Associations between accelerometer-measured physical activity and sleep duration and health indicators in children and adolescents with attention-deficit hyperactivity disorder: A systematic review and meta-analysis

Abstract

Objective

This systematic review and meta-analysis synthesized available studies to explore relationships between accelerometer-measured physical activity (PA) levels or sleep duration and health-related outcomes in children and adolescents with attention-deficit hyperactivity disorder (ADHD).

Methods

Following the Preferred Reporting Items for Systematic Review and Meta-analyses guidelines, the six databases APA PsychInfo, CINAHL Ultimate, Embase, MEDLINE, SPORTDiscus with Full Text and Web of Science were searched from inception through July 2024. The correlation coefficient (r) was employed to determine the effect size in the meta-analysis. A random-effects model was utilised to estimate the potentially heterogeneously distributed effect sizes with a 95 % confidence interval (CI) between groups. Statistical heterogeneity was assessed using I², with a p-value derived from Q statistics.

Results

Thirteen eligible studies were included, involving 738 children and adolescents with ADHD (71.1 % boys). The health indicators assessed were executive functions (EFs), motor development, psychological health, and core ADHD symptoms. We found that higher PA levels (i.e., moderate-to-vigorous physical activity [MVPA]) (k = 9, r = 0.36, 95 %CI 0.22 to 0.49) with medium heterogeneity (Q = 13.006, I² = 38 %, p = 0.112), and longer sleep duration (k = 6, r = 0.20, 95 %CI 0.06 to 0.35) with medium heterogeneity (Q = 7.495, I² = 33 %, p = 0.186), were significantly correlated with better overall EFs. Additionally, MVPA was negatively associated with psychopathology (k = 5, r = −0.19, 95 %CI -0.35 to −0.03) with medium heterogeneity (Q = 8.058, I² = 50 %, p = 0.089) in children and adolescents with ADHD.

Conclusion

The findings provide support for significant relationships between PA levels (i.e., MVPA) and sleep duration with overall EFs and MVPA with psychopathology in children and adolescents with ADHD. It is recommended that future studies conduct moderation analyses to examine factors that affect the generalisability of the findings, such as age, sex, and ADHD subtypes. However, these analyses could not be conducted in the current meta-analysis due to the limited number of included studies.

Highlights

• •The study evaluates the association among physical activity, sleep duration, and health indicators in children with ADHD.
• •Higher levels of physical activity and increased sleep duration are closely linked to improvements in executive function.
• •Higher levels of physical activity are positively associated with reduced psychopathology.
• •Longer durations of MVPA are associated with improved executive function and d psychopathology in children with ADHD.

1. Introduction

Attention-deficit hyperactivity disorder (ADHD) is characterized by the core symptoms of inattention, impulsivity, and hyperactivity.¹ It is a major global public health concern,² with an estimated global prevalence of 7.2 % in all age groups and 5 % in school-age children.³ Psychiatric comorbidity is common in patients with ADHD, particularly in children.⁴ ADHD-related symptoms include executive dysfunction,⁵ mental health problems,⁶ sleep disturbances,⁷ and physical inactivity,⁸ which lead to adverse health outcomes.

The relationship between physical activity (PA) and ADHD is complex, and associations in both directions have been reported. Physical inactivity is significantly more common in children with ADHD, affecting over 67 % of this population compared to healthy controls. The PA data from subjective reports indicate that children and adolescents with ADHD tend not to meet the World Health Organization's (WHO) minimum guidelines of 60 min per day of moderate-to-vigorous physical activity (MVPA).8, 9, 10 However, accelerometer-measured data showed that children with ADHD had higher mean activity levels than children without ADHD.¹¹ Additionally, recent preliminary studies have yielded inconsistent results in actigraphy-measured MVPA for children with ADHD, ranging from 25.27 min per day¹² to 84.19 min per day.¹³

