Article Text

Download PDFPDF

Physical activity and mental health in children and adolescents: a review of reviews
  1. Stuart J H Biddle,
  2. Mavis Asare
  1. School of Sport, Exercise and Health Sciences, Loughborough University, Loughborough, UK
  1. Correspondence to Professor Stuart Biddle, Physical Activity and Public Health, School of Sport, Exercise and Health Sciences, Loughborough University, Loughborough, Leicestershire LE11 3TU, UK; s.j.h.biddle{at}


  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

View Full Text

Statistics from

Mental illness is a serious public health issue. It is expected to account for 15% of the global burden of disease by 2020, which would make it the leading disease burden.1 Suicide, depression, eating disorders and anxiety are some of the conditions that affect young people in disproportionate rates in comparison to many other population groups.2 Moreover, there is widespread belief that physical activity is inherently good for young people in respect of varied psychosocial outcomes, such as self-esteem and cognitive functioning. The majority of studies in this area, however, are cross-sectional and therefore causality cannot be established because the temporal relationship between exposure (physical activity) and outcome (mental health) has not been tested or shown with any consistency. Thus, although there is evidence that physical activity can enhance psychological well-being, such an outcome may not be inevitable3 or may be dependent on certain conditions existing.

The effect of physical activity on mental health in children and adolescents has received significantly less attention than in adult populations.4 Where it has been investigated, the work has primarily focused on depression, anxiety and self-esteem. In addition researchers and education professionals are showing increasing interest in the effects of physical activity on cognitive functioning. For these reasons, and space restrictions, our review is delimited to the outcomes of depression, anxiety, self-esteem and cognitive functioning.

The purpose of this article, therefore, is to synthesise evidence on chronic physical activity participation and mental health in children and adolescents mainly through a review of reviews. Such a method has been adopted to optimise the ability to cover four areas of mental health and to synthesise 18 reviews. It would have been unwieldy to attempt individual systematic reviews of the literature in each of the four mental health domains. Moreover, a review of reviews is accepted practice in the medical and health behaviour literatures.5 In addition a brief analysis will also be provided on relationships between sedentary behaviour and mental health from primary research studies.


To find reviews of studies on chronic physical activity and mental health in young people, the following electronic databases were searched up to October 2010: PubMed, SPORTDiscus, PsychINFO, Web of Science, Medline, Cochrane Library and ISI Science Citation Index. We searched using terms that reflected exposure variables of interest (eg, sport, exercise, physical activity), mental health outcome variables (eg, depression, anxiety, self-esteem, cognitive functioning) and methods (only reviews, systematic reviews, meta-analyses). Only review articles were included. Additional searches of personal files supplemented the electronic sources.

Articles were selected for detailed analysis if they met the following inclusion criteria (1) were review articles (narrative, systematic or meta-analytic); (2) reported the relationship between physical activity and at least one mental health outcome that included depression, anxiety/stress, self-esteem or physical self-worth and cognitive functioning; (3) reported chronic physical activity studies, including interventions (reviews of acute studies were excluded); (4) were on school-aged children or adolescents up to and including 18 years with no known physical health limitations. If a review contained some data on young people, but the primary emphasis was on adults, the data were scrutinised and reported where appropriate.

To balance the work on physical activity, a brief narrative review was conducted on sedentary behaviour and mental health. This is not meant to be a comprehensive systematic review but rather a commentary on recent articles.


Five review articles were retrieved that reviewed evidence concerning physical activity and depression, four for anxiety, three for self-esteem and seven for cognitive functioning. Two articles covered more than one mental health outcome variable, with Larun et al6 addressing anxiety and depression, and Calfas and Taylor7 reviewing depression, anxiety and self-esteem.


Evidence from adult studies demonstrates that physical activity is inversely associated with symptoms of depression,8 9 and there is some evidence that this relationship is causal.10 There is, however, much less evidence for this relationship in children and adolescents. We found five reviews that synthesised data for young people,6,,9 11 although the article by Dunn and Weintraub primarily provides a methodological critique of articles used in the review by Larun et al. Reviews of depression are summarised in table 1, excluding the critique by Dunn and Weintraub, leaving four reviews.

Table 1

Reviews of physical activity and depression in young people

The study by North et al9 was the first meta-analysis investigating associations between physical activity and depression and included all research designs. An overall effect size (ES) of −0.53 was reported, with five studies involving only young people reflecting a similar value (ES = −0.49). However, acute and chronic studies were included as well as non-intervention designs.

