Edited by: Gao-Xia Wei, Chinese Academy of Sciences (CAS), China
Reviewed by: Marinella Coco, Università di Catania, Italy; Ming-Qiang Xiang, Guangzhou Sport University, China; Tao Huang, Shanghai Jiao Tong University, China
This article was submitted to Movement Science and Sport Psychology, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
A sedentary lifestyle and physical inactivity (insufficient exercises time) are prevalent among children and adolescents (Sisson et al.,
Diamond proposed a three-factor model of executive functions and stated that inhibitory control, working memory, and cognitive flexibility are the three core executive functions (EF); the three aforementioned cognitive abilities work together to influence higher-order executive functions such as reasoning, planning, and problem solving (Diamond,
Emerging evidence indicates that physical activity and exercise can influence executive functions such as inhibitory control, working memory and cognitive flexibility both acutely and chronically (Rathore and Lom,
Our research follows the requirements of the international meta-analysis writing guidelines (the PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health-care interventions: explanation and elaboration) for selecting and utilizing research methods (Shamseer et al.,
The databases PubMed, Web of Science, Scopus, The Cochrane Library, China National Knowledge Infrastructure (CNKI), and Wan Fang were searched from January 1 2009 to December 31 2019. Two reviewers independently searched articles published in Chinese and English, supplemented by a manual search, and retrospectively included references if necessary. The following two sets of search terms were used: “physical activity” or “exercises” or “physical fitness” or “physical endurance” or “motor activity” or “physical education” or “sport” or “basketball” or “football” or “running” or “cycling” or “jumping” or “dancing” or “tai chi” or “yoga” or “aerobic” and “executive function” or “inhibition” or “inhibiting ability” or “self-control” or “working memory” or “updating” or “refreshing” or “cognitive flexibility” or “task-switching” or “shifting” and “child” or “student” or “toddler” or “preschooler” or “adolescents.” If an article was incomplete or unavailable, we contacted the corresponding author by email to obtain detailed information. For literature tracing, based on the retrieved literatures or related references listed in the review, we used Baidu Scholar and Google Scholar to search for them retrospectively.
Two reviewers independently screened the articles. When there was a disagreement between the two reviewers, a third reviewer evaluated the original study to reach a consensus. Any potentially relevant research needed to meet the following inclusion criteria: (1) children and adolescent participants aged 5–18 years with right-hand dominance, corrected or normal vision and a healthy body were deemed eligible; (2) the exercise group was the primary intervention measure (e.g., aerobic-based, motor skill-based, combining aerobic, muscular activity, yoga, basketball), compared with different types of control groups (i.e., no-intervention control group, waiting list, and routine care) and all the intervention measures were motor skill-based or aerobic-based and clearly defined in terms of the exercise protocol; (3) preliminary studies were randomized controlled trials, and the randomization was done either on an individual or on a group (e.g., classroom) basis; (4) outcome indicators included test data on executive function (working memory, inhibition and cognitive flexibility), with a minimum of one outcome with quantitative data for calculating the pooled effect size. Conditions for exclusion from the study included: (1) ambiguous explanations of exercises interventions; (2) irrelevant outcomes; and (3) studies for which the full text could not be obtained.
Two reviewers independently extracted data according to a predefined protocol. If there were any differences or inconsistencies, they would discuss the study with a third reviewer. They gathered the following information: (1) literature information, including author name, year of publication and country; (2) sample size (male sample); (3) socioeconomic status; (4) age of subjects, mainly used to divide the type of population; (5) intensity of exercise intervention, duration of intervention, time of intervention, frequency of intervention; (6) intervention program; (7) measurements, mainly including the three dimensions of working memory, inhibition and cognitive flexibility; and (8) adverse events and follow-up.
