Effects of aerobic exercise on executive function in children and adolescents with attention deficit hyperactivity disorder: a systematic review and meta-analysis of randomized controlled trials

Abstract

Background

The benefits of aerobic exercise on the executive function of children and adolescents have been confirmed to a certain extent. However, the effects of such interventions on the executive function of children and adolescents with attention deficit hyperactivity disorder (ADHD) require further exploration. This study is a systematic review of randomized controlled trials (RCTs) of aerobic exercise interventions for children and adolescents with ADHD, aiming to provide valuable intervention suggestions to enhance the executive function of children and adolescents with ADHD.

Method

A comprehensive search was conducted across multiple databases, including PubMed, Web of Science, EMbase, Cochrane Library, ProQuest, Scopus, CNKI, Wanfang, and VIP databases, to identify relevant RCTs. We established detailed inclusion and exclusion criteria, followed by literature screening, data extraction, quality assessment, and data analysis conducted by two independent researchers. Literature screening was performed using EndNote X9. Risk of bias assessment (RoB1) and meta-analysis were conducted using Review Manager 5.4. Network meta-analysis was performed using Stata 17.0.

Results

A total of 16 studies were included in the analysis, involving 668 participants diagnosed with ADHD, including 343 in the experimental group and 325 in the control group. The age range of the participants was between 6 and 18 years. The results of the meta-analysis show that aerobic exercise interventions have a moderate effect size positive impact on inhibitory control (SMD = − 0.69, 95% CI: −1.04, − 0.34, p < 0.05), working memory (SMD = − 0.52, 95% CI: −0.82, − 0.21, p < 0.05), and cognitive flexibility (SMD = − 0.64, 95% CI: −0.83, − 0.45, p < 0.05) in children and adolescents with ADHD. Subgroup analyses indicate that the effectiveness of aerobic exercise interventions is moderated by factors such as intervention period, frequency, session duration, intensity and the choice between acute or chronic exercise. The results of the network meta-analysis showed that ball-based aerobic exercise [SMD = − 1.00, 95% CI = (− 1.98, − 0.03)] performed the best in improving inhibitory control among children and adolescents with ADHD, with a SUCRA value of 65.1%. It is important to note that most of the measurement tools for executive function employed reaction time as the evaluation criterion, which means that a more negative effect size corresponds to greater improvement.

Conclusion

Aerobic exercise interventions can significantly improve the executive functions of children and adolescents with ADHD. To achieve the best intervention effects, it is recommended to engage in chronic aerobic exercise, with a period lasting no less than 12 weeks, scheduled 3 to 5 times per week, each session lasting 60 min or more, and maintaining intensity at moderate or moderate-to-vigorous levels. Additionally, different types of aerobic exercise have varying effects on the inhibitory control abilities of this population, which requires attention.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13102-025-01304-1.

Introduction

Attention deficit hyperactivity disorder (ADHD), commonly referred to as hyperactivity disorder, is a neurodevelopmental disorder frequently diagnosed in children and adolescents [1]. Studies have shown that the global prevalence of ADHD among children and adolescents is 8.0% [2]. Individuals with ADHD typically exhibit age-inappropriate inattention, hyperactivity, and impulsive behavior [3], with comorbidities such as learning disabilities, conduct disorders, and emotional disorders [4]. If no effective interventions are implemented during childhood, ADHD symptoms may persist into adolescence and adulthood [5]. The adverse consequences of such conditions include antisocial behavior, substance abuse, criminal activity, psychiatric hospitalization, and accidental death [6]. A meta-analysis of prison inmates with ADHD shows that the prevalence of ADHD in juvenile inmates is 30.1% and in adult inmates 26.2% [7]. Furthermore, individuals with ADHD have higher risks of suicide, unemployment, and divorce compared to the general population [8, 9], leading to significant economic, psychological, and social effects on families and society.

Executive function impairment is a core symptom of ADHD, and the degree of impairment in executive function is positively correlated with the severity of ADHD [10, 11]. Executive function refers to an advanced cognitive process in which individuals can flexibly and efficiently regulate multiple cognitive processes when given complex cognitive tasks, with the ultimate goal of achieving purposeful and orderly behaviors [12]. Executive function includes three core sub-functions: inhibitory control, working memory, and cognitive flexibility [13]. Inhibitory control, also referred to as inhibition, primarily denotes the ability to suppress impulsive or automatic (dominant) responses during cognitive activities [14]. Working memory, also known as refreshing, is the capacity to encode and maintain information over short periods [15]. Cognitive flexibility, also known as shifting, refers to the ability to switch between different operations or mental sets according to the requirements of the current task [13]. Studies have shown that among children with ADHD, the likelihood of impairment in at least one of the three core dimensions of executive function is as high as 89% [16], and they usually exhibit varying degrees of deficits across all three core dimensions [17].

Inhibitory control impairments in children and adolescents with ADHD are closely related to poor emotional regulation, deficits in self-guided speech, and insufficient attention allocation [18–20]. Working memory deficits reflect a lack of sustained visual attention, making it generally difficult to maintain during the learning process [21]. Additionally, they have difficulty processing information and may also experience social impairments [22]. Abnormalities in cognitive flexibility can lead to poor academic performance and diminished problem-solving abilities [23]. Executive function deficits are just one aspect of the cognitive and behavioral problems in children and adolescents with ADHD, but their negative impact is significant. Improving executive function deficits can be beneficial in the treatment of ADHD [24].

In recent years, research has reported that exercise intervention, as a new therapeutic approach, can improve the executive function of individuals with ADHD [25, 26]. Traditional pharmacological treatments are not only expensive but may also lead to adverse reactions such as headaches, nausea, anorexia, and even stunted growth and development [27]. In contrast, exercise intervention can be well integrated into the daily lives of children and adolescents without adverse reactions. Among various forms of exercise, aerobic exercise has the advantages of strong operability and high compliance. It is currently widely applied to improve the executive function of normal children and adolescents [28–30]. Recently, numerous studies have found that aerobic exercise can also improve the executive function of children and adolescents with ADHD [31–33]. However, Diamond and Ling argue that aerobic exercise and resistance exercise are the least effective forms of exercise interventions [34]. Additionally, some studies conclude that aerobic exercise cannot achieve a significant improvement in executive function [35].