Further, sleep disturbances are significantly more common in children with ADHD than their peers without ADHD, with the estimated prevalence rates ranging from 25 % to 50 %.¹⁴ The most frequently cited sleep problems (in parent-reported questionnaires) in children with ADHD include sleep latency delay, bed-time resistance, nocturnal awakenings, and daytime sleepiness.¹⁵ A recent meta-analysis¹⁶ revealed that compared with healthy controls, children and adolescents with ADHD have higher sleep latency (g = 0.399, 95 % confidence interval [CI] = 0.177 to 0.620) and decreased sleep efficiency (g = −0.221, 95 % CI = −0.383 to −0.059]) as measured using actigraphy. Studies that have examined sleep duration in children with and without ADHD using subjective (e.g., parent-reported questionnaires)¹⁵ and objective (e.g., accelerometer and polysomnography) measurements^11,16,17 have yielded mixed results. Despite emerging evidence of differences in PA levels and sleep behavior between children with and without ADHD, few studies have examined the associations between objectively measured PA levels or sleep duration and health indicators in children and adolescents with ADHD. Understanding the implications of insufficient PA levels and sleep duration in individuals with ADHD during childhood is critical for developing promising strategies aimed at mitigating the adverse effects of physical inactivity and sleep disturbances.

One previous review documented a significant association between a short sleep duration and ADHD, especially the hyperactivity component¹⁸; however, the findings were based on parent-reported questionnaires, which are limited because of the parents' potential to over-report sleep problems in children with ADHD.¹⁹ The importance of measuring objective PA and sleep in studies of health-related behaviors in ADHD has gained attention recently. Actigraphy is normally used to measure the objective patterns of sleep and PA by recording the user's activity during the night and during the day in a free-living environment. It has been widely used in individuals with ADHD.^11,16 Despite increasing interest in the link between PA or sleep duration and health-related outcomes, a systematic review and meta-analysis of studies that have examined the influence of PA levels and sleep duration on key health indicators in children and adolescents with ADHD, with objective measurements in a natural environment, is lacking. A meta-analysis of associated effect sizes may help quantify these relationships to better understand how PA levels and sleep duration, measured using actigraphy in natural environments, are related to health indicators. Therefore, this study aimed to evaluate the associations between PA levels and sleep duration in a natural environment and health indicators in children with ADHD.

2. Methods

This study complied with the Preferred Reporting Items for Systematic Review and Meta-Analyses Statement.²⁰ The protocol for this systematic review has been registered with the International Prospective Register of Systematic Reviews (CRD42022363932).

2.1. Literature search

We conducted electronic searches of APA PsychInfo (via Ovid), CINAHL Ultimate (via EBSCOhost), Embase (via Ovid), MEDLINE (via Ovid), SPORTDiscus with Full Text (via EBSCOhost), and Web of Science databases from their inception to July 2024 to identify all relevant published articles pertaining to associations between PA levels or sleep duration and key health indicators in children and adolescents with ADHD. We limited the search to “English” and “human-related” articles. The initial search was performed using the terms “physical activity,” “sleep”, “ADHD,” and “children and adolescents”(Table S1). We also conducted snowball and citation searches to identify additional relevant articles in reference lists.

2.2. Eligibility criteria

Studies were included in the review if they (1) reported accelerometer-measured PA levels and/or sleep duration (total sleep time) in free-living conditions, recorded on at least 3 valid days/nights in children and/or adolescents with ADHD (2) Were the included observational studies (i.e., cross-sectional, case-control, longitudinal, or cohort); specifically focused on examining the associations between physical activity or sleep duration and health indicators, in order to capture habitual PA and sleep patterns rather than the effects of structured interventions; (3) involved participants aged from 5 to 18 years diagnosed with ADHD using standardized diagnostic criteria (e.g., the Diagnostic and Statistical Manual of Mental Disorders [DSM] editions 3, 4 or 5; the 10th revision of the International Classification of Diseases; or the Kiddie Schedule for Affective Disorders and Schizophrenia – Lifetime Version [K-SADS-PL]) by professionals (e.g., clinicians, psychologist, or psychiatrist) or based on parental/school reports; and (4) were published in peer-reviewed English-language articles with the full text available. Studies were excluded from the review if (1) data on physical activity and sleep duration were collected from subjective measurements (i.e., parent-reported or self-reported questionnaires) rather than from accelerometers; (2) they involved intervention research (e.g., clinical or field trials); (3) they were reviews or meta-analyses or reported as care or government reports or conference abstracts; (4) the sample included preschool children (aged 0–5 years) or adults (aged 18 years or older); or (5) they were published in a language other than English.