The first systematic review concerning physical activity and mental health in adolescents was published as part of the American physical activity guidelines process,7 although the authors used a wide age range of 11–21 years. All research designs were eligible for analysis but ES was calculated only for experimental designs. From only four intervention studies, ES = −0.38 favouring physical activity over a control group. By using all research designs, Calfas and Taylor reported that 9 of 11 studies showed a negative association between physical activity and depression.

Craft and Landers8 conducted a meta-analysis on exercise and depression for those with clinical depression. Only three studies provided data for those aged 12–18 years, showing a small non-significant ES of −0.15, much smaller than that for adults and in all studies (−0.72).

Larun et al6 conducted a systematic review of exercise interventions on depression in young people up to the age of 20 years. Five trials were found that investigated whether vigorous exercise (fitness training and weight training) conferred benefits over no intervention. They found a significant moderate effect (ES = −0.66; CI, −1.25 to −0.08) but noted that the trials were of low quality and highly varied in respect of methodological characteristics, such as sampling and measurement. When comparing vigorous with low-intensity exercise, only two trials were found, and these showed no significant effect, a result repeated for two trials investigating children receiving psychological treatment. This might suggest that lower levels of physical activity intensity may also be effective for anxiety reduction.

When comparing exercise with psychosocial interventions, Larun et al found only two trials, and no significant effect was evident. This was also the case for one trial involving children receiving psychological treatment, suggesting that physical activity may be equally effective as psychosocial interventions. There was no evidence for intervention effectiveness after 8 weeks.

Limitations of findings include the use of broad inclusion criteria that allows participants in some interventions to have rather mild depression over a short time frame (ie, they may not be particularly depressed at all). Studies also fail to specify the exact nature of the physical activity interventions in respect of frequency, intensity, duration and type of activity. Group-based physical activity interventions often fail to control for the effects of social interaction.

In summary, physical activity over no intervention seems to be potentially beneficial for reduced depression, but the evidence base is limited. Intervention designs are low in quality, and many reviews include cross-sectional studies that may distort associations or fail to rule out reverse causality.


Active adults report fewer symptoms of anxiety than inactive adults.12 13 However, the amount of evidence for young people is considerably less. Our search revealed four reviews (see table 2). The meta-analysis of exercise and anxiety reduction conducted by Petruzzello et al14 also reported ESs for those younger than 18 years. Results showed a small-to-moderate effect for physical activity programmes when assessing trait anxiety (ES = −0.47), although only three studies were available for review. The review by Calfas and Taylor7 found only three intervention studies on anxiety and reported an ES of −0.15. This review does not completely map onto the studies reviewed by Petruzzello et al and lacks methodological detail.

Table 2

Reviews of physical activity and anxiety in young people

Larun et al6 conducted a systematic review of exercise interventions on anxiety in young people up to the age of 20 years. Six trials were found that investigated whether vigorous exercise conferred benefits over no intervention. They found a non-significant, although small to moderate, trend (ES = −0.48). Studies were of low quality and highly varied in respect of methodological characteristics. When comparing vigorous with low-intensity exercise, only three trials were found, and these showed no significant effect. When comparing exercise with psychosocial interventions, Larun et al found only two trials, and no significant difference was evident.

Wipfli et al15 claimed to address the methodological weaknesses of previous reviews by conducting a meta-analysis of just randomised controlled trials, although this was also performed by Larun et al.6 Three trials were reviewed for those younger than 18 years, revealing a non-significant effect (ES = −0.18). It is not clear why Wipfli et al only found three interventions 2 years after Larun et al analysed six.

In summary, physical activity interventions for young people have been shown to have a small beneficial effect for reduced anxiety. However, the evidence base is limited and in need of development.


Self-esteem reflects the degree to which an individual values himself or herself and is widely viewed as a key indicator of positive mental health and well-being.16 The belief that physical activity, including sport, is associated with the development of self-esteem in young people is a commonly held view. Typically, global self-esteem is seen as the apex of a hierarchical and multidimensional framework, underpinned by different domains of the self, including the perceptions of physical self-worth.17

We found three systematic reviews addressing physical activity and self-esteem in young people7 18 19 (see table 3). One of the first meta-analyses in exercise psychology was that by Gruber (1986). He meta-analysed 27 experimental designs, mainly on children, but gave limited detail on methodology.

Table 3

Reviews of physical activity and self-esteem in young people

Gruber19 reported an overall effect for physical activity on self-esteem of 0.41, representing a small effect. Larger effects were found for children with perceptual, emotional and learning disabilities (ES = 0.57), for better controlled experiments assessing only one dependent variable (ES = 0.65) and for aerobic fitness activities (ES = 0.89). Three studies analysed by Calfas and Taylor7 showed a small effect (ES = 0.12).