Previous systematic reviews and meta-analyses on the effects of aerobic exercise interventions on executive function have mostly focused on individuals without ADHD [36, 37]. There are few review studies on aerobic exercise interventions involving individuals with ADHD, especially children or adolescents. Only Yang et al.’s study explored the effect of aerobic exercise intervention on the executive function of children with ADHD [38]. However, their study did not identify which type of aerobic exercise yields the greatest benefits for the executive function of children with ADHD. Moreover, it did not investigate the effect of aerobic exercise frequency on their executive function. Most importantly, the study included both acute and chronic exercise interventions but did not effectively differentiate between them in the subgroup analysis. Therefore, there is currently a lack of a comprehensive systematic review on the effect of aerobic exercise interventions on the executive function of children and adolescents with ADHD, making further exploration necessary. By conducting the latest and most comprehensive systematic review, we aim to evaluate the impact of aerobic exercise interventions on the executive functions of children and adolescents with ADHD. This study seeks to provide valuable insights and guidance for developing targeted intervention measures for this population.

Methods

Protocol and registration

This systematic review and meta-analysis were conducted under the Cochrane Handbook for Systematic Reviews of Interventions. The results are reported in accordance with the PRISMA statement [39]. The research protocol was registered on September 29, 2024, in the international prospective register of systematic reviews, PROSPERO, with the ID CRD42024595660 (https://www.crd.york.ac.uk/PROSPERO/view/CRD42024595660).

Retrieval strategy

We conducted searches in the following databases: PubMed, Web of Science, EMbase, Cochrane Library, ProQuest, Scopus, CNKI, Wanfang, and VIP. The initial search started on September 10, 2024, and ended on September 11, 2024. The second search started on June 15, 2025, and ended on June 16, 2025. We imposed language restrictions, including only relevant literature published in Chinese and English. Furthermore, to ensure the quality of the research, this study included only literature published in peer-reviewed journals and did not encompass grey literature (such as dissertations, conference abstracts, etc.). The search strategies and keywords for the databases are as follows: Intervention Methods (“Exercise” OR “Physical Exercise” OR “Physical Activity” OR “Aerobic Exercise” OR “Isometric Exercise” OR “Acute Exercise” OR “Exercise Training” OR “Running” OR “Jogging” OR “Marathon Running” OR “Swimming” OR “Walking” OR “Nordic Walking” OR “Stair Climbing” OR “Exergaming” OR “Aerobic Training” OR “Dancing” OR “Cycling” OR “Endurance Training” OR “Qigong” OR “Taichi” OR “Baduanjin” OR “Wuqinxi” OR “Yijinjing” OR “Yoga” OR “Ball” OR “Soccer Ball” OR “Soccer” OR “Football” OR “Basketball” OR “Ping Pong” OR “Badminton” OR “Tennis” OR “Baseball” OR “Volleyball” OR “Softball” OR “Racket Sport” OR “Lacrosse” OR “Racquetball”) AND Population (“Children” OR “Adolescent” OR “Child” OR “Teenager” OR “Adolescence”) AND Disease Type (“Attention Deficit Hyperactivity Disorder” OR “Attention Deficit Disorder with Hyperactivity” OR “ADHD” OR “ADDH” OR “Attention Deficit Disorder” OR “Hyperkinetic Syndrome”) AND Outcome Measures (“Executive Function” OR “Executive Control” OR “Cognitive Function” OR “Cognitive Performance” OR “Inhibitory Control” OR “Shifting” OR “Working Memory” OR “Refresh Function” OR “Cognitive Flexibility” OR “Updating” OR “Inhibition”). The database retrieval was completed independently by two researchers, and the retrieval results were cross-checked. Any discrepancies were resolved through group discussions among team members until a consensus was reached(PW, DL, and FR). Appendix B presents the search strategies for all databases.

Inclusion and exclusion criteria

Inclusion criteria

The inclusion criteria for this study were as follows: (1) Research subjects: Participants aged between 6 and 18 years who have been clinically diagnosed with ADHD or assessed as having ADHD using scales such as the International Classification of Diseases (ICD) and the Diagnostic and Statistical Manual of Mental Disorders (DSM). (2) Intervention type: Any form of aerobic exercise intervention. (3) Control type: No-exercise intervention or daily activities. (4) Trial design: All included studies are RCTs. (5) Outcome measures: At least one outcome measure with data that could be used to calculate the effect size on for executive function (inhibitory control, working memory, cognitive flexibility).

Exclusion criteria

The exclusion criteria of this study include: (1) Non-Chinese and non-English study; (2) Non-randomized controlled trials; (3) No original data provided; (4) The provided original data cannot be used to calculate the effect size of executive function; (5) There were statistically significant differences in the baseline analysis indicators between the experimental and control groups(p < 0.05); (6) Full text not available.

Study selection

This study utilized EndNote X9 for literature management. Prior to the initial screening, duplicates were checked by using the software’s automatic detection feature to identify potential duplicate documents. For suspected duplicates that the software could not identify, manual verification was conducted by comparing the titles, authors, publication years, and study contents of the documents to ensure that no duplicate studies were overlooked. Upon completion of the duplicate check, these duplicates were excluded. The subsequent literature screening was carried out by two independent researchers (PW and DL). The Initial screening was conducted based on titles and abstracts. A study that preliminarily met the criteria was further evaluated through detailed full-text reading. Finally, the study that fully meets the criteria will be included in this study. Upon completion of all literature screening, a PRISMA flowchart will be generated. If there are differences in the process of selecting and excluding studies by the two independent researchers, a consensus will be reached through group discussions among team members(PW, DL, and FR). We calculated the reliability (Cohen’s kappa) for both the initial screening and full-text screening stages to assess the agreement between the two independent researchers [40].