2.3. Health indicators

This study identified nine health indicators based on previous systematic reviews and meta-analyses.21, 22, 23 Specifically, we identified the health indicators from studies that explored their relationships with objectively measured physical activity,²² objectively and subjectively measured sleep duration,²¹ and the combined effects of physical activity, sedentary behavior, and sleep²³ in children and youth. The nine health indicators were assessed as “critical” or “important” outcomes based on the Grading of Recommendations Assessment, Development, and Evaluation framework (GRADE).²⁴ Using the GRADE system, outcomes are classified as either ‘critical’ or ‘important’ for decision-making. Nine health indicators were selected to encompass a broad range of outcomes, building on previous reviews that identified critical and important health indicators.²⁵ These health indicators were: (1) adiposity (e.g., body mass index); (2) motor development (e.g., locomotion and object control skills); (3) psychological health (e.g., psychological well-being and psychopathology); (4) cognitive development (e.g., executive function); (5) physical fitness (e.g., cardiovascular fitness); (6) bone and skeletal health (e.g. risk of fractures, bone density); (7) cardiometabolic health (e.g., blood pressure); (8) growth (e.g. height and weight) (9) harms/injuries (e.g., accidental fall, fracture);.21, 22, 23 According to the GRADE framework, the outcomes were ranked as critical (adiposity, motor development, psychological health, cognitive development, physical fitness) and important (bone and skeletal health, cardiometabolic health, harms/injuries).²⁶ In addition, because ADHD is a lifelong disorder, we also considered core ADHD symptoms as critical indicators that affect children's health.²⁷

2.4. Study selection and data extraction

Two reviewers (XL and HQ) conducted independent multi-step searching and screening and discussed any discrepancies to reach a consensus. The consistency in abstract and full-text screening between the reviewers was measured using the kappa value.²⁸ We developed a standardized data extraction form to extract relevant study characteristics, including bibliographic details (author and year), research design, participant characteristics (sample size, sex, age, medication status, and diagnosis), PA measures and levels, sleep measures and duration, recorded PA or sleep duration time, health-related outcomes, and major findings.

2.5. Quality assessment

Two independent reviewers evaluated the methodological quality of each included study using the McMaster Critical Reviewer Form for quantitative studies.²⁹ Previous reviews related to research on PA and children with special educational needs have found the numerical rating criteria for non-experimental quantitative studies to be a reliable and valid scoring method for assessing methodological quality.^30,31 We evaluated three key criteria in the included studies: sample, measurement, and analysis,^30,31 because these are fundamental aspects that directly affect the validity and reliability of study outcomes.³² Sample evaluated whether the selection bias was reduced (e.g., representative of selected population or convenience sample), whether the sample size was suitable for the research design and questions, and whether the characteristics of subjects were clearly described by the authors. Measurement examined whether the measurement bias was reduced (e.g., reliability and validity of the measurement tool, recall/memory). Analyses examined whether reported analyses were appropriate for the research questions and outcome measures (e.g., reported statistical significance, point estimates, provided variability, and discussed clinical importance). We assigned scores of 1–3 stars, where 1 star indicated no evidence that the study met any of the criteria; 2 stars indicated that certain pieces of evidence met the criteria, or that the report was unclear; and 3 stars indicated that the evidence completely met the criteria.^30,31

2.6. Data analysis

Data were analyzed using Comprehensive Meta-Analysis (v.3) software (Biostat, Englewood, NJ, USA). A meta-analysis was conducted for studies exploring the associations between PA levels or sleep duration and health-related outcomes in children and adolescents with ADHD. The correlation coefficient (r) was used to calculate the effect size in the meta-analysis. All effect sizes derived from single studies that met our inclusion criteria were included and treated as independent effect sizes in the analysis. A random-effects model was used to compute the potential heterogeneously distributed effect sizes with the 95 % CI between groups. All identified effect sizes were included in the meta-analysis and were transformed into Fisher's z values for further analysis. For presentation, the Fisher's z values were transformed back to r values. The r values were interpreted as small (0.10–0.20), medium (0.21–0.35), or large (>0.35).³³ Statistical heterogeneity was examined using I² values, with a p-value for Q-statistics. I² values were used to represent the degree of heterogeneity: small (≤25 %), medium (50 %), or large (≥75 %).³⁴

3. Results

3.1. Study identification

We identified 1658 relevant articles from the six databases searched. Fig. 1 shows the number of articles screened and the number that met the inclusion criteria. After removing 1028 duplicates, 630 underwent title and abstract screening. Thirty-two abstracts met the inclusion criteria, with an inter-rater reliability of k = 0.81, and 12 articles met the inclusion criteria after full-text screening with an inter-rater reliability of k = 0.91. Additionally, we identified one record that met the inclusion criteria through manual searching, resulting in 13 studies (nine assessing PA levels and four assessing sleep duration) included in the systematic review. All 13 studies reported sufficient statistical data to be suitable for inclusion in the meta-analysis.^12,13,35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45

Fig. 1.