Similar to previous studies (Zou et al.,
Stata 12.0 (Stata Corp, College Station, TX) was used as data processing software (Press,
Given that overall effect sizes may be influenced by heterogeneity factors (age, study quality, motor skill type, composite type, intervention duration, intervention frequency, and intervention time), several regression analyses were separately performed. Additionally, subgroup analyses were also performed for age, study quality, motor-skill type, composite type, and intervention duration, frequency, and time: (1) prepubertal children (5–12 years) vs. adolescents (12–18 years) (Cardoso,
The Grading Recommendations to Assess Development and Evaluation system (GRADE) is an evidence evaluation system and is one of the international standards for evidence quality and the classification of recommendation strength (Zhang et al.,
The latest review of electronic searches (as of December 2019) retrieved a total of 455 articles. During the preliminary screening, we excluded 353 studies based on their title and abstract for reasons including duplications (
Flow of the study selection method.
The 36 articles included were randomized controlled trials (14 acute exercises, 22 chronic exercises), including 4,577 healthy participants (1,308 participants of acute exercises), of which 2,227 were in the experimental group (670 participants of acute exercises) and 2,350 were in the control group (635 participants of acute exercises) (
Characteristics of the studies included in the meta-analysis (acute exercises).
Benzing et al. ( |
Bern, Switzerland (English) | 65 (34) | NR | 14.51 ± 1.08 | EG: Run + jump + resistance exercises (aerobic-based) |
15 | 60%−70% | (Inhibition) (Fluency) (Cognitive flexibility) | No |
Pate ( |
Carolina, USA (English) | 96 (37) | NR | 10.70 ± 0.60 | EG: Better ideas through exercises (aerobic-based) |
10–20 | NR | (Digit Recall) | No |
Gothe et al. ( |
Urbana, USA (English) | 40 (20) | NR | 9.50 ± 0.50 | EG: Yoga (motor skill-based) |
20 | 60%−70% | (Stroop Test) | No |
Chen et al. ( |
Yangzhou, China (Chinese) | 130 (53) | NR | 9.40 ± 0.30 | EG: Basketball (motor skill-based) |
15–30 | 65% | (Flanker task) (1-back) (More-odd shifting) | No |
Ellemberg and St-Louis-Deschênes ( |
Montreal, Canada (English) | 72 (38) | NR | 7.75 ± 0.65 | EG: Basketball (motor skill-based) |
40 | 63% | (Flanker task) | No |
Jager et al. ( |
Bern, Switzerland (English) | 219 (112) | 6.90 (1.56) | 11.35 ± 0.65 | EG: Run + bicycle (combining aerobic and muscular activity) |
30 | 70% | Wisconsin Card Sorting Test | No |
Kubesch et al. ( |
Ulm, Germany (English) | 81 (NR) | NR | 13–14 | EG: Run + resistance exercises (aerobic-based) |
30 | NR | (Flanker task) (1-back) | No |
Chun et al. ( |
Taiwan, China (English) | 22 (9) | NR | 15.42 ± 1.47 | EG: Bicycle (aerobic-based) |
30 | 65% | Wisconsin Card Sorting Test | No |
Budde et al. ( |
Berlin, Germany (English) | 60 (0) | NR | 14.37 ± 0.53 | EG: Run (aerobic-based) |
12 | 50%-85% | Digit Span task | No |
Yan et al. ( |
Yangzhou, China (Chinese) | 244 (113) | NR | 9.50 ± 0.30 | EG: Run (aerobic-based) |
30 | 60%−69% | (Go/no-go) (1-back) (More-odd shifting) | No |
Cooper et al. ( |
Nottingham, UK (English) | 41 (NR) | NR | 12.30 ± 0. 71 | EG: Basketball (motor skill-based) |
60 | 60%−70% | (Stroop test) (Sternberg paradigm) | No |
Park and Etnier ( |
Daejeon, Korea (English) | 22 (NR) | NR | 15.90 ± 0. 29 | EG: Better ideas through exercises (aerobic-based) |
20 | 60%−70% | (Stroop Test) | No |
Egger et al. ( |
Bern, Switzerland (English) | 216 (110) | NR | 7.94 ± 0. 44 | EG: Better ideas through exercises (aerobic-based) |
20 | NR | (Stroop Test) | No |
Vera et al. ( |
Netherlands, Amsterdam (English) | 38 (12) | NR | 12.30 ± 0.60 | EG: Bicycle (aerobic-based) |
20-30 | 40%−60% | (N-back) | No |
Characteristics of the studies included in the meta-analysis (chronic exercises).
