cardiorespiratory obese adolescents produced positive effects similar to those in MICE but more lasting, suggesting that HIIE is more beneficial for high-cardiorespiratory obese adolescents. MICE promoted inhibitory control in obese adolescents with low cardiorespiratory fitness, but HIIE impaired inhibitory control in obese adolescents with low cardiorespiratory fitness immediately after exercise, suggesting that low cardiopulmonary fitness obese adolescents may be suitable for MICE rather than HIIE exercise intervention. The shift from balanced HRV to sympathetic dominance after acute exercise reflects increased arousal levels and may be one of the underlying mechanisms by which acute exercise brings benefits to executive function.
... At the group 294 level, the absence of interaction between arousal and conflict effects showed that conflict processing was not hampered 295 during high arousal. These results are consistent with previous literature showing that physical arousal does not hinder 296 performance in stimulus-response conflict tasks during acute exercise, even if at high intensity(Chang et al., 2012, 297 Davranche et al., 2015, Moreau andChou, 2019). Our results mirror those obtained in drowsy participants perform-298 ing an auditory Simon task, in which the behavioural conflict effect was reliably detected even under reduced alertness299 Canales-Johnson et al. (2020). ...
Full-text available
Fluctuations in physical arousal occur constantly along the day and become particularly pronounced at extreme states such as deep sleep or intense physical exertion. While changes in arousal are thought to affect cognitive control, it has been suggested that cognitive control is resilient during the drowsy state as a result of neural compensatory mechanisms. Here, we investigate the higher end of the arousal spectrum by looking at the modulatory effect of high arousal on behavioural and neural markers of cognitive control. We predicted that preserved behavioural measures of cognitive control under high arousal would be accompanied by changes in its typical neural correlates. We conducted an electroencephalography study in which 39 expert cyclists (37 males, 2 females) were presented with an auditory stimulus-response conflict task while cycling on a stationary bike. Participants performed two experimental sessions on different days: one at low intensity and one at high intensity pedalling. Consistent with our predictions, we found no behavioural difference in cognitive conflict measures between the two exercise conditions. However, the typical midfrontal-theta power signature of cognitive control was no longer reliable at high-intensity exercise. Similarly, time-frequency multivariate decoding failed to decode stimulus conflict. On the other hand, we found no difference between intensity levels in whole-brain connectivity measures. Therefore, we suggest that the human cognitive control system is resilient even at high arousal states and propose that the dissociation between behavioural and neural measures could indicate the activation of neural compensatory mechanisms as a response to physiological pressure.
Full-text available
Underwater operations are widely used in diverse fields such as marine exploration, underwater construction and infrastructure maintenance, and military missions. Previous research has emphasized the significance of maintaining cognitive function during these tasks. However, the impact of underwater operations on cognitive function and the underlying mechanisms remain unclear. Hence, this study aimed to investigate the effects of underwater operations on cognitive function and explore the potential molecular mechanisms involved. We accomplished this first by assessing underwater operators’ stress response, anxiety, and cognitive function before and after a single underwater operation of two different durations and found that 30 min underwater operation improved cognitive function while 3 h underwater operation induced significant cognitive decline. Then, an animal model of swimming in a hyperbaric environment at 2.0ATA (atmospheres absolute) for varying durations was applied to simulate underwater operations. Behavioral tests, histological examinations, biochemical assays were conducted, and results indicated that the effect of a single underwater exercise on cognitive function was time-dependent and prolonged underwater exercise caused significant cognitive impairment. Furthermore, RNA-sequencing was conducted for the normal control group and the most significantly impaired group, leading to the focus on neuroinflammation and the identification of C-C chemokine receptor type 3(CCR3) as a potential target for further investigation. Finally, knockdown experiment was performed using an adeno-associated virus (AAV) vector containing shRNA (CCR3)-EGFP injected to the rats’ hippocampus to explore the involvement of CCR3 in cognitive impairment induced by prolonged underwater exercise. Results revealed that CCR3 knockdown significantly alleviated neuroinflammation and cognitive impairment. Moreover, prolonged underwater exercise activated microglia and promotes their polarization towards the pro-inflammatory phenotype. Conversely, CCR3 knockdown switched the activated microglia to the anti-inflammatory phenotype. Taken together, these results highlight the time-dependent effect of a single underwater operation on cognitive function and shed insight on alleviating CCR3-mediated neuroinflammation as potential intervention targets to protect the brain during underwater operations.