Data extraction

Data extraction was conducted by two independent researchers (PW and DL) using a customized Excel extraction worksheet (Version 16.93, Microsoft, Redmond, WA, USA), and they were both blinded to the identities of the study authors. The information they extracted includes: (1) basic information: author, year of study publication, and country where data was collected; (2) characteristics of trial participants: age, gender ratio, ADHD diagnostic criteria, and sample size; (3) Intervention Components: intervention content, intervention duration period, frequency, session time, intensity, intervention type and whether the exercise is acute or chronic; (4) outcome measures and evaluation tools. Specifically regarding the components of the intervention, the intervention periods for inhibitory control were coded as ≥ 12 weeks and 6–8 weeks, and the intervention periods for working memory and cognitive flexibility were coded as 12 weeks and 6–8 weeks. The intervention frequencies for inhibitory control were coded as 1–2 times per week and 3–5 times per week; for working memory and cognitive flexibility, they were coded as 2 times per week and 3–5 times per week. The intervention session duration were coded as ≥ 60 min and < 60 min. Based on the recommended exercise intensity terminology, exercise intensity is categorized as moderate intensity, moderate-to-vigorous intensity, and vigorous intensity. The criteria for moderate intensity are: 3.00–6.00 METs; 40–60% HRR/VO₂_max; 55–70% HR_max; continuous aerobic exercise that allows for uninterrupted conversation, lasting between 30 and 60 min [41]. The criteria for vigorous intensity are: 6.01-9.00 METs; 61–85% HRR/VO₂_max; 71–90% HR_max; aerobic exercise that does not allow for uninterrupted conversation, lasting 30 min [41]. The criteria for moderate-to-vigorous intensity are defined as exercise intensity that spans both moderate and vigorous intensity levels.

Quality assessment

Two independent researchers, based on the Cochrane Risk of Bias assessment tool (RoB1), evaluated the risk of bias in the included studies using Reviewer Manager 5.4 software. Seven items were included: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias. The risk of bias in the study is categorized as “low,” “high,” or “unclear.” If the evaluation results of the two independent researchers (PW and DL) could not reach an agreement, a third independent researcher (FR) was invited to perform the assessment to ensure the fairness and objectivity of the final decision.

Certainty assessment

The credibility of the results is assessed using the GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) framework [42]. The quality of evidence is evaluated across several domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias. Evidence quality is classified into four levels: high, moderate, low, and very low.

Statistical analysis

Meta-analysis was conducted using Reviewer Manager 5.4 software. The data in the included studies are all continuous variables. However, due to the differences in the assessment tools used across studies, we use the standardized mean difference (SMD) as the effect size and calculate its 95% confidence interval (95% CI). Effect size < 0.2 is a small effect, 0.20–0.49 is a small to moderate effect, 0.50–0.79 is a moderate effect, and effect size ≥ 0.8 is a large effect [43]. If the directions of the effect sizes are inconsistent, we multiplied the results by − 1 to ensure consistency [44]. Heterogeneity among studies was assessed using the I² test and the Q test. If p < 0.05 or I²>50%, it indicated that there is heterogeneity among the studies, and a random-effects model was used to combine the effect sizes; otherwise, a fixed-effects model is used for effect size combination. In addition, The level of heterogeneity was quantified using I² statistics and classified as low (I²≤25%), moderate (25%< I²≤50%), significant (50%< I²≤75%), and very high (I²>75%) [45]. When heterogeneity was present, subgroup analyses were conducted to explore its sources. A network meta-analysis was conducted using a frequentist framework with Stata 17.0 software. A network relationship graph was plotted, followed by statistical analysis. We used the node-splitting method to assess global inconsistency, evaluating the direct and indirect comparisons among different aerobic exercise interventions. When the direct and indirect comparison results met the consistency criteria (p > 0.05), a consistency model was employed for the analysis. The Surface Under the Cumulative Ranking Curve (SUCRA) was used to predict and rank the intervention effects of different types of aerobic exercise. The SUCRA values were expressed as percentages, with higher percentages indicating better effectiveness. Publication bias was assessed using both visual funnel plots and quantitative methods, including Egger’s test and Begg’s test. A P-value of 0.05 was used as the threshold: p > 0.05 indicated no publication bias; p < 0.05 indicated the presence of publication bias. In addition, sensitivity analysis was performed by sequentially excluding individual studies to examine the degree of influence of each study result on the combined effect size.

Results

Study selection

To ensure the accuracy of the study retrieval and screening process, two researchers with expertise in attention deficit hyperactivity disorder and exercise science independently screened the titles, abstracts, and full-text literature after the completion of study retrieval and duplicate removal. The reliability of the two screening stages was calculated using Cohen’s kappa for both the title and abstract screening stage and the full-text screening stage. The agreement levels are categorized as follows: fair agreement (0.40–0.59), good agreement (0.60–0.74), and excellent agreement (> 0.75) [40]. We conducted a comprehensive search of nine Chinese and English databases from their inception to September 10, 2024, identifying a total of 1,875 papers. After removing duplicates, 1,538 relevant studies remained. Subsequently, two independent researchers conducted initial screening through titles and abstracts. Among them, 1474 studies were excluded, and 64 studies met the criteria for full-text review. At this stage, the reliability between the two reviewers was rated as “good” (Cohen’s kappa = 0.71). After a full-text review, 52 studies were excluded: 7 studies were non-randomized controlled trials, 7 studies used interventions that did not meet the inclusion criteria, 4 studies lacked access to the full text, 10 studies did not meet the outcome measure criteria, 8 studies had participants who did not meet the inclusion criteria, 16 studies had incomplete data. Therefore, the initial search retained a total of 12 studies that met the inclusion criteria. At this stage, the inter-rater reliability between the two reviewers was rated as “excellent” (Cohen’s kappa = 0.75). To ensure the timeliness of the research, a secondary search was conducted, comprehensively reviewing literature from September 10, 2024, to June 15, 2025, across the nine electronic databases, which identified 314 studies. After the initial screening, 9 studies met the criteria for full-text review. At this stage, the inter-rater reliability between the two reviewers was rated as “good” (Cohen’s kappa = 0.73). Following the full-text review, 5 studies were excluded, and 4 studies were retained. At this stage, the inter-rater reliability between the two reviewers was rated as “excellent” (Cohen’s kappa = 0.77). After these two retrievals and careful screening, a total of 16 studies [33, 46–60] were included in the quantitative synthesis (see Fig. 1).

Fig. 1.