Chen et al. ( |
Yangzhou, China (Chinese) | 40 (20) | NR | 11.36 ± 0.57 | EG: Aerobic dance |
40 | 3 | 8 | (Flanker task) (1-back) (More-odd shifting) | No |
De Greeff et al. ( |
Groninge, Netherlands (English) | 499 (216) | NR | 8.20 ± 0.70 | EG: Aerobic exercises |
20–30 | 3 | 22 | Wisconsin Card Sorting Test | No |
Kval et al. ( |
Stavanger, Norway (English) | 429 (NR) | NR | 10–11 | EG: Jump rope + running + strength training (combining aerobic and muscular activity) |
45 | 2 | 10 | (Stroop Test) | No |
Jiang ( |
Beijing, China (Chinese) | 61 (25) | NR | 5.56 ± 0.35 | EG: Football (motor skill-based) |
35 | 2 | 8 | (Flexible-Item Selection) | No |
Xin ( |
Shandong, China (Chinese) | 40 (20) | NR | 9.10 ± 0.32 | EG: Tennis (motor skill-based) |
40 | 5 | 16 | (Flanker task); (1-back) (More-odd shifting) | No |
Budde et al. ( |
Hamburg, Germany (English) | 71 (32) | NR | 9.35± 0.60 | EG: Run + jump + resistance exercises (aerobic-based) |
45 | 3 | 10 | (Letter Digit Span) | No |
Purohit and Pradhan ( |
Bengaluru, India (English) | 72 (30) | NR | 12.69 ± 1.35 | EG: Yoga (motor skill-based) |
90 | 4 | 12 | (Stroop Test) (More-odd shifting) | No |
Wang ( |
Beijing, China (Chinese) | 30 (14) | NR | 5–6 | EG: Tennis (motor skill-based) |
60 | 2 | 8 | (Flanker task) (1-back) | No |
Yin et al. ( |
Beijing, China (Chinese) | 326 (165) | NR | 7–9 | EG: Run (aerobic-based) |
30 | 3–5 | 20 | (Flanker task) (1-back) (More-odd shifting) | No |
Chaddock-Heyman et al. ( |
Urbana, USA (English) | 26 (11) | 2.32 (1.09) | NR | EG: Bicycle (aerobic-based) |
76.8 | 5 | 22 | (1-back) (More-odd shifting) | No |
Stroth et al. ( |
Ulm, Germany (English) | 35 (NR) | NR | 14.20 ± 0.50 | EG: Aerobic exercises |
40 | 3 | 12 | (1-back) | No |
Lina ( |
Yangzhou, China (Chinese) | 17 (9) | NR | 11.37 ± 1.53 | EG: Run (aerobic-based) |
30 | 4 | 11 | (0-back) | No |
Yan Jun and Chen ( |
Yangzhou, China (Chinese) | 87 (42) | NR | 9.50 ± 0.30 | EG: Aerobic dance |
30 | 3 | 12 | (Flanker task) (1-back) (More-odd shifting) | No |
Keita et al. ( |
Illinois, USA (English) | 36 (19) | NR | 7–9 | EG: Medicine balls + resistance exercises (combining aerobic and muscular activity) |
70 | 5 | 24 | (Reaction time) | No |
Hillman et al. ( |
Illinois, USA (English) | 221 (NR) | NR | 8–9 | EG: Yoga + run (aerobic-based) |
70 | 5 | 24 | (Flanker task) (Cognitive flexibility) | No |
Telles et al. ( |
Uttarakhand, India (English) | 98 (60) | NR | 10.50 ± 1.30 | EG: Yoga (aerobic-based) |
45 | 5 | 12 | (Stroop Test) | No |
Fisher et al. ( |
Glasgow, UK (English) | 64 (29) | 7(1) | 6.10± 0.30 | EG: Run (aerobic-based) |
120 | 10 | (Reaction time) | No |
|
Tarp et al. ( |
Rotterdam, Netherlands (English) | 698 (309) | NR | 12.90± 0.60 | EG: Whole-body movement games |
60 | 4 | 12 | (Reaction time) | No |
Ludyga et al. ( |
Basel, Switzerland (English) | 36 (18) | NR | 12-15 | EG: Medicine balls + relay games (combining aerobic and coordinative exercises) | 5-10 | 5 | 8 | (Stroop Test) | No |
Nie ( |
Nanjing, China (Chinese) | 40 (19) | NR | 13.81± 0.30 | EG: Wuqinxi (aerobic-based) |
45 | 3 | 8 | (Flanker task) (1-back) (More-odd shifting) | No |
Egger et al. ( |
Bern, Switzerland (English) | 142 (70) | NR | 7.91± 0.40 | EG: Better ideas through exercises (aerobic-based) |
20 | 5 | 20 | (Flanker task) | No |
Vera et al. ( |
Netherlands, Amsterdam (English) | 201 (108) | NR | 10.90± 0.70 | EG: Dance (aerobic-based) |
10 | 5 | 9 | (Stroop Test) | No |
The methodological quality of the included studies is presented in
Physiotherapy Evidence Database Scale (PEDro) of the included randomized controlled trials (acute exercises and chronic exercises).