Full-text available
Aging is associated with deterioration in dentate gyrus (DG) and CA3, both crucial hippocampal subfields for age susceptible memory processes such as mnemonic discrimination (MD). Meanwhile, a single aerobic exercise session alters DG/CA3 function and neural activity in both rats and younger adults and can elicit short-term microstructural alterations in the hippocampus of older adults. However, our understanding of the effects of acute exercise on hippocampal subfield integrity via function and microstructure in older adults is limited. Thus, a within subject-design was employed to determine if 20-min of moderate to vigorous aerobic exercise alters bilateral hippocampal subfield function and microstructure using high-resolution functional magnetic resonance imaging (fMRI) during an MD task (n = 35) and high angular resolution multi-shell diffusion imaging (n = 31), in healthy older adults, compared to seated rest. Following the exercise condition, participants exhibited poorer MD performance, particularly when their perception of effort was higher. Exercise was also related to lower MD-related activity within the DG/CA3 but not CA1 subfield. Finally, after controlling for whole brain gray matter diffusion, exercise was associated with lower neurite density index (NDI) within the DG/CA3. However, exercise-related differences in DG/CA3 activity and NDI were not associated with differences in MD performance. Our results suggest moderate to vigorous aerobic exercise may temporarily inhibit MD performance, and suppress DG/CA3 MD-related activity and NDI, potentially through neuroinflammatory/glial processes. However, additional studies are needed to confirm whether these short-term changes in behavior and hippocampal subfield neurophysiology are beneficial and how they might relate to long-term exercise habits.
Full-text available
Recent failed attempts to replicate numerous findings in psychology have raised concerns about methodological practices in the behavioral sciences. More caution appears to be required when evaluating single studies, while systematic replications and meta-analyses are being encouraged. Here, we provide an additional element to this ongoing discussion, by proposing that typical assumptions of meta-analyses be substantiated. Specifically, we argue that when effects come from more than one underlying distributions, meta-analytic averages extracted from a series of studies can be deceptive, with potentially detrimental consequences. The underlying distribution properties, we propose, should be modeled, based on the variability in a given population of effect sizes. We describe how to test for the plurality of distribution modes adequately, how to use the resulting probabilistic assessments to refine evaluations of a body of evidence, and discuss why current models are insufficient in addressing these concerns. We also consider the advantages and limitations of this method, and demonstrate how systematic testing could lead to stronger inferences. Additional material with details regarding all the examples, algorithm, and code is provided online to facilitate replication and to allow broader use across the field of psychology.
Full-text available
The purpose of this study was to investigate the effect of two different acute bouts of high-intensity interval training (HIIT) on the Stroop Color Word task and Critical Flicker Fusion (CFF) threshold test. Twenty-five recreationally active subjects (12 men, 13 women; age 23 ± 2.79 years; body mass index 25.03 ± 3.15 kg m⁻²) were recruited for this study. The acute bouts of HIIT included low-volume (LV) and moderate-volume (MV) sessions. They each had a work to rest ratio of 1:1 min, and exercise was performed at 85% maximal heart rate. The LV session included five intervals, and the MV session included ten intervals. Prior to and after exercise, subjects completed the CFF test and Stroop task. Repeated measures ANOVAs were used to determine the effect of time and volume on test performance. CFF results showed a significant main effect for volume condition across pre/post trials (p = .006), with the MV having a higher average CFF threshold (37.73 ± .63 Hz) compared to LV (36.79 ± .68 Hz). Stroop results showed no main effect of volume condition or interaction (both p > .05) but revealed a significant main effect of time for mean reaction time (RT) total, mean RT congruent/incongruent, and mean RT control (all p < .001). There were no main effects of volume condition or time for proportion correct for all variables (p > .05). The results show that low- and moderate-volume HIIT have similar effects on post-exercise measures of executive function. Our results suggest that the volume of HIIT needed to elicit executive function performance gains may be lower than suggested in previous research.
Full-text available
International Journal of Exercise Science 11(5): 152-167, 2018. This study aimed to investigate the effect of very brief, high-intensity interval exercise (HIIE) on selective attention in university students. As a follow-up, we investigated whether HIIE performed prior to a university lecture improves retention of lecture material. A single-group counterbalanced post-test only design was used for this study. 22 university students (19 females; age = 20.0 ±1.0 years) performed a HIIE and control visit on separate days. During the HIIE session, participants performed 4 separate bodyweight exercises for 1 set each, consisting of eight 20 s intervals interspersed with 10 s rest, totaling 11 minutes in duration, including rest. 10 minutes following exercise cessation, participants completed the d2 test of attention. The control visit consisted of quiet reading followed by completion of the d2. Selective attention, as assessed by the d2 test of attention was significantly greater following a bout of HIIE compared to the control condition. Effect size analysis revealed a moderate effect in favour of HIIE compared to control (d = 0.459 [0.171, 0.747]). Study #2: 23 university students (17 females; age = 19.0 ±0.5 years) performed HIIE and a rest condition prior to attending an exercise physiology lecture on separate days. A quiz was administered 24-hours post-lecture to assess lecture retention. Quiz performance was not different between HIIE and control conditions (p = 0.18). HIIE is a time-effective exercise stimulus that improves selective attention. However, performing HIIE prior to a university lecture did not impact retention of lecture material.