A recent Cochrane meta-analysis by Ekeland et al18 examined whether exercise interventions improved global self-esteem among children and young people aged 3–20 years. This was later published as a journal article.20 The results showed that in the eight trials available for meta-analysis and testing an exercise-alone intervention versus a no-intervention control, there was a small-to-moderate effect in favour of the intervention group (ES = 0.49).

Most of the trials analysed by Ekeland et al18 were small scale and of short duration. No follow-up results were given so the sustainability of changes could not be assessed. Only one of the trials was considered to be of high methodological quality and demonstrated the strongest effects. A further four trials compared the effects of exercise as part of a comprehensive intervention package against no-intervention control groups and showed a moderate positive effect on self-esteem in favour of the intervention (ES = 0.51).

In summary, physical activity can lead to improvements in self-esteem, at least in the short term. However, there is a paucity of good quality research. Moreover, global measures of self-esteem can be affected by many factors beyond physical activity. Hence, measures of physical aspects of the self, such as body image or physical self-worth, important indices of psychological health in their own right, might be better targets for intervention.

Cognitive functioning

A link between physical activity and cognitive functioning has been believed to exist for many years. Blakemore21 reported that the brain is activated during physical activity by increasing blood flow to essential areas that may stimulate learning.22 Moreover, research has suggested that integrating physical activity in the classroom will enhance student learning.23 However, further studies are needed to make school authorities confident that encouraging school physical activity will improve learning rather than disrupt academic time. Cognitive functioning is best defined as a) intelligence, which is the ability to reason quickly and abstractly; b) cognitive skills of concentration and attention; c) academic achievement, usually assessed by overall school grades and performance.

Seven review articles were found (see table 4). More than a decade ago, Shephard24 reported a systematic review to investigate links between physical activity and cognitive functioning using cross-sectional and longitudinal studies. It was found that increasing physical education (PE) time in schools by 14–26% did not have a significant negative effect on academic performance. However, only three studies on routine physical activity were involved, and evidence was not established concerning additional time for physical activity and intellectual enhancement; students who had extra PE time performed equally well as students who had intact classroom curricular time.

Table 4

Reviews of physical activity and cognitive functioning in young people

Sibley and Etnier25 conducted a meta-analysis to investigate the effects of different types of physical activity on cognitive functioning across varied age groups of young people. All research designs were included. The findings revealed that physical activity was associated with better cognitive functioning across all age groups. A small effect for chronic studies was found (ES = 0.29).

Trudeau and Shephard26 conducted a systematic review on physical activity, physical fitness and academic performance. Seventeen studies of varied research designs were involved. Findings indicated that allocating up to an additional hour per day of curricular time to physical activity programmes in school does not affect academic performance of primary school students. Furthermore, physical activity improved classroom behaviour, attention and concentration. A positive but weak relationship was established between physical fitness and academic achievement. However, for experimental studies, the authors did not find strong evidence for an effect of school physical activity on academic performance because children in the experimental group, whose academic tuition was reduced, did not perform better than the control group; that is, children exposed to additional school physical activity achieved equally well as those students who had regular academic tuition.

Tomporowski et al,27 in a systematic review of 15 studies, examined the relationship between physical activity and cognitive functioning among children aged 6–16 years. For prospective and experimental designs, two of four studies showed improvements in intelligence, three of six for cognition and one of six for academic achievement. In contrast, three of four cross-sectional studies on academic achievement showed positive associations. They concluded that children's cognitive functioning can be enhanced through physical activity, but this is mainly in respect of executive functioning tasks (ie, “goal-directed actions in complex stimulus environments, especially novel ones, in which elements are constantly changing”; p. 126).

To update previous reviews, Keeley and Fox28 conducted a systematic review to investigate the relationship between physical activity, physical fitness and cognitive functioning among 4- to 18-year-old students. Eighteen studies with varied research designs were included. Generally studies involved exposing experimental participants to increased school-based physical activities at the expense of academic work time. A weak positive relationship was found between increased school physical activity and cognitive functioning. Specifically, cross-sectional studies indicated that more physical activity was associated with better performance in some subjects (eg, mathematics) but not in others (eg, English). One intervention study reported that increased physical activity significantly improved cognitive functioning; two intervention studies found no association between increased physical activity and academic performance. A weak but positive association was found between physical fitness and cognitive functioning in young people, with the strongest correlations being with cardiovascular fitness.

Although a strong association has not been established between increased school physical activity and cognitive functioning, Keeley and Fox (2009) found that the introduction of more physical activity, at the expense of academic subject time, did not have a detrimental effect on children's academic performance. This reflects conclusions from other reviews.

The Centres for Disease Control and Prevention29 in the United States has also examined the effect of routine physical activity on the cognitive functioning of young people. The review involved 50 studies with cross-sectional (K = 11) and longitudinal (K = 38) research designs, with 32 classified as interventions. It was found that increased PE time in schools had a positive but weak relationship with academic achievement in 11 of 14 studies. Time spent in increased break (recess) play had a small positive relationship with classroom behaviour, including children's attention and concentration (all eight studies found one or more positive associations between recess and indicators of cognitive skills). In addition classroom-based physical activity (5-20 min break during lessons) improved children's academic behaviour and achievement (eight of the nine studies found positive associations). Extracurricular physical activities (participation in school sports) also had a positive association with academic performance (all 19 studies examining after-school sport found one or more positive associations with cognitive performance). Finally the review concluded that school physical activity programmes have a positive impact on academic performance. However, effects are inconsistent and often small.

In a recent systematic review, Best30 examined the relationship between aerobic physical activity and cognitive functioning in young people. Studies used experimental designs; however, only three studies assessed the effects of chronic aerobic exercise on cognitive functioning. Aerobic exercise did not influence children's tasks requiring visual-motor coordination, but it did improve tasks requiring flexible and divergent thinking.

In summary, systematic reviews on physical activity and cognitive functioning have provided evidence that routine physical activity can be associated with improved cognitive performance, classroom behaviour and academic achievement in young people, but these associations are usually small and not entirely consistent.

The major implication arising from these reviews is that integrating physical activity in the school system may help young people to learn better and reduce the likelihood of negative classroom behaviours. However, strong evidence has not yet been established between chronic physical activity and students' cognitive functioning, and this could be due to methodological shortcomings of studies. These include expectancy effects and unblinded intervention designs. Thus, available evidence does not contribute strongly to the proposition that increasing school physical activity time to the detriment of classroom curricular time is beneficial for school children. Studies with more rigorous designs are needed on physical activity, including non-aerobic exercise, and its effect on cognitive functioning.

Sedentary behaviour and mental health

Sedentary behaviour in young people and adults is a rapidly developing area of research. Operationally defined as “sitting time”, sedentary behaviour can be high in the contexts of leisure time (eg, screen time), school and travel (ie, car use). Most of the evidence with young people has centred on screen time, and TV viewing in particular. Although TV viewing remains the most prevalent sedentary behaviour for youth, it is only one behaviour and may not reflect wider patterns of excessive sitting.31 However, the development of attractive home-based or even mobile electronic entertainment has led to concerns about excessive sitting time in young people.

Most of the evidence linking sedentary behaviour to health outcomes has focused on TV viewing and weight status,32 with more recent studies looking at screen time (ie, TV and computers) and aspects of metabolic health more broadly.33 34 Less has been written about sedentary behaviour and mental health in young people. For this brief section of the article, we review recent articles addressing links between sedentary behaviour and mental health, with a summary shown in table 5.

Table 5

Primary studies investigating sedentary behaviour and mental health in young people

Results from the nine primary studies show consistent negative mental health associations with sedentary behaviour, primarily screen viewing. This mirrors a growing literature showing adverse physical health outcomes of high sitting time.35 The one longitudinal study did show that TV viewing, but not computer games, was associated with increased odds of depression after 7-year follow-up.36

The associations across the studies, however, are small and, all but one, are derived from cross-sectional designs. It is plausible that those with poorer mental health choose to be more sedentary at screens—a reverse causality hypothesis. Only half the cross-sectional studies controlled for physical activity in their analyses. Most of the studies are large, with some being very large, and one assessing across multiple countries. Sedentary behaviour needs to be considered alongside physical activity in the study of mental health.

Overall conclusions

In summarising physical activity and mental health in young people, physical activity is likely to have positive psychosocial outcomes. The review of reviews method adopted allows us to see the wider field of several mental health outcomes associated with physical activity in young people. It also allows a view of the gaps and inconsistencies across different outcomes.

The effects seem strongest for self-esteem (at least in the short term), and those who are physically active seem less likely to suffer from mental health problems and may have enhanced cognitive functioning. Evidence on depression is also promising but remains an underdeveloped area of enquiry. Although all participants are likely to gain significant benefits, such effects are likely to be greater in those who have poorer mental health at baseline; however, the evidence is not extensive. Studies are largely cross-sectional (thus unable to rule out reverse causality) and small scale and lack measurement consistency. In addition although physical activity may enhance psychological well-being, it is possible that the prevailing psychological climate and social interactions inherent in such settings will also be crucial. Unfortunately, such factors are rarely accounted for. Higher levels of sedentary (sitting) behaviour are associated with worse mental health.


Thanks to Michelle Taylor for assistance with literature searching and retrieval.


View Abstract


  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.