Benzing et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Pate ( |
1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 1 | 1 | 6 |
Gothe et al. ( |
1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 1 | 1 | 6 |
Chen et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Ellemberg and St-Louis-Deschênes ( |
0 | 1 | 1 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Jager et al. ( |
1 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 1 | 1 | 8 |
Kubesch et al. ( |
1 | 1 | 1 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 8 |
Chun et al. ( |
1 | 1 | 0 | 1 | 1 | 0 | 1 | 1 | 1 | 1 | 8 |
Budde et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 6 |
Yan et al. ( |
0 | 1 | 0 | 1 | 0 | 1 | 1 | 1 | 1 | 1 | 7 |
Cooper et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Park and Etnier ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Egger et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Vera et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Chen et al. ( |
1 | 1 | 0 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 9 |
De Greeff et al. ( |
0 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 6 |
Kval et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Jiang ( |
1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 1 | 1 | 6 |
Xin ( |
0 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 6 |
Budde et al. ( |
1 | 1 | 1 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 8 |
Purohit and Pradhan ( |
0 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 1 | 1 | 5 |
Wang ( |
0 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 6 |
Yin et al. ( |
1 | 1 | 0 | 1 | 1 | 0 | 1 | 1 | 1 | 1 | 8 |
Chaddock-Heyman et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Stroth et al. ( |
1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 1 | 0 | 5 |
Lina ( |
0 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 6 |
Yan Jun and Chen ( |
0 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 1 | 1 | 5 |
Keita et al. ( |
0 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 6 |
Hillman et al. ( |
1 | 1 | 1 | 1 | 1 | 0 | 1 | 1 | 1 | 1 | 9 |
Telles et al. ( |
1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 10 |
Fisher et al. ( |
0 | 1 | 0 | 1 | 1 | 1 | 0 | 1 | 1 | 1 | 7 |
Tarp et al. ( |
0 | 1 | 0 | 1 | 0 | 0 | 0 | 1 | 1 | 1 | 5 |
Ludyga et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Nie ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Egger et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Vera et al. ( |
1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 7 |
Twenty-three studies examined the effect of exercises on inhibitory control (as measured by the Stroop, Go/no-go, or Flanker tasks). The aggregated result showed a significant benefit in favor of acute exercises on the inhibitory control of children and adolescents (SMD = −0.25; 95% CI −0.40 to −0.09,
The effect of acute exercises on inhibitory control (SMD = standardized mean difference; CI = confidence interval).
The effect of chronic exercises on inhibitory control (SMD = standardized mean difference; CI = confidence interval).
Twenty-two studies (27 pairwise comparisons) examined the effect of exercises on working memory (as measured by digit span backward, Tower of London, and N-back tasks). A higher negative value of the mean change score for the reaction time indicated less time being required for working memory. The aggregated result showed a significant benefit in favor of acute exercises on working memory (SMD = −0.72; 95% CI −0.89 to −0.56,
The effect of acute exercises on working memory (SMD = standardized mean difference; CI = confidence interval).
The effect of chronic exercises on working memory (SMD = standardized mean difference; CI = confidence interval).