PRISMA flow diagram of the selection of studies.

3.2. Study characteristics

Table 2 shows the participants’ characteristics and the results of quality assessments of the included studies. All 13 studies were conducted in the last 13 years (from 2011 to 2022). Half of the included studies were conducted in Asia, five were conducted in North America, and two were conducted in Germany. The 13 studies included 738 children and adolescents with ADHD (71.1 % boys), with an age ranging from 8 to 15. Seven studies (53.8 %) clearly classified ADHD subtypes measured using standardized diagnostic methods (e.g., DSM-4 and -5 and K-SADS-PL). Table 1 summarizes the associations between PA levels and sleep duration and four categories of health-related outcomes. Ten studies^12,13,35, 36, 37, 38, 39, 40^,42,43 explored associations between PA levels and four categories of health-related outcomes, namely EF, motor development, psychological health, and ADHD symptoms. Seven studies^{12,13,35,37,38,40,43} measured daily time spent in MVPA as an outcome to represent the PA dose per day, with a range of 27.57–87.34 min and a mean of 57.97 min in children with ADHD, indicating that they failed to meet the WHO-recommended 60 min of daily MVPA.⁴⁶ Four studies^38,41,44,45 objectively recorded a mean of 454.87 min (7.6 h) of sleep per night for children and adolescents with ADHD, which indicates they fail to meet the recommended sleep duration of 9–12 h per 24 h for school-aged children (6–12 years), as advised by the American Academy of Sleep Medicine.⁴⁷ Additionally, four studies^38,41,44,45 explored associations between sleep duration and three health-related outcome categories, namely, EF, psychological health, and ADHD symptoms. The duration of the studies assessing PA levels and sleep duration ranged from 7 to 14 days.

Table 2.

Summary of participants’ characteristics and quality assessment of included studies.

Study Name (Year, Country/Region)	Sample size	Age (M±SD)	Sex (Male%)	Medication use	Diagnostic Methods (Classification of ADHD subtype)	Quality Criteria
						Sample	Methods	Analysis
Becker et al. (2019, USA)	• 162 ADHD	• 13.17 ± 0.41	• M-64.8 %	Yes (59.3 %)	DSM-5 & P-CHIPS (ADHD-I (120); ADHD-C (42))	∗∗∗	∗∗∗	∗∗∗
Gapin & Etnier (2010, Germany)	• 18 ADHD	• 10.61 ± 1.50	• M-100 %	Yes	Medical professional (ADHD-H (5); ADHD-I (2); ADHD-C (8), not reported (3))	∗∗	∗∗∗	∗∗∗
Gawrilow et al. (2016, Germany)	• 38 ADHD	• 14.37 ± 1.88	• M-47.4 %	Yes	School report	∗	∗∗	∗∗
Li et al. (2021, China)	• 86 ADHD	• 8.45 ± 1.40	• M-82.6 %	Yes	DSM-5 & K-SADS-PL (ADHD-H (5); ADHD-I (51); ADHD-C (30))	∗∗∗	∗∗∗	∗∗∗
Liang et al. (2022, China)	• 56 ADHD	• 8.82 ± 1.49	• M-83.9 %	Yes	DSM-5 & K-SADS-PL (ADHD-H (7); ADHD-I (25); ADHD-C (24))	∗∗∗	∗∗∗	∗∗
Liang, Qiu & Sit (2023, China)	• 85 ADHD	• 8.41 ± 1.44	• M-84.7 %	Yes	DSM-5 & K-SADS-PL (ADHD-H (9); ADHD-I (45); ADHD-C (31))	∗∗∗	∗∗∗	∗∗∗
Licht & Tryon (2009, USA)	• 9 ADHD	• 9.33 ± 1.00	• M-88.9 %	N/A	DSM-4	∗∗	∗∗	∗∗
Lin et al. (2013, Taiwan)	• 20 ADHD	• 8.64 ± 2.57	• M-100 %	N/A	DSM-4	∗∗	∗∗	∗∗∗
Mohammadi et al. (2022, Iran)	• 68 ADHD	• 10.22 ± 1.55	• M-58.8 %	N/A	School report	∗	∗∗	∗
Moreau et al. (2013, Canada)	• 43 ADHD	• 10.01 ± 1.80	• M-58.1 %	None	DSM-4 & K-SADS-PL (ADHD-H (2); ADHD-I (23); ADHD-C (18))	∗∗∗	∗∗∗	∗∗∗
Santiago-Rodríguez et al. (2022, USA)	• 23 ADHD	• 9.2 ± 1.91	• M-52.2 %	N/A	NIMH DISC-IV-P	∗∗	∗∗	∗∗
Waldon et al. (2018, Canada)	• 25 ADHD	• 9.02 ± 1.76	• M-80 %	None	Registered psychologists	∗∗	∗∗∗	∗∗∗
Yu et al. (2019, Taiwan)	• 73 ADHD	• 9.96 ± 1.58	• M-94.5 %	N/A	Medical professional & CBCL (ADHD-H (5); ADHD-I (30); ADHD-C (38))	∗∗	∗∗	∗∗∗