Full-text available
Background High-intensity interval training (HIIT) is promoted as a time-efficient strategy to improve body composition. Objective The aim of this meta-analysis was to assess the efficacy of HIIT in reducing total, abdominal, and visceral fat mass in normal-weight and overweight/obese adults. Methods Electronic databases were searched to identify all related articles on HIIT and fat mass. Stratified analysis was performed using the nature of HIIT (cycling versus running, target intensity), sex and/or body weight, and the methods of measuring body composition. Heterogeneity was also determined ResultsA total of 39 studies involving 617 subjects were included (mean age 38.8 years ± 14.4, 52% females). HIIT significantly reduced total (p = 0.003), abdominal (p = 0.007), and visceral (p = 0.018) fat mass, with no differences between the sexes. A comparison showed that running was more effective than cycling in reducing total and visceral fat mass. High-intensity (above 90% peak heart rate) training was more successful in reducing whole body adiposity, while lower intensities had a greater effect on changes in abdominal and visceral fat mass. Our analysis also indicated that only computed tomography scan or magnetic resonance imaging showed significant abdominal and/or visceral fat-mass loss after HIIT interventions. ConclusionHIIT is a time-efficient strategy to decrease fat-mass deposits, including those of abdominal and visceral fat mass. There was some evidence of the greater effectiveness of HIIT running versus cycling, but owing to the wide variety of protocols used and the lack of full details about cycling training, further comparisons need to be made. Large, multicenter, prospective studies are required to establish the best HIIT protocols for reducing fat mass according to subject characteristics.
Full-text available
Background: Exercise-induced cognitive improvements have traditionally been observed following aerobic exercise interventions; that is, sustained sessions of moderate intensity. Here, we tested the effect of a 6 week high-intensity training (HIT) regimen on measures of cognitive control and working memory in a multicenter, randomized (1:1 allocation), placebo-controlled trial. Methods: 318 children aged 7-13 years were randomly assigned to a HIT or an active control group matched for enjoyment and motivation. In the primary analysis, we compared improvements on six cognitive tasks representing two cognitive constructs (N = 305). Secondary outcomes included genetic data and physiological measurements. Results: The 6-week HIT regimen resulted in improvements on measures of cognitive control [BFM = 3.38, g = 0.31 (0.09, 0.54)] and working memory [BFM = 5233.68, g = 0.54 (0.31, 0.77)], moderated by BDNF genotype, with met66 carriers showing larger gains post-exercise than val66 homozygotes. Conclusion: This study suggests a promising alternative to enhance cognition, via short and potent exercise regimens. Clinical trial registration: Protocol #015078, University of Auckland. Funding: Centre for Brain Research: David Moreau and Karen E Waldie (9133-3706255).
Safarimosavi, S, Mohebbi, H, and Rohani, H. High-intensity interval vs. continuous endurance training: Preventive effects on hormonal changes and physiological adaptations in prediabetes patients. J Strength Cond Res XX(X): 000-000, 2018-The aim of this study was to examine the effects of a 12-week high-intensity interval training (HIIT) intervention, or an isocaloric continuous endurance training (CET) intervention on insulin resistance indices and change in irisin and preptin in patients with prediabetes. Thirty-two prediabetic male patients (age = 38.7 ± 4; body mass index = 26.9 ± 1.4 kg·m; and V[Combining Dot Above]O2peak = 2.49 ± 0.22 L·min) were randomly assigned into 3 training groups (N = 8). These groups were matched based on the required energy expenditure (EE) for completing each protocol: (a) HIIT (10 × 60 seconds at 90% peak oxygen uptake [V[Combining Dot Above]O2peak], 1: 1 work to recovery at 50 W), (b) CET at an intensity equivalent to maximal fat oxidation (Fatmax) (CETFAT) (pedaling for a duration that expends an equivalent EE to an HIIT session [E ≈ HIIT]), (c) CET at an intensity equivalent to anaerobic threshold (CETAT) (E ≈ HIIT), and (d) the control group (CON): continued to perform their daily activities. After intervention, blood glucose levels were significantly (p < 0.05) lower in the HIIT group compared with CETAT group. Exercise training improved the insulin resistance index by 35, 28, and 37% in CETFAT, CETAT, and HIIT groups, respectively. Irisin concentrations in the HIIT and CETAT groups was significantly (p < 0.05) decreased compared with the pre-training values. Also, HIIT and CETFAT resulted in significant (p < 0.05) changes in preptin concentration compared with baseline. This study demonstrated that both HIIT and CETFAT protocols had similar effects on the insulin resistance index of prediabetic patients. Also, the intensity and type of exercise were effective factors in changing irisin and preptin concentrations.