Thirteen studies (14 pairwise comparisons) examined the effect of exercises on cognitive flexibility (as measured by more-odd shifting and the Wisconsin card sorting test). A higher negative value of the mean change score for the reaction time indicated less time being required for cognitive flexibility. The aggregated result showed a significant benefit in favor of acute exercises on cognitive flexibility (SMD = −0.34; 95% CI −0.55 to −0.14,
The effect of acute exercises on cognitive flexibility (SMD = standardized mean difference; CI = confidence interval).
The effect of chronic exercises on cognitive flexibility (SMD = standardized mean difference; CI = confidence interval).
Based on the criteria of GRADE, the assessment of the certainty of the evidence regarding the significant impact of exercises on the subsets (inhibitory control, working memory, and cognitive flexibility) of executive functions in children and adolescents was separately evaluated (
Grading Recommendations to Assess Development and Evaluation (GRADE) assessment of the evidence of certainty for exercise effects.
Chronic exercises | Inhibitory control | Yes | No | No | No | No | II (moderate) (1) |
Working memory | Yes | No | No | Yes | No | III (low) (1) (4) | |
Cognitive flexibility | Yes | No | No | Yes | No | III (low) (1) (4) | |
Acute exercises | Inhibitory control | Yes | No | No | No | No | II (moderate) (1) |
Working memory | Yes | No | No | No | No | II (moderate) (1) | |
Cognitive flexibility | Yes | No | No | No | No | II (moderate) (1) |
For chronic exercises, variables (age, type, study quality, composite, frequency, time, and duration) are likely to be the influencing factors for children and adolescents on their inhibitory control and working memory. Moderator analysis using separate models was employed to examine potential sources of variance. All results are presented in
In terms of the composite type of intervention, either multiple exercise interventions or sole exercise interventions were employed in the original studies. There were no statistically significant differences on the ESs between the two types of interventions (Q = 10.25,
Subgroup analysis of inhibitory control (chronic exercises). SMD: standardized mean difference.
5–12 | 11 (1567) | −0.20 | −0.28, −0.09 | 50.8% | 12.23 | 1 | 0.001 |
12–18 | 3 (148) | −0.81 | −1.15, −0.48 | 72.5% | |||
Open motor skills | 9 (859) | −0.56 | −0.73, −0.40 | 60.1% | 14.49 | 1 | 0.001 |
Closed motor skills | 6 (856) | −0.11 | −0.22, 0.00 | 0.0% | |||
Scores more than 6 (>6) | 8 (716) | −0.32 | −0.55, −0.20 | 25.2% | 2.39 | 1 | 0.124 |
Scores less than 6 (≤6) | 7(998) | −0.18 | −0.31, −0.06 | 84.7% | |||
More than 12 weeks | 7 (545) | −0.22 | −0.36, −0.09 | 69.6% | 0.07 | 1 | 0.791 |
Less than 12 weeks | 8(1170) | −0.25 | −0.37, −0.12 | 72.8% | |||
Multiple exercises intervention | 7 (713) | −0.16 | −0.27, −0.06 | 64.2% | 10.25 | 1 | 0.001 |
Sole exercises intervention | 8 (1002) | −0.55 | −0.83, −0.25 | 55.8% | |||
1–3 times/week | 5(442) | −0.14 | −0.25, −0.05 | 69.0% | 1.48 | 1 | 0.224 |
4–7 times/week | 10(1273) | −0.42 | −0.58,−0.27 | 49.8% | |||
≤30 | 5(489) | −0.12 | −0.24, 0.05 | 0.0% | 3.57 | 1 | 0.110 |
>30 | 10(1226) | −0.47 | −0.73, −0.21 | 70.0% |
In terms of intervention classification, a statistically significant difference of the evaluated ES was observed (Q = 20.53,
Subgroup analysis of working memory (chronic exercises).