Note: ∗ = no criteria were met within that component; ∗∗ = only some criteria were met within the component; ∗∗∗ = all criteria were met within that component. ^#: the criteria for medication use are whether the participants with a history of psychotropic or currently used drugs, including non-stimulant medications for ADHD; or with a drug-naïve before study.

CRS-R:L: Conners' Rating Scale-Revised Long Version; CBCL: Chinese version of Child Behavior Checklist; DSM-4 and-5: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition and Fifth Edition; K-SADS-PL: Schedule for Affective Disorders and Schizophrenia for School-age Children-Present and Lifetime Version; NIMH DISC-IV-P: National Institute of Mental Health Diagnostic Interview Schedule for Children, Version IV, Parent Interview. N/A: Not available.

Table 1.

Summary of included studies on the PA levels and sleep duration and its associations with health-related outcomes.

Study Name (Year, Country/Region)	Measures of PA/Sleep (model)	PA levels	Sleep duration (mins)	Recorded time	Health-related outcomes				Major findings
					Executive functions	Motor development	Psychological health	ADHD Symptoms
Becker et al. (2019, USA)	ActiGraph (GT9X Link)	–	• 464.4 ± 43.2 (school night)	14 days	–	–	• Depression • Anxiety	–	• Depression00 (r = −0.06) • Anxiety00 (r = 0.05)
Gapin & Etnier (2010, Germany)	Accelerometer (New Lifestyles NL-1000)	MVPA • 38.47 ± 17.56 min/day (ADHD)	–	7 days	• IC • WM • CF • Planning	–	–	–	• IC00 (r = −0.11) • WM00 (r = 0.37) • CF00 (r = 0.45) • Planning+ (r = 0.57)
Gawrilow et al. (2016, Germany)	Pedometer (OMRON Walking Style)	Step Counts • 8486 ± 4941 per day	–	10 days	–	–	• Depression	–	• Depression- (r = −0.29)
Li et al. (2021, China)	Actigraph (wGT3X-BT)	MVPA • 85.17 ± 30.52 min/day	–	7 days	–	TGMD-3 total	• QoL	–	• TGMD-3 total+ (r = 0.29) • QoL00 (r = −0.06)
Liang et al. (2022, China)	Actigraph (wGT3X-BT)	MVPA • 87.34 ± 38.25 min/day	• 408.83 ± 30.66	7 days	• IC • WM • CF	–	–	–	• PA&IC00 (r = −0.02) • PA&WM+ (r = 0.48) • PA&CF- (r = −0.52) • Sleep&IC00 (r = −0.07) • Sleep&WM00 (r = 0.23) • Sleep&CF00 (r = −0.17)
Liang, Qiu & Sit (2023, China)	Actigraph (wGT3X-BT)	MVPA • 79.06 ± 31.61 min/day	–	7 days	–	–	• Depression • Anxiety • Stress • Resilience	–	• Depression- (r = −0.39) • Anxiety00 (r = −0.06) • Stress- (r = −0.22) • Resilience+ (r = 0.23)
Licht & Tryon (2009, USA)	Actigraph	Activity Units • 411.50 ± 129.5	–	14 days	–	–	–	• Inattention (school & home) • Hyperactivity (school & home)	• Inattention_school (r = −0.17) • Inattention_home (r = −0.24) • Hyperactivity_school (r = −0.07) • Hyperactivity_home (r = −0.24)
Lin et al. (2013, Taiwan)	Actigraph (GT1M)	MVPA • 51.75 ± 17.12 min/day	–	7 days	–	• Fine motor	–	• Inattention	• Fine motor00 (r = −0.19) • Inattention- (r = −0.48)
Mohammadi et al. (2022, Iran)	Actigraph (wGT3X-BT)	MVPA • 36.46 ± 11.52 min/day	–	7 days	–	–	• QoL	–	• QoL+ (β = 0.13)
Moreau et al. (2013, Canada)	Actiwatch (Mini-Mitter 64®)	–	• 465.17 ± 40.02	7 nights	• IC • EFs	–	–	–	• IC00 (r = −0.18) • EFs- (r = −0.52)
Santiago-Rodríguez et al. (2022, USA)	Actigraph (wGT3X+/GT1M)	VPA • 3.7 ± 4.0 min/day	–	7 days	• EFs • WM	–	–	–	• EFs- (r = −0.46) • WM00 (r = 0.18)
Waldon et al. (2018, Canada)	Actigraph (Micro-Mini Motionlogger)	–	• 481.09 ± 70.68	7 nights	• IC	–	–	• Inattention	• IC00 (β = −0.62) • Inattention00 (β = 0.33)
Yu et al. (2019, Taiwan)	Actigraph (wGT3X+)	MVPA • 27.57 ± 16.89 min/day	–	7 days	–	• MABC-2 total	• Depression	–	• MABC-200 (r = 0.171) • Depression00 (r = −0.015)