Objectives: To examine the immediate and sustained effects of intermittent exercise sessions at different intensities on inhibitory control and task-related heart rate variability (HRV) in adolescents. Design: Using cluster-randomization, participants from six classes (n=94) were allocated to groups performing 20 min of intermittent exercise at moderate (MIE) or high intensity (HIE) and a control group (CON). Method: Using a computer-based Flanker task, inhibitory control was assessed in a classroom setting prior to and following exercise as well as 30 and 60 min after exercise cessation. At each time point, HRV was recorded via electrocardiography while participants performed the cognitive task. The intermittent exercise sessions were administered in form of a circuit training with a work to recovery ratio of 1:1 in MIE and 2:1 in HIE. Results: The statistical analysis revealed a time by group interaction for Flanker task performance. Based on further examination, significant reductions in reaction time were found from pre to post, to post-30, and to post-60 in MIE only. A time by group interaction was also found for LF/HF ratio, indicating greater increases in HIE and MIE compared to CON. Correlations between change scores in LF/HF ration and task performance were not significant. Conclusions: In a classroom setting, improvements in inhibitory control and information processing elicited by moderately-intense intermittent exercise are sustained over at least 60 min. Changes in task-related HRV follow a different time course, providing no indication that exercise-induced benefits are due to a facilitation of arousal.
The construct validity and the test-retest reliability of a self-administered questionnaire about habitual physical activity were investigated in young males (n = 139) and females (n = 167) in three age groups (20 to 22, 25 to 27, and 30 to 32 yr) in a Dutch population. By principal components analysis three conceptually meaningful factors were distinguished. They were interpreted as: 1) physical activity at work; 2) sport during leisure time; and 3) physical activity during leisure time excluding sport. Test-retest showed that the reliability of the three indices constructed from these factors was adequate. Further, it was found that level of education was inversely related to the work index, and positively related to the leisure-time index in both sexes. The subjective experience of work load was not related to the work index, but was inversely related to the sport index, and the leisure-time index in both sexes. The lean body mass was positively related to the work index, and the sport index in males, but was not related to the leisure-time index in either sex. These differences in the relationships support the subdivision of habitual physical activity into the three components mentioned above.
Although growing attention has been drawn to attainable, high-intensity intermittent exercise (HIE)-based intervention, which can improve cardiovascular and metabolic health, for sedentary individuals, there is limited information on the impact and potential benefit of an easily attainable HIE intervention for cognitive health. We aimed to reveal how acute HIE affects executive function focusing on underlying neural substrates. To address this issue, we examined the effects of acute HIE on executive function using the color-word matching Stroop task (CWST), which produces a cognitive conflict in the decision-making process, and its neural substrate using functional near infrared spectroscopy (fNIRS). Twenty-five sedentary young adults (mean age: 21.0 ± 1.6 years; 9 females) participated in two counter-balanced sessions: HIE and resting control. The HIE session consisted of two minutes of warm-up exercise (50 W load at 60 rpm) and eight sets of 30 s of cycling exercise at 60% of maximal aerobic power (mean: 127 W ± 29.5 load at 100 rpm) followed by 30 s of rest on a recumbent-ergometer. Participants performed a CWST before and after the 10-minute exercise session, during both of which cortical hemodynamic changes in the prefrontal cortex were monitored using fNIRS. Acute HIE led to improved Stroop performance reflected by a shortening of the response time related to Stroop interference. It also evoked cortical activation related to Stroop interference on the left-dorsal-lateral prefrontal cortex (DLPFC), which corresponded significantly with improved executive performance. These results provide the first empirical evidence using a neuroimaging method, to our knowledge, that acute HIE improves executive function, probably mediated by increased activation of the task-related area of the prefrontal cortex including the left-DLPFC.
Background: High-intensity exercise is generally considered to have detrimental effects on cognition. However, high fitness levels are suggested to alleviate this effect. Objectives: The specific objective of this review was to evaluate the literature on the effect of acute high-intensity exercise on cognitive performance in trained individuals. Methods: Studies were sourced through electronic databases, reference lists of retrieved articles, and manual searches of relevant reviews. Included studies examined trained participants, included a high-intensity exercise bout, used a control or comparison group/condition, and assessed cognitive performance via general laboratory tasks during or ≤10min following exercise cessation. Results: Ten articles met the inclusion criteria. Results indicated that the effect of acute high-intensity exercise on cognitive performance in trained individuals is dependent on the specific cognitive domain being assessed. Generally, simple tasks were not affected, while the results on complex tasks remain ambiguous. Accuracy showed little tendency to be influenced by high-intensity exercise compared to measures of speed. Conclusion: Multiple factors influence the acute exercise-cognition relationship and thus future research should be highly specific when outlining criteria such as fitness levels, exercise intensity, and exercise mode. Furthermore, greater research is needed assessing more cognitive domains, greater exercise durations/types, and trained populations at high intensities.