5–12 | 9 (638) | −0.64 | −0.87, −0.42 | 47.0% | 18.06 | 1 | 0.001 |
12–18 | 3 (773) | −0.30 | −0.49, −0.12 | 12.8% | |||
Open motor skills | 5 (419) | −0.72 | −0.93, −0.43 | 26.9% | 20.53 | 1 | 0.001 |
Close motor skills | 5 (386) | −0.31 | −0.57,−0.25 | 12.4% | |||
Scores more than 6 (>6) | 6 (884) | −0.33 | −0.55, −0.21 | 6.9% | 27.89 | 1 | 0.001 |
Scores less than 6 (≤6) | 7 (593) | −0.86 | −1.13, −0.39 | 0.0% | |||
More than 12 weeks | 6 (336) | −0.36 | −0.79, −0.15 | 0.0% | 0.16 | 1 | 0.694 |
Less than 12 weeks | 7 (1141) | −0.62 | −0.97,−0.26 | 82.8% | |||
Composite type | |||||||
Multiple-exercise intervention | 6 (666) | −0.58 | −1.15, −0.37 | 48.0% | 1.47 | 1 | 0.257 |
Sole-exercise intervention | 7 (811) | −0.44 | −0.83, −0.25 | 23.2% | |||
1–3 times/week | 5 (478) | −0.40 | −0.65, −0.15 | 21.6% | 0.10 | 1 | 0.953 |
4–7 times/week | 8 (999) | −0.61 | −0.61,−0.32 | 79.6% | |||
≤30 | 4 (295) | −0.82 | −1.01, −0.64 | 59.5% | 18.92 | 1 | 0.001 |
>30 | 9 (1182) | −0.35 | −0.47, −0.22 | 13.1% |
In order to examine the effect of chronic exercises on inhibitory control and working memory, meta-regression analyses were performed to determine if the variables (age, type, study quality, composite, frequency, time, and duration) influenced the different indices in
Regression analysis for chronic exercises versus the control group of inhibitory control.
Age |
14 | −0.608123 | 0.241830 | −1.140387, −0.075859 | 2.51 | 1 | 0.029 |
Type |
15 | 0.451645 | 0.137132 | 0.149819, 0.753470 | 1.82 | 1 | 0.007 |
Study quality | 15 | −0.296366 | 0.229172 | −0.800770, 0.208038 | 0.65 | 1 | 0.222 |
Duration | 15 | 0.006166 | 0.243023 | −0.528724, 0.541055 | 1.31 | 1 | 0.980 |
Frequency | 15 | −0.208072 | 0.225305 | −0.703965, 0.287821 | 0.92 | 1 | 0.376 |
Time | 15 | −0.378365 | 0.222389 | −1.330843, 0.923009 | 1.70 | 1 | 0.117 |
Composite | 15 | 0.406159 | 0.197258 | −0.028003, 0.840320 | 2.06 | 1 | 0.064 |
Regression analysis for chronic exercises vs. the control group of working memory.
Age | 12 | 0.293404 | 0.263048 | −0.225295, 0.683069 | 4.57 | 1 | 0.033 |
Type |
10 | −0.375588 | 0.145919 | −0.693518, −0.057659 | 3.57 | 1 | 0.024 |
Study quality |
13 | −0.555877 | 0.105251 | −0.785000, −0.326554 | 5.28 | 1 | 0.001 |
Duration | 13 | −0.036083 | 0.333311 | −1.29879, −0.090172 | 0.11 | 1 | 0.921 |
Frequency | 13 | −0.113366 | 0.225896 | −0.616694, 0.389962 | 0.50 | 1 | 0.627 |
Time | 13 | 0.303556 | 0.177347 | −0.086782, 0.693894 | 1.71 | 1 | 0.115 |
Composite | 13 | −0.235401 | 0.175217 | −0.617166, 0.146365 | 1.34 | 1 | 0.204 |
Regarding the effects of chronic exercises on working memory, both type (β = −0.375588, Q = 3.57, df = 1,
The present meta-analysis suggests that both acute and chronic exercises may be effective for improving executive functions (e.g., inhibitory control, working memory, and cognitive flexibility) in healthy child and adolescent populations. Moreover, in chronic exercise interventions, working memory was moderated by age, exercise type, and study quality, while only two variables (age and exercise type) played a moderating role in inhibitory control.