Note: +: significant statistical positive association; -: significant statistical negative association; 00: no statistically significant association; N/A: not available.

ADHD: Attention-deficit/hyperactive disorders; BRIEF: Behavior Rating Inventory of Executive Function; CF: Cognitive flexibility; IC: Inhibitory control; EFs: Executive functions; MABC-2: Movement Assessment Battery for Children, 2nd edition; MVPA: Moderate to vigorous physical activity; METs: Metabolic equivalents; QoL: Quality of life; TGMD-3: Test of Gross Motor Development, 3rd edition; VPA: Vigorous physical activity; WM: Working memory; PIM: Proportional Integrating Measure (range 0–65535).

3.3. Meta-analysis of the associations between PA levels or sleep duration and health-related outcomes

3.3.1. Sleep duration

Eight independent effect sizes were used to summarize two health indicator

categories associated with sleep duration (Fig. 2). We found a small, but significant, correlation between sleep duration and EF (k = 6, z = 0.207, r = 0.20, 95 % CI = 0.06 to 0.35) with medium heterogeneity (Q = 7.495, I² = 33 %, p = 0.186). However, there was no association between sleep duration and psychopathology (r = −0.005, p = 0.929).

Fig. 2.

Meta-analysis of associations between sleep duration and health indicators.

3.3.2. PA levels

We summarized four health-related categories associated with PA levels from 25 independent effect sizes (Fig. 3). Nine independent effect sizes in three studies explored associations between PA and EF in children with ADHD. Transforming Fisher's z values back to Pearson's r values revealed a significant, moderate correlation between PA and overall EF (k = 9, z = 0.38, r = 0.36, 95 % CI = 0.22 to 0.49), with medium heterogeneity (Q = 13.006, I² = 38 %, p = 0.112). Eight independent effect sizes tested relationships between PA and psychological health, including five for physiological ill-being and three for psychological well-being. As shown in Fig. 3, we found a significant but weak correlation between PA and psychopathology (k = 5, z = −0.20, r = −0.19, 95 % CI = −0.35 to −0.03), with medium heterogeneity (Q = 8.058, I² = 50 %, p = 0.008). The observed moderate heterogeneity may be attributable to the use of different assessment tools for psychopathology, such as parent-reported questionnaires (e.g., CBCL) and self-reported measures (e.g., DASS-21). However, there was no evidence of an association between PA and psychological well-being (r = 0.12, p = 0.215). Three effect sizes reported associations between PA and motor development, but no significant positive relationships were found (r = 0.17, p = 0.118). Five effect sizes identified core ADHD symptoms, but the results were mixed, and no association was identified with PA (r = −0.30, p = 0.061).