PRISMA flow diagram of the study process

Study characteristics

Table 1 lists the characteristics of the 16 included studies. All studies were randomized controlled trials. The publication years of the study range from 2012 to 2025. The research sample included 668 participants diagnosed with ADHD. Among them, 343 participants were assigned to the experimental group, and another 325 participants were assigned to the control group. 6 studies simultaneously focused on inhibitory control, working memory, and cognitive flexibility [50, 51, 53, 54, 56, 60]; 3 studies simultaneously focused on inhibitory control and cognitive flexibility [46, 52, 57]; 1 study simultaneously focused on inhibitory control and working memory [59]; 5 studies only focused on inhibitory control [33, 47–49, 58]; 1 study only focused on working memory [55]. Among the 16 studies, 3 employed acute aerobic exercise interventions [46, 50, 57], and the others utilized chronic aerobic exercises. The duration of individual sessions in chronic aerobic exercise interventions varied from 25 to 120 min; intervention frequencies ranged from once to five times per week, with the most common frequencies being two or three times per week; the intervention periods spanned from 6 weeks to 78 weeks. The specific aerobic exercise interventions included ball aerobic exercise [47, 48, 52, 56], combined aerobic and neurocognitive exercise [54], cycling [53], taekwondo [33], exergaming [51], acute aerobic exercise [46, 50, 57], combination aerobic exercise [49], game-structured aerobic exercise [58, 60], skating [59], and HIIT [60]. Detailed characteristics of the included studies can be found in Appendix C.

Table 1.

Summary table of included reviews

Study	Country	Participant Characteristics				Intervention								Outcome
		Gender-M (%)	ADHD Diagnostic Methods	Sample Size (EG/CG)	Age (EG/CG)	EG						CG		Outcome Measures
						Intervention content		Intervention time, frequency, period	Intensity	Type	Acute or chronic	Intervention content	Type
Liang et al., 2022	China	M-78	DSM-5	39/39	8.37 ± 1.42/ 8.29 ± 1.27	Combined aerobic and neurocognitive-exercise		60 min/ 3 sessions/ 12 weeks	MVI (60–80% HRmax)	Combined aerobic and neurocognitive-exercise	Chronic	Control	Control	IC: Flanker Task WM: Tower of London Test CF: Trail Making Test
Ludyga et al., 2022	Switzerland	M-70	DSM-5	29/28	10.0 ± 1.2/ 10.8 ± 1.2	Judo		60 min/ 2 sessions/ 12 weeks	MI	Judo	Chronic	Control	Control	WM: Change Detection Paradigm Test
Chen et al., 2022	China	M-83	DSM-5	32/32	8.37 ± 1.68/ 7.89 ± 2.13	Cycling		25 min/ 3 sessions/ 12 weeks	MVI (60–80% HRmax)	Cycling	Chronic	Watching cartoons	Control	IC: Stroop Test WM: N-back Test CF: odd Even Size Test
Chang et al., 2022	China	M-81	DSM-5	16/16	8.31 ± 1.30/ 8.38 ± 1.31	Actual table tennis training		60 min/ 3 sessions/ 12 weeks	MI	Ball aerobic exercise	Chronic	Control	Control	IC: Stroop Test CF: Wisconsin Card Sorting Test
Kadri et al., 2019	Kadri	M-90	Psychiatric physician	20/20	14.5 ± 3.5/ 14.2 ± 3.0	Taekwondo		50 min/ 2 sessions/ 78 weeks	MI	Taekwondo	Chronic	Control	Control	IC: Stroop Color-Word Test
Benzing et al., 2019	Switzerland	M-84	ICD-10	28/23	10.46 ± 1.30/ 10.39 ± 1.44	Exergaming		30 min/ 3 sessions/ 8 weeks	MI	Exergaming	Chronic	Control	Control	IC: Simon Task CF: Flanker Task WM: Color Span Backwards Task
Benzing et al., 2018	Switzerland	M-83	ICD-10	24/22	10.46 ± 1.35/ 10.50 ± 1.41	Exergaming		15 min/ 1 session	MVI	Acute aerobic exercise	Acute	Watching video	Control	IC: Flanker Task CF: Flanker Task WM: Color Span Backwards Task
Pan et al., 2016	China	M-100	DSM-4	16/16	8.93 ± 1.49/ 8.87 ± 1.56	Table tennis		70 min/ 2 sessions/ 12 weeks	MI	Ball aerobic exercise	Chronic	Control	Control	IC: Stroop Color-Word Test
Chang et al., 2012	China	M-93	DSM-4	20/20	10.45 ± 0.95/ 10.42 ± 0.87	Treadmill running		30 min/ 1 session	MI (50–70%HRR)	Acute aerobic exercise	Acute	Watching videos	Control	IC: Stroop Test CF: Wisconsin Card Sorting Test
Memarmoghaddam et al., 2016	Iran	M-100	SNAP-4	19/17	8.31 ± 1.29/ 8.29 ± 1.31	Ball exercise (table tennis, bowling, football, basketball, etc.)		90 min/ 3 sessions/ 8 weeks	VI (65–80% HRR)	Ball aerobic exercise	Chronic	Control	Control	IC: Go-No-Go Test
Lee et al., 2017	Korea	M-100	DSM-4	6/6	8.83 ± 0.98/ 8.83 ± 0.98	Rope skipping and ball exercise		60 min/ 3 sessions/ 12 weeks	MVI (45–75%HRR)	Combination aerobic exercise	Chronic	Control	Control	IC: Stroop Color-Word Test
Song et al., 2022	China	M-100	DSM-4	8/8	7.68 ± 0.56/ 7.53 ± 0.79	Football		60 min/ 5 sessions/ 6 weeks	MVI	Ball aerobic exercise	Chronic	Physical education class	Control	IC: Stroop Color-Word Test WM: Complex Figure Test CF: Trail Making Test
Barudin-Carreiro et al., 2024	USA	M-66	Psychiatric physician, psychologists, etc.	7/8	9.69 ± 1.75/ 9.65 ± 1.19	Walking		20 min/1 session	Light	Acute aerobic exercise	Acute	Sitting	Control	IC: Stroop Color and Word Test CF: Wisconsin Card Sorting Task
Liu et al., 2025	China	M-90	Clinical diagnosis	40/40	15.82 ± 1.11/ 13.65 ± 1.21	Mini games and running activities		120 min/1 sessions/ 12 weeks	LMI, MVI	Game structured aerobic exercise	Chronic	Control	Control	IC: Posner test
Huang et al., 2025	China	NR	DSM-4	12/12	10.21 ± 1.27/ 10.21 ± 1.24	Inline skating exercise		80 min/2 sessions/ 12 weeks	MI	Skating	Chronic	Control	Control	IC: Stroop Color and Word Test WM: Spatial Working Memory
Sun et al., 2024	China	NR	Pediatrician, Clinical Psychologist, Psychiatric physician	27/18	8–13	EG1: GameHIIT; EG2: GameSAE (structured aerobic exercise)		EG1:30 min/2 sessions/ 8 weeks; EG2:60 min/2 sessions/8 weeks	EG1: Vigorous intensity; EG2: Moderate Intensity	EG1: HIIT EG2: Game structured aerobic exercise	Chronic	Control	Control	IC: Color Word Stroop Test WM: Corsi Block-tapping Test CF: Colour-Wisconsin Card Sorting Test