Inhibitory control refers to the conscious inhibition or automatic response in the cognitive process (Wright et al.,
It is common that heterogeneity across studies is present in the meta-analysis, but the impact of acute exercises on the inhibitory control has a small heterogeneity, which is not in agreement with other acute exercise intervention review studies (Moreau and Chou,
Furthermore, the subgroup analysis indicated that, from the age perspective, although 5–12-year-old children and 12–18-year-old adolescents showed positive effects in terms of improving inhibitory control, 5–12-year-old children showed a low inhibitory ability compared with 12–18-year-old adolescents. This finding is in agreement with a previous study (Harnishfeger,
In addition, the experimental intervention characteristics involving duration, frequency, and exercise session time were not moderators of the effect of chronic exercises on inhibitory control (
Working memory mainly measures the preservation and update of information, and the digit span forward, digit span backward, letter digit span, Tower of London, and N-back task (1-back and 2-back) measurement tools are the most commonly used to evaluate the response time of working memory (Chen et al.,
In addition, our results show that acute exercises were characterized by non-significant heterogeneity (
Furthermore, experimental intervention characteristics involving duration, frequency, and exercise session time are crucial to investigating the effects of chronic exercise changes in working memory. The moderator analysis indicated that each session time of ≤30 min can improve the response of large ES (−0.82) on working memory; the effect is more significant than for a session time of >30 min. A prior study suggested that the effect of aerobic exercises for 55 min is not as good as for 30 min (Fu and Fan,
Cognitive flexibility refers to the ability of individuals to constantly adjust their thoughts and behaviors in order to adapt to changing situations (Hernández et al.,
Because cognitive flexibility was present across a small number of eligible studies, the present meta-analysis only synthesized WCST and more-odd shifting. With respect to cognitive flexibility, a significant improvement in cognitive flexibility in the present meta-analysis was identified in favor of acute and chronic exercises, but the ES effects of the two types of exercises regarding cognitive flexibility were small (−0.34). It is reasonable that a significant improvement was observed for cognitive flexibility. That is because exercises intervention can change the brain's activation pattern, which specifically manifests as an increased activation of the bilateral upper frontal gyrus, bilateral middle frontal gyrus, and bilateral upper lobules, and an individual is prone to having activated pre-frontal and parietal lobes when exercising cognitive flexibility (Jamadar et al.,
This study has a certain number of limitations and deficiencies. (1) There was a limited number of works on cognitive flexibility. Therefore, we could not obtain an accurate result regarding the effect of executive function interventions on children and adolescents. (2) The high-level executive functions, such as decision-making and reasoning, can be understood with a detailed assessment of the dynamics of EF performance, but the studies reviewed in this meta-analysis did not include this information. We believe that future studies should collect, retain, and ideally share these types of data to allow more detailed analyses. (3) Only three articles in this study explained random sequence generation in detail, and no other work mentioned the method of random allocation and hiding. (4) The intervention method in some of the included studies was aerobic exercises, which has not been explained in great detail, and thus we were unable to confirm which skills were involved in aerobic exercises. (5) Most studies had no long-term follow-up, and it remains unclear whether a potential benefit will emerge after a longer period post-intervention of chronic exercises in executive function. (6) The current meta-analysis only made relevant reports on the reaction time; in the future, we also need to report on the effect of physical activity on the accuracy of executive functions.
The results of the current meta-analysis demonstrate that acute and chronic exercises may have a positive effect on executive function for children and adolescents, especially in terms of working memory. To better understand the effects of acute and chronic exercises on children and adolescents, rigorous study designs are necessary. In addition, the impact of exercise training on cognitive flexibility needs to be further explored. We should explore the impact of long-term physical exercises on cognitive flexibility, which would also provide a reference for improving executive functions through exercises in the future.
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
SL, ZL, and YC: conceptualization. SL and ZL: methodology, software, and resources. PC, QY, and YZ: validation. ZK, WL, YZ, and SC: formal analysis. SL, QY, and ZL: investigation and data curation. SL, ZL, and YC: writing—original draft preparation. QY, PC, YZ, ZK, WL, SC, and YC: writing—review and editing. All authors contributed to the article and approved the submitted version.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.