Fig. 3.

Meta-analysis of associations between PA levels and health indicators.

3.3.3. MVPA

Among the 10 studies assessing PA levels, eight studies investigating associations between MVPA and health-related outcomes in children and adolescents with ADHD (n = 444) were identified. The results of MVPA and health-related outcomes are shown in Fig. 4. A dose–response relationship between MVPA and overall EF was reported in seven independent effect sizes. MVPA showed a moderate positive association with overall EF in children with ADHD (k = 7, z = 0.39, r = 0.37, 95 % CI = 0.21 to 0.51) with medium heterogeneity (Q = 11.942, I² = 50 %, p = 0.063). Five studies reported associations between MVPA and psychopathology, while three studies assessed relationships between MVPA and psychological well-being. The results showed that MVPA was negatively correlated with psychopathology in children and adolescents with ADHD (k = 5, z = −0.20, r = −0.19, 95 % CI = −0.35 to −0.03) with medium heterogeneity (Q = 8.058, I² = 50 %, p = 0.089). These results were identical to those for the association between PA and psychopathology. However, there was no correlation between MVPA and psychological well-being (r = 0.12, p = 0.215). Three studies reported inconclusive results for the association between MVPA and motor skills development, and no association was found in our meta-analysis (r = 0.17, p = 0.118), given the small number of included studies (k = 3) and the inconsistency in outcome measures of motor skills (e.g., TGMD-3, Sensory Challenge Protocol, and MABC-2).

Fig. 4.

Meta-analysis of associations between MVPA and health indicators.

3.4. Quality assessment

We calculated the inter-rater reliability of study quality assessments between the two researchers for sample (k = 0.88), measurement (k = 0.87), and analysis (k = 0.84). All studies used a convenience sample. More than 70 % of the participants were boys, and 50 % of the included studies gave detailed ADHD subtype classifications using standard diagnostic tools. All of the studies objectively measured PA levels and sleep duration. Fifty-eight percent of the included studies were rated as 3 stars for the analysis component, because they fully reported statistical significance and clearly described the study's limitations and implications of the findings (Table 2).

3.5. Publication bias

As less than 10 studies were included in the associations between PA levels (n = 9) and sleep duration (n = 4) and health indicators, publication bias was not investigated. Also, tests for funnel plot asymmetry should only be used when there are at least 10 studies included in the meta-analysis.⁴⁸

4. Discussion

We conducted a systematic review and meta-analysis of the available evidence for the relationship between PA levels or sleep duration and health-related outcomes in children with ADHD. The novel findings demonstrated that increased PA levels (r = 0.38) and sleep duration (r = 0.21) were positively correlated with overall EF. More importantly, MVPA was found to be positively correlated with overall EF (r = 0.37) and negatively correlated with psychopathology (r = −0.19) in children and adolescents with ADHD.

The results of the present study contribute to the existing evidence on the relationship between PA and EF in children.⁴⁹ Specifically, the findings indicate that prolonged MVPA exhibits an inverted U-shaped effect, producing a greater impact on cognition than either light- or vigorous-intensity exercise.⁵⁰ Furthermore, a recent study examining children with neurodevelopmental disorders found that increasing the number of weeks, total sessions, and overall duration of PA interventions was associated with larger effect sizes for improvements in overall EFs resulting from chronic PA.⁵¹ These findings provide additional support for a significant cross-sectional relationship between PA and EF, indicating that increased exposure to PA may be linked to greater cognitive benefits in this population. However, these effects have mainly been confirmed in healthy children.⁵² Even more importantly, a previous study reported that a single session of MVPA induced a general facilitative effect (g = 0.25 to 0.32) on overall EF in children with ADHD.⁵³ The current study extends this knowledge by linking precise estimates of the correlation between MVPA and overall EF (r = 0.37) in children with ADHD.