EG experimental group, CG control group, NR no report, M male, DSM-5 Diagnostic and statistical manual of mental disorders (fifth editio), ICD-10 International classification of diseases (10th revision), DSM-4 Diagnostic and statistical manual of mental disorders (fourth edition), SNAP-4 Swanson Nolan and Pelham rating scale (fourth edition), MI moderate intensity, MVI moderate to vigorous intensity, LMI light to moderate intensity, VI vigorous intensity, IC inhibitory control, WM working memory, CF cognitive flexibility

Quality assessment

16 studies had no selective reporting or other bias. 8 studies [49, 52–55] had a low risk of bias in the random sequence generation section, and the others were all unclear. 5 studies [51, 55, 57, 58, 60] implemented allocation concealment and had a low risk of bias, and the others were all unclear. Since all the studies included in this meta-analysis are exercise intervention studies, all studies had a high risk of bias in the blinding of participants and personnel section. 5 studies [50, 55, 58–60] had a low risk of bias in the blinding of outcome assessment section, and the others were all unclear. In the incomplete outcome Data section, 2 studies [49, 51] were rated as having a high risk of bias, 1 study [46] was unclear, and the other studies all had a low risk of bias. Detailed information on the risk of bias assessment is shown in Fig. 2 and Appendix D.

Fig. 2.

Methodological quality of included studies

Meta-analysis

Effect of aerobic exercise on inhibitory control

A total of 15 studies investigated the effect of aerobic exercise interventions on inhibitory control in children and adolescents with ADHD, involving 611 participants diagnosed with ADHD. As shown in Fig. 3, compared with the control group, aerobic exercise intervention has a moderate effect size (SMD = − 0.69, 95% CI: −1.04, − 0.34, p < 0.05). This effect shows significant heterogeneity (I² = 77%, p < 0.05). Subgroup analysis was performed, and the results are shown in Table 2.

Fig. 3.

Meta-analysis results for inhibitory control

Table 2.

The subgroup analysis of the effect of aerobic exercise interventions on inhibitory control

Subgroup	Cut-off	Inclusion of literature	Effect Model	Heterogeneity test		Effet Size		The Result
				I2	P	SMD	95%CI	Z	P
Period	≥ 12 weeks	8	Random	80%	<0.001	-0.98	-1.49, -0.46	3.69	<0.001
	6–8 weeks	5		20%	0.28	-0.36	-0.71, -0.00	1.96	0.05
Frequency	1–2 weekly	6	Random	84%	<0.001	-0.91	-1.62, -0.20	2.52	0.01
	3–5 weekly	7		59%	0.02	-0.65	-1.05, -0.25	3.17	0.002
Session time	≥ 60 min	9	Random	66%	0.003	-0.67	-1.08, -0.26	3.22	0.001
	< 60 min	4		84%	<0.001	-0.91	-1.71, -0.11	2.24	0.03
Intensity	MI	6	Random	81%	<0.001	-1.17	-1.83, -0.50	3.35	<0.001
	MVI	4		63%	0.04	-0.94	-1.55, -0.34	3.07	0.002
	VI	2		0%	0.39	-0.29	-0.77, 0.19	1.19	0.23
Acute or chronic	Acute	3	Random	90%	<0.001	-0.28	-1.56, 0.99	0.43	0.66
	Chronic	13		74%	<0.001	-0.71	-1.09, -0.33	3.67	<0.001

MI moderate intensity, MVI moderate to vigorous intensity, VI vigorous intensity

Effect of aerobic exercise on working memory

A total of 8 studies examined the effect of aerobic exercise intervention on the working memory of children and adolescents with ADHD, involving 381 participants diagnosed with ADHD. As shown in Fig. 4, compared with the control group, aerobic exercise intervention has a moderate effect size (SMD = − 0.52, 95%CI: −0.82, − 0.21, p < 0.05). This effect shows significant heterogeneity (I² = 53%, p < 0.05). A subgroup analysis was conducted, and the results are shown in Table 3.

Fig. 4.

Meta-analysis results for working memory

Table 3.

The subgroup analysis of the effect of aerobic exercise interventions on working memory

Subgroup	Cut-off	Inclusion of literature	Effect Model	Heterogeneity test		Effet Size		The Result
				I2	P	SMD	95%CI	Z	P
Period	12 weeks	4	Random	57%	0.07	-0.77	-1.20, -0.34	3.49	<0.001
	6–8 weeks	4		13%	0.33	-0.32	-0.70, 0.06	1.64	0.10
Frequency	2 weekly	4	Random	54%	0.09	-0.44	-0.96, 0.07	1.69	0.09
	3–5 weekly	4		41%	0.17	-0.72	-0.10, -0.33	3.65	<0.001
Session time	≥ 60 min	5	Random	62%	0.03	-0.67	-1.17, -0.17	2.62	0.009
	< 60 min	3		33%	0.22	-0.50	-0.91, -0.99	2.40	0.02
Intensity	MI	4	Random	56%	0.08	-0.37	-0.86, 0.12	1.48	0.14
	MVI	3		0%	0.85	-0.90	-1.23, -0.57	5.37	<0.001
	VI	1		/	/	-0.43	-1.13, 0.28	1.18	0.24
Acute or chronic	Acute	1	Random	/	/	0.00	-0.58, 0.58	0.00	1.00
	Chronic	8		50%	0.05	-0.59	-0.90, -0.27	3.62	<0.001

MI moderate intensity, MVI moderate to vigorous intensity, VI vigorous intensity

Effect of aerobic exercise on cognitive flexibility

A total of 10 studies examined the effect of aerobic exercise interventions on cognitive flexibility in children and adolescents with ADHD, involving 425 participants diagnosed with ADHD. As shown in Fig. 5, compared with the control group, aerobic exercise interventions have a moderate effect size (SMD = − 0.64, 95%CI: −0.83, − 0.45, p < 0.05). This effect shows no heterogeneity (I² = 0%, p > 0.05). A subgroup analysis was conducted, and the results are shown in Table 4.

Fig. 5.

Meta-analysis results for cognitive flexibility

Table 4.

The subgroup analysis of the effect of aerobic exercise interventions on cognitive flexibility

Subgroup	Cut-off	Inclusion of literature	Effect Model	Heterogeneity test		Effet Size		The Result
				I2	P	SMD	95%CI	Z	P
Period	12 weeks	3	Random	0%	0.56	-0.74	-1.05, -0.43	4.69	<0.001
	6–8 weeks	5		0%	0.66	-0.64	-0.95, -0.32	3.99	<0.001
Frequency	2 weekly	2	Random	0%	0.62	-0.64	-1.15, -0.12	2.43	0.01
	3–5 weekly	6		0%	0.63	-0.70	-0.95, -0.46	5.64	<0.001
Session time	≥ 60 min	5	Random	0%	0.67	-0.76	-1.06, -0.47	5.10	<0.001
	< 60 min	3		7%	0.34	-0.63	-1.04, -0.22	3.01	0.003
Intensity	MI	3	Random	17%	0.30	-0.61	-0.99, -0.22	3.12	0.002
	MVI	3		0%	0.59	-0.71	-1.04, -0.39	4.33	<0.001
	VI	2		0%	0.99	-0.77	-1.26, -0.28	3.09	0.002
Acute or chronic	Acute	3	Random	0%	0.62	-0.49	-0.88, -0.09	2.39	0.02
	Chronic	8		0%	0.81	-0.69	-0.91, -0.47	6.14	<0.001

MI moderate intensity, MVI moderate to vigorous intensity, VI vigorous intensity

Network meta-analysis

Figure 6 illustrates the network meta-analysis diagram for inhibitory control. Due to the limited number of studies addressing working memory and cognitive flexibility, the network meta-analysis focused exclusively on inhibitory control. The interventions with larger sample sizes included the control group, ball aerobic exercise, game-structured aerobic exercise and acute aerobic exercise. The most frequently compared interventions were ball aerobic exercise versus control, acute aerobic exercise versus control, and game-structured aerobic exercise versus control. The frequency of these comparisons is indicated by the thickness of the connecting lines. Other interventions, such as cycling, skating, and exergaming, have smaller sample sizes and have only been compared once, as indicated by the smaller nodes and thinner connecting lines in the figure.

Fig. 6.

Network diagram of inhibitory control. CANE: Combined aerobic and neurocognitive exercise

This study examines the changes in inhibitory control among children and adolescents with ADHD following various aerobic exercise interventions. Due to the use of different measurement tools, standardized mean differences (SMD) and 95% confidence intervals (CI) were employed to quantify the impact on inhibitory control. We conducted a global inconsistency test, and the results indicated that it did not reach statistical significance (P = 0.88 > 0.05). Therefore, a consistency model was employed for analysis. The network meta-analysis results indicate that, compared to the control group, only ball aerobic exercise [SMD = − 1.00, 95% CI = (− 1.98, − 0.03)] was statistically significant (see Appendix E). Based on the SUCRA rankings for aerobic exercise interventions, taekwondo ranked highest (SUCRA = 82.2%), followed by cycling (SUCRA = 74.4%), and then skating and ball aerobic exercise (SUCRA = 71.7% and 65.1%, respectively), as detailed in Table 5 and Appendix F. Although taekwondo, cycling and skating achieved relatively high SUCRA values, they did not demonstrate statistical significance compared to the control group. Therefore, considering both the SUCRA rankings and statistical significance, ball aerobic exercise are identified as the most empirically supported interventions, potentially offering more reliable improvements in inhibitory control for children and adolescents with ADHD.

Table 5.

Ranking of SUCRA probabilities

Intervention	Sucra	Rank
Taekwondo	82.2	1
Cycling	74.4	2
Skating	71.7	3
Ball aerobic exercise	65.1	4
Combined aerobic and neurocognitive exercise	48.3	5
Exergaming	40.6	6
Acute aerobic exercise	38.7	7
Combination aerobic exercise	37.1	8
HIIT	34.2	9
Game structured aerobic exercise	33.7	10
Control	24.0	11

Publication bias

We conducted publication bias assessments for the meta-analytic results of inhibitory control, working memory, and cognitive flexibility. Visual inspection of the funnel plots indicated some asymmetry in the distribution of studies on inhibitory control, whereas no asymmetry was observed for studies on working memory and cognitive flexibility (see Appendix G1, G2 and G3). Additionally, the results of Egger’s test and Begg’s test revealed no statistically significant publication bias for inhibitory control (P = 0.802 and P = 1.000), working memory (P = 0.542 and P = 0.917), or cognitive flexibility (P = 0.341 and P = 0.161) (see Appendix H1, H2 and H3). Furthermore, a publication bias assessment for the network meta-analysis results of inhibitory control was performed; visual inspection did not show any apparent asymmetry (see Appendix G4). Egger’s test (P = 0.512) also did not indicate significant bias (see Appendix H4). Sensitivity analyses, which involved sequentially excluding individual studies, demonstrated no significant differences between the pooled effect sizes and the original pooled effect sizes. This indicates that the results of the present study are highly stable (see Appendix I1, I2 and I3).

Certainty assessment results

According to the GRADE approach, the detailed GRADE evaluation results are presented in Table 6.

Table 6.

Level of evidence for outcome indicators

Outcome	Number of studies included	Study design	Evaluation of evidence quality level					Quality of evidence
			Risk of bias	Inconsistency	Indirectness	Imprecision	Publication bias
Inhibitory control	15	RCTs	Serious	Serious	No	No	None	Low
Working memory	8	RCTs	Serious	No	No	No	None	Moderate
Cognitive flexibility	10	RCTs	Serious	No	No	No	None	Moderate

RCTs randomized control trials, No no downgrade

Discussion

Main results

In this study, we conducted both traditional meta-analysis and network meta-analysis, including a total of 16 studies, to evaluate the effects of aerobic exercise interventions on the executive functions of children and adolescents with ADHD. The results of the traditional meta-analysis showed that aerobic exercise interventions have a moderately positive effect on inhibitory control, working memory, and cognitive flexibility in children and adolescents with ADHD. In the subgroup analysis targeting inhibitory control, chronic exercise interventions lasting 12 weeks or more, with a frequency of 3 to 5 times per week, individual sessions lasting 60 min or longer, and involving aerobic exercise of moderate intensity had the largest overall effect size. In the subgroup analyses targeting working memory and cognitive flexibility, chronic exercise interventions with a duration of 12 weeks, a frequency of 3 to 5 times per week, individual sessions lasting 60 min or more, and involving aerobic exercise of moderate to vigorous intensity had the largest overall effect size. The results of the network meta-analysis showed that different types of aerobic exercise have varying effects on improving inhibitory control in this population, indicating that not all aerobic exercise interventions have the same effect. Specifically, taekwondo and ball aerobic exercises demonstrated significant benefits in improving inhibitory control in this population.

Comparison with existing study

We further explored the executive functions of children and adolescents with ADHD, expanding upon previous research on aerobic exercise interventions in the ADHD population. Our study is consistent with earlier findings, and the results support the positive effects of aerobic exercise interventions [25, 26]. That is, overall, aerobic exercise interventions can have a positive impact on the executive functions of children and adolescents with ADHD. Notably, the research by Diamond and Ling indicated that aerobic exercise is one of the less effective interventions for improving executive functions [34]. However, Hillman and colleagues opposed this view. They argued that Diamond and Ling’s study had certain methodological errors and that they might have missed several high-quality studies, which led to their conclusion [61]. Additionally, they cited a meta-analysis conducted in 2017, which showed that aerobic exercise can significantly improve cognitive functions (including executive functions) in adults over 50 years old [62]. In 2018, Diamond and Ling responded to Hillman’s objections, stating that Hillman and his colleagues might have misinterpreted their viewpoints. They only stated that aerobic exercise may exhibit weaker intervention effects compared to other forms of exercise, yet this does not imply that aerobic exercise cannot positively influence executive functions [63]. From the discussions among these scholars, it is evident that they still believe aerobic exercise can positively affect executive functions.

Returning to the ADHD population, only the study by Yang and colleagues explored the impact of aerobic exercise on the executive functions of children with ADHD [38]. Their research found that aerobic exercise can have moderate to large positive effects on the executive functions of children with ADHD. However, they did not specifically examine the effects of different frequencies of aerobic exercise on the executive functions of these children. Moreover, the study included both acute and chronic aerobic exercise but did not reasonably distinguish between them in the analysis, which might lead to some bias in interpreting the results. Therefore, our study extends the age range to between 6 and 18 years old, providing valuable information on aerobic exercise interventions for children and adolescents with ADHD.

Mechanism of aerobic exercise on executive function in children and adolescents with ADHD

Aerobic exercise intervention affects the executive function of children and adolescents with ADHD mainly based on the following aspects. First, individuals with ADHD have abnormalities in brain structure and function, such as in the dorsolateral prefrontal cortex, ventromedial prefrontal cortex, parietal cortex, and other regions [64]. The prefrontal cortex is the main brain region related to executive function [65]. Abnormalities in this area will lead to a decline in executive function [4]. However, due to the theory of brain plasticity, the structure and function of the brain will be continuously modified and reorganized as a result of external environmental changes and experiences [66]. This provides the possibility to improve the executive function of children and adolescents with ADHD. When performing aerobic exercise, brain regions related to executive function are activated, such as increased activity in the bilateral prefrontal cortex, remodeling of white matter integrity, and enhanced efficiency of overall neural circuits in the brain [67, 68]. This has a positive effect on improving the executive function of children and adolescents with ADHD. Secondly, performing aerobic exercise will also have an effect on cardiovascular function. Aerobic exercise can enhance the heart’s ability to supply oxygen to various working muscles, increase the blood supply to cortical capillaries and the number of synaptic connections, and promote the development of new neurons [69]. This process leads to a more efficient and more plastic brain. Finally, some scholars believe that dysregulation of catecholamine function and insufficient secretion of neurotransmitters such as norepinephrine and dopamine may also be the pathogenesis of ADHD [70, 71]. However, studies have found that physical activity can cause changes in cerebral blood flow, serotonin, and brain-derived neurotrophic factors and enhance the release of catecholamine neurotransmitters such as norepinephrine and dopamine [72, 73]. These neurotransmitters facilitate communication across different signal transmission pathways in the brain, enhancing overall brain arousal levels, which in turn improves executive function in children and adolescents with ADHD [74].

Discussion of subgroup analysis results

Our subgroup analysis shows that to improve the overall executive functions of children and adolescents with Attention Deficit Hyperactivity Disorder (ADHD), the best intervention effects can be achieved through chronic aerobic exercise of moderate or moderate to vigorous intensity, with each session lasting 50–90 min, a frequency of 3–5 times per week, and an intervention period of 12 weeks or more.

Regarding the subgroup analysis results of acute versus chronic exercise, our study is similar to a previous study [75]. Their research suggested that acute exercise only had a significant intervention effect on inhibitory control in children and adolescents with ADHD, but the effects on working memory and cognitive flexibility were not significant. The reason for this outcome might be that working memory and cognitive flexibility involve more brain regions and require longer periods of training to achieve significant improvement [76, 77]. However, it is worth noting that our study explored the impact of aerobic exercise on the executive functions of children and adolescents with ADHD and included only three acute aerobic exercise interventions. This number is relatively limited, so conclusions about the intervention effects of acute aerobic exercise should still be approached with caution. Further high-quality studies are needed in the future to verify and deepen existing findings.

The subgroup analysis results regarding exercise intensity showed that moderate intensity had the best intervention effect on inhibitory control, while moderate to vigorous intensity produced the best intervention effects on working memory and cognitive flexibility. Our findings are similar to those of several previous studies [26, 62]. They also believed that exercise intensity needs to reach at least a moderate level to produce significant benefits on executive functions. From the perspective of arousal theory, there is an inverted U-shaped relationship between brain arousal levels and exercise intensity. When the exercise intensity is at a moderate load, the arousal level reaches its optimal point, which is most conducive to the development of cognitive functions [78]. Specifically, moderate-load exercise can optimize the release of catecholamines (such as dopamine and norepinephrine), while also enhancing the general biological arousal effect of the central nervous system, thereby allocating cognitive resources reasonably. The reason why moderate-to-vigorous intensity produces the best intervention effects on working memory and cognitive flexibility may be that exercise programs of moderate-to-vigorous intensity can lead to greater increases in brain-derived neurotrophic factor (BDNF), which helps exert its regulatory role on the plasticity of neural structures and functions, thereby promoting improvements in working memory and cognitive flexibility [79].

In summary, aerobic exercise has a positive impact on improving the executive functions of children and adolescents with ADHD. However, the specific effects of the intervention vary depending on the duration period, frequency, session duration, intensity and the choice between acute or chronic exercise forms.

Discussion of network meta-analysis results

In this study, a network meta-analysis was conducted to evaluate the effects of different types of aerobic exercise on inhibitory control in children and adolescents with ADHD. The results show that ball aerobic exercise have shown significant benefits in improving inhibitory control in this population. A significant characteristic of ball aerobic exercises is that they are open-skill activities. The characteristics of open-skill activities are variability and unpredictability [80]. Before performing skill movements, participants cannot decide in advance what the next movement should be. They need to respond according to the stimulation of the external environment, which requires the body to carry out overall organization, coordination, and control of various information processing processes [80]. Training in this dynamic environment can promote the development of neurocognitive function, enhance neuroplasticity, and improve various functions of the brain, including increasing the number of dendritic branches and spines, synaptogenesis, angiogenesis, and the growth of glial cells [81]. In addition, the movement patterns of ball aerobic exercises are relatively complex. According to Picard and Strick, the complexity of exercise changes together with the brain activation pattern and the speed of information processing [82]. The more complex the exercise is, the higher the degree of brain activation will be, and the more activity there will be in the prefrontal cortex [83, 84]. As a result, the ability of inhibitory control will be enhanced.

Heterogeneity

In the inhibitory control and working memory sections of this meta-analysis, the study included showed a very high heterogeneity, which remained significant even after subgroup analysis, with reductions observed only in certain subgroups. After a detailed reading of the original study and discussions among the research team, we identified several potential reasons for the high heterogeneity. First, in different studies, different scholars use different measurement tools when measuring inhibitory control and working memory, which may be the most important reason for the high heterogeneity. Second, there are certain differences in aerobic exercise types, such as ball aerobic exercise, cycling, acute aerobic exercise, etc. Therefore, the specific effects of aerobic exercise intervention on the executive function of children and adolescents with ADHD can be explored in more detail in subsequent studies.

Strengths and limitations

This study has the following strengths. Firstly, the studies included only contain randomized controlled trials and exclude non-randomized controlled trials. Secondly, the objects of studies included are limited to children and adolescents between 6 and 18 years old, which allowed for more targeted and relevant findings. Thirdly, this study separately examined the effects of duration period, frequency, session duration, intensity and whether the aerobic exercise intervention was acute or chronic on the executive functions of children and adolescents with ADHD. Finally, this study conducted a network meta-analysis to investigate the effects of different types of aerobic exercise interventions on the executive functions of children and adolescents with ADHD.

This study still has certain limitations. Firstly, the number of included studies is limited, and the assessment methods for executive functions vary. This results in high heterogeneity in inhibitory control and working memory during statistical analyses, which may affect the statistical power of the meta-analysis. Secondly, only published studies in Chinese and English are included, and studies in other languages are excluded, which may introduce a selection bias. Thirdly, the number of available studies is relatively small, and the number of studies included in some subgroup analyses is insufficient, resulting in a lack of expected results in certain subgroups. Fourthly, although this study includes two groups of children and adolescents with ADHD, due to the scarcity of studies on adolescents, subgroup analysis between children and adolescents was not conducted. Considering that being in different age stages may have a certain effect on the results. Therefore, the specific effects of aerobic exercise intervention remain to be explored further in the future. Fifthly, due to the limited number of literature, when conducting a network meta-analysis of aerobic exercise interventions, only inhibitory control was analyzed. In the future, analyses of working memory and cognitive flexibility need to be conducted when sufficient studies are available.

Conclusion

Our study indicates that aerobic exercise interventions have a positive impact with a moderate effect size on inhibitory control, working memory, and cognitive flexibility in children and adolescents with ADHD. However, the effectiveness of the intervention is influenced by factors such as the intervention period, frequency, session durations, intensity and the choice between acute or chronic exercise. Specifically, chronic aerobic exercise interventions lasting 12 weeks or longer, with a frequency of 3 to 5 sessions per week, session durations of 60 min or more, and intensities that are moderate or moderate-to-vigorous, have the greatest overall effect. In terms of developing inhibitory control, engaging in ball aerobic exercises may yield the best intervention outcomes. Although we have explored the most beneficial aerobic exercise intervention program for the development of executive function in children and adolescents with ADHD, caution should be exercised when interpreting these findings due to the significant heterogeneity in inhibitory control and working memory. Additionally, the implementation process may also be limited by certain objective conditions, such as school facilities and teacher resources.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ADHD

Attention deficit hyperactivity disorder

CI

Confidence interval

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

RCTs

Randomized controlled trials

SMD

Standardized Mean Difference

Author contributions

All authors contributed to the study conception and design; P.W. and D.L. conceived and designed the study; P.W. and D.L. collected the data; P.W., F.R., K.X., Y.G. and Z.X. analyzed and interpreted the data; P.W., F.R. and D.L. drafted the manuscript; P.W., F.R. and D.L. revised the manuscript; All authors have read and agreed to the published version of the manuscript.

Funding

Liaoning Social Science Planning Fund Major Entrusted Project (L24ZD020).

Data availability

Data is provided within the manuscript or appendix files.

Declarations

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent to publish

Not applicable.

Competing interests

The authors declare no competing interests.

Norm or standard

Not applicable.