In line with a previous meta-analysis that reported a significant positive relationship between sleep duration and cognitive performance in TD school-aged children,^54,55 we found a significant but weak correlation (r = 0.21) between sleep duration and EF, suggesting that a longer sleep duration is associated with better EF in children with ADHD. A previous meta-analysis showed that sleep-restricted children exhibit significantly worse cognitive performance than sleep-extended children, and the impact of restricted sleep on attention may follow a dose–response pattern, indicating that shortened sleep may result in impaired attention (d = −0.37, 95 % CI = −0.55 to −0.19).⁵⁶

Reflecting the inverse association between increased levels of PA and psychopathology in TD children and adolescents,⁵⁷ a negative association between MVPA and psychopathology (r = −0.19) was also found in the current meta-analysis. Children with ADHD commonly show high rates of mild depression and anxiety,^58,59 and childhood ADHD symptoms are predictors of psychological dysfunction in young adulthood,⁶⁰ with increased symptom levels contributing to psychopathology worse.⁶¹ As such, it is important to note that increased PA levels may be a protective factor against psychopathology in children. For example, one longitudinal study indicated that more physically active children with ADHD showed fewer mental health problems than children with a medium or low level of PA.⁶² Furthermore, chronic PA interventions with an MVPA level showed a reduction in anxiety and depression scores in children with ADHD.^63,64 Therefore, it seems that participation in PA, especially high levels of PA (e.g., MVPA), may be associated with reduced psychopathology in children and adolescents with ADHD.

Nevertheless, it should be noted that the associations between PA levels and motor development were inconclusive in the current meta-analysis (r = 0.17, p = 0.118). Motor impairments are pervasive in children with ADHD, affecting approximately 30–50 % of school-aged children.⁶⁵ Findings obtained from healthy children indicate that greater levels of MVPA are associated with better motor proficiency, and MVPA has been significantly related to motor development since childhood.^66,67 However, more evidence is urgently needed to support the reciprocal associations between MVPA and motor development in children with ADHD. Although our meta-analysis did not find significant associations between PA and ADHD symptoms, previous research suggests potential benefits of chronic exercise interventions. A prior study reported that chronic exercise interventions could improve overall core symptoms (SMD = −0.39, 95 % CI: −0.64 to −0.14) and inattention (SMD = −0.32, 95 % CI: −0.63 to −0.004) in children and adolescents with ADHD.⁶⁸ Notably, closed-skill exercises such as walking, running, or jumping demonstrated a large effect size in reducing core symptoms (SMD = −0.83, 95 % CI: −1.30 to −0.35).⁶⁸ These findings underscore the need for further high-quality RCTs to determine the most effective exercise prescriptions for managing ADHD core symptoms.

There are several limitations of this review. First, only a limited number of relevant studies were identified, which may have resulted in inconclusive associations between PA and health-related outcomes. Furthermore, the meta-regression on age and subgroup analyses of measurement tools and geographic regions are limited in their ability to identify potential moderators due to the limited number of included studies. Future research is recommended to perform these analyses when sufficient data becomes available. Second, only two studies recruited adolescents with ADHD^36,45; thus, it was difficult to explore the potential role of age. Third, the current study focused on cross-sectional data, which did not allow us to determine causality in the association between PA or sleep duration and EF across time. Fourth, most of the included studies used a convenience sampling approach, which may have caused sampling bias. Fifth, in our review, all effect sizes were treated as independent; however, this may be a limitation because multiple effect sizes from the same study are not truly independent and are grouped within it. This can cause a 'unit-of-analysis error,' which may bias the results.⁶⁹ Future analyses should consider methods to address such dependency, such as multilevel modelling, to obtain more accurate estimates. Sixth, most participants in the current study are boys (70 %), suggesting that the results have limited generalisability to girls. Finally, only one study simultaneously measured PA and sleep duration, making it challenging to examine their joint effect or their interaction on health-related outcomes.

5. Conclusions

The present study revealed that increased PA levels (i.e., MVPA) and a longer sleep duration were positively correlated with overall EF in children and adolescents with ADHD. Moreover, MVPA was negatively correlated with psychopathology in this clinical population. Therefore, increased engagement in PA (i.e., MVPA), along with longer sleep duration, is recommended for children with ADHD to improve EFs and reduce psychopathological symptoms.

Ethical statement

This review article does not involve human or animal experiments, and it is based exclusively on published literature. Therefore, a Statement of Ethics is not applicable.

Funding source

No sources of funding were used to assist in this article.

Declaration of competing interest

The authors declare that they have no competing interests or conflict of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article: